Public Health Ethics Key - Concepts and Issues in Policy and Practice

Public Health Ethics
Key Concepts and Issues in Policy and Practice
Public Health Ethics

Key Concepts and Issues in Policy and Practice
Edited by
Angus Dawson
CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town,

Singapore, So Paulo, Delhi, Tokyo, Mexico City
Cambridge University Press
The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by
Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521689366
# Cambridge University Press 2011
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.
First published 2011
Printed in the United Kingdom at the University Press, Cambridge
A catalogue record for this publication is available from the British Library
Library of Congress Cataloging-in-Publication Data
Public Health Ethics : Key Concepts and Issues in Policy and Practice /
edited by Angus Dawson.
p. cm
Includes bibliographical references and index.
ISBN 978-0-521-68936-6 (pbk.)
1. Public healthMoral and ethical aspects. I. Dawson, Angus.
RA427.25.P8285 2011
362.1dc22
2010050335
ISBN 978-0-521-68936-6 Paperback
Cambridge University Press has no responsibility for the persistence or
accuracy of URLs for external or third-party internet websites referred to
in this publication, and does not guarantee that any content on such
websites is, or will remain, accurate or appropriate.
Every effort has been made in preparing this book to provide accurate and
up-to-date information which is in accord with accepted standards and
practice at the time of publication. Although case histories are drawn from
actual cases, every effort has been made to disguise the identities of the
individuals involved. Nevertheless, the authors, editors and publishers can
make no warranties that the information contained herein is totally free
from error, not least because clinical standards are constantly changing
through research and regulation. The authors, editors and publishers
therefore disclaim all liability for direct or consequential damages
resulting from the use of material contained in this book. Readers are
strongly advised to pay careful attention to information provided by the
manufacturer of any drugs or equipment that they plan to use.
Contents
List of contributors vi
Preface vii
Acknowledgements viii
Section 1 Concepts
Vaccination ethics
Angus Dawson
Resetting the parameters: public

health as the foundation for public
health ethics 1
Angus Dawson
Environment, ethics and public

health: the climate change
dilemma 154
Anthony Kessel and Carolyn Stephens
Health, disease and the goal of

public health 20
Bengt Brlde
Selective reproduction, eugenics

and public health 48
Stephen Wilkinson
10 Public health research ethics: is nonexploitation the new principle for

population-based research
ethics? 174
John McMillan
Risk and precaution

Stephen John
67
Section 2 Issues
5
Smoking, health and ethics

Richard Ashcroft
Infectious disease control

Marcel Verweij
Population screening
Ainsley J. Newson
85
143
11 Equity and population health:

toward a broader bioethics
agenda 191
Norman Daniels
12 Health inequities
James Wilson
211
100
Index
231
118
Contributors
Richard Ashcroft
Queen Mary, University of London,
London, UK
Bengt Brlde
University of Gothenburg, Gothenburg,
Sweden
Norman Daniels
Harvard University, Boston, MA, USA
John McMillan
Flinders University, Adelaide, Australia
Ainsley J. Newson
University of Bristol, Bristol, UK
Carolyn Stephens
London School of Hygiene and Tropical
Medicine, London, UK
Angus Dawson
Keele University, Staffordshire, UK
Marcel Verweij
Utrecht University, Utrecht,
the Netherlands
Stephen John
Cambridge University, Cambridge, UK
Stephen Wilkinson
Keele University, Staffordshire, UK
Anthony Kessel
London School of Hygiene and Tropical
Medicine, London, UK
James Wilson
University College London, London, UK
vi
Preface
This book serves to introduce a number of central concepts and key issues in public health
ethics. Each chapter is written by an expert in the field and they seek to both introduce and
discuss critically at least some of the relevant literature related to their topic or theme.
However, this book is not merely a textbook, as each chapter also seeks to advance (or in
some cases re-frame) academic debate and practical reflection upon these topics in public
health. This book will be of interest to everyone working in public health and related fields,
policy makers, as well as students in philosophy, ethics, public health ethics, medicine,
public health and critical social science.
vii
Acknowledgements
First, I would like to extend my warmest thanks to the many colleagues and friends that
contributed their chapters to this collection of papers. Public health ethics is a relatively
young discipline but the explosion of high-quality activity in this area in recent years has
been extraordinary. I am very grateful that many of the best minds working on these topics
were able to contribute some of their work to this volume.
Second, I would like to thank everyone at Cambridge University Press who made this
collection possible. Their patience, support and expertise have all been exemplary. Particular thanks are due to Nick Dunton.
Third, I would like to thank Melissa Williams, Ross Upshur and all at the Centre for
Ethics and the Joint Centre for Bioethics, for providing the opportunity to spend time at the
University of Toronto and complete work on the manuscript.
Fourth, I would like to thank Trevor King and Bev Sykes for their help in preparing the
manuscript for publication.
Finally, I would like to thank the editor of the Hastings Center Report and Wiley for
their generosity in granting permission to re-publish material that had previously appeared.
Chapter 8 is a revised and extended version of: Dawson, A. (2007) Vaccination ethics.
In Principles of Health Care Ethics, 2nd edn, ed. R. Ashcroft, D. Dawson, H. Draper and
J. McMillan. Chichester: Wiley and Chapter 11 is a very slightly modified version of:
Daniels, N. (2006) Equity and population health: toward a broader bioethics agenda,
Hastings Center Report, 36(4): 2235.
Angus Dawson
viii
Section 1
Chapter
Concepts
Resetting the parameters

Public health as the foundation for public
health ethics
Angus Dawson
Introduction
In this chapter I introduce a number of different approaches to public health ethics.
However, I do this in a deliberately provocative way. I argue that we need a revolutionary,
rather than evolutionary, approach to the development of public health ethics: in other
words, we ought to reset the parameters that frame this area of applied ethics. I attempt to
argue for this conclusion in the three sections of this chapter. First, I outline and defend
what I consider to be a necessary condition to be met by any adequate theory of public
health ethics. Second, I suggest what I call the traditional liberal approach, currently
dominant in much medical ethics, fails to meet this condition because of the primacy it
accords the idea of non-interference. I also suggest that various proposed alternatives,
although offering some welcome broadening to this traditional liberal position, ultimately
remain restricted by their implicit or explicit acceptance of the parameters set by the liberal
approach. Third, I briefly outline a range of areas where I argue that future work ought to
be directed as a means of developing a sufficiently rich account of public health ethics: a
substantive account that meets my condition. I suggest that such an account must accept a
view of human interests as intrinsically social. My primary focus in this chapter is a general
argument in favour of the re-orientation of the field of public health ethics. I do not defend
any particular theoretical perspective, beyond a general defence of what I term substantive
accounts of public health ethics.
I begin this discussion with the observation that if you approach public health from
the perspective of much contemporary medical ethics, many public health policies and
activities are likely to be viewed as ethically dubious. This is for a number of reasons but
will include the following: public healths primary focus on populations rather than
individuals; public healths assumptions about necessary features of the human good;
and a broader focus on other values beside non-interference. Consider just a few
examples of core public health activities: cancer screening programmes are designed to
reduce the number of cases in a given population, through the early discovery of
asymptomatic cases. This focus might mean that informed consent and individual
decision making are less of a priority than in some other areas of health care. Most
preventive vaccination programmes seek to reduce the risk of individuals being infected
with harmful diseases, through the creation and maintenance of a population effect
called herd immunity. Such programmes aim to maximize participation because if
Public Health Ethics, ed. Angus Dawson. Published by Cambridge University Press. # Cambridge
University Press 2011.
Section 1: Concepts
insufficient individuals contribute, then this important protection cannot be achieved.

A healthy environment with adequate sanitation, clean water and good air quality
requires the coordinated activity of the whole community, through the day-to-day
action of relevant civil or public agencies. This may impose significant cost on some
industries and individuals. Health promotion can seek to change peoples preferences in
relation to issues such as smoking, exercise and food choices, with the aim of reducing
the chronic disease burden in a population through the promotion of healthier lifestyles.
Many public health research activities are focused on populations, where epidemiological work to determine risk factors for disease may require the analysis of personal
health information without the consent of individuals. Preparing for and responding to
public health emergencies may require infrastructure for disease surveillance and legal
structures to compel behaviour and seize property (in at least some circumstances).
Health inequities are the result of many different socio-economic determinants and can
often be addressed only through structural and societal level policy initiatives (that may
in turn restrict or negate individual choice).
If the currently dominant views in contemporary medical ethics are applied to these
public health activities there is a danger that such routine public health actions will be seen
to be wrongly prioritizing population over individual interests. One possible response is to
accept this critique and argue that much public health activity is actually unethical because
it fails to prioritize individuals and their choices, as well as the moral principles that have
evolved within the field of medical ethics to protect these considerations such as informed
consent and patient confidentiality. However, an alternative approach is to argue that public
health is a vitally important activity and that its ends are legitimate and can only be attained
through such population-level interventions. On this view, the problem, assuming it is
one, lies with the perspective derived from traditional medical ethics, not with public health
practice itself. On this latter view, it is certainly possible for public health actions to be
unethical, but the mere fact that they do not easily fit within a medical ethics framework
does not make them unethical; and we certainly have no a-priori reason to hold public
health to be intrinsically unethical. In this chapter, I argue that we face an important
dilemma. Either we explicitly accept the consequences of our liberal framework (and damn
much routine public health practice) or we choose to re-set our parameters and rethink our
ethical theories, thereby ensuring that public health activities and their justification move
closer to the core of ethics. In this chapter I argue that it is time for us to move towards the
latter view.
A necessary condition for any adequate account

of public health ethics
In any clash between a chosen moral theory and public health practice and policy we
have no reason to assume that it is the practice or policy that is problematic rather than
the theory. In this section I will argue that we ought to accept a necessary condition for
something being an adequate theoretical perspective in relation to public health, and
that this condition ought to be applied as a filter in choosing an appropriate ethical
theory for public health. I will call this condition the nature of public health condition.
I have suggested that meeting this condition is a necessary feature of an adequate theory.
What this means is not that we can rule out those moral views that fail to meet it, but
that the consequences of embracing any view that does not meet the condition is likely
Chapter 1: Resetting the parameters
to be too great (in the sense that it will entail excluding a substantial amount of public
health policy and practice, which we otherwise find appealing and arguably is necessary
to establish the conditions for living a good life). What is the nature of public health
condition? The main idea behind it is the thought that any adequate account of public
health ethics must begin with a clear articulation and defence of a concept of public
health. I suggest this is the case because the aims, nature and methods of public health
appear to be different from those of much clinical medicine. If this is true, then we have
good reason to be cautious in simply applying our methods and results from clinical
medicine (the main focus of most discussions in contemporary medical ethics) in the
sphere of public health. However, this focus on the definition of public health is more
difficult than might at first appear because the concept is itself a contentious one. What
do we mean when we talk of public health? What are the legitimate aims of public
health (Munthe, 2008)? There is, of course, a huge literature discussing the concept of
health and a smaller but growing literature discussing the concept of public health
(Brlde, Chapter 2).
In previous joint work with Marcel Verweij we explored many of the most influential
definitions of public health that have been offered. Our own rough account of public
health focuses on this area of activity as being characterized by collective interventions
that aim to promote and protect the health of the public (Verweij and Dawson, 2007:
21). This is further articulated in terms of two different but equally important senses of
public being contained in the notion of public health. These two senses of public in
public health can be structured around, first, the idea of the health of the public as a
social entity or a target for an intervention (that is, a population, community or group)
and, second, public as a description of the mode of intervention, which requires some
form of collective action. As a result, on this approach, we end up with a particular view
of the elements of public health, but also with an agenda for some of the issues to
discuss within public health ethics. There are many issues that require clarification, but
I will consider only three important central issues here. First, public health is focused
on populations (not just individuals). Second, much public health work is preventive
rather than curative. Third, most public health improvements cannot be brought about
by individuals on their own: the attainment of public health ends requires collective
efforts.
First, public health activity is concerned not just with the health of individuals but,
rather, primarily focuses on the health of the population or community. What does this
mean? When we think of a population it is tempting to consider it as merely a collection or
aggregation of particular individuals. There is an important sense in which this is true. If we
have a population and we keep taking away individuals, we will ultimately have nothing left:
the population does not exist as an entity independently of the individuals that collectively
constitute that population. However, while this is important, it cannot be the whole story.
First, this model fails to take into account such things as the way that unequal socioeconomic determinants influence health. You are not just an individual, but an individual
with a particular position in a society; and that position has a significant impact upon your
health status. Second, we can think of different populations or societies as having better or
worse health than others and we can think of a particular populations health improving
or worsening over time (perhaps because they have lower or higher inequities or a lower or
higher overall disease burden). Both of these examples might be used to express a concern
about the publics health in these populations, and may provide a basis for aiming to
Section 1: Concepts
improve a populations health through population-level interventions. Third, population

health cannot merely be viewed as an additive correlate of the health of the constituent
individuals. For example, it looks as though, paradoxically, in targeting population health
we can improve the health of the individuals in that population, but if we target the
individuals as individuals then we may not improve that populations health. This is one
of the many things we can learn from the work of Geoffrey Rose (1992) and his focus on the
complexities of the relationship between population health and individual health. One of his
examples focuses on the influence of drinking cultures on individual behaviour in relation
to alcohol consumption and associated adverse medical consequences. The relationship
between the population and individual is a complex one. While no one would deny that
there may be individual factors such as an individuals genetic makeup that contribute to
the risk of being a heavy drinker, it is important to see that the behaviour of individuals in
relation to alcohol consumption is very strongly influenced by population factors such as
social attitudes to alcohol and the resultant legal and political climate. The more that is
drunk by the average citizen in a population, the more heavy drinkers there will be. Indeed,
even more specific correlations can be drawn between alcohol and disease at the population
level in some situations, such as the drop in cases of cirrhosis in France during times of
reduced access to wine during the two world wars (Rose, 1992: 85). What this suggests is
that lifestyle choices are not simply within the control of individuals. If this is true, then we
should take care in attributing causation (and therefore responsibility for lifestyle behaviour) to individuals alone.1
Second, public health activity focuses on seeking to prevent, reduce or ameliorate harm,
not just treat patients after a negative event has occurred. This is an intuitively powerful
idea. Such interventions require an inference to be drawn from known population risks and
applied to the lives of individuals. This, in turn, will require interventions focused on
asymptomatic individuals, and this can generate anxiety and other harms (Newson,
Chapter 7). Such harms, then, need to be weighed against the benefits of prevention. The
scope for prevention is vast, and it is important for public health to intervene only where it
is appropriate. However, deciding when this is the case is difficult. For example, what limits
ought we to place on the idea of harm? What kind of harms are relevant (Dawson, 2007;
Verweij, Chapter 6)? Public health operates with a broader notion of harm than that
commonly employed in contemporary medical ethics. For example, public health is concerned not just with the immediate factors that impact upon peoples lives but also with the
prevention and reduction of harm as well as the wider determinants of health and many of
the factors that shape the kind of society within which we wish to live. Such influences upon
our health are often best described in terms of probabilities and risks, and so public health is
often motivated by a concern for uncertainty and precaution (John, Chapter 4). The more
complex or broader the notion of harm as the focus of public health, the more likely it is
that the benefits and burdens calculations will become increasingly difficult (and contentious). If it is true that much public health can only be performed through collective activity,
then the bringing about of such ends will entail coordinated action (and, if the end is judged
to be sufficiently important, perhaps, in some cases, coercion too).
See, also, Paul (2009) for an excellent discussion of how Roses work may help us rethink HIV
prevention.
Third, public health requires collective action, as many desired public health ends are
impossible to achieve for individuals by themselves. In reality, collective activity usually
means state action on behalf of society as a whole. This fact can often result in the charge of
paternalism, with the state taking an active decision-making role in relation to the best way
that people should live their lives. Of course, we might question if such activities are really
paternalistic, if this is the only way to secure these ends for the good of the whole population
(Nys, 2008). But, even if the charge of paternalism is fair, we can still ask the question
whether such paternalistic action is always wrong. We have no reason to just assume that all
cases of paternalism are wrong by definition (in the absence of an argument to establish this
rather odd conclusion).
How will these considerations work out in practice? Consider a brief example related to
the key contemporary policy concern of rapidly rising rates of obesity in many parts of the
developed world. First, with this approach it is important to see that this is a population
problem (as well as being a concern for individuals). It is a population problem in a
straightforward epidemiological sense, as we can measure obesity at the population level
by analysing the differences between countries and within countries and relate these
differences to other population features such as socio-economic factors. I do not mean to
suggest that obesity is simply a matter of poverty: it is not. But it looks as though there is an
association between obesity and socio-economic status, in that more affluent individuals are
less likely to be obese (probably for very complex reasons: better access to information,
better quality and variety of food, greater opportunities to exercise in stimulating ways,
etc.). Second, prevention is central to obesity because we have enough empirical evidence to
suggest that while prevention of obesity is difficult, treatment is virtually impossible (except
for surgery, which carries risks significant enough for this to be an option only for those at
very high risk from their obesity). Third, and relating back to the fact that obesity is a
population problem, collective interventions will be vital if we are serious about tackling the
issue of obesity. Such interventions will have to be at the societal level and are likely to
include profound changes to a number of factors including the nature of work and
schooling, the built environment, transport policy, the regulation of the food industry
and the possible restriction to food advertising (especially in relation to children). We
might argue that such interventions do not count as paternalism (as the focus is on the
collective) or that they do (but they are still justifiable). In either case, it might be argued
that such collective interventions may be permissible.
In conclusion to this section, it is vitally important to be clear about what we mean by
public health before we begin to explore public health ethics. Any theoretical perspective
orientated towards public health must, I argue, be responsive to the aims and nature of
public health: too often discussions on public health ethics fail this test. Attempting to
construct a public health ethics without a substantive notion of public health will inevitably
result in error. Setting the correct parameters is the first step in trying to attain the correct
perspective upon this vital area of health care practice.
The inadequacy of liberal medical ethics as a means of thinking

about public health ethics
One important aim of this chapter is to argue that much of the work that has been done on
public health ethics so far, even that explicitly aware of the need for something more than a
liberal approach, has remained locked within the parameters set by the traditional medical
Section 1: Concepts
ethics framework. I argue here that we need to re-frame the way that we think about public
health ethics and move away from the assumption that public health ethics ought to be
structured in terms of debates about non-interference and the subsequent central preoccupation with apparent conflicts between individual and population. This approach,
which I will label as liberal, encourages the idea that a particular value or set of values,
primarily attached to individuals and their decision making (such as liberty and autonomy)
has priority in our moral deliberations. Approaching things in this way places the onus on
those seeking to argue that such values should not always hold sway in public health to
justify situations where it is appropriate to restrict or interfere with an individuals liberty.
In this chapter I argue that this way of conceptualizing things is part of the problem and is
so dominant that it tends to be assumed without argument. I will begin by discussing what
I term narrow or pure liberal views and then move on to what I call moderated liberal
views.
Narrow liberal views
The exact meaning of the term liberal can, of course, be disputed. However, I will take it
here to imply a set of commitments, long dominant in contemporary medical ethics, that
draw upon a particular and narrow reading of the harm principle taken to be derived from
John Stuart Mills On Liberty. Within this view, the only ground for coercive interference in
the decision making of individuals is when their actions may have negative consequences
for others. Any action to reduce or prevent harm to an individual, once they are aware and
informed of the relevant danger, is held to be a case of paternalism and thereby morally
wrong. Within this view, liberalism is seen as centrally concerned with non-interference.2
However, there are a number of problems with this view. First, it is not clear that this is
really Mills considered view. He explicitly includes action to preserve public goods within
the list of acceptable reasons to restrict liberty. In this sense, Mill is not a Millian liberal in
the way that many imply (Dawson and Verweij, 2008). Second, as mentioned before, when
we talk of the concept of harm in relation to public health practice and policy we are
interested in much more than harm to others as traditionally conceived. Public health
actions are designed not merely to prevent harm, but also to reduce or ameliorate it. The
relevant notion of harm implicit in routine public health activities is much broader, more
contextual, more interested in the social reality of actual lived lives, and more about the
conditions that are required to live a healthy life. As a result the traditional distinction
drawn in much medical ethics between beneficence and non-maleficence is less obviously
relevant (Dawson, 2007), and if this is true, this makes it more difficult to defend the very
coherence of the idea of non-interference. At the very least advocates of the Millian view
need to defend such a narrow conception of harm, and it is not clear that this is possible.
Even if it is, we might still argue that discussion of public health requires a broader
conception (that is one that permits harm prevention and harm reduction) to make sense
When I write of liberalism in this chapter I just mean the dominant Millian strand that dominates
much contemporary medical ethics. There are, of course, ways of formulating richer versions of
liberalism, such as that due to Raz (1986). If anyone is offended by my characterization of liberalism
they can read my text as referring to a very particular type of liberalism (perhaps, one to be thought
of as liberalism*). I will also leave to one side the possibility of other options in choosing ones
political philosophy, such as varieties of republicanism (Jennings, 2007a).
of the very idea of public health.3 So, for both these reasons, an appeal to the harm
principle will not result in as clear a policy directive as may be assumed: public health
ethics (and, in my view, medical ethics in general) cannot be built upon such shaky
foundations.
However, let us assume, for the sake of argument, that this is not the case and the idea of
non-interference (allegedly derived from the harm principle) is robust. The first thing to
note is that quite a lot is built upon the idea of non-interference. For example, priority will
be given to individual freedom or autonomous decision making because it is for individuals
to decide what they should do. Liberals tend, also, to support the importance of maintaining
a clear distinction between the private and the public. The private represents an area where
the state or its representatives have no legitimate reason to trespass. With this approach, it is
easy to see how public health ethics can get conceptualized as being about protecting the
sphere of the individual from the interference of state power. An example of this might be a
view of the limits of ethical health promotion as being related to the provision of information as a means for individuals to make their own decisions about what to do. Peoples
existing preferences are to be respected because they are their preferences, and as a result the
state ought to be neutral and not promote particular views of the good life.4
Liberals need not see liberty as the only value that matters or the value that always takes
precedence (perhaps it is that which distinguishes them from libertarians). However, there
is a problem here for the advocate of non-interference that I will briefly explore. The
problem arises from a failure to recognize that non-interference or neutrality towards all
other values is itself a value. There are two coherent options here. First, assume that noninterference itself can be given a different status to other values. The advocate of noninterference looks as though they ought to embrace this option, but then we need an
argument to establish why we ought to see non-interference as a higher or second-order
value (and this account must also explain how such a higher value can cohere with other
values). Second, and alternatively, while non-interference and its cognates, such as autonomy and liberty, are seen as important values, we have no good reason to assign them any
special status. In this view, we have a range of important values of equal status that can be
weighed against each other. Each of these values may take priority over the others in some
contexts. Sometimes liberty is the winning value, but at other times it is not. With this view
it makes no sense to frame the discussion of these issues in terms of only liberty. A nonprivileging account of values in public health ethics will allow liberty to be legitimately
defeated on at least some occasions, perhaps because there will in turn be more liberty
further down the way or because other values are just more important on that particular
occasion. If this second approach is true, as I think it must be, then non-interference as a
privileged value is not a coherent option.
3
Public health is, of course, often concerned with what we can think of as the background conditions
for living healthy lives. It is quite a stretch (perhaps even incoherent) to think of many of these
activities as coercive, and so it might be argued, once again, that non-interference is too narrow a
principle to capture all that is relevant to public health. Thanks to Adrian Viens for discussion on
this point.
Dan Wiklers work (1978) is a good example of a Millian approach to the ethics of health
promotion. Holland (2007) provides another well-worked out position defending a liberal approach
to public health ethics more broadly. See Jennings (2007b) for a critical perspective upon liberal
approaches to public health ethics.
Section 1: Concepts
In addition, a common inference from the liberal non-interference approach to framing

debates in public health ethics is an assumption about responsibility. It is usually assumed
that individuals have the freedom to make choices (whether or not they in fact do), and
therefore responsibility for the consequences of those choices is attached to the individual
choice-makers. The danger is that we are offered a rather simplistic view of both choice and
responsibility (one that tends to ignore the way that many choices are partly or even largely
the product of factors beyond the individuals direct control [such as socio-economic,
historic, geographic or cultural factors]). The liberal faces a dilemma in such a case. Either
the individual is not free because they do not make choices in the relevant sense (but this
may mean that no one is free) or we begin to take seriously the social construction of choice
(in which case, it turns out that the degree to which we can be free is restricted in some
sense. Perhaps there are only a sub-set of choices that are free, or each choice is only free to
some extent). Of course, it has long been recognized that the relationship between causal
and moral responsibility need not be a straightforward one. However, if it looks as though
the causes of behaviour are not merely the result of an individuals choice, then it is clearly
not appropriate to attribute responsibility for the consequences of such lifestyle choices to
individuals in any meaningful sense. What this raises is the possibility that many liberals are
working with a deeply implausible view of human psychology and a potentially morally
problematic view of responsibility attribution.
So leaving these more theoretical concerns about non-interference to one side, I now
turn to some possible reasons for why the liberal framing has been so powerful. I think the
liberal tradition in contemporary medical ethics has been supported by at least three
features: the history of medical ethics as a discipline; its relation with the law; and a set of
assumptions about pluralism. These features are partial explanations as to why the parameters are set in their current position. However, I suggest that none of these three reasons
provide any convincing justification for why we must remain locked within such a
framework.
First, as many people have now noted, the history of medical ethics from its early years
was focused very much on dyadic clinical relationships between doctor and patient and a very
narrow set of issues either related to such a relationship (for example, consent and confidentiality) or a view of ethical theory and principles focused on individual patient rights and
autonomous decision making. The consensus that health care was too paternalistic resulted
in the de facto establishment of respect for individual autonomy as the dominating principle
in medical ethics. Other areas of bioethics, related to animals and the broader environment,
tended to be downplayed. The other factor that has driven much ethical discussion is the
apparent glamour of new technologies and cutting-edge medicine. Much contemporary
medical ethics can be seen as dwelling in one of two camps: those with a tendency to see
technology as providing solutions and those suspicious of it. More recently, many writers in
medical ethics have started to shift their focus, and there is now growing interest in issues
relating to infectious disease (particularly due to SARS, pandemic influenza, tuberculosis,
etc.), the impact of social disparities upon health at the national and international level and
the renewed interest in global justice, particularly in relation to arguments surrounding the
impact of intellectual property issues upon access to medicines. As I have already mentioned,
many others have said that medical ethics must be revised to accommodate these issues.
However, I want to go further and suggest that many of those that have argued that medical
ethics needs to expand, often apply traditional frameworks to issues in public health, and
thereby fail to capture what is special or different about public health ethics.
The second feature that has tended to support the liberal approach is a set of assumptions about the relationship between law and ethics. There is a common tendency to
confuse the two, and this may relate to the apparent obsession that many working in
medical ethics seem to have with the issue of regulation of health care practice (often with
the assumption that anything is and ought to be permitted unless it is explicitly squashed by
law, resulting in the focus, too easily, becoming one of ensuring that regulation is minimal).
The relation between law and ethics is a complex one, but the main point is that the two are
distinct, although they may be related. The problem with confusing the law and ethics in
relation to public health is that the law too often works with narrow accounts of both
causation and responsibility, with a focus on individual action. This can be seen, for
example, in relation to both tort and crime (Coker and Martin, 2006; Martin, 2009),
although the law may also be used in other ways to promote public health (Gostin and
Stone, 2007).
Third, there is an assumption in much contemporary medical ethics that as we cannot
agree in our moral judgments we, therefore, ought to be committed to pluralism in ethics in
general. It is then concluded that we must focus on process values rather than pursuing
substantive answers to ethical questions. The relation of these ideas to liberalism is the
thought that we can remain neutral in terms of values and allow individuals to make their
own decisions and pursue their own view of what is morally appropriate. However, all of
these commitments can be contested. First, the fact that different perspectives exist upon an
ethical issue does not on its own have any implications for our normative views. There
needs to be further substantive argument to establish such a claim. Second, we need to take
care when talking about pluralism. This is not value pluralism (there is more than one
morally relevant value) but judgment pluralism (there is more than one answer to a moral
issue). The former does not imply the latter, and it is the latter that the supporters of such
relativistic pluralism need to establish. It should also be noted that you can be both a value
pluralist (there is more than one value) and a moral realist (there are objective answers to
moral questions) at the same time. Third, one thing that drives the liberal neutrality view
here is a commitment to tolerance. However, it often seems to be missed that judgment
pluralism cannot easily be combined with a coherent defence of a value such as tolerance.
Indeed, a commitment to tolerance is most easily defended from a realist tradition (that is
we ought to be tolerant, even if other people think differently). Fourth, a commitment to
procedure over content is not remaining neutral about values, but just choosing to adopt a
particular account of ethics: one committed to procedural values as though this were not
just a commitment to a particularly thin set of substantive values.5
Whether or not the liberal approach is an appropriate one for public health, and
therefore for public health ethics, it is certainly the case that if we adopt this kind of liberal
framing of public health ethics, many aspects of routine public health practice will be ruled
out as unethical and this approach will not be able to capture the more substantive notion of
public health outlined earlier. In other words, narrow liberal views will fail to meet my
suggested condition for an adequate theory of public health ethics.
Of course, all of these issues are much more complex than I suggest here. My intention is just to
illustrate how the easy moral relativism of our times fits with the alleged neutrality of liberalism.
10
Section 1: Concepts
Moderated liberal positions
Are moderated liberal views any more successful? I call such views moderated liberal
positions as they clearly suggest dissatisfaction with a simple liberal position (for
example, one built solely upon an appeal to non-interference or the harm principle)
in regard to an adequate public health ethics. However, I suggest that these views are
still too cautious or modest. They remain locked within liberal parameters: with, despite
their apparent pluralism, an implicit commitment to giving priority to the liberal
values of freedom and autonomy. The first three views that I discuss here, Upshur
(2002), Childress et al. (2002) and Gostin (2005) are principled approaches. They are
essentially attempts to highlight a useful and pragmatic set of issues, with a clear focus
on practical implementation for those working in public health practice and policy. This
is a laudable aim. However, I suggest that there are specific problems with each view and
there are general problems for any principled approach. Lastly, in this section, I outline
and discuss the recent proposal for a stewardship model described by the Nuffield
Council of Bioethics (2007). I classify the latter view as a modified liberal view because,
like these three principled approaches, it seems deeply committed to working within
liberal parameters.
I will begin by just stating the three principled views. First, Upshur (2002) offers us a
set of four principles for the justification of public health interventions as follows:
1. harm principle;
2. least restrictive or coercive means;
3. reciprocity principle;
4. transparency principle.
Second, Childress et al. (2002) in a paper involving ten authors, many of them wellknown names in public health ethics, are much more ambitious, in that they are
interested in sketching out an account of public health ethics, in the course of which
they outline a set of moral considerations generally taken to instantiate the goal of
public health as follows:
1. producing benefits;
2. avoiding, preventing, removing harms;
3. maximizing utility.
They then offer five justificatory conditions for interventions to promote such public
health goals:
1. effectiveness;
2. proportionality;
3. necessity;
4. least infringement;
5. public justification.
Third, Gostin (2005) outlines a set of what he terms public health values as follows:
1. transparency;
2. protection of vulnerable populations;
3. fair treatment and social justice;
4. the least restrictive alternative.
11
In my comments I will concentrate on general remarks about this kind of approach,

rather than focusing on each account in any detail. The first thing to notice is that while
there is some overlap between the features picked out, there are also some significant
differences. This is partly due, no doubt, to the different aims of each of these three
perspectives but also the different way they are presented. Upshur explicitly calls them
principles, Childress et al. call them considerations and Gostin calls them values. So
this leads to the first set of questions related to the status of these different elements. Are
they in fact principles at all? They are clearly supposed to be normative in some way
but how? All three accounts combine both procedural (transparency, proportionality,
public justification, etc.) and substantive (harm principle, least restrictive, social justice,
etc.) elements. Do these different elements have equal status or value? Are they prima
facie principles or values or more substantive than this? These questions then lead on
to the second set of issues related to how they are to be used. Are they to be literally
applied to cases or are they just things to be kept in mind? Are they to be used as
justifications for actions as suggested by Upshur and Childress et al.? How are they to be
combined? What happens when they conflict? If a potential action that would bring
about greater social justice is more restrictive is this permissible or not (Gostin)? If a
satisfactory implementation of our obligation of reciprocity seems to require a more
restrictive option is this permissible (Upshur)? If a possible option is likely to be more
effective but will be a greater infringement is this justifiable or not (Childress et al.)? Is
the order of the principles significant? Upshur places transparency last, whereas Gostin
has it first. There are many different possible options when it comes to ordering and
ranking principles and we are owed some account of the justification of the weighting to
be given to the relevant principles and an explanation about how deliberation ought to
occur (Dawson and Garrard, 2006).
Of course, it will be argued that this discussion is unfair as these approaches are not
supposed to be well worked out theories of public health ethics. This is true. Perhaps
they are best thought of as frameworks for thinking about issues.6 But, even so, the
problem is that such approaches are taken very seriously. They are taken up by busy
people and influence the way that people operate and think about these issues. In the
end, of course, they perhaps say most about the dominant values in society, and in this
sense, I would argue, they illustrate my point about the dominance of liberal
assumptions. So, for example, all three views mention the idea of least restrictive or
coercive means/alternative/infringement. This clearly points at the high value to be
assigned to individual freedom (in the sense of both autonomy and liberty). There is
no corresponding mention of the common good, public good or community benefit.
There is no statement that captures the two senses of public in public health as outlined
6
Although they do have more normative content than some other proposed frameworks, such as that
proposed by Kass (2001). Kass seems to think that her approach, and perhaps frameworks in general,
can be normatively neutral. However, her framework incorporates many values and assumptions
(mostly ones clearly deriving from a liberal approach). The distinction between a theory and a
framework is not a clear one, but we might, roughly, think of a theory as providing normative
justification and a framework as being more focused on aiding deliberation. The framework must, of
course, be supported ultimately by normative commitments, even if they can be put to one side at the
moment of deliberation. The best framework will capture all of the relevant issues: ones that are
partial will not produce the appropriate answers. For more on frameworks, see Dawson (2009).
12
Section 1: Concepts
above: public as a social entity (to which every individual belongs) and as requiring
collective action. These three accounts seem to broadly accept the minimalism and
individualism of liberal approaches, except for Upshurs invocation of reciprocity and
Gostins mention of social justice (and, perhaps, his idea of protecting vulnerable
populations). So while each of these three views has details specific to it, and to be fair,
they should be considered in more detail, they all share a particular approach. Although
they are an improvement upon the narrow liberal view, they inherit and share too much
of the general liberal approach, and for that reason they fail my condition for an
adequate public health ethics.
The last view I will consider in this section is explicitly a modified liberal view. It is
called the stewardship model and is proposed by the Nuffield Council of Bioethics (2007).
They argue that an approach built upon a narrow reading of the harm principle is
inadequate, as they draw attention to the importance of being able to justify support for
government action in relation to health care provision to ensure a framework for adequate
equality of health care access. They suggest that a fuller liberal approach can be supported
by what they term the stewardship model. They do not really explain what this view is in
any detail and provide little support for it in a theoretical sense. They focus instead on
sketching out what such a model would support in practice. It is explicitly said to be a
revised liberal approach. It is certainly wedded to strong liberal assumptions (visible in
their intervention ladder and their discussion of suggested policy options) and it clearly
fails my condition. There is some discussion about what they mean by the concept of public
health, but it does not seem to be carried over into the development of the stewardship
model. Does it move beyond the Millian paradigm that it is critical of? I think not (Dawson
and Verweij, 2008).
I have argued that, despite their differences, both narrow and moderated liberal views
fail to meet my condition of adequacy because they share a common set of assumptions that
frame the way that ethical decision making is viewed. Progress can only be made in public
health ethics by seeking to overcome the hold that this framing has on our deliberations.
Fortunately, there are many different options that are still open to us, and I will explore
some of these in the next section.
Resetting the parameters: in defence of a substantive

public health ethics
A revolutionary or what we might call a substantive account of public health ethics will
begin elsewhere: not with the individual and the assumption of liberty as our prime value,
but with a set of values that try to capture the public nature of public health. This approach
will begin with some defence of the different normative factors that might be used to take
seriously what we share in common as human beings and social creatures. Taking this as
the basis of public health ethics will lead to a re-orientation of much of the field of
discussion and to the re-setting of the parameters as envisaged by the core metaphor
of this chapter. This change in perspective will, in turn, set a clear agenda for the work of
public health ethics.
If we are serious about defending public health activities from the traditional criticisms
drawn from the direction of medical ethics, then we need to think about how we can justify
a more population- or community-orientated approach to ethics. However, this does not
just mean an easy flip into some form of communitarianism. Although communitarianism
13
is one option, there are many other ways of attempting to capture the publicness of public
health through the exploration of a series of theoretical questions. How does public health
relate to such concepts as social justice, solidarity and reciprocity? Can we seek support
from moral theories such as contractarianism, welfare forms of consequentialism, or
accounts of human flourishing and capabilities? Can we capture the idea of public goods
and common goods in our approach? In other words, reflection upon the nature of public
health practice, and the ethical issues that arise there, is also likely to result in a welcome
development of our moral theories. However, before we proceed any further, it is worth
sounding two notes of caution.
First, of course, (at least some of) these views may be combined with liberal values or
even be integrated into a more sophisticated version of liberalism.7 The account of
liberalism that I have criticized is a fairly simple, if influential, one. The claim about
using the substantive approaches discussed here for a re-setting of the parameters for
public health ethics is based on the idea that such approaches (to varying degrees) make
no assumption that we ought to begin with individuals or individual decision making,
nor that liberty or individual autonomy ought to be treated as having a special status in
a set of values. Of course, this does not mean that liberty or autonomy are not important
values. The argument is about the special or privileged status that these values are often
given in discussions. I argued above that approaches to public health ethics that focus on
such values alone will fail to meet my condition for adequacy, but this does not mean
that these values are without merit and ought not to play a role in a developing public
health ethics.
Second, it is worth stressing, again, that in the course of this chapter, I offer no defence
of any particular substantive approach to public health ethics. What I do want to do is
briefly show how there is plenty of potential for exploring how we can use these theories
and ideas to begin to ground an account of public health ethics that meets the criterion for
public health ethics I outlined above. It should also be noted that these approaches are not
mutually exclusive and that they may well, eventually, cohere into a well-rounded theory of
public health ethics.
Substantive notions of the good

First, I begin with a broad family of theories that support the idea of an approach to
ethics based around a substantive notion of the good. Perhaps the most discussed such
theory in relation to public health is consequentialism. Consequentialism is really a
family of theories with a common focus on the goodness of outcomes of actions. This
family will include the classical utilitarianism of Bentham, with its focus on maximizing
pleasure and minimizing pain, but is not limited to it, as it will include preferencesatisfaction and objective list forms of consequentialism (Parfit, 1984: Appendix I;
Griffin, 1986). This is not the place to discuss the relative merits of these approaches,
but we should note that both the last two options can be used to capture a complex and
pluralistic account of welfare that may satisfy many of the requirements we are looking
for in the field of public health ethics. Furthermore, some non-consequentialist theoretical positions may be seen to be related to such accounts because of their commitment
7
For one discussion of what such a broader liberalism might mean for public health ethics, appealing
to the work of Raz (1986), see Coggon (2008).
14
Section 1: Concepts
to a substantive notion of the good such as human flourishing or a particular notion of

the good life or the capabilities that are necessary to live a good life.8
It remains to be seen how these views might be used in the context of public health
ethics, but they surely offer a rich source of enquiry because they are all seeking to promote
a substantive notion of the good by providing the conditions not just for health, but for a
good life. Public health as an activity is concerned with exactly such a notion by seeking to
prevent, minimize or remove harms, including those that can best or only be tackled
through collective action. This means that such accounts can meet my suggested condition
for an adequate theory for public health ethics.
Collective values: interests, goods and harms

Second, many of the values that are the focus of public health activity are collective in nature.
It is important to see that such collective values are of different types, including collective
interests, goods and harms. I suggest that at least some of these can only be pursued and
attained through joint or common activity, but this is a complex notion. In this section,
I draw on the work of Postema (1987) to offer a taxonomy of the different ways that we might
use the idea of what we have in common.9 This is a necessary simplification, as a first step to
exploring in more detail what normative implications such distinctions might have.10
I begin by defining three related terms: goods, harms and interests. The goods I am
interested in here are those things that are good for humans (or allow human flourishing or
the good life).11 They will include some of the same things that are often appealed to during the
formulation of arguments over the objective list forms of consequentialism (for example,
autonomy, pleasure, health, etc.). Whatever these goods are (and this will be a matter of vigorous
debate), all human beings have an interest or a stake in attaining a sufficient level of each good to
ensure the basis of a flourishing life. Any setback to such interests will count as a harm
(Feinberg, 1984).12 It should be noted that some interests in certain goods will remain even if
they are not desired by the individual. In other words, desiring something may create an interest,
but not all interests require that they be desired for them to exist as interests for that person.
I suggest that when we think of collective values we can distinguish at least three senses of
interest and three correlated notions of goods. Public health (as an activity) will be interested in
all three pairings, even if (as I suggest below) only one is genuinely a common good, because all
relate to the idea of protecting and promoting the health of the public through directed
activities that will require collective action as a means of attaining the satisfaction of the
8
9
10
11
12
There are many options to explore here, but they will include: Sen (1999), Nussbaum (2000) and
Kraut (2007).
I have found Postemas (1987) paper inspiring, and in this brief discussion I borrow much of his
terminology. However, I find at least some of his discussion rather confusing, so I deliberately
simplify and adapt his ideas. I do not explore how exactly my views contrast with his.
Note that my interest here is a far broader one than that of public goods in the economists sense.
They are perhaps just one particular form of the collective entities I am exploring here. I think it is
to be regretted that sometimes people use the terms public goods and common goods
interchangeably. Hardin (1968), for example, is really interested in the provision of public goods
when he discusses the commons.
I leave other kinds of possible goods (and bearers of goods) to one side.
Although I agree with Postema (1987) that we need to allow for the notion of collective harms that
is only weakly present (at best) in Feinbergs work.
15
Figure 1.1 Congruent interests. The ellipse

represents the structures that need to be in place for
our individual interests (represented by the thin arrows)
to be attained.
Figure 1.2 Convergent interests. The ellipse

represents the shared end that needs to be in place for
our individual interests (represented by the thin arrows)
to be attained through the public good (represented
by the thick arrow).
relevant interest. I build upon Postemas work to offer a hierarchy of these collective values and
attempt to make clear the degree of commonness required to create and maintain them. I will
discuss three sets of collective interests: congruent, convergent and common interests.
The first two of these senses of interest are well described by Postema (1987) in his
contrast between congruent and convergent interests. Congruent interests are the same
kind of interests that we have as individuals in the same kind of thing. Such interests
essentially run in parallel with each other and require some minimal collective response as a
means of meeting them (see Figure 1.1). An example might be the enforcement of safe food
preparation techniques in restaurants and food processing plants. Such congruent interests
are not merely a group of separate private interests, as they require public structures,
processes or regulations to be in place for these individual ends to be attained. However,
they are only very weakly public. It is convenient or more efficient to meet such parallel
interests through the action of public health departments.
Convergent interests, by contrast, are interests that converge on the same end and are
more strongly public (see Figure 1.2). Indeed, they will include the kind of things that
economists mean when they talk of public goods. What the relevant criteria are for
something to count as a public good is disputed. However, they often include non-excludability (people cannot be excluded from the enjoyment of the public good once it is
created); indivisibility (the relevant public good cannot be broken up into individual
goods, without destroying the good itself). They often, also, depend upon the cooperation
of a large number of people to create and maintain them (Dawson, 2007). Convergent
interests may be expressed in terms of our individual interests, but as it turns out the only
way to bring about their fulfilment is through the production and maintenance of such
public goods. As a result, the attaining of such goods will often require some potential
means of enforcement to avoid the problems of free-riding. A good example of a public
good, and therefore of a convergent interest, is the creation and maintenance of herd
protection through sufficient vaccination within a community as a means of protecting
individuals from harm (Dawson, 2007). Such convergent interests require a coming
together in a more substantive way than congruent interests: they require the attainment
of a collective end as well as public provision. Contrast vaccination for tetanus (a disease
that cannot be transmitted person-to-person and therefore herd protection is impossible)
and measles (a highly transmissible disease where we can attain herd protection). Public
health will be interested in both, even though the former is fulfilling a congruent interest
16
Section 1: Concepts
Figure 1.3 Private interests. The circles represent the

chosen ends for each individual private interest
(represented by the thin arrows).
and the latter a convergent interest. The former requires the infrastructure to be in place to
satisfy the interest, but the latter can only be satisfied if 90% of the population participate
through individual vaccination.
Just before I move on, it might also help to say a brief word about private interests (that
is, those that are in no way public) (see Figure 1.3). It looks as though, in the end, the only
private interests we can make sense of will be quite trivial wants (for example, the fact that
you want strawberry and I want vanilla ice cream). This is because it is hard to think of the
satisfaction of any more substantive interests, without them collapsing into one of the forms
of collective interests (particularly congruent interests). For example, if I decide to aim to
become a professional ice hockey player or a scientist, the satisfaction of either interest will
require substantive public provision of infrastructure (for example, ice rinks, laboratories
and schools). In these cases although I might claim to be pursuing my own private interests,
I can actually only do so through taking advantage of various collective goods. What this
suggests is that the satisfaction of many of our interests, and certainly most of our healthrelated interests, actually requires essentially collective provision. If this is true, then it pulls
public health closer to clinical medicine, and so, public health ethics closer to clinical ethics:
but it is public health ethics, not clinical ethics, that ought to be dominant and frame the
debate (because of public healths core concern with the satisfaction of the various relevant
collective interests).13
However, I now turn back to the third type of non-private interests: common interests.
They are an example of the strongest form of collective values, and it is they that are related
to what I think of as genuine common goods (see Figure 1.4). Collective interests relate to
the common goods that that we necessarily and irreducibly share as a group or community.14 It will be remembered that I defined goods as those things that are good for humans.
Common goods will include the shared norms and behaviour that create and maintain rich
and substantive social values, such as solidarity and trust. Such goods are good for all
humans. Neither the relevant goods nor the corresponding interests are the result of
individual consent and so do not need to be consciously adopted by each individual; indeed,
many people might not even be aware of the operation of such common interests (and the
value of the corresponding common goods) in their lives despite the fact that they benefit
from them. Some of these aspects are the result of the type of society within which we live, but
we do not have reason to hold them to be mysterious. Examples might include the way that
one society may be more or less accepting of diversity than another; or the fact that a society
has lower HIV rates as a result of particular attitudes to the disease; or, to return to my earlier
example, that changes in relation to societal attitudes, food production and consumption,
work patterns and lifestyle changes result in greater rates of overweight and obesity. Many of
13
14
I think the best expression of something close to my views here is that to be found in Cribb (2005).
Once again, much of the way this is phrased is borrowed from Postema (1987).
17
Figure 1.4 Common interests. Our individual

interests represented by the thin arrows are part of, and
jointly constitute, the common interests we have.
these factors can be captured by the kind of population effects that were explored in the work
of Rose (1992) as outlined above. Our attitudes and behaviour as a society can result in
positive or negative results for individuals. The causal mechanisms for this are complex and
seeking change (if needed) will be difficult. But we need to see that such common goods (and
the corresponding harms) are largely emergent properties of the relevant social group, rather
than the product of consensual creation or assent. This is an important justification for
collective action as an individual does not have total control of all the factors that influence
their life. Public health interventions can reinforce, shape and change some of these factors
and this fact provides a potential normative justification for such interventions when they
seek to prevent or reduce harm or bring about or reinforce common goods. An awareness of
the importance of such goods, that are neither a mere aggregation of individual goods nor
public goods in the traditional sense, is one reason why the notion of community is a concept
that is often invoked in such discussions. Those with parameters set by liberal medical ethics
are often dismissive of such a notion, but the concept of community captures the vital
collective and emergent aspects of such important social and common goods.15
Conceptions of social justice

Third, social justice is at the core of much public health activity, both nationally and
internationally (Anand et al., 2004; Faden and Powers, 2006). This concern is an option
for the development of substantive public health ethics because of the way that inequalities
(and therefore their impact on health) can be thought of as a population effect. Such a harm
is not brought about by individual action (although, of course, this may contribute) but
partly (perhaps even largely) by ones place within a society (Marmot and Wilkinson, 2005).
Again, this is an excellent illustration of the reason why population-level analysis is vital to
public health, and therefore, why it ought to be central to public health ethics (see Daniels,
Chapter 11, and Wilson, Chapter 12). It can also be argued, as it is by ONeill (2002), that a
concern for justice in public health will have clear implications for the kinds of interventions that may be justified. She argues that the obsessive focus on individual autonomy in
medical ethics may result in a negative impact upon the publics health, and that compulsion, even coercion, is necessary at times to achieve the desired aims of public health policy.
In other words, a concern for justice may well result in actions that a Millian liberal may
find uncomfortable because in those situations liberty is not the core value. Sometimes a
concern about equality may result in the sacrifice of individual liberty. A judgment about
whether this is appropriate will have to be made in an individual situation, but the point is
15
For further discussion see, for example, Cribb (2005), Jennings (2007b) and Baylis et al. (2008).
One area for potential fruitful development in the future is to consider how social ontology and
public health ethics may be mutually supporting.
18
Section 1: Concepts
that there is no a-priori reason to give priority to liberty. Of course, there is a huge literature
on notions of justice, but I raise it here because of its importance in public health work and
the fact that accounts of public health ethics committed to the idea of social justice will be
able to meet my suggested condition.
Conclusion
Public health ethics is enjoying a surge of interest. The work of this discipline is not only to
help to explore the ethical issues that emerge in public health practice, policy and research,
but also to consider how medical ethics and ethical theory may need to be rethought and
revised as a result of taking public health and public health ethics seriously. There is
no a-priori reason for thinking that public health ethics ought to develop as a separate
discourse, distinct from that of medical ethics or bioethics more generally, and on the whole
I think we have good reasons to retain or seek theoretical unification. While there is a lot of
interesting work going on at the moment in the field of public health ethics, there is much
still to be done. It is almost certainly too early to outline a satisfactory account of what
public health ethics ought to be. Despite this we can say something about the areas that we
need to explore in the future. The core values will be some of those I have identified above as
belonging to the list of factors that might be used as the basis for building a substantive
public health ethics: the concepts of group, community, population, public goods, common
goods, solidarity, reciprocity, welfare, well-being and justice. This predominantly social
focus need not mean that we can ignore individual concerns, or that individuals are
sacrificed for the sake of populations. The way forward is surely a sufficiently rich and
pluralistic account of public health ethics that does not feel the need to embrace such
dichotomous thinking. Individuals are neither just individuals, nor merely constituent parts
of groups, societies, communities and populations. What we need is a public health ethics,
and in turn a bioethics, that is both sensitive to and grounded in the reality of relevant social
relationships.
Acknowledgements
I am very grateful to the following for their critical comments on a previous version of this
chapter: Cecile Bensimon, John Coggon, Alan Cribb, Ross Upshur and Adrian Viens.
References
Anand, S., Peter, F. and Sen, A. (eds). (2004)
Public Health, Ethics & Equity. Oxford:
Oxford University Press.
Baylis, F., Kenny, N. P. and Sherwin, S.
(2008) A relational account of public
health ethics. Public Health Ethics, 1, 3:
196209.
Childress, J. F., Faden, R. R., Gaare, R. D., et al.
(2002) Public health ethics: mapping the
terrain. The Journal of Law, Medicine &
Ethics, 30, 2: 1708.
Coggon, J. (2008) Harmful rights-doing?
The perceived problem of liberal paradigms
and public health. Journal of Medical Ethics,

34: 798801.
Coker, R. and Martin, R., eds (2006) Public
Health, 120, Supplement 1: The Importance
of Law for Public Health Policy and Practice.
Cribb, A. (2005) Health and the Good Society.
Oxford: Oxford University Press.
Dawson, A. (2007) Herd protection as a public
good: vaccination and our obligations to
others. In Ethics, Prevention and Public
Health, ed. A. Dawson and M. Verweij.
Dawson, A. (2009) Theory and practice in
public health ethics: a complex
relationship. In Public Health Ethics &

Practice, ed. S. Peckham and A. Hann.
London: Polity Press.
Martin, R. (2009) The role of the law in public

health. In The Philosophy of Public Health, ed.
A. Dawson. Aldershot: Ashgate.
Dawson, A. and Garrard, E. (2006) In defence of

moral imperialism: four equal and universal
prima facie duties. Journal of Medical Ethics,
32, 4: 2004.
Mill, J. S. (1859) On Liberty. (Many editions).
Dawson, A. and Verweij, M. (2008) The steward

of the Millian state. Public Health Ethics,
1, 3: 1935.
Nuffield Council on Bioethics. (2007) Public

Health: Ethical Issues. London: NCB.
Feinberg, J. (1984) Harm to Others. Oxford:

19
Munthe, C. (2008) The goals of public health: an

integrated, multidimensional model. Public
Health Ethics, 1: 3952.
Nussbaum, M. C. (2000) Women and Human

Development. Cambridge: Cambridge
University Press.
Gostin, L. O. (2005) Public health

preparedness and ethical values in pandemic
influenza. In The Threat of Pandemic
Influenza: Are We Ready? Institute of
Medicine, ed. Washington: National
Academies Press.
Nys, T. (2008) Paternalism in public health care.

Public Health Ethics, 1: 6472.
Gostin, L. O. and Stone, L. (2007) The health of

the people: the highest law. In Ethics,
Prevention and Public Health, ed. A. Dawson
and M. Verweij. Oxford: Oxford University
Press.
Parfit, D. (1984) Reasons and Persons. Oxford:

Griffin, J. (1986) Well-Being: Its Meaning,

Measurement and Moral Importance. Oxford:
Clarendon Press.
Hardin, G. (1968) The Tragedy of the
Commons. Science, 162, 3859: 12438.
Holland, S. (2007) Public Health Ethics. London:
Polity Press.
ONeill, O. (2002) Public health or clinical

ethics: thinking beyond borders. Ethics &
International Affairs, 16, 2: 3545.
Paul, C. (2009) The common good argument

and HIV prevention. In The Philosophy of
Public Health, ed. A. Dawson. Aldershot:
Ashgate.
Postema, G. J. (1987) Collective evils, harms,
and the law. Ethics, 97: 41440.
Powers, M. and Faden, R. (2006) Social Justice:
the Moral Foundations of Public Health and
Health Policy. Oxford: Oxford University
Press.
Jennings, B. (2007a) Public health and civic

republicanism: toward an alternative
framework for public health ethics. In Ethics,
Press.
Raz, J. (1986) The Morality of Freedom. Oxford:

Jennings, B. (2007b) Community. In Principles

of Health Care Ethics, ed. R. Ashcroft,
A. Dawson, H. Draper and J. McMillan,
2nd edn. Chichester: Wiley.
Upshur, R. E. G. (2002) Principles for the

justification of public health interventions.
Canadian Journal of Public Health, 93, 2:
1013.
Kass, N. E. (2001) An ethics framework for

public health. American Journal of Public
Health, 91, 11: 177682.
Verweij, M. and Dawson, A. (2007) The

meaning of public in public health.
In Ethics, Prevention and Public Health, ed.
A. Dawson and M. Verweij. Oxford: Oxford
University Press.
Kraut, R. (2007) What is Good and Why: The

Ethics of Well-being. Harvard, MA: Harvard
University Press.
Marmot, M. and Wilkinson, R., eds (2005) Social
Determinants of Health, 2nd edn. Oxford:
Rose, G. (1992) The Strategy of Preventive

Medicine. Oxford: Oxford University Press.
Sen, A. (1999) Development as Freedom. New
York: Knopf.
Wikler, D. I. (1978) Persuasion and coercion for

health: ethical issues in government efforts to
change lifestyle. Millbank Memorial Fund
Quarterly, 56, 5: 30338.
Section 1
Chapter
Concepts
Health, disease and the goal

of public health
Bengt Brulde
Introduction
What is the ultimate goal of public health activities, such as health promotion, health
education or health protection? Or rather, what should these activities aim at, what values
should public health workers try to realize or maximize?1 To suggest that the goal of public
health is to promote or improve population health is not very informative, at least not
before we establish what is meant by population health in this context, or rather, what the
phrase should mean in this context. Moreover, there might be some appropriate goals that
are not really possible to subsume under this heading, at least not without stretching the
ordinary meaning of the phrase a little too far. That is, the central question of this chapter
cannot be regarded as conceptual.
To come up with a reasonable view on what the ultimate goals of public health are, we
need to look at what more specific suggestions have been made. Among the different goals
that have been suggested, there are at least three that are worth considering more closely,
namely (1) to improve the average level of health (or the like) in the relevant population,
(2) to reduce health inequalities between groups or individuals and (3) to create certain
types of opportunities for health, most often equal opportunities. It is possible that we
should add (4) to improve the health level of the worst-off groups.2,3
To assess the quality and effectiveness of different public health strategies (practices, interventions or
efforts) is, to a large extent, a matter of determining to what extent these ultimate goals have been
attained. It is worth noting that most operative public health goals, for example, to reduce the
consumption of alcohol or the percentage of smokers, are not ultimate or final in this sense, but
merely instrumental.
In my view, the question of what the ultimate goals should be is one of the three most central
issues in public health ethics. The other two are (i) the question of what interventions are morally
acceptable, for example, in what ways it is morally permissible or legitimate to influence peoples
choices or to manipulate the social determinants of health; and (ii) the question of how the resources
available should be allocated or distributed, for example, whether the allocation of resources
should be based on considerations of need or on considerations of cost effectiveness. Both these
questions are intimately related to the question of the ultimate goals.
Like (2), this is a distributive concern, but it is not identical to (2). It might be combined with (2),
however, for example, if the goal is to reduce health inequalities through improving the health
of the worst-off groups. In the following, I will discuss (4) in connection with (2), that is, I will,
for the time being, not treat it as a goal in its own right.
20
Chapter 2: Health, disease and the goal of public health
21
Assuming that (1)(3) constitute a pretty comprehensive view of the goals of public
health, let us now take a closer look at how exactly they should be understood.
To improve the average level of health

To improve the average level of health in a population is to improve population health in
the most narrow sense of the term, or alternatively, to make an improvement in what
Verweij and Dawson (2007) refer to as the aggregative dimension of population health.4
This goal reflects the plausible idea that public health activities should aim to benefit the
individuals in the relevant population, and the more these people are benefited the better.
However, it is less clear in what way the individuals should be benefited, that is, in what
dimension or dimensions. For example, to suggest that public health should promote or
maintain individual health is far from sufficient. To prevent disease and injury (to reduce
morbidity) is a second way in which public health should benefit people and to prolong
peoples lives (to reduce mortality) is a third possibility. There are more views than these of
what the relevant individual outcomes should be, for example, empowerment (peoples
ability to control the determinants of health or to make healthy choices),5 responsibility,
autonomy, dignity, integrity (Buchanan, 2000), well-being (quality of life) or health-related
quality of life. In short, it is an open question how exactly public health should benefit the
individuals in the relevant population, that is, what individual health outcome it should
aim to improve or maximize. I return to this question below.
To reduce health inequalities

To reduce health inequalities is to make an improvement in what Verweij and Dawson
(2007) refer to as the distributive dimension of population health. This suggestion also gives
rise to a large number of questions. First, we need to ask what inequalities should be
reduced, that is, what the relevant individual health outcome is. In my view, we can safely
assume that it is the same outcome as above, that is, the outcome that public health should
also try to improve or maximize. (The concern with social inequality as a means to equality
in health does not belong here, but in (3).) Second, we need to ask what distributive
Many of the standard epidemiological measures are strongly related to (1), properly construed
(see below). For example, most traditional morbidity and mortality measures tell us to what extent (1)
has been attained. However, there are also a number of traditional measures that are, at best,
indicators of population health. It might be worth considering whether public health should try to
improve these measures as well, for example, whether it should also try to reduce the utilization of
health services, the number of sick days per capita and year, or the amount of early retirements due to
illness.
Few theorists regard this as the only goal of public health, but Rose (1992) seems to be an exception.
It is worth noting that if this were the case (that is, if there were no distributive goals), then
health care and public health resources should simply be allocated so as to maximize the health
benefits they produce, and it would be sufficient to use cost effectiveness analysis alone to set
priorities (Brock 2004). It is also worth noting that if maximization of aggregate health were the only
goal, then we should be indifferent between small benefits to a large number of persons and large
benefits to a small number of persons (regardless of peoples initial health states).
In this context, it is worth noting that empowerment has a rather complex role in this context: the
term does not just denote a goal (a state to be achieved), but also a certain kind of process, viz. a
process characterized by a high level of client influence and participation (Tengland, 2007).
22
Section 1: Concepts
considerations are most plausible. There are other possible distributive considerations
besides the aim to reduce health inequalities, for example, (4) to give priority to those with
the worst health or shortest life expectancy, to reduce health inequities (those inequalities
that are unjust) or to ensure that no one falls below a certain critical level. It is not quite
clear how these other considerations are related to the goal of reducing inequality, and what
we should do if different distributive considerations pull in different directions. Maybe we
should abandon egalitarianism altogether, and replace it with some version of (4), such as
the priority view. In any case, we need to ask what inequality is, and how it should be
measured in this context. Without doing so it is hard to formulate a plausible distributive
goal for public health activities.
To create (equal) opportunities for health

This goal corresponds, at least in part, to Verweij and Dawsons (2007) third dimension of
population health. If this dimension is formulated as a goal, we get the suggestion that
public health should improve those conditions that are relevant for the health of everyone
(2007: 26), for example, that it should try to contain and control those environmental (and
perhaps social) risk factors that might affect the health of everyone, that it should aim to
exclude these factors from the population. (An example of this is to achieve herd immunity
against certain diseases by vaccination.) The idea that public health activities should aim to
create equal opportunities for health seems to include more than this, however, viz. that
these activities should also attack other central risk factors or determinants, risk factors
that might affect the health of many but not everyone. For example, it might be argued that
public health should try to improve the social conditions of the worst-off groups, or to
reduce socio-economic inequalities. (This is a central goal of what Seedhouse (1997) refers
to as social health promotion.) The idea that health promotion should aim to empower
groups or communities (not just individuals) might belong here as well.6
Regardless of what exactly is included under this heading, it is clear that success in
this area does not automatically manifest itself in improved aggregative or distributive
figures: a decrease in average health or an increased inequality in health is not necessarily a
sign of failure.7 Now, it is rather obvious what this might mean on the individual level,
6
This goal is sometimes qualified in different ways. For example, Munthe (2008) thinks that this goal
has weight only when the average level of health is acceptable. Another possibility is that the
goal to create equal opportunities for health is not really ultimate, for example, that it should only be
pursued when this is expected to result in better population health.
According to Verweij and Dawson (2007), this suggests that the health of a population can improve
even if no one becomes healthier and even if the inequality in health is not reduced. I prefer a
somewhat different language, however. In my terminology, public health work may have other goals
besides improving population health, for example, to protect population health, and can thus be
successful even if there is no improvement in population health. In my view, the health of a
population is solely a function of average level and distributive features. (Which implies that
population health can be improved by selectively aborting everyone with a below average life
expectancy.) Or alternatively put, the health of a population is a function of the actual health states
(life expectancy, etc.) of its members: if we know how healthy the relevant individuals are
(including how individual health is distributed in the population), we can (in principle) determine
the health of the population. It can also be argued that the question of how population health
should be measured is really an evaluative question, that the question of how different health
distributions should be ranked with regard to population health cannot be distinguished from the
23
for example, the fact someone has a good opportunity to stop smoking does not imply
that they will in fact stop smoking. However, what counts as an opportunity in this context,
how do we determine how big a certain opportunity is, and how big a gap can there be
between peoples opportunities for health and their actual health states? (Does being
informed about the dangers of smoking count as an opportunity for health?) I will not
pursue these questions any further. Neither will I discuss to what extent the third goal is
reasonable, but it is worth noting that considerations of autonomy suggest that we cannot
ignore this goal.
Towards a conception of the goals of public health

It is important to note that the above three goals can conflict in several different ways. For
example, aggregative considerations can conflict with distributive considerations: that the
health of the best-off group is improved is desirable from an aggregative perspective but
might be undesirable from a distributive perspective. A theory of the goals of public health
cannot be complete unless it tells us how we should deal with these conflicts. One way to do
this is to create a goal structure (Brlde, 2001) and another (perhaps more promising) way
is to create a summary measure; to combine the different concerns into a single measure.
In summary, several questions need to be answered if we want to arrive at a wellfounded and reasonably complete view of the goals of public health. In the following, I will
restrict myself to two of these questions: (1) In what respect or respects should public health
try to benefit the relevant individuals? What individual health outcomes are most relevant
in this context, that is, what exactly should public health try to maximize and distribute
fairly or equally in the relevant population? In connection with this I discuss central
concepts such as health, disease, quality of life, health-related quality of life and
quality-adjusted life years. I will also offer a few reflections on how valuable different
health gains and life extensions are from a public health perspective, and whether it is
possible to combine the relevant variables (for example, health and life expectancy) into a
single measure. (2) What is the most plausible distributive goal for public health activities?
For example, is it to reduce all health inequalities, or only those inequalities that are unjust
or unfair? Is the priority view more plausible than strict egalitarianism? Is there a critical
level, such that it should be regarded as especially important that people do not fall below
this level? There will also be a short discussion on how aggregative concerns might be
weighed against distributive concerns. I will devote more space to the first question than to
the second. The two questions will, to a certain extent, be treated separately, but not fully:
we can hardly arrive at a reasonable individual summary measure unless some distributive
considerations are incorporated into this measure.
The relevant individual outcomes

In what dimensions should public health activities try to benefit the members of the
relevant population? There are several different possibilities here, for example, to promote
health, to prevent disease and injury, to prolong lives, to promote well-being or healthrelated quality of life, and to promote (individual) empowerment, responsibility or autonomy. In my view, goals such as individual autonomy and empowerment are either
question of how they should be ranked with regard to value. This suggests that there is little or
no point in trying to define population health, that it is far more important to focus on what the
legitimate goals of public health activities should be.
24
Section 1: Concepts
constitutive parts of well-being and/or mental health, or they belong to the third goal (to
create opportunities for health) and thus they need not be treated separately in this context.
(It can also be argued that these outcomes are only desirable from a public health perspective if they are conducive to better health or a longer life; Tengland, 2006.) The other
possibilities are worth looking into, however. Let us start with the perhaps most frequently
used measure of the health status of a population, namely life expectancy at birth.
To prolong lives
How long we live obviously matters, not just from our own perspective but also from a
public health perspective. But how much does it matter, and in what way? A key issue here
is whether we should always attribute the same value to an additional life year (whether we
should regard each life year as equally important) or whether we should give different moral
weights to life years at different ages. In my view, we can safely assume that each life year is
equally valuable to the person who is living the life (with one possible exception, see 5 below),
but it is not clear whether public health professionals ought to evaluate each life year equally
(or whether each life year gives an equally large contribution to the health of a population).
Should each life year be given the same weight, regardless of when during the life cycle it
occurs? There are at least five simple answers to this question (see also Figure 2.1):8
1. Each life year added is equally valuable (given an acceptable level of health, etc).
2. The older the person is, the less is the value of an additional year, that is, the
marginal value of a life is diminishing with increasing age.
3. After a certain age, an additional year is worth nothing from a public health perspective.
Years of potential life lost is a measure that reflects this view, since it only counts
the years lost before a certain (arbitrarily determined) age, for example, 75 years.
4. After, for example, the age of 75, each additional year still has value, but it is worth
less (from a public health perspective) than it was before this age.
5. Each of these four views can be modified as follows: a life year has less value in the
beginning, when we are infants, for example, because were lacking in crucial capacities.
There are also more complex views on this matter, for example, the age-weighting used in
the so-called GBD (Global Disease Burden) formulation of the DALY measure (Murray,
1996; Anand and Hanson, 1997; Murray and Acharya, 1997).9 In this measure, age is
weighted according to a (somewhat) bell-shaped curve whereby a year of life during young
adulthood counts for more than one, peaking at 1,5 during ones early twenties, and a year
of life counts for almost zero when one is close to birth or death (Selgelid, 2008).10
9
10
It is worth noting that these answers are, to a large extent, distributive considerations. Or more
specifically, that they are combinations of aggregative and distributive considerations. They can
therefore be regarded as partial answers to question (2) above.
This particular DALY measure was only used in the first GBD study (in 1990), there was no ageweighting in the DALY measure that was used in the 2001 study.
It is important that the question of age-weighting be carefully distinguished from another temporal
question, namely whether it is appropriate to discount future life years when making decisions in the
present, that is, to regard future life years as less and less important as the years in question fall
further into the future. Time discounting is often used in health economics, and future life years
were discounted by 3 per cent in both the 1990 and the 2001 GBD studies. This can hardly be
Accumulated
value
25
(1)
(2)
Accumulated life years
(3)
(4)
(5)
Figure 2.1 Graphical representations of positions (1)(5). Note that the measure on the x-axis is longevity
(accumulated life years) and that the measure on the y-axis is accumulated value.
This kind of age-weighting can only be justified instrumentally. A healthy life year at
the age of 30 is regarded as more valuable than a healthy life year at the age of 70 because
30-year-olds are typically more productive and because others typically depend on them
economically. As Murray and Acharya (1997: 719) put it, the health of some groups is
instrumental in making society flourish: therefore collectively we may be more concerned
with improving health status for individuals in these age groups.11
To justify a certain age-weighting in this manner is ethically problematic (Brock, 2004).
Moreover, it leads to double-counting (Anand and Hanson, 1997). When assigning value to
different life years, we should not evaluate these years instrumentally, but intrinsically. Does
this mean that all age-weighting is unjustified, that is, that we should accept (1) above?
I think not. In my view, a combination of (2) and (5) seems most plausible, and for the
11
justified. The usual arguments for discounting money do not apply to discounting healthy life years,
and it seems absurd to discount public health measures that reap their health benefits years into the
future, like vaccination programmes. Future benefits are appropriately discounted when they are
more uncertain than proximate benefits, but (even) this consideration does not require the use of a
discount rate (Anand and Hanson, 1997; Brock, 2004; Selgelid, 2008). There is also a third
temporal question that needs to be addressed when assessing disease burden or cost effectiveness,
namely what ideal life expectancies should be used (Brock 2004; Selgelid 2008). For example,
should we use the same uniform measure (for example, life expectancy in Japan) when calculating
disease burden as in resource allocation contexts, for example, when calculating the benefits of life
saving interventions (that is, the number of life years gained)?
It has also been pointed out that life years lived as a young or middle-aged adult are valued
more by people in general: one QALY for a 30-year-old is regarded as equal in value to three QALYs
at age 50 or to nine QALYs at age 70. We should be grateful that these people are not running
the health care system!
26
Section 1: Concepts
following reasons. The reason why we should accept (5) is that life of an infant has less
quality. Now, it might be argued that we may deal with this on the qualitative side, but
since full quality for an infant is not as high as full quality for an adult, it is more
manageable to incorporate this on the temporal side: in this way, we can regard full
quality at the age of 1 as equal in value to full quality at the age of 30, which makes it much
easier to construct plausible summary measures (see below). To see why we should accept
(2), let us first note why (4) (and perhaps (3)) is more plausible than (1). A good reason for
this is the fair innings argument (Williams, 1997). Given that our resources are limited,
equality of opportunity requires that society gives lesser weight to a year of life extension
beyond the normal life span. The reason why we should prefer (2) to (3) and (4) is simply
that it involves no arbitrary cut-off point.
To differentially weight life years in this way might be irrelevant in a public health
context, however, we can probably stick to life expectancy at birth here. In any case, to
prolong life is certainly a plausible individual goal of public health.
Summary measures that combine quality and quantity

Now, it is quite clear that life expectancy is not the only relevant individual outcome in this
context. How long we live is not all that matters: it also matters how healthy our lives are.
And we cannot assume that most or all life years are healthy, especially not in rich societies,
where we are good at prolonging unhealthy lives, and where increases in life expectancy are
not always accompanied by increases in health. This suggests that we need a complex
measure that does not just take the duration of life into account, but also some measure of
its quality (broadly understood).
The most straightforward summary measure is to simply count the number of years with
full health or full quality. Here, equal weight is given to health and life expectancy, that is,
one year in full health is given the same value as two years lived in 50 per cent of full health
(whatever this means). There are several summary measures of this type, for example,
QALYs (quality-adjusted life years), HALEs (healthy life expectancy), DFLE (disabilityfree life expectancy) and DALE (disability-adjusted life expectancy). What is measured here
are the years of life expected to be lived in full health, in full quality or without disability. It is
also possible to measure the number of healthy life years lost through illness, disability or
death, for example, as in the case of DALYs (the number of years lost because of death and
disability in relation to some arbitrary standard age, for example, 82.5 for women and 80 for
men). The basic idea is the same in all these cases: Public health should not just maximize
peoples life expectancy, but also try to make sure that the life years lived are healthy life
years of good quality. To illustrate this idea, we can represent each life graphically, where
the number of life years (or weighted life years) are represented on the x-axis, and where the
persons health state (or quality of life) is represented on the y-axis (Figure 2.2).
What should be maximized is (roughly) the area below the curve. This idea gives rise to
two important questions, however. First, what exactly should be represented on the y-axis,
that is, what are the relevant qualitative individual outcomes? And second, once this has
been decided, how should the scale be calibrated, for example, what level of health should
count as full health, or as 50 per cent of full health? In this contribution, I focus on the first
question, but first a few remarks on what might be meant by full health in this context.12
12
Thanks to Niklas Juth, who made me aware of the importance of this question.
'Quality'
27
'Quality'
Longevity (accumulated life years)
Accumulated weighted life years
Figure 2.2 An example of what a graphical representation of a life can look like in a health context. Note that
the measure on the x-axis is the number of weighted life years in the graph to the right, and that these weights can
be represented by making the more valuable life years broader (and thus more weight when the area below the
curve is calculated).
What is full health? And is there a notion of full health that makes the HALE view (the
view that public health should try to maximize the number of years with full health)
plausible? One possibility is to conceive of full health as maximal health, but this
possibility is hardly plausible. Such a view would imply that few people are in full health,
and that public health should try to benefit people who are very healthy (but below the
maximum). Such a view would also make it quite difficult to make trade-offs between
health and longevity. Another possibility is to conceive of full health as acceptable health,
where acceptable health may vary over time and across nations. Since many peoples health
is better than acceptable, this view implies that if these peoples health is improved, there is
no corresponding improvement in HALE, that is, that such improvements have no value
from a public health perspective (assuming that HALE is the relevant individual
outcome).13
However, to present the problem as a choice between different notions of full health is
quite misleading. The question is really how we should value different improvements in
health from a public health perspective, a question that is analogous to the question how we
should value different life years depending on when during the life cycle they occur.
(Moreover, we want to calibrate the health scale (functioning scale, or the like) in a way
that makes it possible to compare life years lost to premature death with life years lived in a
reduced health state, for example, in a way that makes it possible to translate a number of
years lived with a certain disability into an equivalent time loss. Only then does it make
sense to say that we ought to maximize the area below the curve.) There are at least four
possibilities here, possibilities that correspond to (1)(4) above:
1. Each improvement in health of a certain magnitude is equally valuable (from a public
health perspective), regardless of where on the scale it occurs. (This is the view we are
likely to adopt if we combine the HALE view with the idea that full health equals
maximal health.)
13
This is not to deny that it can have instrumental value to improve peoples health above this
level, however, since this might prolong their lives, or postpone the time when they sink below
acceptable health. The observation is Per-Anders Tenglands.
28
Section 1: Concepts
Accumulated
weighted quality
intervals
Accumulated weighted life years

Figure 2.3 An example of how a life can be represented graphically if we are looking for a measure of weighted
QALY or HALE. Note that the measures on both axes are weighted, which can be represented by making
the more valuable life years and the more valuable quality intervals broader: they are thus given more weight
when the area below the curve is calculated.
2. The better a persons health is, the less is the value of a certain improvement, that is, the
marginal value of health is diminishing.
3. Once a certain level of health has been attained, a further improvement is worth
nothing from a public health perspective. Public health should do nothing for those
people who have reached this level (and who can be expected to stay on this level).
(This is the view we get if the idea that full health equals acceptable health is
incorporated into the HALE view.)
4. Improvements beyond acceptable health still have value, but they are worth less (from a
public health perspective) than improvements below this level.
Now, the HALE view cannot be made sufficiently precise unless we take a stand on this
issue. In my view, (2) is most plausible. If this idea is accepted, we get the following
modification of the HALE view: We represent each life graphically, in the form of a curve
(Figure 2.3). The number of weighted life years are still represented on the x-axis, but we no
longer measure peoples health states on the y-axis. The y-axis is no longer a straightforward
heath scale, but a health value scale that consists of weighted health units. It is the area
below this curve that we should try to maximize, that is, weighted HALEs (QALYs, or the
like). But again, this maximizing idea only makes sense if the health value scale is calibrated
in a way that makes it possible to make trade-offs between health and longevity, for
example, to determine how many life years lived with a certain disability that is equivalent
in value to one life year lived in full health. The calibration question will be discussed later.
Let us now return to the first question, that is, what exactly should be represented on the
y-axis. The perhaps most natural answer to this question (an answer that has been taken for
granted in the above) is health. There are several views of what health is, however, one of
which is the absence of disorders or maladies (that is, diseases, injuries and defects). To
suggest that public health should promote health in this sense is to suggest that it should
aim at preventing disease and injury. (As we will see, this goal is central regardless of what
notion of health is chosen, but if health is defined in terms of functioning or well-being,
this is because disease is regarded as a common cause of ill health.) Let us therefore start
with the goal to prevent disease or injury.
29
To prevent disease and injury

This is obviously a plausible goal of public health. But to find out what it means more
exactly, and how it relates to the other goals (for example, whether it is an instance of the
goal to promote health or a means to promoting health), we need to know how the term
disorder (or malady) can and should be defined.
What is a disorder? Or alternatively put, what conditions should be categorized as
disorders, mental or somatic? How should we draw the line between pathological and nonpathological conditions?14 There are many reasons why it might be of practical importance
how we define the concept (Brlde, 2003; Brlde and Tengland, 2003). In this context, the
most important practical purpose is that a definition of disorder can help us specify the
proper goals of medicine, and maybe also of public health activities. It seems reasonable to
assume that one of the central goals of public health is to prevent diseases and other
disorders, which suggests that a well-founded definition of disorder can help us specify
the goals of public health, and that it can perhaps help us (at an institutional level) to
demarcate the special responsibilities of public health professionals from those of other
professionals.
Disorders as internally caused undesirable conditions

It is generally assumed that disorders are physical or mental states or processes (for
example, underlying anatomical or physiological pathologies or abnormalities) that typically manifest themselves in different kinds of undesirable symptoms. On this view, a
condition is a disorder if and only if (1) it is undesirable or bad (either in itself or because
of its consequences) let us call this the value component of the concept and (2) the
proximate cause of the symptoms is some type of internal state or process (for example, a
lesion or a part dysfunction), that is, something inside the individuals body or mind let us
call this the factual or explanatory component.15
This mixed or hybrid view has been challenged in three different ways. First, there is the
view that the concept of disorder (particularly mental disorder) is a purely evaluative
concept, that is, that there is really no need for a factual component (Wakefield, 1992).
Second, it has been argued that the concept is a purely factual or scientific concept, that the
presence of the right kind of internal cause is not just necessary for disorder, but also
sufficient (Boorse, 1975). For example, on the so-called traditional medical model (or
machine-fault model) of disorder, disorders are machine-faults, for example, underlying
structural or functional abnormalities. The presence of a machine-fault is not just necessary
for disorder, but also sufficient, and it is often assumed that the presence of such an
14
15
The main reason why the question is best phrased in terms of disorder rather than in terms of illness
or disease is that the most practically relevant category is a broader category that also includes
injury, retardation, and so on. The practically important thing is obviously how we distinguish
disorder from non-disorder, and not how we draw the line between for example, disease and injury,
or between illness and disease.
This rudimentary conceptual theory of disorder offers us truth conditions for disorder
statements, that is, it tells us under what conditions a disorder is present. This leaves it open
what kind of thing a disorder is, however, for example, whether it is (1) the condition that is caused
by an underlying dysfunction (or the like), that is, some kind of syndrome, (2) the internal cause
that produces the symptomatic manifestations or (3) the whole complex, that is, the underlying
pathology plus the symptoms.
30
Section 1: Concepts
abnormality can be established in an objective and scientific way. Third, it has been
suggested that the value component and the factual component should be supplemented
by a third component, for example, the idea that a disorder is, by definition, a condition
that health professionals treat, or a condition that (we think) should be treated by health
professionals or by medical means (Reznek, 1987).
All these suggestions fail, but for slightly different reasons (Brlde, 2003). That is, it
seems reasonable to accept the general idea that disorders are undesirable conditions
immediately caused by internal states and processes. This assumption gives rise to two
questions: (1) How should the evaluative content of the concept of disorder (its value
component) be characterized? (2) What kind of internal proximate cause is essentially
involved, that is, how should the factual component of the concept be characterized?
The value component: harm and other bad things

With the possible exception of a few machine-fault theorists and a number of physicians, it
is generally agreed that we have to rely on value judgments to distinguish disorders from
other conditions. It is not quite clear how, however. To clarify this, we need to know more
than to what extent attributions of disorder are dependent on values. In particular, we have
to know what kind of evaluations we have to rely on to identify the class of disorder.
If disorders are by definition bad, in what way are they bad?16
The most obvious answer is that we have to rely on considerations of harm. Disorders
typically involve some kind of harm to the individual who has the disorder, for example,
distress or disability. As Wilkinson (2000: 289) puts it, we have a prima facie reason for
believing that a condition is a disorder if it is a state of persons which causes them to be
harmed (for example, through death or pain). This suggests that the connection between
disorder and harm is conceptual rather than contingent, that is, that we rely rather heavily
on considerations of harm when we want to determine whether a certain condition should
be regarded as a disorder. The specific harms that are most often involved in disorder are of
three different kinds, viz. (1) displeasure (for example, pain), (2) disability and (3) significantly increased risk of suffering the harms mentioned in (1) or (2), or of suffering
premature death.
Many theorists (for example, Wakefield, 1992) claim that harm to the individual is a
necessary condition for disorder. Some theorists (including Wakefield) also claim that we
dont have to rely on any other kinds of evaluative considerations to delineate the class of
disorder, that is, that harm for the individual is also sufficient for disorder given that the
condition has the right kind of proximate cause. However, it seems that we sometimes need
to rely on other evaluative considerations, particularly in the case of mental disorder. In my
view, there are at least two additional evaluative considerations that are of relevance when
we want to determine whether a certain type of condition should be regarded as a mental
disorder, namely (1) judgments about what is bad or harmful for others, and (2) judgments
16
In my view, there are at least two more questions that any reasonably complete theory of the
evaluative content of the concept of disorder need to answer, namely (1) whether attributions of
disorder contain any implicit reference to some specific evaluative standard and (2) whether the
concept of mental disorder is value laden in the definitional or in the epistemic sense. In Brlde
(2007a) I argue (1) that there should be no references to any specific evaluative standards at all, and
(2) that even though disorders are necessarily undesirable, disorder may well be a descriptive
phrase.
31
about abnormal functioning on the holistic level, for example, the idea that the persons
behaviour deviates from some standard of good or normal functioning (where this standard
is not fully derived from or based on considerations of harm, for example, as in attributions
of severe irrationality). In this view, the presence of harm for the individual is neither
necessary nor sufficient for disorder (given that the factual component is in place). The
main reason for this is there seem to be mental disorders that are either not associated with
such harm, or not considered disorders primarily in virtue of being associated with such
harm, for example, paedophilia and antisocial personality disorder (Brlde, 2007a). If these
conditions are correctly classified as mental disorders, then this is not by virtue of being
harmful for the individual who has the condition, but rather because these conditions are
abnormal and/or harmful to others.
The factual component: machine faults and other

internal causes
On the hybrid view, the undesirable symptoms are by definition caused by the right kind of
internal factor. But what kind of proximate internal cause is essentially involved in
disorder? There are at least five possible answers to this question.
The first two views are both based on the assumption that the relevant internal cause can
be fully characterized in objective, scientific or essentialist terms: (1) In the lesion view, all
disorders are essentially lesions, that is, structural or anatomical abnormalities. A person
can only have a disorder if some structural part of the organism (like a cell, a tissue, or an
organ) is damaged. (2) In the dysfunction view, disorders are essentially part dysfunctions,
for example, physiological or biochemical abnormalities or disturbances that may or may
not be caused by lesions. These two ideas are both versions of the so-called traditional
medical model of disorder. Disorders are regarded as machine-faults, that is, underlying
abnormalities (structural or functional) that tend to cause problems on the holistic level.
This view has two central features, namely: (1) The reductionist idea that the symptoms are
caused by some kind of underlying or lower-level pathology. The machine as a whole
does not function because there is a fault somewhere in the apparatus. (2) The view that
these machine-faults are abnormalities, that is, deviations from normal structure or function, where normality is understood in terms of species design.
Since the dysfunction view is more plausible than the lesion view (see for example,
Brlde, 2003), let us restrict our attention to this view. There are at least two different
views on the nature of dysfunction: the first idea is that dysfunctions are statistical
abnormalities that give rise to biological disadvantage (where the notion of biological
disadvantage is understood in evolutionary terms, viz. in terms of reduced survival or
lowered reproductive fitness [Boorse, 1975; Kendell, 1975]). If this notion of dysfunction
is incorporated into the concept of disorder, we get the idea that a disorder is (roughly) a
condition that tends to reduce longevity or fertility. The second view of dysfunction
(Wakefields view) is also inspired by evolutionary theory. Here, a dysfunction is regarded
as a failure of some internal mechanism to perform its natural function, where a natural
function of a mechanism is an effect of this mechanism that explains why it was naturally
selected (Wakefield, 1992, 2000). Both dysfunction views have been heavily criticized, for
example, Wakefields evolutionary view has been criticized by Lilienfeld and Marino
(1995), Murphy and Woolfolk (2000a, 2000b), Nordenfelt (2003) and many others. Most
of the objections to Wakefields theory purport to show that dysfunction (as defined by
32
Section 1: Concepts
Wakefield) is not necessary for disorder, that is, that someone may well suffer from a
disorder even when there is no evolutionary malfunction. In my view, all traditional
medical (essentialist) dysfunction analyses of disorder are highly problematic and should
be rejected (Brlde, 2003).
A major problem with the traditional medical dysfunction view is that it is most
probably too narrow. If dysfunction (defined in evolutionary terms) is necessary for
disorder, this seems to exclude too many consensual disorders from the category of
disorder. Conversely, it seems quite clear that if someone has a disorder, this means that
something has gone wrong with the person. So how should we specify the central notion of
going wrong, now that we can no longer appeal to dysfunctions? There are at least three
non-essentialist answers to this question: (1) A dysfunction view that appeals to failed
Cummins-functions, where a so-called Cummins-function of a structure or mechanism is
its causal contribution to the overall operation of the system that contains it (Cummins,
1975). (2) The idea that any internal cause will do, or more specifically, that any type of
internal cause will do. On this type of view, diseases and other disorders are, by definition,
internal conditions that tend to compromise peoples health, where health is, as a rule,
defined in terms of functional ability. (3) In Culver and Gerts (1982) view, disorders are
(roughly) internally caused harmful conditions, and the phrase internally caused is specified in terms of the absence of a distinct sustaining cause. In this view, a person can only
have a disorder (malady) if the distress or disability he is suffering does not have a
sustaining cause which is clearly distinct from the person.
The appeal to failed Cummins-functions tends to collapse into (2), the idea that any
internal cause will do (Brlde, 2003). In this type of view (adopted by for example,
Nordenfelt, Prn, Reznek, Whitbeck and Sedgweck), a disorder is, by definition, an internal
condition that tends to compromise peoples functional ability. This view gives rise to
several difficulties, for example, how to distinguish the internal from external, or more
specifically, how to determine whether a certain incapacitating condition is caused by
internal factors or whether it is merely a reaction to environmental factors. How can
internally caused conditions be distinguished from externally caused conditions in a way
that helps us to delineate the concept of disorder? Culver and Gert (1982: 72) offer an
interesting answer to this question when they specify the phrase internally caused in terms
of the absence of a distinct sustaining cause: We could say that to be a malady, the evilproducing condition must be part of the person. However, for reasons of conceptual rigor, a
more formal negative statement is preferable: the person has a malady if and only if the evil
he is suffering does not have a sustaining cause which is clearly distinct from the person.
The notion of a distinct sustaining cause is defined as follows: X is a distinct sustaining
cause of a condition C if and only if (1) X is a cause of C, (2) X is not part of the person
with C, that is, it is distinct from this person and (3) if X were removed, C would cease to
exist almost immediately, that is, X is necessary for sustaining C (Wilkinson, 2000: 301).
That is, the idea is that if a condition has a distinct sustaining cause, then it is externally
caused (in the relevant sense) and cannot be a malady (disorder). This is not to deny that a
malady may have been caused originally by factors distinct from the person. The important
thing is that to be a malady, a condition is not now in a state of continuing dependence on
those [external] factors; rather, it is present even in their absence (Culver and Gert, 1982: 72).
This may well be the most adequate view of how the factual component of disorder
should be characterized. However, Culver and Gert do not draw the line between the
pathological and the non-pathological where most people intuitively want to draw it. For
33
example, their theory might be considered too inclusive, because it seems to imply that
normal grief, unhappy love, ignorance or fanatical beliefs should be classified as mental
disorders. It is hard to see on what grounds they can argue that fanatical beliefs or unhappy
love are disorders. It is of course possible to refer to what is considered normal in the
culture in which the condition appears, but this would really turn it into another theory,
which includes a new evaluative element. Another possibility is to argue that ignorance or
fanatical beliefs are only harmful in certain environments, and that they should therefore be
conceived of as suboptimal behaviours (or the like) rather than as disorders. However, this
strategy cannot explain why conditions like unhappy love or normal grief should not be
regarded as pathological. It is of course possible to accept the implications, and to regard
grief as a mental injury and fanatical beliefs or infatuation as viruses of the mind, and that
they should therefore be classified as mental disorders.
To conclude, Wakefields idea that disorders are, by definition, caused by dysfunctions
(defined in evolutionary terms) is a well-founded theory, but it tends to exclude some
consensual disorders from the category of disorder. The most promising alternative to this
analysis is the idea that a disorder is a harmful condition that does not have a distinct
sustaining cause. However, this tends to include some more or less consensual nondisorders in the category of disorder. In my view, this strongly suggests that the different
desiderata for a good definition are pulling in different directions. For example, it seems
that if we rely on our linguistic intuitions, then we probably have to give up the idea that
disorder can be defined in terms of necessary conditions that are jointly sufficient. In fact,
it seems likely that if we want to capture the concept of disorder as it is actually used in
everyday speech, we should probably settle for some kind of family resemblance analysis or
prototype analysis.
Disorder and public health

It can hardly be doubted that the prevention of different disorders is an appropriate goal of
public health. The reason for this is partly that it is, by definition, bad to suffer from a
disorder. To repeat, the harms that are typically associated with disorder are of three
different kinds, viz. suffering, disability, and significantly increased risk of future suffering
or disability, or of premature death. This strongly suggests that the severity of a disorder
should be measured along these lines. The goal to prevent premature death has already been
considered, however, which suggests that it is only the other two harms that can be added to
the summary measure we are trying to construct. That is, it seems that the qualitative part
of the relevant individual goal consists of trying to reduce (internally caused) suffering and
disability. This idea is remarkably similar to the idea that public health should aim at
promoting health (on the lower end of the health scale).
To promote health (physical and mental)

To understand what it means to promote peoples health, we need to know how the concept
of health can and should be defined in this context. We should not forget that a reasonable
definition should take both somatic and mental health into consideration. It is easy to forget
mental health, perhaps especially in public health contexts.
The concept of health can either be understood in terms of normal biological and
physiological function, in terms of physical and mental functioning of the person as a
whole, in terms of well-being, or pluralistically.
34
Section 1: Concepts
Biomedical definitions
It is sometimes claimed (for example by Boorse, 1977) that good health is the same as
absence of disease or disorder. This definition is normally supplemented by a biomedical
conception of disorder, that is, the machine-fault model. On this view, a person is in good
health if there are no structural or anatomical abnormalities, and if there are no part
dysfunctions. The objection that this leaves no room for variations in positive health is
easily met. Even if there are no machine-faults, for example, even if all the parts of the
apparatus are functioning normally, the apparatus can be in better or worse shape. This
suggests that we can simply stipulate that a person (an organism) is in full health if its parts
are in good order, and if the different physiological systems are working well. This is a
reductionist view, since the health of the whole is regarded as wholly dependent on the
functioning (or health) of its parts.
If this view is adopted, how can a persons health be measured? Let us start with the
negative case: If someone suffers from a disorder, it is often the severity of the symptoms
that determine how severe the disorder is, that is, how much the person is suffering and to
what extent they are disabled. (We should not forget the risk element, however, for example,
a disorder can be severe in virtue of giving rise to a high risk of premature death). In the
positive case, we can of course measure the ability to absorb oxygen or the like, but it seems
easier to measure for example, athletic performance, that is, how the person is functioning
as a whole. In short, the biomedical definition cannot help us to measure health; for this
purpose, we have to rely on other criteria. Consider also the fact that many diagnoses are
symptomatic in character, especially in psychiatry, which suggests that health is often the
same as absence of symptoms. And since most symptoms are either sufferings or disabilities, this line of reasoning also suggests that we measure health in terms of functioning and/
or well-being. On the next two definitions, the clinical status of organs, tissues and cells is
not conceptually related to health, but a determinant of health.
Functional definitions
On the functional view, to be in good health is (roughly) to function well as a whole person.
However, this should not be taken to imply that every improvement in functioning is also
an improvement in health. After all, there are abilities (such as the ability to read or to
swim) that have nothing to do with health. A functional definition must therefore contain
some idea of what abilities are health-related, for example, what it is that a healthy person
can (by definition) do better than an unhealthy person. It must also contain some reference
to the circumstances involved. That a person cannot walk is often considered a disability,
but that he cannot walk in a hurricane is not. So, the question is under what kind of
circumstances a person should be able to function well to be regarded as healthy.
Regarding the relevant circumstances, there have been four different suggestions: It has
been suggested that a person cannot be regarded as fully healthy unless they are functioning
well (1) in their actual environment, (2) under normal circumstances, (3) under reasonable
or acceptable circumstances or (4) in many different types of circumstances (for example,
the better they are functioning in adverse circumstances, the healthier they are). There is
good reason to believe that the last two suggestions are most plausible (Brlde and
Tengland, 2003; Chapter 9).
In response to the question of what abilities are health related, there are at least eight
different suggestions. A persons level of health is at least partly dependent on to what
35
extent they have the ability to (1) satisfy their basic needs, (2) realize their goals, (3) realize
their vital goals, (4) acquire the ability to realize their vital goals, (5) live normally, that is, do
the things people normally do, or (6) live a life that is normal for themselves. In my own
view, a persons health is conceptually dependent on (7) what basic abilities they have, on the
one hand, and (8) their ability to perform, that is (roughly) the ability to make use of the
abilities one has, on the other (Brlde, 2000a, 2000b; Brlde and Tengland, 2003).17
Well-being definition
Health is rarely defined exclusively in terms of well-being, WHOs (1948) classical definition is an exception. Well-being is often regarded as one of several components of the
concept, however. In this view, a persons level of health is, in part, conceptually dependent
on their level of well-being. Or alternatively put, some improvements in subjective wellbeing are also, by definition, improvements in health. This should not be taken to imply
that every improvement in well-being is also an improvement in health. If a person becomes
happier as a result of winning the lottery or meeting a new partner, this has nothing to do
with health. Just as in the functional realm, we need to know what feelings are health
related. In my view, a pleasant or unpleasant mental state is health related if it is, to a
considerable extent, internally caused, and associated with the ability to perform. This is a
tricky issue, however, since a persons hedonic level is typically caused by both internal and
external factors. Anyone can be happy under optimal circumstances, which suggests that
circumstances should be normal (whatever that means). We should also remember that
some people put themselves in positions that would make everyone suffer. In this case,
however, the problem is really functional.
Pluralistic definitions
A definition is pluralistic (or multi-factorial) if two or more of the above considerations
are combined into a single definition. In my own tentative view (Brlde, 2000a, 2000b),
a persons position in the health-illness dimension is a function of their respective
positions in five different dimensions, viz. (1) their clinical status, or organ function, (2)
their ability to perform, to use the abilities they have, (3) to what extent they have certain
relevant basic abilities, (4) what their health-related mood state is like and (5) how
pleasant-unpleasant their health-related bodily experiences and emotions are. The first
of these dimensions is biomedical, the following two are functional, and the last two
concern the persons well-being.
Another way to accommodate all three definitions is to introduce a distinction
between fundamental health and manifest health. A persons manifest health can vary
from day to day, it gets worse when we catch colds, and it can be temporarily improved by
different forms of doping or recreational drugs. To have good manifest health is roughly
to be in good shape. A persons fundamental health is of a more dispositional or resourcelike kind, and it remains intact when she is struck by the flu. Fundamental health is to a
considerable extent a matter of how resistant and resilient one is, and it determines how
17
Nordenfelt (1995: 148) defines health in terms of functional ability only. In his view, A is healthy
if, and only if, A has the second-order ability, given standard circumstances, to realize all his
vital goals [that is, the set of goals that are necessary and together sufficient for his minimal
happiness].
36
Section 1: Concepts
disposed one is to fall ill or to die prematurely. Now, it seems that manifest health is
mainly a matter of (health-related) functioning and well-being, whereas fundamental
health is more of a biomedical matter. In this way, all three definitions can be
accommodated.18
Health and public health

To promote health is clearly an appropriate individual goal of public health, at least up to a
certain limit (see above), and as long as the individuals involved give their consent.19 But is
it an ultimate goal? It seems that subjective well-being is a better candidate than functioning
in this regard, since it is good as an end whereas functioning is most likely merely good as a
means. This suggests that functioning should only be promoted to the extent that it can be
expected to have positive effects on well-being. Subjective well-being is more closely related
to quality of life than to health, however, and this suggests that the ultimate individual goal
of public health is to promote good and long lives rather than to promote long and
healthy lives. But is it really a plausible goal of public health activities to promote peoples
quality of life? Is it a good idea to engage in what Seedhouse (1997) refers to as good life
health promotion?
A preference-based health index?

Before we take a closer look at this possibility, let us first look at the preference-based health
index approach. (This approach can be regarded as something between the health approach
and the quality of life approach.) By letting individuals themselves rank possible health
states with regard to value, we can (for example, by calculating the average) create a
preference-based health value index in which a utility number is assigned to each possible
health state. Here, it is possible to make use of the Rosser classification (Rosser and Kind,
1978), a two-dimensional system where health states are ranked with regard to disability
(8 levels ranging from 1 normal functioning to 8 unconsciousness) and suffering
(4 levels ranging from A no suffering to D severe suffering). We can assign the utility
number 1 to health state 1A, and then make use of the individual preference rankings to
calculate a utility number for the other 28 possible health states (8 BD are not possible).
This utility number may well be negative, for example, in the case of 7D, bedridden with
severe suffering (Levin et al., 1991).
The individual rankings can be generated in several different ways, for example, by
using Standard Gamble (SG), Time Trade-Off (TTO), Rating Scale or Magnitude
Estimation (Karlsson, 1991). Let us call the health state we want to evaluate H. In SG,
the individual is supposed to choose between H (with certainty) and a lottery (or gamble)
with full health and death as prices. The utility of H (for the individual) is calculated by
18
19
It should be noted that all pluralistic approaches make it rather tricky to measure health. For
example, how should ones position in the functioning dimension be combined with ones position
in the subjective well-being dimension to yield a single measure? Maybe we should let each
individual attach a weight to each dimension, assume that the dimensions are independent, let each
individual place a value on the different levels in each dimension, and then use the average value
to calculate the health level?
The question of to what extent it is acceptable to promote health by paternalistic means falls outside
the scope of this chapter.
37
finding at what odds the individual is indifferent between H and the lottery. In TTO, the
choice is between being in H during a certain time and being fully healthy during a
shorter time, for example, between 5 years in full health and x years in H. By finding the
x-value where the individual is indifferent between the alternatives, we can assign a utility
value to H. This method provides us with a way to balance health gains against life years
gained, since this is exactly the kind of balancing the individuals are asked to make. This
may well be the best solution to the calibration problem formulated above.
Utility values of different health states (or severity weights for different disabilities) can
also be generated by the Person Trade-Off (PTO) method. Here, people are asked how
many outcomes of one kind they consider equivalent in social value (for example,
measured in terms of claims on resources) to x outcomes of another kind, where the
outcomes are for different groups or individuals with different conditions (Brock, 2004).
That is, people are asked to adopt the perspective of a policy maker. There are several
versions of PTO. Let us assume that we want to determine the preference weight for a
health state H, for example, a state of severe disability. One way to do this is to make
trade-offs between life extensions for people in H and equal life extensions for healthy
people, for example, by asking how many one year life extensions for people in H are
equivalent in value to 1000 one year life extensions for healthy people (Murray and
Acharya, 1997). Suppose we regard 8000 life extensions for people in H as equivalent in
value to 1000 life extensions for the healthy. In this case, the value of H is 0.125. Another
way to determine the value of H is to make trade-offs between raising those in H to
perfect health for one year and extending life for healthy individuals for one year, for
example, by asking how many improvements for people in H that are equivalent in value
to 1000 one-year life extensions for healthy people (Murray and Acharya, 1997). Suppose
we regard 5000 one-year long improvements for people in H (to full health) as equivalent
in value to 1000 one-year life extensions for the healthy. In this case, the value of H is 0.2.
It is worth noting that the PTO approach is designed to permit people to incorporate
concerns for equity or distributive justice into their judgements about the social value of
alternative health programmes (Brock, 2004: 221). The values generated by this approach
are (most probably) strongly affected by distributive concerns, and they are thus unsuitable for assigning utility values to individual health states. It is better to use some other
method to determine the relevant utility values, and then use the PTO method as a purely
distributive device (Brock, 2004; and below).
There are several general objections to the preference-based health approach. First, we
can ask how informed peoples preferences are in this area. The value of a certain health state
is mainly instrumental (it is good as a means rather than as an end), and it can be quite
difficult to determine the consequences of different health states. One might try to avoid
this problem by disregarding the preferences of non-disabled ordinary citizens, and instead
appeal to the preferences of people who are well acquainted with the relevant health state, for
example, the disabled. A problem with this approach is that the disabled tend to adapt rather
well to their condition, which may result in evaluations that are too positive. Another option
is to rely on the preferences of independent health experts, but this suggestion is also
problematic (Brock, 2004). It seems quite hard to determine whose preferences should
be used to determine the values of different health states, and since different evaluative
standpoints tend to give rise to different evaluations, this is rather problematic.
Second, we may ask what kind of value is assigned to the different health states.
With the exception of the PTO approach, it seems quite clear that it is value for the
38
Section 1: Concepts
individual rather than value from a public health perspective (the kind of value public
health is supposed to promote or maximize). This suggests that the health-related utility
values that are generated by for example, SG or TTO have a diminishing marginal value in
a public health perspective (see above). A third objection is that the restriction to health
states is somewhat arbitrary. If we decide to appeal to peoples preferences when we
construct the relevant quality measure (the y-axis), why should we restrict ourselves to
how people evaluate their health states? Why not start from how people evaluate their lives
as wholes? This suggestion is very closely related to the idea that the most relevant
qualitative individual measure is quality of life.
To promote quality of life

The question of the good life (well-being, eudaimonia or quality of life) is one of the
classical questions in philosophy. This question has been formulated in somewhat different
ways, for example, What makes a life good for the person who lives it? or What does
ultimately make a life worth living?. In order to make this question more precise, philosophers have formally defined the notion of well-being (quality of life) in terms of what has
final value for a person. There are three important aspects of this formal definition. First, to
claim that a certain life is good, or that it is of high quality, is to evaluate it in a positive way.
That is, the question of the good life is a purely evaluative question, for example, it is not
empirical. Second, the type of value that is of relevance in this context is value-for (or
prudential value). When we say that someone has a good life, we do not mean that their life
is morally or aesthetically good, but that it is good for them. That is, the question of wellbeing is a question about what kind of life is good for the person who lives it. And third, the
relevant prudential values are final values rather than instrumental values. On this view, the
question of well-being is really a question of what is good for us as ends rather than as
means. To have a good life is simply to have a lot of positive final value (and little or no
negative final value) in ones life (Brlde, 2007b).
The different theories of the good life that philosophers have formulated and defended
over the years can be classified in somewhat different ways. Most of the modern discussion
of well-being is based on Parfits (1984) distinction between three kinds of conceptions of
the good life (or theories of self-interest), viz. hedonistic theories, desire-fulfilment theories, and objective list theories.20
According to the hedonistic theory, the good life is identical with the pleasant or happy
life. Well-being is the same as subjective well-being, to feel good. In this view, the
prudential value of a life is a function of how much pleasure and displeasure (for example,
suffering) this life contains. The more pleasure it contains, the better, and the more
displeasure it contains, the worse.
There are other mental-state theories besides hedonism, however, such as the happiness
theory. On the pure version of this theory, a persons quality of life is dependent on one
thing only, viz. how happy that person is. Whether this theory is not just a mental state
20
Parfits (1984) classification is not the only possible classification, however. An alternative
classification (closely related to Parfits) has been proposed by Kagan (1992), who suggests that
theories of the good life can be classified as subjective versus objective, on the one hand, and as
internalist versus externalist, on the other.
39
theory, but also a hedonistic theory, depends on what conception of happiness is incorporated into the theory, for example, whether happiness is regarded as an attitudinal state (life
satisfaction), an affective state (feeling good), or as a complex mental state consisting both
of an affective and a cognitive component.
According to the desire-fulfilment theory (desire theory, or preferentialism), a person has
a good life if and only if they have the kind of life that they themselves want to have. The
only thing that has positive final value for a person is that their intrinsic desires are fulfilled
(that they get what they want for its own sake), and the only thing that has negative final
value for a person is that their (intrinsic) aversions are fulfilled, that is, that they get what
they do not want. This is not a mental state theory, since whether a desire is actually
satisfied depends in part on the state of the world.
Finally, according to objectivist pluralism (the objective list theory, or the substantive
good theory), there are several objective values (besides pleasure or happiness) that make a
life good for a person, independently of what they themselves think of the matter, and to
have a good life is to have these values present to a high degree. Classical examples of such
alleged objective values are knowledge, contact with reality, friendship, love, freedom,
autonomy, to function well (for example, virtuously), personal development, meaningful
work and rational activity. Health or empowerment are other possibilities, but these values
are most often regarded as instrumental.
There are also mixed theories. The modified happiness theories are good examples of
such theories, for example, the idea that a life cannot be really good for the person who lives
it unless the cognitive part of their happiness (the positive value judgement) is informed
and autonomous (this is Sumners [1996] view), or the idea that a state of happiness is more
valuable for the happy person if the relevant value judgement is based on true beliefs about
her own life (this is a central part of my own view; Brlde, 2007c). On these mixed views
(which are all based on the hybrid view of the nature of happiness), the presence of a happy
mental state is necessary but not sufficient for maximal well-being, that is, these theories
require that the happiness in question satisfies certain further criteria.
Well-being and public health

Now, is the promotion of well-being (on any of these views) a plausible goal of public
health? It might be argued that it is. After all, that we live long and good lives is the only
ultimate goal, and institutional arrangements should be created with this in mind. So, why
not attribute this goal to public health activities? Here, it might be objected that a certain
division of labour is necessary, for example, that society is best divided into separate
spheres, and that each sphere (for example, the health care system) should focus on its
own goals (Brock, 2004). It might also be objected that some kind of bottom-up method is
most appropriate in this context, that is, that we should start from existing institutional
arrangements and their existing goals, and then reflect on these arrangements in order to
come up with suggestions for reform. For these reasons, I reject the idea that the
promotion of well-being is a plausible goal of public health. The suggestion that public
health should promote quality of life can only be plausible if this idea is qualified in
several ways. For example, it would be rather unreasonable to suggest that public health
should try to improve the quality of life in any respect (by relieving midlife crises) or by
any means (by offering luxury holidays). It seems that public health can and should
primarily promote quality of life through promoting health or through preventing disease
40
Section 1: Concepts
or injury. But we may add that public health should (ideally) only try to promote health
when this is expected to have positive effects on quality of life and/or life expectancy, since
this is what matters ultimately.21
To conclude, it seems that health is the most plausible qualitative measure we can find,
especially if it is taken to include a well-being component. To promote health is not just a
more plausible goal than preventing disease (since it includes the positive part of the health
scale), it is also a more plausible goal than promoting quality of life. This suggests that
something like HALE (healthy life expectancy) is the best individual summary measure we
can find. To repeat, we can represent the relevant aspects of a life graphically, as follows. On
the x-axis, weighted life years are represented, but the number of life years will serve as a
good approximation. A persons position on the y-axis depends on her weighted (healthrelated) functioning and well-being. It is hard to combine a measure of each into a single
measure, however. It is also hard to determine how much weight should be attributed to
this combined (and weighted) health measure compared with the (weighted) longevity
measure. In practice, it may well be necessary to make use of peoples preferences on both
counts, that is, first to create a single y-axis (as it were), and then to calibrate the resulting
health scale in a way that makes it possible to weigh, health gains against life extensions.
Only when this has been accomplished does it make proper sense to say that the individual
goal of public health is to maximize the area below the curve. All other possible goals of
public health, for example, to promote empowerment, are merely instrumental in relation
to this complex ultimate goal.
What distribution?
I have suggested that it is HALEs that count. The higher the HALE average in a population,
the better (from a public health perspective). The weighted HALE measure I have suggested
already contains some distributive considerations (a certain version of the idea that we
should give priority to the worst-off), but we may still ask how the HALE-levels of different
individuals (or groups) should be distributed. Besides the idea that we should give priority
to the worst-off (the priority view that has been incorporated into my weighted HALE
measure), there are at least three possible suggestions in this area (given a certain average
21
This suggests that health-related quality of life (HRQOL) might be the most appropriate
individual goal of public health. This is a rather tricky notion, however, for example, does it refer to
the part of a persons quality of life that depends on her health state, or to the part of a persons
health that has a positive effect on her quality of life? In either case, it seems unnecessary to
complicate matters by introducing such a notion, since we can say whatever we want to say in terms
of health and/or well-being. Moreover, as the category HRQOL is normally understood, it is far too
heterogeneous to be of any use. The instruments that purport to measure HRQOL normally contain
both functioning and well-being variables, and sometimes also certain features of external
circumstances, for example, vocational status. In fact, the phrase sometimes seems to denote
everything that medicine should aim at, at least everything of a psycho-social nature. As
ODonoghue et al. (1998) point out, [health-related] quality-of-life models typically consist of a
mixture of physical symptoms (for example, pain or fatigue), assessments of emotional state (for
example, happiness, anxiety and depression scales) and aspects of occupational and social
functioning. Treating these disparate consequences of disease within a single framework causes
difficulties. They also add that the very complexity of the notion of HRQOL suggests that it may be
helpful, when trying to understand the impact of therapeutic interventions, to focus on different
levels separately (1998).
41
level): to reduce health inequalities; to reduce health inequities; and to minimize the
number (or percentage) of people below a certain critical level.
Now, there are at least two ways we can go to find out what distributive considerations
are most plausible. The first way is to study the distributive aspect in and by itself, that is, to
keep the average level (and population size) constant and then try to rank different (nonweighted?) HALE-distributions where the HALE-levels of the different individuals are
represented on the y-axis and where the population is represented on the x-axis, for
example, from worse to better with regard to value or population health. To focus
exclusively on the distributive aspects of different HALE-distributions only takes us so
far, since we also want to know how the most plausible distributive principles should be
weighed against our aggregative concerns.
A more preferable approach is to study the distributive aspect in combination with
average level, population size, etc. In this view, the central question is how we should rank
different HALE-distributions with regard to value. (The reason why I phrase the question in
terms of value rather than population health is to make it explicit that the question is to a
large extent evaluative, which should come as no surprise given that its relevance is purely
normative anyway.) Or alternatively put, how should we measure population health, for
example, what changes should be regarded as improvements in population health?
Some distributive principles (for example, the priority view or the idea of a critical level)
are straightforward (but partial) answers to this second question, whereas others (for
example, to reduce inequality or inequity) are purely distributive. Let us start with the idea
that the central distributive goal of public health is to reduce health inequalities.
To reduce inequality
It can hardly be doubted that health and other values are very unequally distributed in this
world. What is inequality, and how can it be measured? And is the reduction of inequality a
plausible goal of public health? If so, is it the only distributive goal? And how should it be
weighed against the goal to increase the average?
On the standard egalitarian view, it is bad in itself that some individuals are worse off
than others (through no fault of their own). This also holds when the worst-off are pretty
well off in absolute terms. Should we also be concerned with equality between groups?
I think not. In practice, we often need to concern ourselves with group inequalities, but this
is mainly because some health inequalities between individuals are caused by social
injustices, and because it is a more manageable political goal to try to reduce the latter.
But we should also be concerned with inequality within groups.
So, how should we determine just how bad a certain unequal distribution is, that is, how
should we rank different distributions in this regard? Temkin (1993) has found six relevant
egalitarian considerations, three views of complaints and three principles of equality, where
each view of complaints can be combined with each principle of equality to yield a specific
ranking of distributions with regard to how good or bad they are with regard to equality.
The three possible views of how individual complaints should be measured, of how
serious a complaint is, are: (1) the relative to the average view of complaints (AVE); (2) the
relative to the best-off person view of complaints (BOP); and (3) the relative to all those
better off view of complaints (ATBO). The first view assumes that only those worse off than
the average has a complaint, whereas the last two views assume that everyone worse off
than the best-off person has a complaint.
42
Section 1: Concepts
There are also three principles of equality that tell us how distributions should be ranked
with regard to value (given a certain view of complaints): (1) According to the maximin
principle (MP), the larger the (average) complaint of the worst-off group is, the worse the
distribution (or world) is with regard to equality. If the complaint of the worst-off is
equally large in two worlds, the world where the worst-off group is smaller is better. If there
is still a tie between two worlds, the procedure should be repeated for the next worse-off
group, and so on. (2) On the additive principle (AP), the larger the sum total of complaints
is, the worse is the inequality. (3) On the weighted additive principle (WAP), we should
attribute larger weights to larger complaints before we add all the (weighted) complaints.
Temkin also points out that different traditional measures of inequality appeal to
different views of complaints, on the one hand, and different principles of equality, on
the other. For example, there are a number of measures that appeal to AVE and WAP, for
example, the variance and the standard deviation, the coefficient of variation, the standard
deviation of the logarithms and Atkinsons measure. To get the variance, we calculate the
difference between each value and the mean, we then square these differences (so that
larger deviations from the mean get more weight) and then add these squared differences
up. In this measure, deviations above the mean and deviations below the mean are given
the same weight. In the standard deviation of the logarithms, deviations from the mean
are weighed before they are squared and added, so that deviations below the mean are
given more weight than deviations above the mean. This measure has several plausible
features, for example, it seems reasonable to attribute more weight to larger deviations
from the mean (that is, to accept WAP) and to attribute more weight to deviations below
the mean than deviations above the mean. The problem with the standard deviation of the
log is that this is done in a rather arbitrary way.22
This is not the place to argue for any particular version of egalitarianism, or any
particular view of how inequality should be measured. However, I strongly suggest that a
plausible measure be based on moral considerations, for example, on some of the six
egalitarian considerations listed above. That is, we first need to take a view on what the
morally relevant aspects of inequality are. We can then find a way to measure each aspect. If
there are more aspects than one, we also need to determine the relative importance of each
aspect, give each (partial) measure a weight, and then combine them all into one single
measure.
A few words need to be said on the relative importance of equality, however. How much
does equality matter compared to for example, the average level? It is clear that equality is
not all that counts: if it were, we could improve the value of a distribution by making things
considerably worse for the best-off, for example, by injuring them. This is not plausible,
however, and we therefore have to admit that the average level matters too, and that tradeoffs are possible. But how should the level of equality be weighed against the average level,
for example, is it ever possible to improve the health of a population by making the
healthiest people a little less healthy? One way to find principled answers to such questions
22
In a similar way, it can be shown that the range appeals to BOP and MP, the relative mean deviation
appeals to AVE and AP, and the Gini coefficient appeals to ATBO and AP (Temkin, 1993).
There are many possible equality measures besides these, however, where it is not as clear which
egalitarian considerations are appealed to. Examples of such measures are the semiquartile range,
the interquartile range, and comparisons between the top third (1/5, 1/10, etc.) with the bottom
third (etc.) with regard to how much they have of the total share.
43
is to simply combine an aggregative measure and an egalitarian measure into one single
summary measure. For example, Veenhoven and Kalmijn (2005) have suggested a measure
of happiness in nations that combines utilitarian and egalitarian concerns (IAH inequality adjusted happiness). Here, equal weight is given to the average level of happiness and the
standard deviation (as a measure of disparity in happiness).
Egalitarianism versus the priority view

It is also possible to question egalitarianism altogether, that is, by arguing that peoples
relative standings are not morally relevant at all, that the important thing is that people live
as long and healthy lives as possible, not how long or healthy these lives are relative to
others. On this type of view, our legitimate egalitarian concerns are best captured by the
idea that we should give priority to the worse-off. There are several possible versions of this
idea. On the maximin version, we should give absolute priority to the worst-off, that is, the
value of a distribution is almost wholly dependent on how well off the worst-off group is.
The better off this group is, the better is the distribution, and the only way to improve
population health is to improve the health of the most unhealthy group. This view is not
plausible, since it suggests that we spend all resources on those who suffer most even when
this is of little benefit.
A much more plausible version of the priority view is the idea that a certain improvement has more value (or moral weight) if it befalls the worse-off: The worse someones
health is, the more important it is to improve his health. Health has diminishing marginal
value, that is, a given increase in health is better (makes a larger contribution to population
health) the further down the health scale it occurs. Every increase in health has value,
however, for example, a large improvement for the healthy may well have more value than a
small improvement for the unhealthy. (This idea is identical with (2) on page 28 above.)23
The view combines distributive concerns with aggregative concerns, and it thus partly
avoids the problem of how the two concerns should be weighed against each other. In this
view, it is not really possible to distinguish these two concerns from each other, but the
problem is still there (internal to the view), since the shape of the value function partly
depends on how the two concerns are weighed against each other. Once this problem is
solved, the idea is that we should simply maximize peoples weighed HALE-levels.24 No
further distributive considerations are necessary, for example, we need not be concerned
with questions of justice. Is this a plausible view?
To reduce inequity
Those who think there is still something to be said for egalitarianism (that is, that relative
levels matter) might consider the idea that not all inequalities matter, but only unjust or
23
24
It is not easy to determine who should count as the worst off, for example, those with the worst
health or those with the worst overall well-being. Moreover, we have to decide whether our
concern for the worst off should focus on who is worst off at a certain point in time or on who is
worst off over an extended period of time, such as a lifetime (Brock 2004). In my HALE view,
public health should focus on who has the worst health over a lifetime, but in the case of health
care, I tend to think that it is more appropriate to focus on who has the worst health at a certain
point in time.
It may not be possible to solve this problem in a principled way, however. In practice, it may well be
necessary to use the PTO approach; see below.
44
Section 1: Concepts
immoral inequalities. In this view, public health should try to reduce health inequities, that
is, those inequalities in health that are caused by unjust or morally unacceptable circumstances (Rogers, 2007). But what counts as an unjust inequality, that is, in virtue of what is
an unequal distribution unjust or morally unacceptable?
In public heath contexts it seems rather common to regard an inequality as an inequity
if it is due to the fact that the worse-off groups are poor, unemployed or discriminated
against, or that they have less access to education or health care. This view seems to assume
that all inequalities in health that are caused by social or economic inequalities should count
as inequities. On this view, only those health inequalities that are caused by natural
biological variations or freely chosen behaviours are morally acceptable (Rogers, 2007).
But are all socio-economic inequalities really unjust or morally wrong, and if so why?25 As
far as I can tell, the idea that all or most socio-economic inequalities are unjust is often
derived from the egalitarian idea that all inequalities are unjust unless they can be justified in
terms of for example, need or desert (something which is rarely the case).26 This view has
been challenged by certain kinds of liberals, however, who argue that a distribution (equal
or unequal) is morally acceptable if it has been caused by a process in which no negative
rights have been violated, for example, if it is a result of agreements between free individuals. This is a substantive issue in political philosophy that should not be avoided, but there
is no space to pursue it further in this chapter.
So, is it a good idea that public health should only concern itself with those inequalities
that are unjust? This of course depends on what makes an inequality unjust, but I think the
idea is practically useless in all its forms. In practice, it can be quite hard to distinguish those
inequalities that are unjust from those inequalities that are, for example, caused by freely
chosen behaviours. And even if this could be done, it would probably have no practical
importance, since public health work is not concerned with targeting individuals. To the
extent that egalitarianism and considerations of justice are relevant at all, public health
should try to reduce the inequalities it can reduce, and not just some inequalities.
A critical level or a non-principled alternative?

The fourth distributive idea is another version of the general idea that we should give
priority to the worst-off. On this view, there is a critical level analogous to the poverty line
(for example, minimally acceptable health and life expectancy), such that it should
have high priority to move people from below this line to above the line. The percentage
of the population above or below a certain level is a central measure when we measure the
poverty in nations (the poverty line). The question is whether this idea can work in health
contexts as well, and if so, how it can be incorporated into a plausible summary measure.
25
26
And are naturally caused inequalities morally acceptable? Many political philosophers would argue
that social justice requires that we compensate everyone who is worse off through no fault of her
own, for example, because she is born with a handicap. On this rather widespread view, we act
unjustly if we do not support these people, or more precisely: our institutions are unjust if they do
not support these people.
It is worth noting that on this line of reasoning, there is nothing special about socio-economic
inequalities: naturally caused inequalities are even harder to justify. The idea that socio-economic
inequalities are typically unjust may also be derived from the idea that everyone has a number of
welfare rights, but only on assumption that these rights are rather extensive in scope, for example,
that they entitle the right-holders to more than just a certain minimum.
45
The most promising way to justify the idea of a critical level (a cut-off point) is probably in
terms of rights. It is not unreasonable to assume that we all have certain welfare rights,
including a right to health. The human right to health is sometimes regarded as a right to
the highest attainable standard of physical and mental health, but it is more plausible to
regard it as a right to a certain minimum or adequate level of health care and other forms of
health support, for example, the level of support needed to receive a fair chance in life, or
why not 75 life years at a certain level of health (a fair innings as it were, given that
circumstances are normal)? But in my view, there is no need for any critical level. This level
would be pretty arbitrary, it would depend on the amount of available resources (that is,
vary from country to country), and it is doubtful whether there is any natural line that is
sharp enough to force us to modify the priority view.
There are also non-principled ways to evaluate different distributions. For example, it
might be possible to develop a quantitative tool that measures the specific weight people give
to equity concerns in comparing interventions that raise issues of distributive justice or
benefit individuals differently (Brock, 2004: 220). In Brocks view, the PTO approach may
well be useful for this purpose. Let us assume that we have managed to assign utility values to
different health states. We can then use PTO to determine the social value of whole distributions. For example, we might ask people how many patients treated with A (which will
improve the value of their health state from 0.25 to 0.45) would be equivalent in social value to
treating 1000 patients with B (which will improve them from 0.60 to 0.90). This will tell us in
quantitative terms how much importance people give to treating the sickest when doing so
conflicts with maximizing aggregate health benefits (Brock, 2004: 220). If we do a similar
thing in terms of HALE, it will help us construct the kind of function a prioritarian needs.
Conclusion: a summary measure of population health

To summarize the tentative suggestions I have made: The relevant individual outcome
measure is a form of weighted HALE, where health is, roughly, regarded as a combination
of (health-related) functioning and (health-related) well-being. We get this measure by
creating a function that assigns a value to each health level-life expectancy pair (according
to the priority view). In practice, it might be necessary to use some preference-based
approach, such as TTO or PTO, to determine these values. To obtain a decent summary
measure of population health, we simply have to aggregate these (weighted) HALE values.
The higher the average level of this value is (in the relevant population), the better, that is,
this is the value that public health should aim to maximize. It is possible that we can get an
even better measure if we incorporate some further egalitarian (justice-based) considerations as well, but we must not forget that some egalitarian concerns are already embedded
in the priority view. In short, the HALE version of the priority view may take us pretty far,
and we may have less need for egalitarianism than we think.
References
Anand, S. and Hanson, K. (1997) Disabilityadjusted life years: a critical review. Journal of
Health Economics, 16: 685702.
Boorse, C. (1975) On the distinction between
disease and illness. Philosophy and Public
Affairs, 5: 4968.
Boorse, C. (1977) Health as a theoretical

concept. Philosophy of Science, 44: 54273.
Brock, D. W. (2004) Ethical issues in the use of
cost-effectiveness analysis for the
prioritisation of health care resources. In
Public Health, Ethics, and Equity, ed.
S. Anand, P. Peter and A. Sen. New York:
Oxford University Press, pp. 20123.
46
Section 1: Concepts
Brlde, B. (2000a) On how to define the

concept of health: a loose comparative
approach. Medicine, Health Care and
Philosophy, 3: 3058.
Brlde, B. (2000b) More on the looser
comparative approach to defining health: a
reply to Nordenfelts reply. Medicine, Health
Care and Philosophy, 3: 31315.
Brlde, B. (2001) The goals of medicine:
towards a unified theory. Health Care
Analysis, 9: 113.
Brlde, B. (2003) The concept of mental
disorder. Web Series: No. 29. Gothenburg:
University of Gothenburg.
Brlde, B. (2007a) Mental disorder and values.
Philosophy, Psychiatry, and Psychology, 14:
93102.
Brlde, B. (2007b) Happiness and the good life:
introduction and conceptual framework.
Journal of Happiness Studies, 8: 114.
Brlde, B. (2007c) Happiness theories of
the good life. Journal of Happiness Studies,
8: 1549.
Brlde, B. and Tengland, P.-A. (2003) Hlsa och
sjukdom en begreppslig utredning [Health
and Disease: A Conceptual Investigation].
Lund: Studentlitteratur.
Buchanan, D. (2000) An Ethic for Health
Promotion. Oxford: Oxford University Press.
Culver, C. M. and Gert, B. (1982) Philosophy in
Medicine: Conceptual and Ethical Issues in
Medicine and Psychiatry. Oxford: Oxford
University Press.
Cummins, R. (1975) Functional analysis. Journal
of Philosophy, 72: 74165.
Kagan, S. (1992) The limits of well-being. In The
Good Life and the Human Good, ed. E. F.
Paul, F. D. Miller Jr and J. Paul. Cambridge:
Cambridge University Press.
Karlsson, G. (1991) Vrdering av hlsotillstnd
och individuella preferenser [Valuation of
health states and individual preferences]. In
Hlsa, sjukdom, livskvalitet [Health, Disease,
Quality of Life], ed. P.-E. Liss and
L. Nordenfelt. Linkping: Linkping
University, pp. 15966.
Kendell, R. E. (1975) The concept of disease and
its implications for psychiatry. British Journal
of Psychiatry, 127: 30515.
Levin, L-., Karlsson, G. and Carlsson P. (1991)

Vrdering av hlsotillstnd: en pilotstudie
[Valuation of health states a pilot study]. In
Hlsa, sjukdom, livskvalitet [Health, Disease,
Quality of Life], ed. P.-E. Liss and
L. Nordenfelt. Linkping: University of
Linkping, pp. 16781.
Lilienfeld, S. O. and Marino, L. (1995)
Mental disorder as a Roschian concept:
a critique of Wakefields harmful
dysfunction analysis. Journal of Abnormal
Psychology, 104: 41120.
Munthe, C. (2008) The goals of public health: an
integrated, multidimensional model. Public
Health Ethics, 1: 3952.
Murphy, D. and Woolfolk, R. L. (2000a) The
harmful dysfunction analysis of mental
disorder. Philosophy, Psychiatry, and
Murphy, D. and Woolfolk, R. L. (2000b)
Conceptual analysis versus scientific
understanding: an assessment of Wakefields
folk psychiatry. Philosophy, Psychiatry, and
Murray, C. J. L. (1996) Rethinking DALYs. In
The Global Burden of Disease, ed. C. J. L.
Murray and A. D. Lopez. Cambridge, MA:
Harvard School of Public Health, on
behalf of the WHO and the World Bank,
pp. 198.
Murray, C. J. L. and Acharya, A. K. (1997)
Understanding DALYs. Journal of Health
Economics, 16: 70330.
Nordenfelt, L. (1995) On the Nature of Health.
An Action-Theoretic Approach, 2nd rev. edn.
Dordrecht: Kluwer.
Nordenfelt, L. (2003) On the evolutionary
concept of health: health as natural function.
In Dimensions of Health and Health
Promotion, ed. L. Nordenfelt and P.-E. Liss.
Amsterdam: Rodopi Press.
ODonoghue, M. F., Duncan, J. S. and Sander,
J. W. (1998) The subjective handicap of
epilepsy: a new approach to measuring
treatment outcome. Brain, 121: 31743.
Parfit, D. (1984) Reasons and Persons.
New York: Oxford University Press.
Reznek, L. (1987) The Nature of Disease.
London: Routledge.
47
Rogers, W. (2007) Health inequities and the

social determinants of health. In Principles of
Health Care Ethics, 2nd edn, ed. R. Ashcroft,
A. Dawson, H. Draper and J. McMillan.
Chichester: Wiley.
Veenhoven, R. and Kalmijn, W. (2005)

Inequality-adjusted happiness in nations:
egalitarianism and utilitarianism married in
a new index of societal performance. Journal
of Happiness Studies, 6: 42155.
Rose, G. (1992) The Structure of Preventive

Medicine. Oxford: Oxford University Press.

meaning of public in public health. In
Ethics, Prevention, and Public Health, ed.
University Press, pp. 1329.
Rosser, R. and Kind, P. (1978) A scale of

valuations of states or illness: is there a social
consensus? International Journal of
Epidemiology, 67: 34758.
Seedhouse, D. (1997) Health Promotion:
Philosophy, Prejudice and Practice.
Chichester: Wiley.
Wakefield, J. C. (1992) The concept of mental

disorder: on the boundary between biological
facts and social values. American Psychologist,
47: 37388.
Sumner, L. W. (1996) Welfare, Happiness, and

Ethics. Oxford: Clarendon Press.
Wakefield, J. C. (2000) Aristotle as

sociobiologist: the function of a human
being argument, black box essentialism, and
the concept of mental disorder. Philosophy,
Psychiatry, and Psychology, 7: 1744.
Temkin, L. S. (1993) Inequality. Oxford: Oxford

University Press.
WHO (1948) Official Records of the World

Health Organization, Vol. 2. Geneva: WHO.
Tengland, P.-A. (2006) The goals of health work:

quality of life, health and welfare. Medicine,
Health Care and Philosophy, 9: 15567.
Wilkinson, S. (2000) Is normal grief a mental

disorder? The Philosophical Quarterly, 50:
289304.
Tengland, P.-A. (2007) Empowerment: a goal or

a means for health promotion? Medicine,
Health Care and Philosophy, 10: 197207.
Williams, A. (1997) Intergenerational equity:

an exploration of the fair innings argument.
Selgelid, M. J. (2008) A full-pull program for the

provision of pharmaceuticals: practical
issues. Public Health Ethics, 1(2): 13445.
Section 1
Chapter
Concepts
Selective reproduction, eugenics

and public health
Stephen Wilkinson
Introduction
In the wake of Nazi barbarities associated with eugenics, it has become impolitic to speak openly
about improving population health and well-being by trying to influence the kinds of people who
will be born. But is this goal in fact ethically unacceptable if dissociated with other immoral
features of many historical eugenics programs?
(Wikler and Brock, 2007: 87)
This question, posed in a recent paper by Wikler and Brock, is the main concern of my chapter.
Given recent and impending developments in reproductive and genetic technologies, it is likely
that intervening before birth, before embryo-implantation, or even before conception, to alter
the composition of future populations, will become an increasingly effective and cost-effective
strategy for improving public health. In essence, ensuring that healthier future people are born,
and conversely that unhealthier (possible) people are not born, may be a better way of delivering
improved public health than the more usual biomedical and environment methods. Or even if
it is not a better way, it may nonetheless be a useful addition to public health policys toolbox.
The first part of this chapter deals with a number of conceptual or definitional issues, all
of which concern, in one way or another, the relationship between selective reproduction
and public health. The second part of the chapter considers what I shall term eugenics
arguments against using selection, ultimately concluding that worries about the spectre of
eugenics do not give us strong enough reasons to avoid using selective reproduction to
achieve public health goals.
Selective reproduction
By selective reproduction I mean the attempt to create one possible future child rather
than another (Wilkinson, 2010). The reason for wanting to practise selective reproduction is
usually that one possible future child is, in one way or another, more desirable than the
alternatives. The kinds of desirability that people have in mind are many and varied, and the
question of what counts as desirable is controversial. There is however one relatively
uncontentious example: selection to avoid disease. If one possible future child would have
a disabling, excruciating and life-shortening disease, while another would not, then many of
us will think that ensuring that the disease-free child is created is the sensible thing to do.
Disease-avoidance is the most prevalent and widely accepted rationale for selective reproduction, at least within Western reproductive medicine and biomedical ethics.
48
Chapter 3: Selective reproduction, eugenics and public health
49
Recent biotechnological developments (notably, the advent of preimplantation genetic

diagnosis, the possibility of determining sex by sperm sorting and the ever increasing
sophistication of prenatal tests) have made selective reproduction a pressing issue for
regulators, academic bioethicists and the news media. However, selective reproduction
need not be high tech and has been around in technologically unsophisticated forms for
many years. Perhaps the most obvious example of this is using contraception or sexual
abstinence to delay conception. For instance, a (bioethically inclined) 15-year-old girl might
think to herself:
If I have a child now, then it will have a substantially lower quality of life than a different child
conceived when I am 25. Therefore, Ill wait for a decade and create a better-off child instead.
This is an instance of selective reproduction: choosing an apparently better off possible

future child over one that would be less well off. Interestingly, there is very widespread
agreement that this sort of selective reproduction (avoiding teenage pregnancy through
abstinence or contraception) is not merely morally unproblematic, but to be encouraged.
Another kind of low tech selective reproduction is where a woman chooses a sperm
donor with certain desirable characteristics (or indeed chooses to have sex with such a man)
in the hope that his advantageous features will be passed on to her children. A good
example of this is the renowned Repository for Germinal Choice. This was set up during
the 1980s and dubbed the Nobel Prize Sperm Bank because it was thought to contain
several Nobel Prize winners sperm. Similarly, in 1996, The Times reported that a sperm
bank had been established for members of MENSA who wish to help create a master breed
of super-intellectuals (Rogers, 1996). Members of MENSA, The High IQ Society, are
required to have an IQ in the top 2% (MENSA, no date).
Same/different number scenarios and the non-identity problem
In this section, I briefly explain two important distinctions. First, selective reproduction (as
I use the term here) can cover both choices between different possible future children and
decisions about how many children to have (if any). Choices of the latter kind have been
termed different number (because we are choosing between one number of children and
another) while those of the former kind are called same number (because we are choosing
not how many but rather which possible future children to create).1 The same number
versus different number distinction is often theoretically and ethically important but is
rarely clear-cut when it comes to questions of policy. This is because many policy decisions
affect both who comes to exist and how many people come to exist: both the constitution
and the size of the future population. For instance, vigorously discouraging teenage
pregnancy may well result in different children being born, but also in fewer children being
born.
The second, more complicated, distinction is between choices that are identity-affecting
and those that are not. Some choices alter the characteristics of a determinate future person
(making someone more or less healthy, for example) whereas others amount to choosing
1
This terminology originates in Parfits seminal work, Reasons and Persons. He writes: Different
Number Choices affect both the number and the identities of future people. Same Number Choices
affect the identities of future people, but do not affect their number. Same People Choices affect
neither (Parfit, 1984: 356).
50
Section 1: Concepts
between the creation of distinct possible future persons (choosing a healthier one over one
with a genetic disorder, for example). Selective reproduction (or at least the kind of selective
reproduction that is the subject of this chapter) involves the latter rather than the former.
The choices that we make when we practise selective reproduction are identity-affecting.
They are also existential, causing people to exist who otherwise would not, or preventing the
creation of (possible future) people. To make this distinction clearer, think about the
following actions or policies:
Vaccinating infants
Encouraging breast feeding
Discouraging smoking and drinking during pregnancy
----------Discouraging teenage pregnancy
Preimplantation genetic diagnosis (PGD) and embryo selection
Sperm sorting for sex selection
Having unprotected sex with numerous MENSA members (for reproductive purposes)
rather than ones intellectually mediocre husband.
The first set of policies are not (or at least not directly or intentionally) identity-affecting.
These policies, if successful, would lead not to the creation of different possible future
people but rather to improvements in the health and longevity of people (or foetuses) who
already exist, or whose future existence is not dependent on the action or policy in question.
Of course, even these policies may have indirect effects that are identity-affecting or
existential, affecting which (possible future) people come to exist and/or whether certain
(possible future) people come to exist. Take the following case, for example:
Grace, a young woman, smokes and drinks heavily during pregnancy. Tragically, as a result of this
behaviour, Graces baby, Olivia, dies during infancy. A year later Grace proceeds to have another
child, Jack, who is in good health. Grace is certain that, had it not been for Olivias death, she would
not have had Jack.
If the population contains a number of women like Grace, it is safe to assume that
implementing policies that discourage smoking during pregnancy will have identityaffecting side-effects. They will lead to the survival of babies like Olivia (which is not,
considered in itself, identity-affecting) but also to the non-conception of (possible future)
babies like Jack (an existential effect). But at least we can say of the policies in the first group
that, while they may indirectly affect who comes to exist, this is not their main aim.
The actions and policies in the second group, conversely, are directly identity-affecting
and altering the constitution of the future population is part of their aim (although those
involved may not conceptualize it in quite these terms). These policies determine which
(possible future) people will come to exist. For example, if we successfully encourage
prospective teenage mothers to delay their pregnancies by ten years then we are getting
them to have different children from the ones they would otherwise have had (and/or to
have fewer children). When, following PGD, we choose between embryos for possible
implantation we are choosing between one (possible future) child and others. When we
separate X-chromosome sperm from Y-chromosome sperm and choose to use one or the
other for sex selection, we are choosing between different (possible future) children and
may, rather obviously, end up with a boy instead of a girl. And the same goes for a woman
51
who is inseminated by MENSA members, rather than her husband: she is choosing to have
a different (possible future) child.
Why do we generally regard such policies as identity-affecting? It may be because we
accept what Parfit calls the Time-Dependence Claim:
If any particular person had not been conceived within a month of the time when he was in fact
conceived, he would in fact never have existed.
(Parfit, 1984: 352)
However, this looks less credible than when his Reasons and Persons was first published
because of the possibility of freezing and later using gametes. So perhaps what Parfit calls
the Origin View is what underpins these intuitions about identity:
. . . each person has this distinctive necessary property: that of having grown from the particular pair
of cells from which this person in fact grew.
(Parfit, 1984: 352)
According to the Origin View, coming from the particular sperm and the particular ovum
from which I in fact originate is a necessary condition for somethings being me. So, if those
particular gametes had never existed, or if a different sperm had got to that egg first, then
I would never have existed. A different (numerically distinct) person would have come to be
instead.
In the case of preimplantation genetic diagnosis and embryo selection (and a fortiori in
the case of selective abortion) there may be an additional reason for regarding selective
reproduction as identity-affecting. In principle, each of the available embryos could be
implanted and become a person (Embryo A would become Person A, Embryo B would
become Person B, Embryo C would become Person C, etc.). And it seems to follow that,
just as we could decide to implant them all and thereby create A, B and C, so we can
choose between (possible future) Person A and Person C, by choosing between Embryo
A and Embryo C. Perhaps one of the main underlying thoughts in the embryo selection
case is simply that Embryo A and Embryo C are distinct organisms, spatially distinct
physical objects. Thus Person C, qua Adult Body C, could not have been Embryo
A because Embryo A and Adult Body C lack physical spatio-temporal continuity (whereas
Adult Body C and Embryo C are spatio-temporally continuous). Two things (or thingstages) are spatio-temporally continuous, in this sense, if they are connected by a
continuous path through space-time; spatio-temporal continuity is widely regarded as
being essential to the identity or persistence of physical objects over time.
Something like the Origin View (which we may also term gametic essentialism)
appears to be the bioethics orthodoxy, but it is not entirely uncontroversial and has
been questioned by some scholars (Wrigley, 2006). However, even if we were to eschew
a fully fledged gametic essentialism (and to accept, for example, that there is a possible
world in which I came from a different spermatozoon) we could nonetheless retain the
view that embryo selection (and a fortiori foetus selection) involve choosing between
different numerically distinct (possible future) persons. This is because of the point just
made: that in such cases we are choosing between existent spatio-temporally distinct
physical objects. So since practices of this kind are my main concern in the rest of this
chapter, it is safe to assume (for the present purposes) that selective reproduction is
identity-affecting.
52
Section 1: Concepts
Table 3.1 Ruby and Emily: summary table of benefits and harms
Who benefits?
Who is harmed?
Decide to create Ruby
Possibly Ruby, but only existential benefit
Neither
Decide to create Emily
Possibly Emily, but only existential benefit
Neither
The main reason for dwelling on the fact that selective reproduction is identity-affecting
is that this has implications for the way in which the ideas of benefit and harm can be
applied. For example, say we have a choice between bringing into existence:
RUBY: a (possible future) healthy child, and
EMILY: a different (possible future) child with a genetic disorder, Z Syndrome; and that, according to existing
people with Z Syndrome, it is extremely debilitating and painful but consistent with living a fulfilling and
worthwhile life.
If we choose to create Ruby, who will have been benefited and who will have been
harmed (leaving aside for the present the interests of third parties, such as parents and
the National Health Service)? Arguably Ruby benefits, but the benefit for her is existential,
the gift of existence; crucially, it is not being born without Z Syndrome, because the choice
facing us was not Ruby with or without Z Syndrome, rather it was Ruby or Emily. Of course,
Ruby might feel glad not to have Z Syndrome and is better off than she would be if she had
Z Syndrome, but this is true of all healthy children, not just selected ones. Turning now to
Emily, she does not appear to benefit at all, both because she does not exist and because
(even if we can make sense of non-existence being beneficial) her life, had it existed, would
have been worth living and of some positive value to her. Has Emily been harmed by the
decision to select Ruby? Probably we would say no because of the problematic (arguably
absurd) implications of allowing the idea that we harm a merely possible future person when
we fail to bring her into existence.
What if we had chosen Emily instead? Similar considerations apply. If Emily has a
worthwhile life overall we might say she had been benefited existentially (just as Ruby
would have been if she had been chosen). And, while she would be better off without
Z Syndrome (in that sense, it is a harmful condition), creating Emily has not harmed her
because she would not be better off not existing. As for Ruby, we probably want to say (as
we did of Emily in the previous scenario) that her non-existence neither benefits nor harms
her (Table 3.1).
The relevance of this is that often, when practising selective reproduction, while there is
an overall welfare gain (because of the absence of disease), there may be no particular
individual who benefits (or is protected from harm) in any straightforward way because the
welfare gain is distributed across different possible populations.
Selective reproduction and health

Is selective reproduction a public health issue? This question can be broken down into two
parts. How does selective reproduction relate to health? And are those aspects of health to
which it relates public health?
Selective reproduction does not necessarily have anything to do with health and it
can be used for a variety of non-health purposes. Sex selection (when not used to avoid
53
sex-linked disorders) is an obvious example of this. However, as I mentioned at the

outset, using selective reproduction to avoid disease, to bring about the birth of a
healthy (or healthier) child is (at least in the UK and many other Western countries)
the most widely accepted rationale for selective reproduction. The link between selective
reproduction and health is strongest (though still contingent) when we look at law
and policy (and here my focus will be almost entirely on the UK). Two areas in
particular stand out: the law on abortion, and the regulation of preimplantation genetic
diagnosis (PGD).
Abortion
The Abortion Act 1967 allows for termination to be lawful where one of a number of
conditions is met. One of these grounds, contained in section 1(1)(d), is that abortion may
be authorized by two doctors who agree that there is a substantial risk that if the child were
born it would be seriously handicapped. Since 1990, there have been no time limits for
terminations performed on this ground, a policy criticized both by pro-life moral conservatives and by many activists and scholars from a disability rights perspective (Sheldon
and Wilkinson, 2001). Section 1(1)(d) of the Abortion Act thus enshrines in legislation the
health rationale for selective reproduction (in this case, selective termination) by giving a
privileged status to abortions which are performed in order to avoid the birth of a child
with a disability or disorder (a serious handicap).
The claim that this legislation is based upon the health rationale for selective reproduction does however require a couple of qualifications.
First, someone might object that the rationale for section 1(1)(d) is not obviously
health (or not obviously just health). It may, for example, be concern for the welfare of
prospective parents, or about the costs of health and social care. This may well be true
and, in an earlier paper, Sally Sheldon and I explored some of the different possible
justifications and motivations for this piece of legislation (Sheldon and Wilkinson,
2001). That said even if the fundamental justification for section 1(1)(d) is saving
the National Health Service money, or concern for the welfare of the prospective parents
health, or rather disease-avoidance, is nevertheless the intended means of delivering
these more fundamental goods.
The second qualification is that some people, especially (though not exclusively)
adherents of the social model of disability, may want to say that the rationale for
section 1(1)(d) cannot be health (or cannot just be health), because many of the
characteristics selected against are not states of unhealth, but rather characteristics
which are intrinsically neutral in health terms, but which cause people to suffer disadvantage and discrimination.2
We can distinguish three different critiques here.
The first, what we might term social modelism, says that none of the so-called
serious handicaps cited as grounds for abortion should be regarded as disorders; rather,
they are all simply bases for social discrimination. On this view, being selected out
(aborted) for having a serious handicap is morally exactly like being selected out for
being black, female or left-handed. This position seems implausible to me because at
2
For exposition and critique of such views see, for example: Harris (2000), Oliver (1990; 1995),
Shakespeare (2006) and Terzi (2004).
54
Section 1: Concepts
least some of the conditions labelled serious handicap do involve serious physical
impairment, including pain and/or premature death. For example, the Abortion Statistics
for England and Wales: 2006 (Department of Health, 2007a) tell us that section 1(1)(d)
was used to justify 166 terminations for anencephaly (all before 24 weeks) and 44 for
hydrocephalus (of which 10 were after 24 weeks). And, even if one accepts (as we
probably should) that people with these conditions suffer from social discrimination,
it is hard to deny that they are states of unhealth, given their very obvious effects on
length and quality of life (effects that would, crucially, exist even without social
discrimination).
The second critique concedes that some of the serious handicaps are disorders but
claims that, in practice, the line is drawn in the wrong place and that some things are
misclassified as diseases. Claims of this kind received considerable attention when, in 2002,
it was reported that a late termination had taken place on the grounds that the aborted
foetus had a cleft lip and cleft palate. Joanna Jepson, at the time studying to become an
Anglican clergywoman, complained to the police about the case (which came to light in the
official abortion statistics). Jepson is reported to have said:
This is eugenics . . . I am not prepared to sit back and let it happen. I want people to know that this is
the way society is going, and we need to do something about it.
(Rogers, 2002: 13)
One issue was whether cleft palate is a health problem or whether it is merely a cosmetic
state. Jepson, writing in the Sunday Telegraph, comments:
It is as though such a child though the cleft palate condition is trivial and easily corrected by surgery
somehow represents a threat to our society, increasingly obsessed as it is by a harsh, unattainable
notion of cosmetic perfection.
(Jepson, 2003)
Another example raising similar issues is Downs syndrome. This is a relatively common
reason for abortion under section 1(1)(d), with 436 such terminations taking place in
2006 in England and Wales (but just 12 after 24 weeks) (Department of Health, 2007a).
As with cleft palate, though, some people want to argue that Downs syndrome is more
of a difference than a disease and that, as such, it does not justify abortion under section
1(1)(d).
Assessing the health status of cleft palate and of Downs syndrome is not the purpose
of this chapter and there is not space here to consider the arguments on both sides
(including the various different accounts of disability and disease that are in play).
I simply note for the time being that both cleft palate and Downs syndrome do often
involve features that we generally associate with disease, most conspicuously loss of
mental and/or physical functioning. Also, as Rev Jepson herself concedes in the above
quotation, cleft palate can be easily corrected by surgery which suggests that it is a
defect standing in need of a cure.
Third, and finally, it might be argued that although most of the so-called serious
handicaps used to justify abortion really are disorders, some of them (and, again, cleft
palate and Downs syndrome will be leading examples) are not sufficiently serious to justify
abortion. This view does not challenge the disease-status of these handicaps but rather
questions whether their impacts on (among other things) quality of life are sufficient to
merit termination. I mention this view just for analytic completeness and will not attempt
55
to assess it here. It is a view that will depend, for each condition, on the facts of the case
(for example, to what extent, if any, does Downs syndrome reduce quality of life?).
Preimplantation genetic diagnosis (PGD)

Turning now to the regulation of preimplantation genetic diagnosis, the Human Fertilisation and Embryology Authoritys (HFEAs) Code of Practice says:
It is expected that PGD will be available only where there is a significant risk of a serious genetic
condition being present in the embryo.
(HFEA, 2003)
Hence, the view that the proper purpose of PGD (and selective reproduction generally) is to
avoid genetic disorders is not merely part of established ethics and practice but is (in the UK
at least) enforced by the regulator.
Interestingly, this view prevails (and is perhaps even slightly strengthened) in the
Human Tissue and Embryos (Draft) Bill which, at the time of writing (2007), is being
debated and scrutinized by the UK Parliament. The present version of the Bill would, if
enacted, allow embryo testing to be licensed by the HFEAs successor-body only for specific
purposes laid down in the draft legislation (Department of Health, 2007b: Schedule 2, 3,
1ZA(1)). In general terms, these are:
to ascertain how likely a particular embryo is to result in a live birth (and to enable the selection of
embryos on these grounds);
to enable the creation of a saviour sibling, in cases where the existing child (the one in need of salvation)
suffers from a life-threatening medical condition which could be treated by umbilical core stem cells;
where uncertainty has arisen, to test the embryo to find out whose gametes it came from;
to test for gene, chromosome, or mitochondrion abnormality or to sex select for the purposes of
avoiding sex-linked genetic disorders.
So, with just one exception (testing to ascertain who the prospective parents are), all of the
grounds for using PGD are health-oriented. PGD must aim either to increase the chances of
a live birth (choosing a healthy embryo), or at ensuring that the child born is free from
genetic disorders, or at improving the health of an existing sibling.
Selective reproduction and public health

So, while selective reproduction is not necessarily a health-oriented activity, it is in practice
often closely tied to health: in particular, to disease-avoidance. Indeed, when it comes to
reproductive technologies like preimplantation genetic diagnosis, striving to create a healthy
baby is (and is likely to remain) pretty much the only permitted kind of selection (with the
small exceptions noted above). Selective reproduction then is closely associated with health,
but is this public health? As with the link between selective reproduction and health, the
answer here is that, while there is no necessary connection between selective reproduction
and public health, there is a strong contingent connection and, in particular, many of the
policy and regulatory issues raised by selective reproduction are public health issues.
My reasons for thinking this are based in part on Verweij and Dawsons (2007) useful
discussion of defining public health. They draw our attention to two senses of public in
public health. The first is the health of the public:
56
Section 1: Concepts
. . . definitions and concepts of public health . . . almost all pick out interventions that aim at
protecting and promoting the health of the public. Talking about the health of the public obviously
involves the health of more than one person (or even a few persons). Public health concerns the health
of populations, or at least larger groups of persons.
(Verweij and Dawson, 2007: 22)
Trying to improve the health of the public, as opposed to merely the health of individuals,
has two distinctive features. First, as Verweij and Dawson put it:
Public health interventions are expected to make a difference on a population level, and this seems to
imply that they should affect the health of many.
(2007: 22)
So there is a quantitative threshold for what counts as a public health intervention. Merely
treating half a dozen patients in my own clinic will not normally be a public health
intervention; although, of course, it could be if, for example, they were carriers of a highly
infectious disease, or if these treatments were part of a wider systematic treatment programme. Public health actions then must impact substantially on the overall health status of
a population: typically (though not necessarily) in ways that would show up in populationbased statistical data (for example, about the prevalence of certain diseases). Verweij and
Dawson also explore various complications concerning, among other things, what counts as
a population, but I shall leave these to one side for the present.
The second relevant distinctive feature of public health actions is that they need not
benefit identifiable individuals:
It might be unclear, even with hindsight, which persons in fact benefited from the intervention. This
is one of the salient dimensions of prevention: effective primary prevention results in things that do
not happen (e.g. the onset of disease in persons). For example, as a result of an effective Hepatitis
B vaccination programme fewer people will get Hepatitis, yet the persons that benefit are not
identifiable, and success exists only in a statistical sense . . .
(Verweij and Dawson, 2007: 22)
Given Verweij and Dawsons remarks, it seems that the relationship between selective
reproduction and public health is as follows.
Many individual decisions made by prospective parents (to go through PGD, or to have
an abortion, for example) have nothing much to do with public health (even where the
motivation is the desire to avoid the creation of a child with a disease or disability); and
much the same can be said of the actions of individual clinicians. This is chiefly because
such decisions, considered in isolation, do not meet the quantitative threshold for public
health: essentially, they only affect (and are only meant to affect) one or a very small
number of persons.
When however we move to the policy level then selective reproduction can, and often
does, have a public health dimension. A fairly clear example of this is policies on prenatal
testing and abortion. (Policy here could cover law, National Health Service policy, or
professional practice.) Imagine, for example, that a government encourages all pregnant
women to have their foetuses tested for Condition X, provides these tests for free, informs
the pregnant women that Condition X is hideous (both for those with it and their carers),
and ensures that abortion is both lawful and readily accessible (at least where the resultant
child would have Condition X). This is a clear case of public health policy (although it may
have non-health aims as well, such as ones to do with the autonomy of women).
57
Why is it a clear case of public health policy? First, because there is an obvious
population health end in sight: reducing the prevalence of Condition X. Second, the policy
would, I imagine, be above the quantitative threshold for public health interventions,
given that all pregnant women are encouraged to have the test (and so the policy applies to
them even if they decline the test, and even if the test result is negative). Finally, in
common with many paradigm public health interventions, many of the benefits of the
policy cannot be attached to determinate individuals in a straightforward way. Obviously,
some of the benefits here can be attached to individuals. For instance, the testing
programme will generate cases in which women abort foetuses with Condition X and go
on to have other healthy children instead (or choose to remain childless). Given the
dreadfulness of Condition X, probably many such women will live happier and healthier
lives than they would have lived if they had been the mothers of children with Condition X.
However, even for these women, there will be residual doubts about which ones have
truly benefited, because we will not know for sure how well they would have coped with a
Condition X child. The other benefits of reducing the prevalence of Condition X will be
much harder to attach to any determinate individual. As Verweij and Dawson point out
this is to a certain extent just a general feature of preventative measures. But one
additional consideration here is that the policy is identity-affecting. Thus, if it is successful,
many parents will in effect choose to replace a (possible future) child without Condition
X for one that would have had Condition X; in other words they will choose to abort the
Condition X foetus and to have a different healthy child later. Who benefits from this?
I have already mentioned possible third-party benefits to the parents and there may also
(depending on how the health economics works out) be some savings for public service
budgets. But what about direct benefits? The candidate direct beneficiaries seem to be the
abovementioned substitute children, those who would not exist were it not for the
governments Condition X policy. If these children really are beneficiaries of the policy
then the benefit involved is what I earlier termed existential. For these children have not
been cured of, nor even prevented from having, Condition X. Rather, they have been
created instead of someone else who would have had Condition X. So the benefit of which
they are recipients (if any) is existing-rather-than-not-existing, not being free from Condition X. This of course raises various philosophical conundrums, ones that I will not get
into here. For the present, it will suffice to say that it is difficult to assign some of the
benefits of the policy to determinate individuals in a straightforward way, and that this is
a typical characteristic of public health measures.
Policy concerning PGD can also be part of public health policy. For instance, the
government could encourage couples at risk of having a Condition X child (or indeed all
couples) to conceive using IVF and PGD. And, possible ethical differences notwithstanding,
from a public health point of view, this would have essentially the same features as avoiding
Condition X through prenatal testing and selective termination. In practice, PGD probably
does not yet meet the quantitative threshold for being a public health measure since, as
recently as 2004, there were only around 200 treatments per year in the UK (HFEA, 2005: 7).
We can certainly envisage it and similar selection practices becoming more prevalent and
eventually being viable public health interventions. As Wikler and Brock put it:
New reproductive methods present some choice among potential offspring, and more choices lie just
over the horizon.
(2007: 87)
58
Section 1: Concepts
Eugenics
. . . in activist literature, genetics becomes a coherent and consistent plot to eliminate disabled people.
(Shakespeare, 2006: 856)
In their direct form, eugenics arguments say that using PGD (for example) to select out
disability and/or disease is an instance of eugenics and so since (it is supposed) eugenics is
wrong then so is this use of PGD. There are also several indirect eugenics arguments. One of
these is the claim that, although using PGD may not itself be eugenic, it is likely to lead
down the proverbial slippery slope to other practices that are. For reasons of space and
focus, I shall concentrate for the present on a direct version of the eugenics argument, one
which takes the following form.
1. Using PGD (and/or other selection techniques) to select out disability and/or disease
is a case of eugenics.
2. Eugenics is morally wrong.
3. Therefore: using PGD (and similar techniques) to select out disability and/or disease
is morally wrong.
This argument is logically valid: that is, (1) and (2) do jointly entail (3). However, questions
can be raised about both of the premises: specifically, is PGD an instance of eugenics and is
eugenics, in any case, wrong? Not surprisingly, both of these questions depend on what we
mean by the word eugenics and this is a contested issue with different sides in the debate
using the term rather differently. I have dealt with some of the definitional complexities
relating to eugenics in more detail elsewhere (Wilkinson, 2007; Wilkinson, 2010). So for
the present I shall offer a working definition before proceeding to tackle the substantive
questions of whether premises (1) and (2) above are plausible.
The term eugenics dates back to 1883, when it was coined by Francis Galton (Coutts
and McCarrick, 1995: 163; Oxford English Dictionary). Galton defines eugenics as the study
of the conditions under which men of a high type are produced and as the science which
deals with all influences that improve the inborn qualities of a race (Oxford English
Dictionary; Galton, 1909: 35). Other definitions include those found in the Oxford English
Dictionary, which defines eugenic as pertaining or adapted to the production of fine
offspring, esp. in the human race, and in the Routledge Encyclopaedia of Philosophy, which
defines eugenics as the attempt to improve the human gene pool (Chadwick, 1998). This
last definition is particularly relevant, since our primary concern is genetics and reproductive technologies, and it will therefore be adopted as my provisional definition. However, as
I have stated, this is only a working definition and much more can be said about what
eugenics is and about the different possible kinds of eugenics.
Eugenics is then an attempt to improve the human gene pool and thus has intention
built into its definition (although we may also allow that systems are eugenic insofar as there
are systemic aims that cannot readily be attached to individuals). However, as with public
health programmes, the aims of the whole system (or the intentions of those running the
system) may be quite different from those of individual actors. Thus, I suspect that very few
individual parents (or prospective parents) could fairly be labelled eugenicist even when
they opt to terminate or select out a (possible future) child with a genetic disorder. This is
because such parents will usually have little interest in improving the gene pool as a whole.
Rather their concern is with making sure that their (future) child has less pain and suffering
in its life, or perhaps with their own inability to cope with looking after a severely disabled
59
child. These prospective parents are not eugenicists; at most, they are colluding with or
tolerating eugenics. Eugenics may nonetheless be a property of the complete system (of
prenatal testing and termination) provided that improving the gene pool is one of the
systems aims, or is intended by whoever designs or runs the system. Hence, it certainly
makes sense to suggest, as some commentators have done, that eugenics may be an
emergent property of contemporary prenatal screening and testing practices, even if individual doctors and patients at the coal face do not have a eugenic intent. (Shakespeare,
1998: 667; Hampton, 2005: 553). (I do not claim that such commentators are correct
merely that their positions are coherent. Whether they are correct about any given system
would be mainly an empirical matter.)
Authoritarian eugenics
It is also worth making a distinction, within eugenics, between what have been termed
authoritarian and laissez-faire eugenics. This is best seen as a continuum, with some
eugenic policies and practices being more or less authoritarian than others. At the authoritarian end of the range sit Nazi eugenics, compulsory sterilization programmes and the like.
These are to be contrasted with what Nozick terms the genetic supermarket, in which
children are produced according to the individual specifications (within certain moral
limits) of prospective parents: a social setup which, for Nozick (1974: 315), has the great
virtue that it involves no centralized decision fixing the future of human type(s).
Laissez-faire eugenics occurs when private individuals practice eugenics with no, or
minimal, state involvement. Authoritarian eugenics is a little harder to characterize. Its
main defining feature is that prospective parents are compelled to behave eugenically and
are forced to reproduce, or to refrain from reproducing, or to reproduce in a particular way:
normally, though not necessarily, by the state. A practice is authoritarian eugenics then only
if either (1) prospective reproducers are physically forced to reproduce, or prevented from
reproducing, in certain ways (for example, a woman might be kidnapped and forcibly
inseminated or sterilized, or a pregnant woman might be subjected to prenatal testing and
selective termination without her consent) or (2) prospective reproducers are allowed to
decide for themselves whether or not to reproduce in particular ways, but their decisions are
somehow involuntary (resulting, for example, from coercion or manipulation).
Whether a particular practice or policy falls into category (2) will often be a controversial and difficult question. One (putative) example of this relates to the levels of service
offered to people with disabilities and their carers and an argument can be reconstructed
along the following lines. Prospective parents (through prenatal testing and, to a lesser
extent, PGD) are offered the choice of whether to have a disabled child or not. This decision
could in principle be voluntary. However, prospective parents are on the receiving end of a
systemic form of coercion. It works like this. The state (or society generally) ought to
provide a certain level of support to people with disabilities and their carers. However, it in
fact provides much less support than this and this position is unlikely to change in the
foreseeable future. The state is then in effect coercing people into refraining from having
children with disabilities by threatening them with poverty and social disadvantage
poverty and disadvantage that it has a moral responsibility to ameliorate, and for which it
is therefore at least partly responsible. I have a great deal of sympathy with this form of
argument, although whether it in fact applies to our present society is an open question that
depends on two questions that I am unable to answer here. First, there is the thorny political
60
Section 1: Concepts
question of what precisely the state owes its disabled citizens and their carers. Second, there
is the equally tricky empirical question of whether this posited standard of care (whatever
that is) is in fact met.
Another possible case of subtly authoritarian eugenics concerns the way in which
prenatal tests are offered and presented to pregnant women. As Modra notes:
Numerous commentators point to the conflict between the . . . systems commitment to promoting
autonomous PNGD [prenatal genetic diagnosis] choices on the one hand, and its commitment to
the prevention of genetic disability on the other. They suggest that this leads to a covert yet decisive
pressure on women to terminate affected pregnancies.
(2006: 259)
Thus if pregnant women offered PNGD were routinely pressured (perhaps subtly and/or
unintentionally) into accepting the test and into going for a termination (if faced with an
adverse result) then the prenatal testing system, considered as a whole, may be a sort of
authoritarian eugenics (provided that the prevention of genetic disabilities and disorders is
one of its aims). I should however reiterate that whether a practice is authoritarian is a
question of degree, and also note that this would be a relatively mild form of authoritarianism compared to (for example) the compulsory sterilization of adults with disabilities.
Nonetheless, other things being equal, it would be better if the fully voluntary (and
sufficiently informed) consent of pregnant women could be obtained for prenatal testing
and termination, and if similar outcomes could be achieved without anyone being
pressured.
The significance of the preceding discussion for the eugenics argument is that many
concerns about eugenics are really about authoritarian eugenics and so it is important to
highlight the fact that eugenics is not necessarily authoritarian or coercive. As Caplan et al.
tell us, while it is admittedly:
. . . morally objectionable for governments or institutions or any third party to compel or coerce
anyones reproductive behavior . . . the goals of obtaining perfection, avoiding disease, or pursuing
health with respect to individuals need not involve coercion or force.
(1999: 1284, my emphasis)
Thus, in this respect at least, eugenics may be morally unproblematic. So when it comes to
an assessment of eugenics moral status we must be careful not to assume that it will be
authoritarian. For, as Caplan et al. note, it is perfectly possible to support broadly eugenic
aims while eschewing coercion and authoritarianism just as one might support reducing
the prevalence of obesity without thinking (for example) that the state has a right to micromanage peoples eating or exercise behaviours.
The moral standing of eugenics
Ought we to define eugenics such that being eugenic is always a wrong-making, or morally
bad, property of actions or policies? Is eugenics what Bernard Williams (1985: 129) terms a
thick moral concept one such as treachery and promise and brutality and courage, which
seem to express a union of fact and value.
The main reason for thinking of eugenics as a negative moral term is that most
people who use it in contemporary debates do so to express condemnation, and it is rare
for people who support embryo selection to describe it as eugenic. While, on the other
side, those with pro-choice or pro-biotechnology views generally avoid the word. So
61
nearly everyone agrees that eugenics is hugely emotive and negative. As Raanon Gillon
(1998: 219) puts it, Eugenics is widely regarded as a dirty word. But while, for some, this
is a reason to avoid it; for others, this makes the term eugenics a good way of getting
their message across. A useful comparator here is the expression unborn child. Many
people with pro-life views use this to draw attention to (alleged) similarities between
foetuses and children (or indeed to express the view that foetuses are children); while
those with pro-choice views generally avoid this expression since they have the opposite
aims and views. The same goes for eugenics. Those hostile to embryo selection often use
eugenics to draw attention to (alleged) similarities between practices such as PGD and
historical atrocities associated with eugenics movements of the past; while those who
support PGD generally avoid the word eugenics in order to avoid drawing attention to
these (alleged) similarities.
One implication of all this is that defining eugenics is itself complex and controversial.
In particular, the working definition suggested above was largely descriptive, the attempt to
improve the human gene pool, so how can this be squared with the view that eugenics is a
negative moral term?
One option is to treat eugenics as a moral term by defining it as wrongfully attempting
to improve the human gene pool; on this view, permissible attempts to improve the gene
pool do not count as eugenic. Alternatively, we could insist that eugenics be descriptively
defined and therefore that there is at least a theoretical distinction, within eugenics, between
permissible and wrongful eugenics. In this latter view, we could still take account of the
point made earlier about linguistic politics, about eugenics being a dirty word, and there
still being good reasons not to use it except in cases of wrongful eugenics, but at least in
principle permissible eugenics would be a possibility.
I have argued elsewhere that this definitional issue cannot be resolved prior to an
assessment of the substantive moral arguments against eugenics (descriptively defined)
(Wilkinson, 2008). And this is my next task: to look at one particular attempt to show that
improving the human gene pool is morally problematic. If this is successful then we will
have some reason to view eugenics with distrust and there may also be reason to build
wrongness into the meaning of the word eugenics.
Is the very idea of genetic improvement flawed?
As has been noted, some programmes that aim to improve the gene pool are morally
wrong because they use unacceptably authoritarian means. These however will be disregarded from henceforth and I shall focus now on whether there is something wrong with
the end of genetic improvement.
Obviously, some things that are touted as genetic improvements are not really improvements. If someone suggested using selective reproduction to increase the incidence of tall
blonds, we should be sceptical about whether this would be an improvement. Similarly, the
history of eugenics is full of cases in which (for example) behavioural, cultural and racial
differences are wrongly seen as defects (Kevles, 1999; Baker, 2002; Barnett, 2004). So we
should approach improvement claims with caution. However, it does not follow from the
fact that many people have misused the idea of genetic improvement that it is necessarily
flawed. We must though be clear in each case what kind of improvement we are talking
about and demand reasons why the thing in question is not merely a change but an
improvement, a change for the better.
62
Section 1: Concepts
The most obvious type of genetic improvement that we can make sense of, and one that
is especially relevant to our present concern with public health, is health improvement:
using selective reproduction to reduce the prevalence of disease and impairment in future
populations. On the face of it, this looks like a kind of genetic improvement since most of us
would agree that, other things being equal, a less diseased and disabled future population is
preferable to a more diseased and disabled one because of the positive effects that good
health generally has on welfare.
So is reducing the prevalence of genetic disease and impairment through selective
reproduction a genuine instance of improving the gene pool? (Gillott, 2001; Human
Genetics Alert Newsletter, 2001; Savalescu, 2001: 413; Holland, 2003: 402). It seems that
it is, on account of the strong connection that exists between disease and disability and
reduced welfare. However, this view is challenged by what I have elsewhere termed the
Equal Value Principle (Wilkinson, 2006: 2651).
Shortly after its creation in 2000, the Disability Rights Commission (DRC, n.d.) was
asked by LIFE for its view on section 1(1)(d) of the Abortion Act 1967 (as amended by the
Human Fertilisation and Embryology Act 1990) which, as we saw earlier, permits (without
any time limit) termination on the ground that there is a substantial risk that if the child
were born it would suffer from such physical or mental abnormalities as to be seriously
handicapped. In response, the DRC said that while Section 1(1)(d) is not inconsistent with
the Disability Discrimination Act it is nonetheless:
. . . offensive to many people; it reinforces negative stereotypes of disability; and there is substantial
support for the view that to permit terminations at any point during a pregnancy on the ground
of risk of disability, while time limits apply to other grounds set out in the Abortion Act, is
incompatible with valuing disability and non-disability equally.
(DRC, n.d., my emphasis)3
In the same statement, the DRC informs us that:

Throughout its programme of work on ethical issues, the DRC will be guided by two principles:
valuing disability and non-disability equally, and the right of individuals to make informed,
autonomous choices.
(DRC, n.d., my emphasis)
It is the first of these principles, the Equal Value Principle, that I shall focus on here. For it
seems as if this principle could underpin an ethical argument against any reproductive
practice that selects against disability. After all, if disability and non-disability were valued
equally then why would anyone have reason to select the latter and deselect the former?
Furthermore, if disability and non-disability were valued equally then we could not regard
reducing the prevalence of genetically-based disability and impairment as an improvement,
since improvement means moving from something less valuable to something more
valuable. It may be argued then that using PGD (and similar techniques) to select out
disability is wrong because it is not consistent with the Equal Value Principle.
Two questions are raised by the Equal Value Principle. First, does screening out
disability necessarily involve failing to value disability and non-disability equally? And
second, how plausible is the Equal Value Principle itself?
For a more detailed discussion of this issue see Sheldon and Wilkinson (2001).
63
Taking the first question first, one view is that clinicians who offer embryo testing and
selective implantation to prospective parents are not themselves making value judgements
about disability. Rather, they are offering choice to the parents. So, as far as the ethics of
clinical practice is concerned (as opposed to the personal morality of the prospective
parents) the doctors have not acted badly because they are merely facilitating choice.
This could be true in some possible cases. However, there are reasons to doubt whether
this degree of neutrality exists in many actual cases, and whether the general policy context
within which clinical practice takes place can support such a neutral stance. One such
reason is that clinicians offering specific preimplantation genetic tests usually know what
the prospective parents preferences and values are and, more often than not, are only
carrying out the tests because they know that the parents wish to select against a particular
genetic disorder. This does not of course entail that the clinicians involved do not value
disability and non-disability equally. Nonetheless, it does at least raise the question of
whether a clinician who subscribes to the Equal Value Principle should really support
and collaborate with parents attempts to avoid disability or disease.
Another more policy-oriented reason is that, as we have seen, the HFEA only licenses
PGD for certain purposes. In nearly all cases, the aim of these licensed procedures is to
prevent the birth of a child with a disease or disability and, as we have already seen, the
HFEAs Code of Practice tells us that it is expected that PGD will be available only where
there is a significant risk of a serious genetic condition being present in the embryo
(HFEA, 2003: 24. s. 14.22). (One notable exception to this is PGD with tissue typing
where the aim, or one of the aims, is to save an existing childs life [HFEA, 2004].) It
cannot therefore be argued convincingly that the aim of the overall system of PGD is
merely choice, because the system is so clearly oriented towards disease-avoidance and
disability-avoidance (which presumably stem from valuing non-disability more highly
than disability, and health more highly than disease). So, rightly or wrongly, the general
policy context in which selective reproduction occurs is very much shaped by the
medical goals of disease-avoidance and disability-avoidance and it is hard to see why
someone would have such goals unless she valued non-disability and the absence of
disease more highly than disability and disease.
Turning now to the second question, how plausible is the Equal Value Principle? It
seems to me that the principle is not as compelling as the DRC has suggested, mainly
because its proponents neglect an important distinction between valuing disability and nondisability equally, and valuing disabled and non-disabled people equally. Clearly, we should
value disabled and non-disabled people equally and give them equal respect and rights as
human beings, but it does not follow from this that we must value disability and nondisability equally.
This is clearest when the disability in question involves a straightforward physical
injury. Take, for example, people with injured and permanently non-functional legs.
Obviously, we should value people with and people without functioning legs equally and
grant them equal respect; indeed, we should go further and grant people without functioning legs additional resources to compensate for their reduced mobility, and (where practicable) modify built environments to facilitate access. None of this however requires us to
value functioning and non-functioning legs equally. On the contrary, most of us have a
strong and rational preference for functioning legs. So it is tempting to say that, while
people with and people without disabilities should be treated as equals in terms of their
moral status, given a choice, non-disability is usually preferable to disability.
64
Section 1: Concepts
This point is bolstered by another argument against the Equal Value Principle, which
claims that the principle has absurd, or at least unpalatable, consequences. For commitment
to this principle entails that we should not seek to cure people when they acquire disabilities, and certainly should not use public resources to do so, because this would suggest that
disability is less valuable than non-disability. Indeed, if we do not value non-disability more
highly than disability then (with the exception of non-curative interventions such as pain
relief) it is hard to see what rationale there could be for having a medical profession and for
making people pay taxes to fund the National Health Service (Edwards, 2004). So the Equal
Value Principle appears to have unacceptable implications and should be rejected.
So is the very idea of improving the gene pool fundamentally flawed? It seems to me
that, with some qualifications, the answer is no. There have of course been many versions of
eugenics that have incorporated dangerously flawed ideological and pseudo-scientific
beliefs, such as Nazi racial science. However, there is no need to assume that all attempts
to improve the gene pool will be similarly flawed, and we can imagine versions of gene
pool improvement, such as the attempt to create healthier future populations, that would
and should attract very widespread support. So perhaps (for example) improving the gene
pool in ways that improve future public health would be morally acceptable (and even
desirable). Against this, some people (notably proponents of the Equal Value Principle)
have argued that even the attempt to improve population genetic health is unethical because
it is premised on valuing non-disability more highly than disability. However, it seems to
me that this argument (along with the Equal Value Principle itself) is unsound and should
be rejected. There is nothing wrong with assigning a negative value to the functional
impairment aspects of disability and disease and this negative valuation of impairment
does not entail and need not be accompanied by any negative valuation of the person with
the impairment.
Conclusion
Given recent developments in reproductive and genetic technologies, it is likely that
intervening before birth or before embryo-implantation to alter the composition of future
populations, will become an effective strategy for improving public health. The ethics of
selective reproduction is thus a part of public health ethics, insofar as selection is practised,
promoted or regulated with a view to achieving public health goals. In the first part of this
chapter, I suggested that selective reproduction is already being treated in this way, citing
abortion law and the regulation of preimplantation genetic diagnosis as leading examples.
One objection to using selective reproduction as a means of achieving better public
health is that doing so is eugenic. How should we respond to such claims? The view
that I have argued for here (and at greater length elsewhere) is that we should concede,
though with some important caveats, that public health-oriented selective reproduction may
be eugenic, although perhaps the term eugenics is best avoided (especially outside academic work) because of its extremely negative connotations and historical associations
(Wilkinson, 2008; 2010).
So what are these important caveats? First, it must be noted that the term eugenics is
being used here non-morally to mean improving the gene pool; and while clearly there is a
sense in which even this definition is evaluative (for we are talking about improving genetic
health, not merely changing it) the evaluation appealed to is not a moral one as such, but
one concerning either physiological functionality or welfare, or both. Second, it has to be
65
conceded that certain sorts of eugenics are generally wrong. In particular, authoritarian
eugenics is usually objectionable because of the harms and violation of autonomy that it
entails. Similarly, eugenics (pseudo-eugenics perhaps) that passes off mere differences
(such as real or perceived racial differences) as defects and then attempts to eliminate
them is morally objectionable. Having made these qualifications then, we can assert that
some forms of eugenics may be acceptable, particularly those that genuinely aim at public
health and which are robustly non-authoritarian, requiring the valid consent of any
individuals concerned (especially prospective parents). Hence, using selective reproduction
to achieve public health goals may (under certain circumstances) be permissible even if this
is eugenic.
Finally, though, I should add that there are of course many other arguments
against selective reproduction that I have not had chance to consider here. Hence, insofar
as this is not a comprehensive survey of all possible arguments against using selective
reproduction as a public health policy instrument, any conclusions must remain somewhat
provisional.4
References
Baker, B. (2002) The hunt for disability: the new
eugenics and the normalization of school
children. Teachers College Record, 104(4):
663703.
Barnett, R. (2004) Keywords in the history of
medicine: eugenics. The Lancet, 363: 1742.
Caplan, A., McGee, G. and Magnus, D. (1999)
What is immoral about eugenics? British
Medical Journal, 319: 1284.
Edwards, S. (2004) Disability, identity, and the

expressionist objection. Journal of Medical
Ethics, 30: 41830.
Galton, F. (1909) Essays in Eugenics. Honolulu,
HI: University Press of the Pacific.
Gillon, R. (1998) Eugenics, contraception,
abortion and ethics. Journal of Medical
Ethics, 24(4): 21920.
Gillott, J. (2001) Screening for disability:
a eugenic pursuit? Journal of Medical Ethics,
27(Suppl. II): 213.
Chadwick, R. (1998) Genetics and ethics. In

Routledge Encyclopaedia of Philosophy, ed.
E. Craig. London: Routledge.
Hampton, S. J. (2005) Family eugenics.

Disability and Society, 20: 55361.
Coutts, M. and McCarrick, P. (1995)

Eugenics. Kennedy Institute of Ethics Journal,
5(2): 16378.
Harris, J. (2000) Is there a coherent social

conception of disability? Journal of Medical
Ethics, 26: 95100.
Department of Health (2007a) Abortion

Statistics for England and Wales: 2006.
London: Department of Health. Available at:
http://www.dh.gov.uk/en/
Publicationsandstatistics/Publications/
PublicationsStatistics/DH_075697 (accessed:
19/7/08).
Holland, S. (2003) Selecting against difference:

assisted reproduction, disability, and
regulation. Florida State University Law
Review, 30(Winter): 40110.
Department of Health (2007b) Human Tissues

Embryos (Draft) Bill 2007. London:
Department of Health.
Disability Rights Commission (n.d.) DRC
Statement on Section 1(1)(d) of the Abortion
Act 1967. Manchester: DRC.
Human Fertilisation and Embryology Authority

(HFEA) (2003) Code of Practice, 6th edn.
Available at: http://www.hfea.gov.uk
(accessed: 19/7/08).
(HFEA) (2004) HFEA agrees to extend policy
on tissue typing. Press release, 21 July.
Available at: http://www.hfea.gov.uk
Many of these other arguments are reviewed in Wilkinson (2010).
66
Section 1: Concepts

(HFEA) (2005) Choices and Boundaries.
Available at: http://www.hfea.gov.uk/docs/
Choices_Boundaries.pdf (accessed: 19/7/08).
Human Genetics Alert (2001) Newsletter. Issue
1, December.
Jepson, J. (2003) Herod is not alone in his
fear of the helpless. Sunday Telegraph,
21 December.
Kevles, D. (1999) Eugenics and human rights.
British Medical Journal, 319: 4358.
MENSA (n.d.) Mensa information. Available at:
http://www.mensa.org/index0.php?page=10
Modra, L. (2006) Prenatal genetic testing kits
sold at your local pharmacy: promoting
autonomy or promoting confusion?
Bioethics, 20: 25465.
equality. Disability and Society,

13: 66581.
Shakespeare, T. (2006) Disability Rights and
Wrongs. London: Routledge.
Sheldon, S. and Wilkinson, S. (2001)
Termination of pregnancy for reason of
foetal disability: are there grounds for a
special exception in law? Medical Law
Review, 9: 85109.
Terzi, L. (2004) The social model of disability:
a philosophical critique. Journal of Applied
Philosophy, 21(2): 14157.
Verweij, M. and Dawson, A. (2007)
The meaning of public in public health.
In Ethics, Prevention, and Public Health,
ed. A. Dawson and M. Verweij.
Oxford: Oxford University Press,
pp. 1329.
Oliver, M. (1990) The Politics of Disablement.

Basingstoke: Palgrave Macmillan.
Wikler, D. and Brock, D. W. (2007) Populationlevel bioethics: mapping a new agenda. In

Oliver, M. (1995) Understanding Disability:

From Theory to Practice. Basingstoke:
Palgrave Macmillan.
Wilkinson, S. (2006) Eugenics, embryo

selection, and the equal value principle.
Clinical Ethics, 1: 2651.
Oxford English Dictionary. Oxford: Oxford

University Press.
Wilkinson, S. (2007) On the distinction between

positive and negative eugenics. In Arguments
and Analysis in Bioethics, ed. M. Hayry, T.
Takala and P. Herissone-Kelly. Amsterdam:
Rodopi.
Nozick, R. (1974) Anarchy, State and Utopia.

Oxford: Blackwell.
Parfit, D. (1984) Reasons and Persons. Oxford:

Rogers, L. (1996) Mensa sperm bank set up to
create super humans. The Times (London),
11 February: 1.
Rogers, L. (2002) Police to probe late abortion of
harelip baby. Sunday Times (London),
27 October: 13.
Wilkinson, S. (2008) Eugenics talk and the

language of bioethics. Journal of Medical
Ethics, 34: 46771.
Wilkinson, S. (2010) Choosing Tomorrows
Children: The Ethics of Selective Reproduction.
Savalescu, J. (2001) Procreative beneficence: why

we should select the best children. Bioethics,
15(56): 41326.
Williams, B. (1985) Ethics and the Limits of

Philosophy. London: Fontana.
Shakespeare, T. (1998) Choices and rights:

eugenics, genetics and disability
Wrigley, A. (2006) Genetic selection and modal

harms. The Monist, 89(4): 50525.
Section 1
Chapter
Concepts
Risk and precaution

Stephen John
Introduction
The concepts of risk and precaution are central to public health policy, and therefore to
philosophical and ethical reflection on such policy. A wide range of public health activities
such as health and safety legislation, food standards monitoring and the emerging field of
public health genomics are explicitly framed in terms of risk-reduction and risk-management. Furthermore, we can use the concept of risk to understand public health policies
that are not normally framed in these terms. For example, draining a malarial swamp can be
understood as eliminating health-risks, as can a policy of compulsory vaccination. The
control, minimization or elimination of health-risks can, then, be seen as the shared
concern of the heterogeneous activities that comprise public health policy.
Normally, we understand the reduction of risk as instrumentally valuable. That is to say, we
view it as a tool for ensuring a particular distribution of health-outcomes (or, perhaps, a
distribution of opportunities for health). Arguably, we might also view the reduction of risk as
directly valuable, because living in an environment where risks of serious physical harm or
suffering have been reduced is a constituent of what Amartya Sen (1992) calls our capabilities,
our real freedoms to achieve valuable functionings. Someone who lives in an environment in
which malaria has been eradicated, for example, is not only less likely to suffer ill-health than
someone who lives in a malarial environment, but may also be said to enjoy greater positive
freedom; the range of activities she may reasonably choose to pursue is greater than the range
of activities her counterpart can reasonably choose to pursue.1 A narrow focus on the healthoutcomes that follow risk-reduction can, perhaps, hide from us the more general benefits that
follow from creating risk-free environments, pointing towards an important, if little-explored,
distinction between the value of prevention and of cure.
Just as we can frame an account of the means and the ends of public health policy in
terms of risk, we can also pose the challenges of public health ethics in these terms.
Examples of such problems include the following. How ought we to choose between public
health policies that will have different effects on the overall distribution of health-risks in
the population? For example, should we focus on alleviating the plight of the most
vulnerable or on minimizing expected overall mortality and morbidity? How ought we to
decide when the risks associated with some activity are tolerable? Is there an important
See Olsaretti (2005) for a complex argument for these kinds of claims.
67
68
Section 1: Concepts
difference between policies that impose risks on individuals and policies which merely fail to
reduce the risks individuals suffer? For example, if we must decide whether or not to allow a
drug, with known benefits for many but potential side effects for ex-ante unidentifiable
others, onto the market, is there a morally significant distinction between not allowing the
drug onto the market, leaving some at risk, and allowing the drug onto the market, placing
some at risk? Is there a difference between voluntary and involuntary risk? If so, is
regulating risky activities performed by consenting adults unacceptably paternalistic?
Unfortunately, developing an ethics of risk is no easy task. In the literature, we find
battle-lines drawn between a scientific or statistical account of risk associated with the
policy tool of risk cost-benefit analysis (RCBA) described by Adams (1995) as the
Royal Society view of risk and a constructivist or sociological account of risk, often
associated with the precautionary principle.2 Proponents of the first approach accuse
proponents of the second of being un- or anti-scientific; proponents of the second
approach accuse proponents of the first of technocracy.3 Those who combine aspects
of both approaches are attacked from all sides. These debates are both heated and
theoretically complicated because they involve four issues that cross the empirical/normative boundary: how we ought to define risk; how members of the public actually think
about risk; how people ought to think about risk; and how we ought to handle risks in a
democratic society.
On the one hand, it seems clear that an ethics of risk is necessary for ethical debate of
much public health policy, and may have wider ramifications for public health ethics as a
whole. On the other hand, the ferocity and complexity of the debates mentioned above
make the task of constructing an ethics of risk indeed, even understanding what risk is
extremely difficult. In this chapter, I will set out one way in which we might understand
these problems, by looking at the arguments of those who promote the precautionary
principle as an approach to both risk-assessment and to risk-management. In the first
section, I set out the standard view of risk, which the precautionary principle is often
claimed to challenge. In the second and third sections, I set out ways in which we might
understand a precautionary approach to public health policy, and in the fourth section,
I show how these claims relate to the debate between technocratic and constructivist
accounts of risk. In the fifth section, I suggest that perhaps the most interesting feature of
the precautionary principle is that it leads us to rethink the relationship between ethical
norms and scientific testing. It should be stressed that this paper is not intended to provide
a substantive account of the ethics of risk for purposes of public health policy. Rather, the
aim is to provide the reader with a guide through a complex debate, and, in so doing, to
outline some of the key issues for and constraints on any substantive theory of the ethics of
risk. In the sixth section, I do, however, show how the arguments sketched in this chapter
might help us to construct a new ethics of risk.
See, also, Ling and Raven (2006) for a standard account of these disputes. The standard text for a
social-constructivist account of risks is Beck (1992). See Douglas and Wildavsky (1980) for an
altogether more interesting anthropological approach to these debates, which might seem to suggest
that both sides of the dispute can be explained from an anthropological perspective.
See Sunstein (2002) for an example of the first sort of charge, and Jasanoff (1999) for an example of
the second. These debates are well summarized in Lewens (2007).
Chapter 4: Risk and precaution
69
The precautionary principle and RCBA

A precautionary approach to policy is often understood to be opposed to the standard
model of risk identification and risk-management, identified with the tools of RCBA.4 In
this section, I shall outline this approach. To understand the concept of risk and the
methodology of RCBA, it is useful to consider the model of rational choice theory from
which RCBA derives. Imagine that I want to buy a sandwich. I can either go to Shop A next
door, which makes awful sandwiches, or to Shop B on the other side of town, which makes
delicious sandwiches. These two courses-of-action have different benefits associated with
them an awful sandwich or a delicious sandwich and they have different costs associated
with them, a short walk or a long walk. The alternative courses-of-action might also be
associated with different probabilities of suffering harm: going to Shop B involves crossing
a busy road, thereby running a small probability of serious injury, whereas popping next
door does not. The term risk, as used in its Royal Society sense, is typically used to denote
such possibilities of suffering harm: we say that there is a risk associated with going to the
far-off shop, and no risk (or comparatively low) risk associated with popping next door.5
Sometimes we say that we would not be willing to run a risk of serious harm for any
potential benefit. However, our behaviour often runs contrary to this platitude. I, for one,
regularly run all sorts of risks of serious harm for the sake of minor benefits, although
I would not run much higher risks of harm for the same benefits. I regularly run a small risk
of being run over in return for a tasty sandwich, but would not run a much higher risk of
harm for the same sandwich. So doing does not seem irrational. Indeed, not to do so would
be to lead to a dreadfully impoverished life. One way in which to model these kinds of
everyday decisions is to say that I act (or ought to act) in accordance with the goal of
maximizing my overall expected utility. For example, we can model my decision thus:
Expected utility of going to Shop B (value of tasty sandwich) plus (negative value of walking so far)
plus (probability of being hit by a car multiplied by negative value of being hit by a car).
The greater the probability of serious injury attached to my walk to the shop, the lower my
expected utility, and the more attractive the option of popping next door instead (where the
expected utility of doing so is simply value of awful sandwich minus cost of walking a short
distance).6
RCBA applies a version of this model at the social level as a guide to policy. We choose
between different policies by asking what the certain costs and benefits and risks associated
with each of those policies are. Our choice, then, is guided by asking which action has the
best expected outcomes. Of course, the use of RCBA in policy-making is not straightforward. However, very roughly, the idea is simple: first, we construct a single scale along
4
5
6
For useful introductions to RCBA see Shrader-Frechette (1980) and Sunstein (2002). Note that,
strictly, there are differences between the claims made by proponents of RCBA, by proponents
of cost benefit analysis and by proponents of risk-trade-off-analysis. For our purposes, however,
these differences are relatively unimportant.
Although we should note that the term risk can be used in other ways, leading to some confusion.
For a useful overview, see Hansson (2007).
This is a very simplified model of the kind of theory proposed by such writers as Luce and Raiffa
(1958). For a useful history of the relevant arguments (which makes clear the link between
developments in the theory of risk and the collection of epidemiological statistics) see Hacking
(1975) and Hacking (1990).
70
Section 1: Concepts
which we can compare all possible benefits and costs of potential outcomes of action
(normally, a monetary scale is used for these purposes). We can then compare different
policies by comparing the expected costs and benefits of those actions (by multiplying the
probability of those outcomes by their magnitude). It is important to recognize here that
many people find the idea of constructing a single scale along which we might compare all
outcomes of action to be morally suspect. I will return to these issues below. However,
before going further, it is important to note that treating disparate outcomes as commensurable seems to be consistent with everyday life: someone who regularly refused to tradeoff risks of death against even such minor pleasures as a tasty sandwich would seem very
strange.
I have identified two substantive aspects of RCBA. First, there is a claim about how we
ought to reason: we ought to aim to maximize our expected utility. Second, a claim that we
can place all costs, benefits and risks on a unitary scale, such that we can rank risks from
most to least serious by multiplying the extent of harm associated with some event by the
probability of that event. This way of thinking about policy-making is extremely powerful
as a tool for thinking about public health policy, most notably health-and-safety policy.
After all, we do think that there are limits on which public health policies we ought to
pursue; even if obtainable, safety at any cost would not necessarily be reasonable: RCBA
seems to provide us with a way in which to quantify such judgments.7
There are two necessary conditions for any philosophical defence of some government
policy tool: a condition of practicability and a condition of justifiability. By practicability,
I mean that it should be clear how we ought to apply any proposed tool for policy-making.
Although sometimes fraught in practice, as noted above, it is clear how RCBA should be
applied in practical contexts. However, a defence of any policy tool should also meet a
justifiability constraint. This task is particularly difficult in societies such as ours where
there is widespread ethical disagreement. Therefore, I suggest that a defence of a policy tool
should show how that tool relates to the political principles which constitute what Rawls
calls a free-standing module, on which there can be an overlapping consensus. That is, we
should be able to defend use of a policy tool in such a way that different members of our
community can agree to use of this tool for policy-making, regardless of their own views of
the good life (Rawls, 1993). One key advantage of standard models of RCBA is that they are
supposed to be ethically neutral, and, thus, to meet this publicity condition.
This may appear puzzling. After all, RCBA obviously involves placing some kind of
valuation on different outcomes, and the justification of such valuations say, treating a
human life as costing x million may seem extremely difficult. However, typically, the
value-inputs into RCBA are supposed to be generated by adopting citizens own valuations
of different outcomes. The value claims necessary for comparing risks are generated either
by looking at individuals behaviour or by a process whereby their preferences between
different outcomes are elicited. In effect, then, by assuming that everyone always reasons so
as to maximize expected utility, in the way set out in the sandwich case, we generate an
account of what people do value, and base policy on such valuations. RCBA is supposed to
be ethically neutral in that (at least in its pure form) it does not rest on an objectivist
account of good and bad, but generates claims about value on the basis of what people
For a useful analysis of these tools as they apply in the context of railway safety, see Wolff (2002;
2006).
71
actually do value. This process leads Cass Sunstein (2005a) to suggest that such procedures
are justifiable because they promote citizens well-being use of such tools avoids forcing
individuals to pay more for some good such as risk-reduction than they would be willing
to pay and respect citizens autonomy by respecting their valuations of outcomes.
The precautionary principle and value

Having identified some of the key aspects of RCBA, I will now turn to the precautionary
principle. It is common to distinguish two forms of bioethics: environmental ethics and
medical ethics, and to note that these are very distinct areas of applied philosophy (ONeill,
2002). An interesting feature of public health ethics, however, is that, although it is clearly
related to medical ethics, insofar as both focus on normative problems related to individuals health, it also shares concerns with environmental ethics. The pursuit of public health
policy often resembles environmental policy both involve structuring features of a shared
environment with the end of preserving certain sorts of goods and similar normative
problems, most notably problems of free-riding, occur in both contexts. Therefore, it is no
surprise that the precautionary principle, much debated in recent environmental ethics,
has increasingly been appealed to in public health ethics (Martuzzi and Tickner, 2004).
The canonical version of the precautionary principle states that where there are threats
of serious or irreversible damage, lack of full scientific certainty shall not be used as a reason
for postponing cost-effective measures to prevent environmental degradation (United
Nations, 1992). It is clear that, in and of itself, this principle is not directly relevant to
public health policy (unless there is always some link between serious or irreversible
environmental degradation and public health). However, proponents of a precautionary
approach to public health claim that, by replacing serious environmental damage with
some set of outcomes such as serious negative public health problems, we can generate a
precautionary principle, which ought to guide public health policy.8
Precautionary thinking has had an impact on UK public health policy, most notably in
the 2000 Stewart Report which recommended on precautionary grounds that mobilephone masts should not be placed near schools (Burgess, 2004). Furthermore, the importance of adopting precautionary approaches to public health policy has been stressed by EU
directives (European Commission, 2000). There is much disagreement over what precisely
is involved in adopting a precautionary approach to public health policy. However, at least
in the environmental context, it is assumed that there is a tension between the use of RCBA
as a guide to policy and a more precautionary approach. Therefore, we can expect that an
interesting public health version of the precautionary principle would also be in tension
with the tools of RCBA.
What is implied by the claim that we ought to adopt an alternative precautionary
approach to policy? It might seem as if a precautionary approach to policy involves denial
of the first aspect of RCBA: the claim that we ought always to seek to maximize expected
outcomes. Within the literature on rational-choice-theory, some have argued that in
8
Indeed, an alternative influential version of the principle, the Wingspread formulation, explicitly
focuses on health issues: when an activity raises threats of harm to human health of the
environment, precautionary measures should be taken even if some cause and effect relationships are
not fully established scientifically (see http://www.sehn.org/wing.html). For purposes of
exposition, I shall, however, focus on the UN formulation.
72
Section 1: Concepts
circumstances where there is uncertainty over the likelihood of certain sorts of outcomes,
we ought to adopt a strategy of maxi-min: we should adopt the policy with the least bad
outcomes. In turn, some defenders of the precautionary principle have claimed that it can
best be understood as a formulation of a maxi-min principle: that is, where there is good
reason to believe that there is some risk of serious harm associated with some policy, then
we should not adopt that policy, even if, for some reason, we cannot establish the risk
precisely.9
However, I suggest that this way of reading the precautionary principle is orthogonal to
many of the concerns expressed by proponents of precautionary reasoning. Claims that we
ought to adopt maxi-min strategies normally assume a framework where all harms and
benefits are, ultimately, to be viewed as commensurable. Interesting versions of the precautionary principle, by contrast, seem to involve a more general claim: that the assumption of
commensurability implicit in RCBA is suspect.
In light of these comments, one interpretation of the precautionary principle is that we
(rightly) possess a tripartite division of the outcomes of action. Unlike normal bad
outcomes (ugly mobile-phone masts), which might be outweighed by good outcomes
(greater ease of communication), we ought to treat certain outcomes (the avoidable death of
innocent children) as special bad outcomes. In some sense, risks of such outcomes are not
subject to the standard tools of RCBA, and, for some related reason, claims about such risks
ought to be given credence even when they have not been established with scientific
certainty. Before deciding precisely how to interpret these claims that is, answering the
practicality worry above we need to answer a prior question, related to the publicity
condition: are there non-controversial reasons to think that certain sorts of bad outcomes
ought to be treated as special?
I shall distinguish two kinds of argument for such a tripartite distinction of
outcomes: standard-consequentialist and non-standard-consequentialist. I use standardconsequentialist here to mean ethical theories which assess actions in terms of the statesof-affairs produced by those actions, and which assume that there is commensurability
between different states-of-affairs. It looks as though it is precisely the assumption of
commensurability underlying RCBA to which proponents of the precautionary principle
object. However, it is possible to argue from a standard-consequentialist standpoint, which
views all outcomes as ultimately commensurable, to a view that certain sorts of outcomes
ought to be treated with particular care. To do so, we must show that a policy which does
not treat those outcomes as special is, over time, likely to lead to worse outcomes than is a
policy which treated that class of outcomes as special. We might, as it were, treat certain
outcomes as outside the purview of standard RCBA not because those outcomes are
really special, but because we have good long-term reason to treat those outcomes as if
they were special.
Such an argument might, in turn, be motivated by inductive evidence. Some of the
worst public health crises of the twentieth century would have been avoided had we
treated the threat of certain sorts of serious or irreversible damages as special, and,
therefore, we ought now to treat such threats as special. Perhaps the most famous
examples of such catastrophes born of misplaced certainty are DDT in the USA and
9
For a version of this kind of account of the precautionary principle in the environmental context see
Gardiner (2006). Arguments for maxi-min strategies are, of course, famous in political philosophy
because of their use in Rawls (1999).
73
thalidomide in the UK. This form of argument seems to be in-line with repeated
arguments for adoption of the precautionary principle in an environmental context, where
inductive evidence of past failures of science policy is used to cast doubt on contemporary
risk-analysis (Harremoes et al., 2002).
Does a standard-consequentialist argument meet the publicity condition? Defenders of
precautionary reasoning might respond to this question in a hypothetical mode: if you
think that risk-cost-benefit-analysis, the methodology to which precautionary reasoning is
normally opposed, meets the publicity condition, then precautionary reasoning also meets
the publicity constraint as it is grounded on RCBA. The two approaches stand or fall
together.
Even if we think that the standard consequentialist argument for treating certain
outcomes as special fails, we might simply adopt a non-standard-consequentialism which
denies the commensurability of outcomes. We might argue that certain sorts of outcomes
cannot be traded-off against other sorts of outcomes. The deaths of innocent children, for
example, might be thought simply to fall into a different category of outcome than the
annoyance felt by many people at being unable to get mobile-phone reception. In turn, we
might think that such incommensurability implies we ought to accord the avoidance of the
relevant special bad outcomes some kind of priority over all other consequences in our
moral reasoning. Whenever there is a possibility of special bad outcomes, we ought to seek
to avoid those outcomes irrespective of normal costs and benefits. (A similar sort of
argument is sometimes used in the environmental context, where it is claimed that the
intrinsic value of nature is such that it is a mistake to think that we can ever trade-off
environmental degradation against other kinds of harms and benefits [see, for example,
Elliot, 1997].)
Deontological defences of the precautionary principle

and the argument over lay-rationality
I shall discuss one way in which we might defend non-standard consequentialism in the
next section. First, however, I shall outline a more interesting route for defending the
precautionary principle, by appeal to deontological considerations. What is a deontological
approach to the precautionary principle? One common complaint against the precautionary principle is that it seems to tell us to regulate against new technologies or innovations,
when those technologies or innovations pose some threat of serious or irreversible damage,
even when we have good evidence that those technologies will probably help many.
Defenders of the precautionary principle are accused of a peculiar form of myopia
(Sunstein, 2002, 2003, 2005b).
However, we might claim that the precautionary principles focus on avoiding harm,
rather than alleviating suffering, can be justified by appeal to the distinction between
doings and allowings. For example, in the environmental context, Marion Hourdequin
(2007) has argued that the precautionary principle represents or incorporates a distinction
between doing environmental damage and merely allowing environmental damage to
occur (Hughes, 2006). Of course, this way of reading the principle poses a new challenge, as
some past harms may have been the result of human actions: taking seriously a doing/
allowing distinction is not the same as the conservative principle that action always requires
more justification than inaction. However, Hourdequins key point is simple and compelling: in everyday thought and political discussion, we distinguish between human-caused
74
Section 1: Concepts
environmental damage and environmental damage which occurs naturally. This distinction, which we overlook if we think of policy solely in terms of outcomes, is one which the
precautionary principle aims to respect. If we take seriously some kind of deontological
intuition, then, we are likely to think that RCBA is a deeply flawed methodology, because it
treats the risks which would, in some sense, be imposed by action as equivalent to the risks
which action might aim at relieving. Of course, we need not deny that there is some
perspective from which we might say that the risks associated with action and inaction
are equal. However, we might deny that all we do, or ought to, care about is this kind of
information. Rather, when considering action, we need to ask whether any resultant risks of
harm would arise as a result of our actions or as a result of our inactions.
The deontological sense of a distinction between doings and allowings is, arguably, one
which is reflected not only in environmental contexts, but in contexts of health-and-safety
regulation. Consider, for example, whether we ought to allow a drug onto the market. We
know that the drug will save many lives, but we are worried that it might have special bad
effects for a small (but ex-ante unidentifiable) sector of the population. What does the
precautionary principle tell us to do in such cases? It seems to tells us that we ought not to
allow the drug onto the market because there is a small chance that so doing will lead
to serious or irreversible damage (as a result of the possible side-effects). However, it also
seems to tell us that we ought to allow the drug onto the market, as not to do so also
involves running a risk of serious or irreversible damage (that is, the deaths of many
people whose lives would have been saved had the drug been put on the market). Such
arguments against the precautionary principle seem strong. However, a deontological
account of the principle can show why they may be mistaken. The key difference between
allowing the drug onto the market and not allowing the drug onto the market is that, in the
first case but not in the second, the deaths of some people would be a result of our actions,
whereas, in the second case, the deaths would be something we allow, rather than do.
Conservatism might not be a form of indifference or myopia, but, rather, an expression of a
certain conception of the extent of (and limits to) our obligations.
I have suggested, then, that we might want to base a defence of the precautionary
principle on the deontological distinction between doings and allowings. This claim may
seem vulnerable to a familiar worry already hinted at above that the doing/allowing
distinction is extremely difficult to defend; deciding precisely when health-and-safety
regulation would prevent risk-imposition (and, thus, is demanded to stop a social doing)
and when such regulation would merely reduce natural risks seems likely to be an
impossible task. It is worth noting here, then, that a strong doing/allowing distinction need
not be the only way in which we could incorporate deontological concerns into discussions
of the regulation of risk. In recent work, for example, Thomas Pogge (2004) has argued that
accounts of justice and health have systematically overlooked the important differences
between different ways in which the same health problem might have been generated (for
example, through deliberate discrimination, through lax implementation of laws, through
the results of natural processes, through individuals behaviour, and so on). In turn, he has
suggested that the strength of the demand that we prevent or cure illness turns, in part, on
these kinds of considerations. This is not the place to develop or to defend Pogges claims.
However, there is some plausibility in the claim that the strength of the demand that we do
something to prevent individuals falling ill (that is, the strength of the demand that we
reduce the risks they face) turns, in part, on the source of the risks they face. To the extent
that a version of this claim can be defended, then it might serve to explain not only why the
75
precautionary principle seems to focus on a particular class of outcomes, but also why it
seems to pay particular attention to ensuring that human agencies are not (causally)
responsible for those outcomes, even if avoiding such (causal) responsibility is, in other
ways, costly.
Scientific and constructivist views of risk

It is one thing to say that we can distinguish between different kinds of outcomes or
different kinds of obligations, either of which views might underlie precautionary
approaches to policy, another to say that either of these distinctions is conceptually sound,
and yet another to say that such a distinction can be part of a free-standing module on
which there might be an overlapping consensus. How might the very general claims about
how we might understand the philosophical urges behind some formulations of the
precautionary principle be defended in a pluralist society?
To understand our options here, it is useful to turn to some of the issues that arise in the
heated controversy over risk. There often seem to be differences between how experts and
members of the public talk and reason about risk: risks which the experts claim are
intolerable are tolerated by the public, whereas risks which the experts think the public
should, if they are to be consistent, tolerate are not tolerated (Slovic, 2000; House of
Commons Science and Technology Committee, 2006). I shall focus on a distinction
between two ways in which we might make sense of the apparent gap between expert
and public views of risk (these two approaches do not jointly exhaust the sociological
approaches to risk, but are the most interesting accounts for our purposes).
One approach is to suggest that public views of risk are, quite simply, misguided. For
example, it is clear that some public views about risk say, the risks associated with genetic
technologies are based on scientific misunderstanding. Furthermore, there are experiments
which seem to show that people often have difficulty with understanding probability, and that
their reasoning is influenced by misfiring cognitive heuristics. These factors can, then, be
appealed to in an attempt to explain how the public, who are assumed to think about risks and
how to deal with risks in terms of maximizing their expected utility, are often led astray
(Sunstein, 2002). This approach to thinking about lay views of risk is, of course, consistent
with the project of RCBA: RCBA provides a way in which to overcome the cognitive failings
inherent in public discourse about and understanding of risk-based policy. (Although it
should also be noted that such results also warn us that if we want to gauge what peoples
preferences are as the basis for RCBA, we need to be extremely careful how we do so, as both
what people do and what they say may not be a good guide to what they really value.)
However, there is a second way of understanding public reasoning about risk. Rather
than say that the public are simply confused about risks, some theorists have argued that
lay reasoning about risk displays an alternative, lay or thick rationality, which differs
fundamentally from the kind of consequentialism implicit in much risk-assessment.10
According to these views, when members of the public say that some Risk A is far worse
than a second Risk B (although both represent the same probability of similar harm), then,
their fears may simply reflect a cognitive failure, but it may also be the case that they are
expressing a view about the moral status of these different risks. Arguments to the effect
10
Versions of the view suggested here maybe found in Slovic (2000); Wolff (2006) and, in a very
different form, Irwin (1995).
76
Section 1: Concepts
that lay reasoning about risk is not irrational, but in fact grounded on complex (if perhaps
inchoate) ethical theories, have been mounted in a range of different public health policy
debates, most notably in debates over the BSE crisis, and in debates over health-and-safety,
such as railway safety (Irwin, 1995; Wolff, 2006). Sometimes such accounts of lay
reasoning about risk are described as social constructivist. However, this is slightly
misleading insofar as to say that some concept is socially constructed is normally understood to debunk that concept (Hacking, 2000). Furthermore, this appellation may be
confusing as there are some social constructivist writers who seem to suggest that there
is no distinction between the public perception of risk and what risks there are (Beck, 1992).
Arguments that lay risk-assessment is value-laden, by comparison, do not necessarily deny
any distinction between risks and risk-perception, but, rather, they suggest that what
might appear from the viewpoint of RCBA to be a cognitive confusion is often better
understood in terms of evaluative disagreement.
A better way in which to understand these approaches, then, might be to say that they
argue that identification of and reasoning about risks is highly ethically loaded. Whereas
the standard view of risk associated with RCBA assumes that we can rank all risks from
most to least severe by multiplying the probability of outcomes against the badness of those
outcomes, lay reasoning involves a far more complex process, where risks are ranked not
merely in terms of expected harm, but in terms of their wrongness, or in terms which deny
that all harms are commensurable.
In turn, we can use such claims to argue against the acceptability of standard RCBA in
two different ways. First, the fact that public attitudes towards risk seem to incorporate
such concerns can help to strengthen an argument that RCBA rests on a particular moral
theory, and a moral theory that requires further defence (Cranor, 2007). That is to say, we
can treat public views about risk as reminding us that what appears to be an ethically
neutral way in which to guide health-and-safety policy by assessing policies in terms of
expected outcomes actually rests on assumptions, such as that all outcomes are commensurable or that doings and allowings are morally symmetrical, which reasonable individuals
can (and, in fact, do) dispute.
Second, we might adopt a kind of democratic approach to argue against the use
of RCBA in policy contexts. Rather than defending the correctness of nonstandard-consequentialism or some deontological view of risk, which we claim to be
implicit in everyday thought, we might instead argue that members of the public hold such
views, the views are not clearly false, and, as such, they ought to be respected in the
decision-making process. Just as the proponent of RCBA claims simply to be reflecting
social values when she constructs an account of value, the proponent of some kind of
precautionary approaches to policy might claim simply to be reflecting social values when
she denies the legitimacy of RCBA.
Precaution in practice: questioning the ethical/epistemic divide

In the previous sections, I argued for two ways in which we might interpret the motivations
behind precautionary thought: either as reflecting the view that a certain class of outcomes
are (or should be viewed as) special, and thus outside the ambit of RCBA, or, more
interestingly, as reflecting a deontological distinction between the wrongness of different
sorts of risks. It should at this point be stressed that many of the writers I have cited
throughout my discussion of the precautionary principle would not normally be thought of
77
as defenders of the precautionary principle. My aim has not been so much to provide an
exposition of standard precautionary thought, but to provide an exposition of alternatives
to RCBA, paying particular attention to the question of justifiability.
I have suggested that a second constraint on any account of the precautionary principle
is that it ought to show us how and when we are to adopt the principle. This practicality
constraint is particularly important as it has seemed to many that the precautionary
principle is too vague to be of any real use in policy-making (Sandin et al., 2002). This
might explain why even those who deny RCBA do not necessarily commit themselves to the
precautionary principle.
The precautionary principle faces multiple problems as a guide to action. Showing that a
tripartite distinction of outcomes is defensible is not the same as specifying precisely how
the importance of certain outcomes ought to be treated as special. Similarly, claiming that
the avoidance of imposing certain sorts of risks ought to have priority over allowing similar
risks to continue naturally does not tell us how to choose between policies. However,
rather than focus on all of these questions here, for purposes of exposition, I will focus on a
rather puzzling feature of the standard formulation of the precautionary principle; the claim
that we ought to act to forestall the risk of certain kinds of harms even when we lack full
scientific certainty that there is a link between some policy and a risk of such harm. This
widens the informational bases on which our choices are to be made. However, so doing
seems to leave the principle open to charges of being, at best, irrational, obscurantist and
anti-scientific, and, at worst, of being impracticable.
To see why, let us return to the case of placing mobile-phone masts next to schools.
Imagine that all agree that childhood cancer is incommensurably worse than other
outcomes, such that if an official policy, say allowing masts to be placed anywhere,
poses a risk of such an outcome, then we have strong precautionary health-and-safety
reasons not to adopt that policy. Even if we can agree with the claim, we might be
reluctant to reject a laissez-faire policy regarding mobile-phone masts, thus potentially
foregoing many benefits, when there is no scientific certainty over whether there is such
a risk at all. The Stewart reports recommendations seem, then, to rest on two dubitable
premises: first, any risk of childhood cancer associated with locating mobile-phone
masts next to schools would make it illegitimate to locate masts next to schools; second,
we should act as if there is some risk of such harm even when there is no scientific
certainty that there is any such risk. Opponents of the principle object that it seems to
allow hunches, rather than knowledge, to guide policy.11
Furthermore, opponents of the precautionary principle note that allowing hunches or
mere possibilities to guide our reasoning reopens the possibility that the principle will
lead to paralysis. Just as there is no scientific certainty that mobile-phone masts pose a
potential risk to children, so there is no scientific certainty that mobile-phone masts do not
promote good health. If so, the non-standard-consequentialist defender of the precautionary principle seems to have the following choice: allow masts to be placed next to schools or
do not allow masts to be placed next to schools. Adopting the first course-of-action is
associated with a non-ascertained risk of harm, but so, if we are to base policy on any old
claims, is the second course-of-action. Even if we have a principle for choosing between
11
Resnik (2003) provides a useful overview of many of these debates, paying particular attention to
the context of environmental policy.
78
Section 1: Concepts
courses-of-action in situations where both are associated with some risk of special harm,
our principle will be inapplicable in this situation, as we have no risk-data on the basis of
which to choose, as there is no scientific certainty that either option poses any risk at all.
Admittedly, a deontological reading of the precautionary principle, one which stresses the
idea that there are certain sorts of actions we ought not to perform might not be vulnerable
to the paralysis problem. However, it will certainly seem vulnerable to attack on the
grounds that a policy of never being involved in the imposition of risk, regardless of how
certain we are that there are such risks, seems to threaten extreme conservatism in action.
Of course, a proponent of the precautionary principle could say that it makes sense to
think that there is a possibility that mobile-phone masts pose a risk and no reason to think
that there is a possibility that they might potentially benefit children, even if science has not
shown either claim to be true (or false). However, we need some way in which to distinguish
those claims that have not been proven with scientific certainty, which we should take
seriously, from claims not thus proven, which we should not take seriously. As stated, the
principle does not help with this task.
This epistemic objection to the precautionary principle is extremely strong. One option
would be to give up on the aspect of precautionary thought which seems to set extremely
low epistemic thresholds on when precaution is warranted.12 However, in the rest of this
section, I shall suggest one way in which a defender of the principle might motivate the
claim that we ought to act as if courses-of-action pose some threat even in the absence of
full scientific certainty that they do so, and suggest one way in which this general claim
could lead to a practicable principle. In the arguments above, I implicitly assumed as do
many contributors to this debate that, when generating the factive inputs into policymaking, our choices are either to rely on those risk-claims we have established with
scientific certainty or to rely on hunches and guesswork.
However, matters are not as simple as this sharp dichotomy suggests. To see why,
consider what happens in epidemiological testing. In standard scientific statistical testing,
we wish to discover, on the basis of the evidence available, whether we ought to say that
there is some connection between some kind of event and some kind of outcome (the
relevant connection may be expressed in probabilistic terms: the epidemiological evidence
does not show that if you smoke, then you are bound to develop lung cancer, but that if you
smoke, then there is a higher-than-normal chance that you will develop lung cancer). In
standard scientific testing, we choose a null hypothesis that there is no connection
between two events, for example, smoking and developing lung cancer and an alternative
hypothesis that there is some connection (perhaps a probabilistic connection) between
the two events, for example that smokers are at greater risk of developing lung cancer than
are (comparable) non-smokers.13
12
13
Note that the problem raised here is not unique to the interpretation of the precautionary
principle as involving either a tripartite distinction between outcomes or a focus on deontological
concerns. Even defenders of the precautionary principle who make it a version of maxi-min thought
need to be able to distinguish those cases where we ought to worry about some threat, the
probability of which is not established with scientific certainty, from cases where a non-established
threat is not worth worrying about (see Gardiner, 2006).
Note that I leave to one side the metaphysics of risk regardless of whether we think of risk as
epistemic or as ontological, we can be more or less accurate in terms either of how the world is or
what we ought to believe about the world. For more discussion of these topics, see Mellor (2006).
79
A statistical test can produce one of two results: either, it tells us that we ought to
reject the null hypothesis (that there is no link) and accept the alternative hypothesis,
or it tells us that there is insufficient evidence to reject the null hypothesis. In testing,
however, we can make two kinds of errors: we can generate false positives (we accept
the alternative hypothesis when it is false) and false negatives (we do not accept
the alternative hypothesis when it is true). It is impossible to minimize both of these
kinds of errors simultaneously (Levi, 1962). In standard scientific testing, it is thought
that false positives are a more serious kind of error than are false negatives. Therefore,
investigators routinely set very high p-values for their statistical tests, thus minimizing
their chances of generating false positives (asserting there is some significant connection when there is no connection), but also running a risk of generating false negatives
(failing to assert that there is some connection when there is, in fact, a connection
(Levi, 1964).14 Setting a high p-value means that if we claim that there is some
connection between two events, we are warranted in an extremely high degree of
certainty that we are right in our positive claim, as the chance that we have (falsely)
made such a claim has been minimized. As the term confidence level suggests, the
higher the p-value we adopt for statistical testing, the more confident we can be that a
positive result is correct.
A failure to show that the alternative hypothesis is true is not the same as showing that
the null hypothesis is true. Rather, the proper conclusion to be drawn when we fail to
prove the alternative hypothesis is that there is insufficient evidence to reject the null
hypothesis. After all, as we have seen, with a high p-value, it is possible that we will often
generate false negatives, that is, we will not say that there is a link when there is a link
between the relevant events. There is, then, an important distinction, perhaps elided by our
choice of symmetrical terms, between false positives and false negatives. To generate a
false positive is to make a claim that is a false, whereas to generate a false negative is,
strictly, merely to fail to assert a claim that is, in fact, true.
However, it is clear that when scientific results are transferred from journals and
laboratories to the context of policy and decision-making, failure to show a link is often
treated as proof of no link. This is a simple epistemological error the claim that we have
insufficient reason to believe some proposition p (there is insufficient reason to believe that
there is a link between X and Y) is confused with the claim that we have reason to believe
not-p (there is no link between X and Y).
Such confusion can have grave consequences for action, because acting as if it is the
case that not-p, when all that we can reasonably say is that p is not proven, can lead us
to act on false beliefs. This can have disastrous results. For example, return to the
question of whether or not we ought to allow mobile-phone masts to be placed next to
schools. Imagine that we cannot show, with scientific certainty, that the location of
masts is linked to childhood cancer. Therefore, we go ahead with placing masts next to
schools, something that we would not have done had we thought that so doing poses an
excess risk however small that children might, as a result, develop cancer. However,
also imagine that there is in fact a link between the location of mobile-phone masts and
schools: our claim that there was no link was a false negative. In this case, a policy of
14
Indeed, these standards are often set in computer programmes as default.
80
Section 1: Concepts
epistemic caution only acting on those claims about probabilistic relations which we
have very good reason to believe are true might lead us to a course-of-action which,
had we been less epistemically cautious, we would have rejected.15
These facts help us to understand what may be at stake in (at least some, but not all)
discussions of the precautionary principle. Many proponents of the precautionary
principle list examples of cases where scientific evidence that some activity was safe
turned out to be wrong, thus leading to catastrophic effects for the environment or for
human life (Harremoes et al., 2004). Of course, we can claim that there are many
different kinds of lessons that we ought to learn from such past mistakes. For example,
at least in the environmentalist context, it is often claimed that we have overwhelming
evidence that we misunderstand the complexity and underestimate the fragility of nature
(Carson, 1962). However, a more modest lesson we might draw from the past record is
that we ought to think more about the burden of proof; we ought to change our
standard testing methodologies at least in some cases to minimize our risk of false
negatives even at the cost of generating many false positives. Of course, setting out
which p-values we ought to adopt, and how such asymmetric testing methods might
work is no easy task. However, philosophers of science, most notably Carl Cranor
(1993), have worked on alternative testing methodologies which incorporate such
concerns.
We can, then, generate a sympathetic reconstruction of the precautionary principle
as a complex combination of a decision principle with an epistemic rule. The nonstandard-consequentialist proponent of the precautionary principle can be interpreted as
claiming that certain sorts of outcomes represent special bad outcomes, and, as such,
avoidance of these outcomes has lexical priority (of some kind) over other sorts of
considerations (such as the expected sum-total of normal harms and benefits associated
with some potential action). However, as well as this decision rule, she also suggests an
epistemic rule for the generation of inputs into policy-making. Even if standard statistical testing procedures ought to be used to ascertain the possible connections between
various courses-of-action and normal good or bad outcomes, special bad outcomes
are somehow importantly different, such that we ought to lower our normal scientific
standard of proof we ought to lower our p-values when investigating the possible
connections between some course-of-action and these outcomes. In effect, special bad
outcomes are doubly special: first, avoiding such outcomes has lexical priority when
choosing between different courses-of-action; second, the factive claim that there are
such possibilities can legitimately be asserted in policy-making contexts even when such
a claim has not been demonstrated to the level of certainty we normally require of
scientific claims.16
15
16
Similar points were raised in the epidemiological context by Austin Bradford Hill, one of the
founders of modern epidemiology. See Hill (1965), especially the final section.
How do these claims relate to the deontological version of the precautionary principle? I cannot
answer this question fully here, but it is interesting to note that we do often seem to vary our
standards-of-proof in proportion to what seems natural to call the moral costs of acting wrongly:
consider, for example, the difference between the standards of proof in the criminal and in the civil
law. Therefore, it seems reasonable to assume that the deontologist defender of the precautionary
principle might reach a similar conclusion.
81
Risk, tolerability and action: the way forward

for an ethics of risk?
The arguments in the preceding sections suggest that there are serious challenges to an
unthinking use of the tools of RCBA to risk-assessment and risk-management. One way in
which to frame these challenges would be to say that RCBA overlooks at least two important
issues: first, it assumes that all outcomes are ultimately commensurable, thus ignoring
important ethical distinctions between different kinds of outcomes; second, it fails to
recognize the complex web of obligations which might relate to the pursuit of public health
policy. Furthermore, although strictly a reliance on standard statistical testing procedures is
not an essential feature of RCBA, a proponent of the precautionary principle might suggest
that most RCBA rests on an implausibly idealized view of the value of normal models of
scientific inquiry. However, this view of science ignores the fact that the values of scientific
caution may be in conflict with our other ethical norms and worries.
There are serious ethical challenges to both our standard way of assessing risk and our
standard way of managing risk. However, RCBA has one extremely important feature,
arguably missing from many accounts of the precautionary principle: It recognizes the
unavoidability of trade-offs between the ends of public health policy and other kinds of
social goods. We might find something distasteful about placing a monetary value on
human life, but it is unclear that we always think that the fact that some policy or human
activity poses some risk that lives will be lost is a decisive argument in favour of ending that
policy or banning that activity. Rather, sometimes we do think that some system or policy is
safe enough, because the costs of making that system any safer would be excessive. An
advantage of RCBA is that at least it makes such judgments explicit. As I have already
indicated, it is unclear quite which public health outcomes are relevant to precautionary
thought, but there seems a danger that a precautionary approach involves commitment
either to the claim that no level of safety is ever safe enough or to the claim that we can
always avoid trade-offs. Neither of these claims seems at all plausible.
Therefore, constructing an ethics of risk for public health policy is no easy task. I do,
however, suggest that the arguments above provide us with one way in which we might
construct a plausible account of how we ought to deal with risks. The key claim of RCBA is
that we can legitimately impose risks on individuals or allow them to suffer certain sorts of
risks because of the social costs of not imposing or allowing those risks. An ethics of risk for
public health policy should, of course, allow that some risks are tolerable because those
risks are somehow necessarily related to the pursuit of other goods (Cranor, 2007). Of
course, RCBA sometimes seems to miss the complexity of such trade-offs. It would seem
odd, for example, to think that a trade-off between limiting liberty and reducing health risks
should be framed in monetary terms. However, at least, RCBA seems to provide a
framework for trading-off risks, costs and benefits of policies in a way in which the claims
that we should treat certain bad outcomes as special or that we should avoid doing certain
sorts of harm, even at the cost of allowing similar harms to occur, do not.
It seems, then, that RCBAs great advantage as an account of risk tolerability is that it
recognizes the unavoidability of trade-offs, whereas evocations of the precautionary
principle do not. However, reflection on some of the arguments for the precautionary
principle discussed in the third and fourth sections above can show a fact that proponents
of RCBA often overlook: what makes risks tolerable or intolerable is extremely contextsensitive, and related to complex ethical considerations. To make this point vivid, let us
82
Section 1: Concepts
return to an example where RCBA may seem at its strongest. Let us imagine, for example,
that mobile-phone masts do pose a risk of childhood cancer.17 However, we also know that
increased ease of communication might save lives (for example, because if it is easier to use
mobile phones, then people will find it easier to telephone the emergency services). In this
situation, a proponent of RCBA would say that a reasonable policy with regard to the
regulation of mobile-phone masts would have to balance the risks that children will develop
cancer against the risks that some people will be unable to phone the emergency services
when considering legislation on the location of mobile-phone masts.
However, this seems far too simplistic a way to view this decision. After all, if we place
mobile-phone masts next to schools, then we are placing children at risk, through our
adoption of a policy, and, furthermore, we are placing them at risk in an environment
which is one which they are legally compelled to attend. If we fail to place mobile-phone
masts next to schools, then we are leaving some at risk (that is, those who would, otherwise,
be able to contact emergency services using a mobile phone). Of course, all things considered, it is normally a good thing to save lives. However, we might think that, in this case,
there is an important distinction, overlooked by RCBA, between imposing risks on some
and merely failing to reduce the risks some suffer. This sense might be strengthened if we
discover that most of those who need to use mobile phones to call emergency services end
up in this situation because of their own actions. Of course, the issues here might get even
more complex again, but my point is simple: reasoning about this case (and, I suggest, any
case which involves risk reduction) requires not only that we balance outcomes, but that we
balance an extremely complex set of moral judgments, concerning both magnitudes of
harm and the causation of harm.
The key point that emerges from discussion of the precautionary principle, then, is that
the tolerability of risk may be extremely context-sensitive, and that this context-sensitivity
might relate to ethical considerations in a manner which cannot necessarily be captured in
monetary terms. Furthermore, an advantage of this context-sensitive view of risk tolerability is that it allows us to make sense of, at least some, features of how the public seem to
view and think about risk. When the public express fears over some risk which RCBA tells
us is not a serious risk, their fears may not be cognitively confused, but, rather, might reflect
a genuine ethical worry. Note, however, that this is not to say that the public cannot be
wrong about what the risks are. Rather, it is to say that public claims over risks can be a
useful guide to ethical considerations for risk policy.
The idea that we can use public attitudes towards risk as a guide to the ethical issues in
risk-management is also related to the issues discussed above, concerning the relationship
between ethical and epistemic norms in the identification of risks. Public scepticism about
scientific claims that some policy is safe enough need not be understood as expressing an
unwarranted anti-scientific world view. Rather, in at least some cases, they may be understood as pointing us towards an important question: whether the costs of acting on a falsenegative in some particular case are so great that we ought to change our standard statistical
norms in this context.
These comments on risk-tolerability are sketchy. However, I hope that they help the
reader to understand how we might hope to construct an ethics of risk which takes account
17
I should stress at this point that there seems to be very little evidence of any possible link between
mobile-phone masts and childhood cancer. The example here is chosen purely for expository
purposes.
83
of the fact that risk policy cannot be isolated, either from the pursuit of other important
social goods (as the precautionary principle would seem to suggest) or from complex webs
of obligation (as RCBA would seem to suggest). Rather, to identify and to manage risks
requires us to construct an account of when risks are tolerable, a highly contextual affair.
Furthermore, as the fifth section suggests, there may be reasons to be suspicious of using
standard statistical tools for the purposes of finding out what risks there are. Therefore, an
ethics of risk which still requires a complete articulation might provide a useful guide
not only to public health ethics but to the ethics of epidemiology.
References
Adams, J. (1995) Risk. London: UCL Press.
Beck, U. (1992) Risk Society: Towards a New
Modernity. London: SAGE.
Burgess, A. (2004) Cellular Phones, Public Fears
and a Culture of Precaution. Cambridge:
Carson, R. (1962) Silent Spring. Boston, MA:
Houghton Mifflin.
Cranor, C. (1993) Regulating Toxic Substances:
A Philosophy of Science and the Law. Oxford:
Cranor, C. (2007) Towards a nonconsequentialist approach to acceptable risks.
In Risk: Philosophical Perspectives, ed.
T. Lewens. London: Routledge.
Douglas, M. and Wildavsky, A. (1980) Risk and
Culture. Berkeley, CA: University of
California Press.
Elliott, R. (1997) Faking Nature. London: Taylor
& Francis.
European Commission (2000) Communication
from the Commission on the Precautionary
Principle. Brussels: European Commission.
Available at: http://eurlex.europa.eu/LexUriServ/
LexUriServ.do?uri=COM:2000:0001:FIN:EN:
PDF (accessed: 10/1/08).
Gardiner, S. (2006) A core precautionary
principle. Journal of Political Philosophy, 14
(4): 3360.
Hacking, I. (1975) The Emergence of Probability.
Cambridge: Cambridge University Press.
Hacking, I. (1990) The Taming of Chance.
Hacking, I. (2000) The Social Construction of
What? Boston, MA: Harvard University
Press.
Hansson, S.-O. (2007) Risk. In Stanford

Encyclopaedia of Philosophy. Available at:
http://plato.stanford.edu/entries/risk/
Harremoes, J., Gee, D., Macgarvin, M., Stirling,
A., Keys, J., Wynne, B. and Vaz, S. G., eds
(2002) The Precautionary Principle in the
Twentieth Century. London: Earthscan.
Hill, A. B. (1965) The environment and disease:
association or causation? Proceedings of the
Royal Society of Medicine, 58(5): 295300.
Hourdequin, M. (2007) Doing, allowing and
precaution. Environmental Ethics, 29(4):
33958.
House of Commons Science and Technology
Committee (2006) Scientific Advice, Risk and
Evidence Based Policy Making. London: The
Stationery Office.
Hughes, J. (2006) How not to criticize the
precautionary principle. Journal of Medicine
and Philosophy, 27(3): 35568.
Irwin, A. (1995) Citizen Science. London:
Routledge and Kegan Paul.
Jasanoff, S. (1999) The songlines of risk.
Environmental Values, 8(2): 13549.
Lewens, T. (2007) Introduction: risk and
philosophy. In Risk: Philosophical
Perspectives, ed. T. Lewens. London:
Routledge, pp. 120.
Levi, I. (1962) On the seriousness of mistakes.
Philosophy of Science, 29(2): 4765.
Ling, T. and Raven, A. (2006) Pharmacogenetics
and uncertainty: implications for policy
makers. Studies in the History and Philosophy
of the Biological and Biomedical Sciences, 37
(3): 53349.
Luce, R. D. and Raiffa, H. (1957) Games and
Decisions. New York: Wiley.
84
Section 1: Concepts
Martuzzi, M. and Tickner, J., eds (2004)

The Precautionary Principle: Protecting
Public Health, the Environment and
the Future of Our Children. Copenhagen:
WHO.
Mellor, D. H. (2005) Probability: A Philosophical
Introduction. London: Routledge.
Olsaretti, S. (2005) Endorsement and freedom in
Amartya Sens capability approach.
Economics and Philosophy, 21: 89108.
ONeill, O. (2002) Autonomy and Trust in
Bioethics. Cambridge: Cambridge University
Press.
Pogge, T. (2004) Relational conceptions of
justice. In Public health, Ethics and Equity, ed.
F. Anand, F. Peter and A. K. Sen.
Oxford: Oxford University Press,
pp. 13562.
Rawls, J. (1993) Political Liberalism. New York:
Columbia University Press.
Rawls, J. (1999) A Theory of Justice, rev. edn.
Resnik, D. B. (2003) Is the precautionary
principle unscientific? Studies in the History
and Philosophy of the Biological and
Biomedical Sciences, 34(2): 32944.
Sandin, P., Peterson, M., Hansson, S. O., Ruden,
C. and Juthe, A. (2002) Five charges against
the precautionary principle. Journal of Risk
Research, 5(4): 28799.
Sen, A. K. (1992) Inequality Reexamined.
Shrader-Frechette, K. (1980) Nuclear
Power and Public Policy: The Social and
Ethical Problems of Fission Technology.

Dordrect: D. Reidel.
Shrader-Frechette, K. (1991) Risk and
Rationality: Philosophical Foundations for
Populist Reforms. Berkeley, CA: University of
California Press.
Slovic, P. (1987) Perception of risk. Science, 236
(4799): 2805.
Slovic, P. (2000) The Perception of Risk. London:
Earthscan.
Sunstein, C. (2002) Risk and Reason. Cambridge:
Sunstein, C. (2003) Beyond the precautionary
principle. University of Pennsylvania Law
Review, 151(3): 100358.
Sunstein, C. (2005a) Environmental
protection and cost-benefit analysis. Ethics,
115(2): 35185.
Sunstein, C. (2005b) Laws of Fear: Beyond the
Precautionary Principle. Cambridge:
United Nations (1992) Rio Declaration on
Environment and Development. Report of the
United Nations Conference on Environment
and Development, Rio de Janeiro, June 314.
New York: United Nations.
Wolff, J. (2002) Railway Safety and the Ethics of
the Tolerability of Risk. London: Railways
Safety Research Programme. Available at:
http://www.rssb.co.uk/pdf/policy_risk.pdf
Wolff, J. (2006) Risk, fear, blame, shame and the
regulation of public safety. Economics and
Philosophy, 22(3): 40927.
Section 2
Chapter
Issues
Smoking, health and ethics

Richard Ashcroft
Introduction
The moral status of smoking is much disputed. Because smoking is known to be
harmful to the health of the smoker, and to those exposed to their smoke, and because
smoking is addictive, it is natural for public health practitioners and policy makers to
consider smoking to be a public health problem, and to seek ways to eradicate it, or at
least bring it under control (Boyle et al., 2004). Because smoking is for many smokers a
pleasure, and an activity rich in social and cultural meaning, it is natural for such
smokers (and some non-smokers) to resist these attempts to control smoking. For those
smokers for whom their smoking is no longer a pleasure, or for whom it may be a
pleasure but who accept that it is a pleasure they wished they could forego, attempts to
control smoking can be both aid and irritant. And even unrepentant smokers would
concede that encouraging young children to take up smoking would be discreditable at
least, immoral at worst.
In many respects, then, smoking provokes as complex social, moral and legal reactions
as those other risky, scandalous and usually pleasurable activities sex and drinking. We can
begin to get an understanding of the moral issues by distinguishing those moral issues that
arise in connection with the smokers self-regarding behaviour, and those which arise in
connection with her other-regarding behaviour. Another way to approach the issues would
be to look at the manufacture, supply and marketing of tobacco products, and to consider
the ethical aspects of regulating the tobacco market. From a public health practitioners
point of view this might actually be the more logical approach, since it uses a kind of
population perspective. Yet this would be to neglect the crucial feature of tobacco regulation
(as distinct from say environmental pollutant regulation), which is that there is a buoyant
market of tobacco users who for various reasons want to consume the regulated product.
We need to understand the nature of that want, and its moral standing, if we are to
understand the grounds for regulating the product and its markets. For this reason, I will
proceed from the tobacco smoker up, rather than from the market down.
The morality of smoking itself

The painter David Hockney has been widely interviewed on his opposition to banning
smoking in public places. In a letter published in the Guardian newspaper on 1 June 2004
he wrote:
85
86
Section 2: Issues
[. . .] Smoking is a great pleasure. If it knocks some time off your life, its only at the end of it. So what!
I admit people might not like smoky places, but you have to accept that some people dont mind
them, its a price you pay for some excitement, but the dreary dont want anybody near smoke.
I loathe the view of life that thinks some people have so much more knowledge of life that their advice
must be listened to. Im the expert on my life, not the doctor. He can advise, make suggestions, but
ultimately I decide what will give me the excitement I want from life.
I take exercise. I walk in Holland Park, here in west London, every morning. One morning I was
watching a peacock strut around with two rabbits scampering about, and then a black-and-white bird
arrived. The scene was magical to me. I smoked a cigarette.
Three girls came by, jogging. One looked at me, shaking a finger and tut-tutting at me for smoking.
OK, but they didnt see the peacock or the rabbits or the bird they were so obsessed with their
own bodies, and they thought their activity was healthier than mine. I do not think it was, but I dont
want to ban jogging in Holland Park.
(Hockney, 2004)
In an open letter published in the Guardian on 25 February 2006, he wrote:

[. . .] Gordon Brown is a prig P.R.I.G., a dreary atheistic Calvinistic prig, who Im sure will never be
elected in England. He goes along with a health lobby whose view of life itself I detest.
I have utter contempt for it. I feel I am entitled to my opinion. I dont mind prigs but when they want to take
my little corner as well, I have a right to argue against their dreary view of life contaminating mine.
(Hockney, 2006)
On 15 May in the following year, he said in a further article in the same newspaper I smoke
for my mental health. I think its good for it, and I certainly prefer its calming effects to the
pharmaceutical ones (side effects unknown), going on to identify the motives behind the
smoking ban as being anti-smoker, being led by mean-spirited and dreary people and a
political and media elite. In particular, he sees this as part of the loss of a sense of
messiness. Both the ban, and the signs required in shops and offices to proclaim it, are
signs of the uglification of England (Hockney, 2007).
None of Hockneys interventions were distinguished by much in the way of careful
argumentation, but he highlights an important concern about the regulation of smoking.
While proponents of smoking regulation see this as regulation in the interests of health,
Hockney suggests that it is more properly understood as the regulation of pleasure. He thinks
that ones pleasures are no one elses business, and that the interference of the state in his
pleasures is illegitimate. Its illegitimacy lies both in the extension of the reach of government
outside its proper sphere, and in the transfer of personal dislikes into public policy by prigs
with power, and in the colonization of the messy but beautiful world by bureaucratic tidiness
in all its (unintended?) ugliness. I will return to the question of the proper reach of
government later, but here I want to concentrate on the charge of priggishness.
Smoking, virtue and vice

Consider the activity of smoking itself. Is there anything we can say about it, morally,
insofar as it affects only the smoker himself? Set aside for the time being the questions about
possible harms to others. Conceivably, one could argue that smoking is morally
praiseworthy. This would fit most naturally into a virtue ethics framework. Hockney would
no doubt like us to see him as a genial fellow, who takes his pleasures where he finds them,
Chapter 5: Smoking, health and ethics
87
does not impose his views or himself on others, and has sympathy for the underdog and an
appreciation of the ramshackle. The opposite of a prig, in fact. And his smoking could be
understood in relation to his character as a genial fellow.
The difficult part here would be in establishing whether the activity of smoking was
intrinsic to his character as genial fellow, or accidental. If intrinsic, this would presumably
mean that he would be less genial without his smoking, and one could inquire as to why this
would be so. Most likely, it would be because he would be tenser, or more irritable, or less
tolerant of the failings of others and their own particular foibles and habits. Disentangling
this from the consequences of dependency on nicotine, either chemical or psychological or
both, would be difficult. Evaluating the moral significance of this from a virtue ethical point
of view would also require us to consider the interdependence of the virtues his smoking
enables him to actuate with other virtues (such as virtues of moderation and self-denial,
even-temperedness and so on). If a consequence of stopping smoking was to become less
genial, then we might say that the smoking was a necessary condition for the geniality;
equally, we might say that while we value the geniality, we disapprove of the irritability
displayed when he is deprived of his tobacco. And this would connect the tobacco as much
to a vice as to a virtue. In other words, it is not that smoking is morally praiseworthy as
such, but rather that it points us to particular moral features of Hockney (and people
generally) underlying the habit and practice of smoking. The same would apply to other
activities, such as eating good food or drinking fine wine.
Conversely, if we take features of tobacco smoking that are intrinsic to the substance and
the practice of smoking it, we do know that it is both harmful to health and addictive. It is also
the case that, unlike alcohol or most foods, it is harmful even in small quantities. So we might
say that smoking is morally blameworthy in itself, because it involves deliberate harm to self.
While most harms to self are regarded as beyond the proper reach of law and government, at
least where these harms are inflicted knowingly and voluntarily, virtue ethics and most
varieties of natural law theory would hold that self-harm is immoral. However, this is also
difficult to cash out in detail. Many most? pleasures and indeed many morally obligatory
activities involve a degree of taking risk to ones own welfare. It would be difficult to
distinguish consistently between smoking and other risky pleasures. Moreover, if we frame
the issue in terms of self-harm, it is not clear that the argument goes through. Hockney (2007)
claims that he smokes for his mental health, and if so, we could see the choice to smoke as an
example of a health-health trade-off (Sunstein, 1997: ch. 12). And even if it were a trading off
of health against some other personal good (such as pleasure), it would not be easy to say why
an adult cannot or should not choose to make this trade-off.
The addiction issue is different. From a virtue ethics point of view, it is questionable
whether a self-harming behaviour that is addictive could be virtuous, even if it was
contingently a condition for the exercise of some virtue in this individual, since as an
addiction it would interfere with virtues of self-possession and self-control. Nonetheless,
this is a fiddly argument to make out, since the connection between virtue and habit, and
virtue and character trait, is complex.
Are voluntary, self-inflicted harms to health special?

One factor that may confuse us here is the connection between smoking and health. If
smoking were not hazardous to health, but simply something that was pleasurable while
attracting some other kind of cost for instance, making the smoker smelly, or look
88
Section 2: Issues
ridiculous there would be no question of paternalistic regulation. There is of course a

question of ensuring that people make choices that are both genuine choices and adequately
informed, especially since smoking is addictive. It may be that this is purely a question of
ensuring that choice is autonomous; once that is done if it can be is there anything else
to say? There is an interesting question to be asked about why it is health that attracts
special attention in terms of (attempted) paternalistic regulation. Another way to think of
this comes from the observation, frequently made, that if tobacco were to be introduced to
the market today, no government would allow it to be sold. This would be because it is
addictive and harmful to health, even in small quantities, and, in addition, causes harm to
third parties exposed to second-hand smoke. Again, we will return to the harm to third
parties arguments later. The observation we are considering here suggests that governments
are keen to regulate hazards to health, including hazardous consumer goods. But this does
not explain why hazards to health are particularly important to governments, nor the
source of the normativity of health concerns for government. Moreover, the observation
points out a feature of public policy that matters: it does not have to be consistent. We
might, with hindsight, regret the introduction of tobacco smoking. But that regret does not
provide a strong basis for removing tobacco smoking, if we set it against the other interests
relevant in this case, in terms of the interests smokers have in smoking. The same goes for
alcohol, gambling and so on. Unless there is something about health that justifies paternalistic intervention to protect it, the basis of the paternalist regulation of smoking would seem
to be mere disapproval. And this would be precisely the kind of tyranny of the majority,
which classical liberalism tried to control. As we shall see below, health hazards to third
parties, considered as involuntarily incurred harms, are a justification for the regulation
of smoking, but this is, if paternalistic, of a different kind to the paternalistic regulation of
first-party harmpleasure trade-offs.
Moral uncertainty and moralism

Given these difficulties in sorting out the morality of smoking insofar as it bears on the
agents own self-regarding behavior, it is reasonable to decide to set the issue aside. We
might set it aside on one of the following grounds. First, we might hold that in a situation of
moral uncertainty, we should withhold judgement. Second, we might decide that while
there may be a moral question involved, it is not our place to decide on it as it applies to the
conduct of another. Or rather, third, we may think what we like, but it would be rude to
express our view when the behavior has no impact on us. To do so would be to be priggish
or moralistic (Fullinwider, 2005). I feel some sympathy with Hockney confronted by the
joggers in Holland Park.
None of what I have sketched here necessarily involves an appeal to respect for autonomy,
first person privileged access to her preferences, or similar liberal premises. In starting off by
considering the moral standing of smoking from a self-regarding point of view, I eschewed
the assumption that the autonomous choices of the individual cannot be evaluated, or that
personal preferences are not subject to rational or moral correction. The argument so far has
led us to the point where we have a qualified defence of an autonomy principle, to the effect
that taking the agent as an integrated person, and his projects and preferences in the round,
we cannot form a settled judgement on the morality of smoking from a self-regarding point
of view. We should leave the smoker to work out for herself whether she thinks of her
smoking as a moral issue at all and, if she does, what she makes of it.
89
Suppose then, with Hockney, we see the citizens of England as being oppressed by a
moralistic elite intent on denying them a pleasure of which there is no convincing moral
ground for disapproval so far as it affects the smoker alone. We might well think of
activities such as banning smoking in enclosed public places, imposing steep taxes on
tobacco products, and so on as the means by which an elite dislike or disapproval of
smoking is imposed on individuals who like smoking, and groups of individuals who like
smoking together, and on individuals with non-smoking friends and relatives who do not
object to being in the presence of smoking. Absent good moral reason for regulating the
behavior of individuals, this would be the imposition of elite preferences without justification. Even if there were good moral reason for disapproving of the conduct of smokers from
a self-regarding point of view, this might not be sufficient to warrant public policy or
legislation to regulate their behavior (Hart, 1963). Even if one took the view that some
moralistic legislation is proper if it secures the commonly shared values in a community
(Devlin, 1965), one could argue, with Hockney, that the relevant values requiring protection
are those of toleration and un-priggishness.
Harms to self, the undeserving patient and taxation

The public health practitioner is by now probably very impatient. The public health
practitioner does not see the regulation of smoking as a matter of moral regulation, but
simply as a matter of health promotion. And more importantly, even if the individual is free
to go to hell in his own way, as Robert Frost put it, we are entitled to regulate the impact
his choices have on others. We will come to the latter part of the argument in due course.
But there is a residual area of concern with the conduct of the individual which would bear
on the public health practitioners interests.
We can argue that our health is our own, and that while we have a legitimate interest in
controlling actions and policies of others which threaten that health, what we do to our own
health is our business. Some religious believers do not accept this, holding that our lives,
bodies and health are gifts from God, which we are to safeguard. Similarly, some
communitarians would argue that care of our health is part of our general collection of
obligations to each other in a moral community: only by maintaining ourselves in good
health, so far as this lies within our power, are we able to discharge our responsibilities to
the weaker members of society and to promote the common good. Neither of these lines of
argument are especially widely shared at present. But it is commonly argued that someone
who deliberately damages his health is less deserving of health care that could fully or
partially remedy that damage to his health. The smoker is often argued to be a good
illustration of this principle. Let us consider the arguments briefly.
Imagine a smoker of 60 years of age, who has advanced heart disease and would be a
candidate for a heart transplant, other things being equal. One clinical approach would be
to consider her for a transplant without regard to the smoking. She might be listed, or not,
purely on the basis of likely chances of success and expected survival post-transplant, in the
same way as any other transplant candidate. Another would be to distinguish between the
smoking-related cause of the illness and the impact post-transplant smoking would have on
the success of the transplant. She could be asked to agree to stop smoking as a condition of
the transplant. This might be understood as a moral requirement (damaging ones donated
heart or ones health could be seen as wilful self-harm or as damaging a scarce resource
someone else could have benefited from), or simply as an evaluation of likely effectiveness
90
Section 2: Issues
of a transplant. She could be assigned to the transplant list with a lower priority than an
equivalent non-smoker. Or she could be refused access to the list. Setting aside general
issues of distributive justice in health care which concern duties to rescue, allocation by
need or allocation by expected outcome, and so on, we can distinguish two claims about
desert and access to treatment in this situation. The first claim is that past health behavior
can affect the degree to which one deserves treatment for ill health resulting from that
behavior. The second claim is that future health behavior can affect the degree to which one
deserves treatment for ill health relating to that behavior. The first claim involves a claim
that the behavior itself is discreditable. Most clinicians would express some discomfort
about applying a retrospective judgement, perhaps because they believe it to be moralistic,
or perhaps because they see it as involving a double jeopardy (you smoked, so you are ill,
but because you smoked, we will not treat your illness). The second claim involves a
judgement of relative desert between candidates for transplant which can be affected by
the fact of smoking. This may be defensible, if the fact of continued smoking is taken on
good evidence as a sign that compliance with other necessary treatment will be less likely.
Although there are some arguments which might explain offering a transplant to a
smoker on conditions, or not at all, which do not depend on moral judgement of their past
or present smoking, it is more plausible to say that we do judge smokers, and to consider
whether this is justified. Consider two people, one our smoker just described, another a
lifelong non-smoker of similar age and ill health. Both need a heart transplant and there is
just one heart available. If we think the non-smoker should have the heart because he is a
non-smoker, rather than because of likely success or greater life expectancy or some other
more general principle of distributive justice in health care, then why is this? It may be that
the smoker is somehow responsible for doing the non-smoker out of the heart, if she gets it.
How would this be? It may be because we think the non-smoker would be ill anyway,
whereas but for her smoking, the smoker would not be. If so, we are considering that the
smoker has caused her ill health, and also that by claiming a scarce resource to restore that
health, she is denying someone else that resource. So she is displacing the consequence of
her behavior onto another, blameless, person. Thus, her smoking cannot be considered a
purely self-regarding behavior in this context, even neglecting (as we have so far) the third
party effects of smoking. This is hard to defend, however. Any allocation of a scarce
resource between parties with a claim on it will involve winners and losers. The losers
suffer no wrong at the hand of the winners, other things being equal. If we admit desert into
our allocation scheme for health care resources, we have three problems. First, we have little
settled agreement on the conception of desert that we ought to use, or what its scope may be.
Second, people rarely come to allocation decisions clean, in the sense that everyone has
some health-related habit which undermines our health to some extent, be it smoking,
drinking, physical idleness, excess workplace stress, or even jogging. Adjudicating between
the claims of differently compromised individuals is at least difficult to do objectively.
Third, we may be happy to transfer technical and professional responsibility to doctors. We
are much less happy to see them as moral experts who can adjudicate between patients
differing degrees of desert.
In passing, another issue arises in connection with publicly funded health care. It is
sometimes observed that the proceeds of taxation on tobacco products exceeds the costs to
the National Health Service of smoking-related illness. This would require some justification. Imagine, however, that these two figures were more or less in balance. Then we could
see this tax as a kind of insurance against eventual tobacco-related illness. In that case, we
91
would have no scarce resource argument for non-treatment of smokers, save as regards
those resources which are scarce on non-financial grounds (such as organs for transplant).
Imagine, next, that the taxation income does not cover the costs to the NHS of smokingrelated illness. Then we might say that smokers deserve lesser treatment because they have
not paid their way. The basis of the NHS is solidary coverage of the nations health care
costs; some pay more than they take out, others take out far more than they can contribute.
What could be objected to in the case of the smoker is deliberate over-drawing. Of course,
the answer here would likely be to increase the duty on tobacco products, rather than to
restrict smokers access to health services. So, if smokers are over-contributing, does this
give them special claim to health resources? This possibility becomes even more probable if
one considers not only the contribution made by smokers through taxation but also the way
that smokers by dying earlier than average draw less than their expected entitlement from
the pensions system over their post-retirement lifetime, given that that section of their life
may well be shorter than average. The main reason for taxation is not to insure smokers
against the risks of their smoking, but to raise revenue for the State, to fund the NHS, and to
discourage smoking. The solidary basis of NHS funding through income taxation and
national insurance does not allow for special claims related to level of contribution. So as
a matter of public policy smokers are not entitled to priority treatment merely because they
have paid more in. What the taxation issue points out is partly that there is indeed a degree
of State moralism regarding smoking, but also that the State has other reasons for wishing
to discourage smoking than the regulation of individuals self-regarding behavior
(Wilkinson, 1999).
Other aspects of smoking and its regulation

The classical liberal approach to the regulation of personal behavior is to avoid doing so
unless that behavior imposes harm on others without their consent. This is a superficially
simple criterion, but its subtleties are quite deep. For instance, we need to know whether
any harm at all should be prevented through regulation, or whether a threshold approach is
needed. Some harms may be trivial in intensity or duration. Second, we need to know
whether harms caused by omission or inaction are as significant as harms caused by action.
Can people be obliged to act so as to prevent a harm? Third, we need to know whether the
burden of regulation itself is proportionate to the harm prevented or controlled. Fourth, we
need to know whether the harm prevented can be characterized objectively. For instance,
liberals try to distinguish harms from nuisances or irritations; the former may well elicit a
legal response, while the latter will do so only in relatively extreme circumstances. Finally,
we need to know whether consent by the injured party is enough to remove government
interest.
The regulation of information about smoking
Until the evidence for the impact of passive smoking became conclusive, smoking could be
conceived as purely a matter of harm to self, combined with potential nuisance to others.
Harm to self would not normally be grounds for paternalistic intervention, as we have seen,
unless it could be shown that this harm was unwitting in some way. Perhaps from this point
of view it is the addictiveness of smoking which is more important than the harm of the
activity of smoking itself harmful though each cigarette is, it is the accumulation of harms
and the difficulty of stopping which creates the most serious problems. Thus the earliest
92
Section 2: Issues
stages of smoking regulation concentrated on measures to improve or regulate information

to smokers. For instance, tobacco advertising is tightly regulated or even banned. Where it
is permitted, the attractive images used in advertising and branding are supplemented by
striking warnings of health hazards; and information about the nicotine and tar content of
tobacco products are also closely supervised. Aggressive public information campaigns
about the dangers of smoking are pursued, and in many countries this negative advertising
may now be the only mass media advertising of tobacco products which is permitted. This
way of regulating the tobacco market achieves a number of functions. It ensures that
smokers (or would-be smokers) have access to accurate information about the risks and
benefits of the product they are using. It provides a constant stimulus to smokers to inspire
thoughts of giving up smoking or to reinforce the will to quit among those who are trying
not to take smoking up again. It provides a discouragement to those who would see
smoking as something worth trying. The essential function of this information regulation
is to improve the quality of consent to tobacco consumption (or its refusal), but it also seeks
to make that choice less attractive, and to alter the cultural presumption in the direction of
the disapproval of smoking its de-normalization.
Now against a backdrop of a long period in which smoking was normal and acceptable,
arguably because people were unaware of how dangerous smoking was to their health and
that of others, this is understandable. But it is at least debatable whether actively discouraging people from smoking, and focusing messages on smoking is bad rather than choices
between different kinds of tobacco product or different brands of cigarette, is justifiably
paternalistic. The natural response to this argument is to point out two features of tobacco
marketing: first, that until tobacco advertising was regulated, people were given sometimes
extremely favourable and misleading information about cigarettes and their health benefits,
and that there was no debate about the paternalism in pro-advertising. Second, because
smoking is addictive, the impact of addiction on the will is to undermine autonomy, so that
paternalistic intervention in terms of countervailing messages is justifiable as a proportionate response to autonomy compromised by addiction (Hansson, 2005).
What cannot easily be defended is the use of informational measures to alter the context
of smoking by promoting a dislike or intolerance of smoking by third parties as a means to
altering the behavior of smokers. And yet it is a known feature of behavior modification
that altering the climate of acceptance of a given behavior is an effective method for
changing that behavior. A good example of this is in teenage health education. We can
try to discourage teenagers from taking up smoking directly, through the restriction of sales
and through health education. But an important means for discouraging teenage smoking is
to get teenagers themselves to despise their smoking peers, for whatever reason. Questions
of teenage autonomy aside, persuading a teenager that smoking is uncool affects both the
self-regarding behavior of the teenager but also their other-regarding behavior. The former
is presumably acceptable, the latter more difficult to justify. Of course, it may well be
justified as a desirable but unintended consequence of the promotion of positive selfregarding health behaviors. But there is something questionable about using teenagers
clique-forming behavior to ostracize smokers as a method of making smoking unattractive.
All of this must be taken in the context of very powerful cultural and social psychological
factors disposing teenagers to smoke, from their tendency to seek risks and identify
behavior disapproved of by authority figures as attractive, to the associations of smoking
with other forms of teenage social life (such as attendance at performances of popular music
and discotheques). What may be hard to justify on its own merits in terms of direct
93
attempts to modify behavior and cultural attitudes needs to be seen against the background
of other, no less direct, means by which behavior and culture are continually shaped with
very little consideration of ethics.
Regulating smoking as a nuisance

In the period before passive smoking became a major focus of smoking regulation, a
regulatory approach focused on smoking as a nuisance to others was also developed. This
would include the establishment of non-smoking carriages on trains, non-smoking sections
of restaurants or non-smoking rooms in hotels, and, more subtly, the gradually rising social
unacceptability of lighting up without asking permission and thence the gradually rising
social acceptability of refusing permission. But notice that all these measures arise out of
more or less explicit contracts between parties, rather than formally imposed state regulation of nuisance.
Regulating smoking as a harm to others

Taking the various approaches to tobacco control we have examined so far, we see that if we
exclude harm to others, tobacco regulation largely turns on smoking being seen as a social
problem. It may be a nuisance to others, and we may be concerned about the quality of
information available to those who in smoking may be more or less unwittingly harming
themselves or becoming addicted, especially where the most vulnerable are those who for
other reasons are imperfect decision-makers. But the appropriate response to these features
of tobacco use would be better information and regulation of the marketing of tobacco,
rather than regulation of its consumption. And we might retain, with David Hockney, a
sense that many of the initiatives in this area do tend to present smoking as a nuisance to
third parties, rather than as a pleasure to first parties, and further, to regulate that nuisance
in a paternalistic way.
However, just as we could not examine smoking regulation for long without recognizing
its third party effects, neither can we consider those third party effects without considering
those which amount to substantial harms. Three cases are of particular importance. The
first case is that of smoking in enclosed public places. The second case is that of smoking in
the home. The third case is that of the impact of smoking on the unborn child.
Smoking in the workplace and in public

Smoking in enclosed public places has received a lot of attention recently as national and
regional governments across the world have begun to implement bans. There are two
broad kinds of justification offered. The first is harm-related; the second is public goodrelated. The harm-related argument focuses on harms to third parties exposed to second
hand smoke involuntarily. The public good-related argument focuses on the way in
which smoking may degrade a public good, clean breathable air. We will return to this
argument later.
Other things being equal, if person A smokes, and person B consents to being in the
presence of that smoke, this releases person A from any responsibility for the harm caused
to person B by that smoke. Obviously, if person Bs consent is uninformed, or not truly
voluntary, this would potentially invalidate the consent. It need not matter whether person
B comes into contact with person A directly. If person B enters a bar, say, knowing that
this is likely to be a place where smoking is permitted, then that may be sufficient to count
94
Section 2: Issues
as consent. Until recently, it may have been accepted that the most efficient way to achieve
a reduction in smoking in public places was via consent. Places wishing to attract nonsmokers would bar smoking, or try to restrict it to separate rooms with strong extractor
fans, for instance. Places wishing to attract smokers would make no such efforts. As more
people favoured smokeless environments, the market would work its magic, and gradually
a pattern of facilities, shops and offices would emerge in which a small number of places
would cater specifically for smokers and their friends, and while most places would
exclude smoking. Some (particularly in the tobacco industry) hoped for technological
innovation which would allow environments to be virtually smoke free without the noise
or drafts associated with extractor fans. For whatever reason, this seems not to be a
possibility at present.
This market based approach continues to attract support, but in the United Kingdom
and elsewhere a more direct approach is now favoured (Royal College of Physicians, 2005).
This is to impose a ban on smoking in public places. A weak argument for this might be
that where ones consent is presumed upon too freely and generally, then it is not consent.
Imagine a society where people are casually rude to each other in a way which is not directly
and deliberately offensive, but merely betokens a sort of mild incivility. Suppose that a
generation before, people were generally civil (if somewhat formal) with each other.
A survivor from that generation might well mark and regret this difference. Nonetheless,
they go about their business in the new world, faute de mieux. We can hardly say that they
consent to this general incivility. It is more that they put up with it. Perhaps consent is
ambiguous here, meaning both a weak sense of agreement (putting up with, going along
with, possibly under protest) and a strong sense of endorsement.
A stronger argument for smoke-free public places points us to the individuals who are
least likely to be giving voluntary consent to smoke exposure, who are the workforce in
smoky environments, rather than the consumers who can come and go as they please.
Labour markets being what they are, workforce members are relatively weak in bargaining
over their conditions of employment, especially where they are in a service industry where
consumer preferences are thought to be crucial. Unlike most consumers in bars, bar
workers (a typical example) will be exposed to smoke all day, every day, without much
say in the matter, and without control over how much they smoke. It could be suggested
that they could change employment, but many bar workers take this sort of work in
circumstances where suitable alternative work is hard to find. Under these circumstances,
one could argue that exposure to second-hand smoke is not free, and thus not consented to.
And thus, the harm argument for regulating smoking to protect non-consenting third
parties is able to get off the ground.
Two objections might arise here. First, one might question the level of risk being
regulated. If we could, by use of fans or equivalent, lower, without eliminating, the risk to
workers, would we not reach a situation in which the exposure to the hazard of smoking is
analogous to other, less regulated workplace hazards? Why should we adopt a zero tolerance
approach to exposure to workplace smoke, but not to other hazards? Second, one could
argue about the proportionality of the regulatory response. While some smoky dive bars are
definitely hazardous to health, a large well-ventilated modern office block does not present
the same exposure, given the relatively small number of office workers who now smoke. It is
accepted that some workplaces (such as prisons and psychiatric hospitals) are exceptions to
regulations, since their inmates are not free to move around. But surely the workers in these
workplaces are just as deserving of protection as bar workers?
95
Objections like these, which have some merit on the face of it, direct us to two features
of public policy. The first is that public policy will be framed partly with an eye to simplicity
and enforceability. A simple ban with limited exceptions is much easier to enforce (and
much fairer, since clear to all regulated under it) than a complex ban. The second is that
consistency is not a necessary component of public policy. That there may be other less well
regulated workplace hazards does not justify regulating workplace smoke exposure more
loosely it just means that they may well be subject to different balancing exercises, or be
candidates for further regulation in their turn in future. Of course, it may well be that there
are wider objectives behind a ban than simply regulating harm to third parties. For public
health minded policy makers, reducing smoking in the name of protecting non-smokers has
the likely useful consequence of reducing smoking by smokers. It is probably accepted that
this is not an appropriate goal of non-paternalistic anti-smoking policy, but no public
health policy maker is going to regret this unintended consequence.
Smoking in the home and during pregnancy

From the point of view of protecting the vulnerable third party, it is possible to argue
that regulating the smoky workplace is a less important task than that of protecting
young children in the home. While spouses and partners of smokers can have their own
ways of persuading smokers to quit, or to regulate their behavior, or can be considered
to consent, children may have relatively little influence, and can be particularly vulnerable both to the health effects of passive smoking and to taking up smoking. However,
what measures could be applied to regulate smoking in the home? A range of methods
could be used at the draconian end of the spectrum: it is not true, as is sometimes
thought, that the home is a sacrosanct place, where the reach of the law does not run.
One could criminalize smoking, and then carry out raids on homes where smoking was
suspected; one could take children of recalcitrant smokers into care or make smoking
into grounds for assignment of custody in divorce cases to the non-smoking parent.
However, we have to ask whether these methods are proportionate in their response to
the risks of smoking, respect other important principles such as the human right to a
private and family life, and are in any event effective in reducing smoking in the home.
Similar arguments apply, with even more force, to the case of smoking during pregnancy (by the expectant mother or by her co-resident family members). There is a
strong moral appeal that can be made not to risk harm to vulnerable others, but given
other ethical and policy arguments in support of womens autonomy and bodily privacy,
it is unlikely that an argument for compelling smoking-cessation (through criminal or
civil law remedies) could be successfully mounted.
New directions in the debates on ethics and smoking

Three recent debates are of particular note. The first, mentioned above, is a new approach to
arguing for the control of smoking which focuses not on harm but on the protection of a
public good (Smith et al., 2003). A public good may be defined as a good which is nonexcludable (that is, anyone can use it) and non-rival in consumption (that is, my use of it
does not diminish its quantity or availability to you). The standard example is clean,
breathable air. Public goods may be destroyed, however. Again, consider clean air.
A release of toxic gas into the atmosphere would significantly degrade the air quality.
Public goods have interesting properties for the economist. In particular, because anyone
96
Section 2: Issues
can use them, it is in no ones personal interest to pay for them, or to preserve them. Hence
there is a need for state action to provide or secure them as public. If we consider tobacco
smoke as a noxious degrading of clean air, harmful to all, but not consented to by all, then
we can mount an argument for the regulation of smoking that is not based on unconsented
harm to identifiable third parties. Recently public health practitioners have suggested that
not only is smoke-free air a public good, so is tobacco control itself (Smith et al., 2003).
Whatever the merits of the clean air argument, this seems to go too far in confusing means
and end. We do not enjoy tobacco control the way we enjoy clean air or an environment
relatively hazardless to health.
An alternative public good approach would be to say that the good is not the smoke-free
environment, but rather an environment free from inducements to harm oneself or others.
Given that smoking is addictive, we would enjoy an environment that lacks both inducements to smoke and signals that smoking is possible or even desirable (thus protecting nonsmokers from the hazard that they may take it up) and triggers that would provoke
recidivism among recovering smokers. Similarly, we might say that the good we enjoy is
public health itself as we all benefit indirectly from the health of others as well as directly
from our own health. These proposals are more plausible than the clean air as public good
proposal, but it seems to me to be a straining of language to consider this a good, even a
public good. Rather, we might say that we value this sort of environment, and that we
assent to the sort of soft paternalism it involves.
In any case, public good arguments are difficult to apply in practice, partly because we
cannot easily perform trade-offs between public goods, or between public and private
goods. For example, Hockney would perhaps agree to see toleration (of smoking) as a
public good. How do we decide which public good is more important to us?
A consequentialist argument would presumably be relatively easy to construct, either on
general welfare principles, or more narrowly on health-related welfare principles, but
consequentialist arguments are relatively difficult to dovetail with rights-respecting liberalism. And even on strict consequentialist grounds, it is far from clear that a Hockney-type
utility function could not be constructed that traded present pleasure against future ill
health and shortened life span in a way which favoured present pleasure. Psychologists of
risk have indeed shown how people do seem to carry out this trade-off in respect of their
smoking behaviours (Slovic, 2001).
A second debate of importance concerns the availability of tobacco products that are
primarily aimed at making stopping smoking easier. There has been a long-standing debate
over the tobacco industrys proposals to market safer products, from lower tar cigarettes,
to smokeless cigarettes, to alternative ways to consume tobacco (such as Swedish snus).
There has been a tendency in the tobacco control movement to argue that there is no such
thing as a safe tobacco product, although this is not uniformly agreed internal to that
movement (Kozlowski and Edwards, 2005). Recently, new tobacco-derived products
designed as medically supervised substitutes for smoking have been widely adopted, such
as nicotine patches. The current debate concerns whether these should be much more
widely available, as over-the-counter pharmacy products, or subject to lighter taxation than
smoking tobacco. The ethical issues here largely turn on standard concerns in the ethics of
harm reduction: do safer products provide a gateway to riskier products, or slow down
smokers attempts to move to complete cessation, or communicate that message that safe
smoking may be possible? One approach, which may be feasible, is to regulate the market in
tobacco products in a more comprehensive way, in the name of public health and safety
97
(Royal College of Physicians, 2002). This approach would need to be underpinned by an

ethical framework that would justify both the continued manufacture and sale of a harmful
and addictive product range, and the concentration on tobacco in contrast to other similar
harmful products, such as alcohol (Fox, 2005). This is a consistent problem for advocates of
harm reduction: while harm reduction proponents see what they propose as ethically
necessary, they are consistently attacked by tougher minded public health and public policy
people for condoning the vice or wrong they are nominally trying to alleviate and remove
(Royal College of Physicians, 2007).
The third current debate concerns health inequalities. It is well known that smoking
prevalence is higher in poorer socio-economic groups, and that there is a linear relationship
between economic wealth or income and life expectancy and health status, and an inverse
linear relationship between economic wealth or income and smoking prevalence. If we are
concerned not only with individual liberty and harm prevention or reduction, but also with
equality, then the liberal approach I have broadly supported in this chapter is inadequate
unless supplemented with egalitarian principles. This is even more obvious when one notes
the role of direct taxation in tobacco control and the further fact about the inverse care law
in access to health care by poorer people. Whole access to health care can be improved by
committing resources to reversing or removing inequalities in access to health care, the
measures that could be taken to reduce smoking rates in poorer socio-economic groups
will, on the one hand, perhaps diminish inequalities in health but, on the other hand,
impose more severe burdens on just those people as they make (or fail to make) the
transition to being non-smokers. This is an area that merits detailed investigation by
ethicists and policy makers.
Conclusion
As we have seen, tobacco smoking provokes heated and complex debates, and in part
this is because while there is a public health consensus about the need to control tobacco
use, and prevent the very large number of preventable deaths and illness related to its
use, there is not a wider social consensus about seeing tobacco use as primarily a public
health problem. While many democratic states are willing to construct a political and
legislative consensus around controlling smoking in public places, especially as the
prevalence of smoking continues to decline, there is a residual concern, even among
non-smokers, about the moralism that sometimes accompanies tobacco control efforts.
While this chapter has done little more than sketch the main issues, I hope the overall
message that smoking is an important hazard to health, and that controlling it in the
interests of the unconsenting and vulnerable is important is clear. One topic has been
given what I accept is quite inadequate discussion, viz. the nature of smoking as
addictive. In my defence, I have tried to concentrate on what is distinctive to smoking,
and the arguments presented here should be complemented with more general arguments about the regulation of addictive products and practices (including gambling,
drinking, overeating, taking addictive psychoactive substances, and perhaps other activities including the use of pornography, abuse of consumer credit and so on). What is
not clear is the wider issue of how we reconcile the public healths laudable concern with
health with other social interests. And this is a problem that will reappear again and
again as problems such as obesity, teenage pregnancy and workplace stress are framed as
health problems (Cribb, 2005; Rose, 2007).
98
Section 2: Issues
References
Boyle, P., Gray, N., Henningfield, J., Seffrin J.
and Zatonski, W. (eds) (2004) Tobacco:
Science, Policy and Public Health. Oxford:
Cribb, A. (2005) Health and the Good Society:
Setting Healthcare Ethics in Social Context.
Devlin, P. (1965) The Enforcement of Morals.
Fox, B. J. (2005) Framing tobacco control efforts
within an ethical context. Tobacco Control,
14(Suppl. 2), 3844.
Fullinwider, R. K. (2005) On moralism. Journal
of Applied Philosophy, 22(2): 10520.
Hansson, S. O. (2005) Extended antipaternalism.
Journal of Medical Ethics, 31: 97100.
Hart, H. L. A. (1963) Law, Liberty and Morality.
Hockney, D. (2004) Smoking is my choice.
Guardian, 1 June.
Hockney, D. (2006) A letter from David
Hockney. Guardian, 25 February.
Hockney, D. (2007) I smoke for my mental
health. Guardian, 15 May.
Kozlowski, L. T. and Edwards, B. Q. (2005) Not
safe is not enough: smokers have a right to
know more than that there is no safe tobacco
product. Tobacco Control, 14(Suppl. 2), 37.
Rose, N. (2007) The Politics of Life Itself:
Biomedicine, Power and Subjectivity in the
Twenty-First Century. Princeton, NJ:
Princeton University Press.
Royal College of Physicians (2002) Protecting
Smokers, Saving Lives: The Case for a Tobacco
and Nicotine Regulatory Authority. London:
Royal College of Physicians.
Royal College of Physicians (2005) Going
Smoke-Free: The Medical Case for Clean Air
in the Home, at Work, and in Public Places.
London: Royal College of Physicians.
Smith, R., Beaglehole, R., Woodward, D. and

Drager, N. (eds) (2003) Global Public Goods
for Health: Health Economic and Public
Health Perspectives. New York: Oxford
University Press.
Sunstein, C. R. (1997) Free Markets and
Social Justice. New York: Oxford University
Press.
Wilkinson, S. (1999) Smokers rights to
health care: why the restoration argument
is a moralising wolf in a liberal sheeps
clothing. Journal of Applied Philosophy,
16: 27589.
Further relevant literature

Caan, W. and De Belleroche, J. (eds) (2002)
Drink, Drugs and Dependence: From Science
to Clinical Practice. London: Routledge.
Elster, J. and Skog, O.-J. (eds) (1999) Getting
Hooked: Rationality and Addiction.
Feinberg, J. (19841988) The Moral Limits of the
Criminal Law (4 vols) New York: Oxford
University Press.
Feinberg, J. (1991) Review of Robert Goodin: No
Smoking. Bioethics, 5: 1507.
Fox, B. J. and Katz, J. E. (eds) (2005) Individual
and human rights in tobacco control.
Tobacco Control, 14(Suppl. 2): 149.
Goodin, R. E. (1989) No Smoking: The Ethical
Issues. Chicago, IL: University of Chicago
Press.
Hall, W., Carter, L. and Morley, K. I. (2004)
Neuroscience research on the
addictions: a prospectus for future
ethical and policy analysis. Addiction,
29: 148195.
House of Commons Health Committee (2006)
Smoking in Public Places; First Report of
Session 20052006 (3 vols). London: The
Stationery Office.
Husak, D. N. (1992) Drugs and Rights.
Royal College of Physicians (2007) Harm

Reduction in Nicotine Addiction: Helping
People Who Cant Quit. London: Royal
College of Physicians.
Mill, J. S. (1989) On Liberty and Other Writings,

ed. S. Collini. Cambridge: Cambridge
University Press.
Slovic, P. (ed.) (2001) Smoking: Risk, Perception

and Public Policy. Thousand Oaks, CA: Sage.
Pope, T. M. (2000) Balancing public health

against individual liberty: the ethics of
smoking regulations. University of Pittsburgh

Law Review, 61: 41998.
Royal College of Physicians (2000) Nicotine
Addiction in Britain. London: Royal College
of Physicians.
Scruton, R. (2000) WHO, What and Why?
Transnational Government, Legitimacy and
the World Health Organization. London:
Institute of Economic Affairs.
Skrabanek, P. (1994) The Death of Humane
Medicine and the Rise of Coercive Healthism.
London: Social Affairs Unit.
99
Verweij, M. (2000) Preventive Medicine: Between

Obligation and Aspiration. Amsterdam:
Kluwer Academic.
Verweij, M. (2007) Tobacco discouragement: a
non-paternalistic argument. In Ethics,
Press, pp. 17997.
Wilson, N. and Thomson, G. (2005) Tobacco
taxation and public health: ethical problems,
policy responses. Social Science and Medicine,
61: 64959.
Section 2
Chapter
Issues
Infectious disease control

Marcel Verweij
Introduction
Infectious diseases have been one of the major factors affecting public health in the past,
and will probably remain so in the future. There was a time that physicians thought that, at
some point in the future, the perils of infectious diseases would have been overcome. In the
middle part of the twentieth century, progress was made in controlling diseases such as
smallpox, measles, typhoid and plague. Better diet and hygiene, improved living conditions
and vaccinations all helped to strengthen peoples immunity against several diseases.
Patients who got ill could be treated effectively with antibiotics. Morbidity and mortality
due to several infectious diseases decreased significantly. However, this success story has
had its limits: advances in medicine and public health have mainly benefited the developed
world, whereas infectious diseases in the developing world have remained high; new viruses
such as HIV and SARS have created new problems in both high- and low-income contexts;
extreme multi-drug resistant forms of tuberculosis are spreading, especially among people
living with HIV and AIDS; and, lastly, at the beginning of the twenty-first century, many
countries and the World Health Organization started preparations for an influenza pandemic that could in the future be reminiscent of the 1918 Spanish Flu.
Medical treatment of patients with clinical symptoms is only one and probably not the
most important way to control infectious diseases. The spread of disease can be reduced
most effectively by controlling infection, for example by improving hygiene, by social
distancing or quarantine, by raising immunity, etc. Many of these interventions require
collaborative public action (Verweij and Dawson, 2007). If human contact is a source for
infection, then preventive measures must aim at reducing or even prohibiting such contacts.
However, such measures can have negative if not detrimental consequences for some.
Measures of control may involve excluding some individuals or groups from public life.
This chapter discusses various examples of measures to protect public health against
infectious diseases. The focus will be on contagious disease that spreads from human to
human, because controlling measures against such diseases can have a deep impact on
personal life and well-being. Moreover, such controlling measures are often compulsory
and set constraints on the liberties of individuals or groups. The largest part of the ethical
discussion in this chapter will focus on the justifications for such compulsory interventions.
Protection against animal diseases (including avian influenza), food-borne diseases (for
example, salmonella) and non-contagious diseases (for example, tetanus) also raise moral
questions but these will not be discussed.
100
Chapter 6: Infectious disease control
101
Historic examples
The phrase social distancing as a measure to prevent infection is a relatively new term,
but it has been practised throughout history; as long as it has been known that certain
diseases are contagious. People have always tried to protect themselves and others by
setting patients with particular diseases apart, and excluding them from the rest of
society. People with suspected symptoms of a disease could be treated as outcasts or
outlaws, or sent to a closed institution, such as a leprosarium. History provides numerous
examples that illustrate how such measures could be harmful and unjust to the victims of
disease. Saul Brody describes how in the Middle Ages, leprosy patients or at least
persons who were thought to have this disease were cast out of society. The first step
would be to detect suspected lepers. They were then presented to special committees of
experts with the job of making a diagnosis. This physical examination might be carried
out by doctors and surgeons, or often certain laypersons, such as gate porters, policemen,
priests, monks or even other lepers (Brody, 1974: 634).
There must have been a considerable risk that diseases were misclassified: any skin disease
or abnormality could be seen as a form of leprosy. If a person was judged to be a leper, the
victim was told that he would be separated from the healthy population. The law could be
very hostile to lepers, and for many, the leprosarium might even be a relatively save haven:
In brief, the law could place a person outside of society by depriving him of his rights to marry or
to stay married, and to own and transmit property. It could simply and effectively deprive the leper
of the right to have a home, and that being so, it could compel him to depend upon the very
society which, out of loathing and fear, wrote those laws. Under such circumstances, the best the
leper could do would be to turn from the world and enter the closed society of the leprosarium.
There, at least, he would have a bed and food. The prison could also be a refuge.
(Brody, 1974: 86)
This may sound like ending the social life of a person, and actually it was meant to be so: the
diagnosis of leprosy could be a reason for authorities to officially declare that the victim was
now judged socially dead. The church played an important role, providing symbols and
rituals that were similar to those of a funeral. A mass was held for the leper, who at some
places was required to stand in a grave in a cemetery. The officiating priest would throw
some earth from the grave over the victims head, explaining that this symbolized the death
of the leper to the world. As harsh as this may sound, the fear of leprosy sometimes led
some to enact even more drastic measures. For example, Henry II of England, as well as
Philip V of France,
. . . chose to replace the religious service with a simple civil ceremony. It consisted of strapping
the leper to a post and setting him afire. [Edward I of England] adhered a trifle more closely to
the letter of the ecumenical decree. Lepers, during his reign, were permitted the comforts of a
Christian funeral. They were led down to the cemetery and buried alive.
(Rouech, 1953: 117)
Leprosy is a devastating and mutilating disease, which explains the phobia-like responses to
lepers throughout history. However, despite this fear, leprosy was not a disease that
disrupted cities or the whole of society like some epidemics of other infectious diseases.
For example, the various pandemics of bubonic plague could easily destroy city life in a
short time, as the panic accompanying such disasters must have been immense. The Great
Plague of 1665 killed almost one-third of the inhabitants of London who did not flee
102
Section 2: Issues
(Moote and Moote, 2004). Whole families passed away. People would become ill and die on
the same day. Inevitably, such panic could lead to drastic measures. Houses of patients
could be sealed until the victims either recovered or died. Guards were posted at the door to
see that no one got out. The guard had to be bribed to allow any food to pass to the inmates.
If patients were not isolated at home but sent to a plague house, they did not only face the
dangers of the plague itself, as unconscious patients were easily confused for those who were
deceased and taken to the cemetery. Although some authorities advised against burying
corpses for 24 hours after death, the fear of infection stimulated caregivers to bury victims
as soon as possible (Noordegraaf and Valk, 1988: 104). If a patient recovered from the
disease, he or she could still be considered dangerous. Like other suspected persons, they
were often required to walk with a white stick, thus motivating others to avoid them
(Noordegraaf and Valk, 1988: 103).
In order to prevent spread of the plague in 1347, the city of Venice and other cities
proclaimed quarantine measures for ships coming from infected areas. Such ships and their
crew were isolated initially for 30, later for 40 days (quaranti giorni). People who did not
develop symptoms of the disease were released after this period and allowed to enter the city
(Biraben, 1976: 1735). The quarantine rules thus went beyond isolating patients from
healthy persons: all persons on a ship coming from a suspect area were sent to quarantine.
In theory, these measures could protect cities against incoming disease, but they involved
clear risks for those held in quarantine: if someone among them became ill, all ran the risk
of being infected. Some may have tried to escape from quarantine, but they would then face
other risks: violations of the quarantine orders could be punished with the death penalty.
Another classic example of isolation and restriction of liberties is the story of Typhoid
Mary (Porter, 1997: 4245; Wald, 1997). Between 1900 and 1907 Mary Mallon was a cook
in New York. Twenty-two people for whom she cooked or cared for developed typhoid
fever. Mallon remained healthy, but she was suspected and later identified to be a carrier of
the disease. Understandably, she could not grasp that she had caused all the cases of disease,
not having been ill herself. Mary Mallon was isolated against her will in a hospital on North
Brother Island in 1907. In 1910 the public health authorities released her, on condition that
she would refrain from taking up her job as a cook. Unfortunately, in 1915 she did return to
cooking, accepting a job in the Sloane Hospital for Women. Twenty-five people developed
typhoid fever and two of them died. Again Mallon was identified as the vector, and this time
she was sent into quarantine for life. She died in 1938, after having been quarantined for
26 years of her life.
Mary Mallons case is especially interesting from an ethical perspective, because she was
not only restricted in her liberties (apparently for good reasons) but also held personally
responsible for the spread of disease. Public health officials possibly contributed to this
accusation, naming her Typhoid Mary in their medical publications. Later newspaper
articles showed cartoons of Mallon adding small skulls in a frying pan (The New York
American, 20 June 1909). For the public it might have been difficult not to see her as an evil
cook who poisoned her employers and clients.
Measures of infectious disease control

The life of Mary Mallon illustrates how public health authorities can have the power
to restrict peoples individual liberties. Effective infectious disease control may require
a large number of coordinated measures, from surveillance to notification, restricting
103
traffic to mandatory isolation or even compulsory medical treatment. Let me review

various measures briefly, with an eye on the values and rights that are at stake. As
will become clear, the impact of such public health interventions on individual well
being and freedom can be immense. After this brief inventory of various measures and
their implications, I will discuss the moral justification of restrictions to freedom
in more detail.
Surveillance, screening and notification

Most preventive measures will only be effective if enacted in a timely manner. A good
surveillance mechanism is therefore indispensable. This involves acquiring data about
infectious diseases including personal information about individual human beings. In
most, if not all countries, physicians and medical laboratories are legally required to
notify public health authorities if they find cases of specific infectious diseases. For
several diseases an anonymous report may be sufficient, but for many others it will be
necessary to mention the diagnosis together with the patients name and address. This is
in contrast to the general requirement to maintain confidentiality, central to Hippocratic
as well as modern medicine.
However, such notification procedures for identified and diagnosed cases will not be
sufficient as a means of surveillance within a population context. Surveillance and control
of infection may be more effective if healthy people are tested on a large scale. For
example, in the Netherlands and many other countries, pregnant women are tested for
HIV and Hepatitis B on a routine basis. Similar practices exist for other diseases,
including tuberculosis. In some contexts people may have the right to opt out if they
do not want to be tested, but in other contexts or countries it may be practically difficult
or even legally impossible to refuse a test. Hence such interventions result in restrictions
to the rights to autonomy and bodily integrity, and moral rules that govern normal
medical practice, such as voluntary informed consent, on the grounds of protecting the
public health. Other ethical issues arise if these tests and their results can have negative
effects on the well-being of the persons involved. For example, a positive HIV-test might
make it difficult for someone to acquire health care insurance or life insurance; it could
lead to stigmatization or even social rejection. These disadvantages may sometimes
outweigh possible benefits of the test result, such as timely access to antiviral treatment.
Routine test policies especially raise ethical issues in low-income contexts, where there
may be insufficient access to health care. In such a case, test subjects may be informed that
they have a disease, but left without the treatment they need. This may for example
happen in countries in which HIV is endemic: routine testing for HIV exists, but
antiretroviral treatment may not be available for those who test positive (Rennie and
Behets, 2006). If there are many reasons for individuals to forego testing, then routine
tests in a health care setting may also have drawbacks from a public health perspective.
People who feel ill may decide to abstain from seeking medical care or postpone their visit
to a clinic. This is not only undesirable for the person herself: if the person is pregnant, or
if she suffers from an infectious disease, then it might be harmful to others if she does not
seek medical care in time.
Screening procedures can also be applied at airports, seaports or other places open to
international traffic. During the SARS outbreak in 2003, countries required inbound and
outbound airport passengers to have their bodily temperature taken, with the aim of
104
Section 2: Issues
detecting those infected and so stop further spread of the disease. One of the problems
for any screening programme is that tests and procedures can be imprecise. Test results
can therefore be unreliable as an indication of disease. Where there is a case of an
alarming outbreak of a highly infectious disease, screening of travellers should not miss
cases. However, if tests are made very sensitive hence minimizing the risk that cases
are missed (false-negatives) this will inevitably lead to large numbers of false-positive
test results. Consequently, many passengers will need to undergo further examination
and tests before they can be allowed to continue on their journey.
Measures after cases have been found: contact-tracing

and further testing
If a case of a contagious disease has been confirmed there may be reason to trace people
who have been in contact with the index-patient. The necessity and urgency of contacttracing, as well as the way to do it, will differ for various diseases. A disease like
tuberculosis spreads easily in the right conditions. Contact-tracing may involve mobilizing whole communities to test family members, neighbours, colleagues or classmates;
or, alternatively, warning the public that everyone who has been at a specific place at a
specific time (for example, a shopping centre or cinema) should report to the health
authorities and have themselves tested. One problem with such a strategy is that it may
raise panic and fear. In some cases however, the commotion and concerns, or even fear,
that are part of panic, may be justified. More important moral issues occur if such panic
arises where the name of the index-patient is known to the public. If they are in fear of
infection, people may easily confuse natural causes and transmission with culpability
and responsibility, thereby stigmatizing and possibly harming the initial victim of the
disease. Contact-tracing and screening programmes should therefore protect privacy
where possible.
Somewhat different problems of privacy and confidentiality arise if public health
professionals try to trace and notify contacts of people with a sexually transmitted
disease. The natural way to do this is to ask the index-patient to contact his or her sex
partners, as it will be difficult for public health nurses to trace contacts without the
index-patients cooperation. However, for most people, sex is a private matter, and if a
patient does not want to cooperate, or contact the partners him/herself, further
attempts to trace contacts will be futile. In some cases however, a physician or public
health professional may have good reason to consider traceable individuals to be at risk.
One typical example is where a physician is treating a person with syphilis and knows
that the patients spouse is not aware of the diagnosis. On the one hand, the physician
could protect the spouse by warning them about the patients condition, but, on the
other hand, the physician has an obligation of confidentiality towards their patient.
Some would argue that a patient who deceives their spouse and refuses to warn them
about the risk they run, forfeits all rights to confidentiality. However, for public health
physicians working on sexually transmitted diseases, the importance of confidentiality
goes beyond the right of the individual patient. For infectious disease control to be
effective, the public should be able to trust public health authorities to keep their
personal health information confidential. This holds especially for prevention and
treatment of sexually transmitted diseases. STD clinics will be most effective if patients
can trust that their private details and stories are held confidential. If confidentiality is
105
not respected in public health, many patients with infectious diseases may postpone or
even forego visiting a STD clinic. Confidentiality has its limits, but partner notification
without consent, as a measure of infectious disease control, may only be justified in
exceptional cases.1
Public health measures: reducing contacts, quarantine and isolation

An important part of infectious disease control is to prevent further spread of disease
during an outbreak. This is especially true for respiratory infections like influenza, where
coughing, sneezing or talking are important means of transmission, as it makes sense to
limit the number of contacts between healthy people, and to avoid large group meetings or
other risk situations. In an outbreak, public health authorities may decide to cancel fairs,
football matches, to close schools, crches, companies or shops. Again this will be a clear
but sometimes necessary restriction of liberties for many people, but moreover, these
measures will have significant impact on their daily life, as well as on their financial
resources or their property. For example, the closing of factories, offices and shops and
cancelling of mass-events will involve financial burdens for companies and citizens.
Employees may run the risk of losing their job if their place of work is closed for a long
time. Such burdens will be faced particularly severely by vulnerable groups who are socialeconomically worse-off, for example employees with temporary jobs or no contracts at all.
If certain groups of people run risks and bear significant extra costs due to public health
measures, the question arises whether they would have a justified claim to receiving
compensation for their losses. Government-based compensation schemes can be considered
as an expression of solidarity: if there is a public health threat then all citizens should be
willing to share in the costs of protective measures after all, if such measures are effective
then everyone will benefit.
Compensation schemes are also an important issue for containment of animal diseases
such as avian influenza. Livestock have, throughout time, played an important role in the
spread and transmission of many human infectious diseases. Nowadays, outbreaks in
livestock are often countered by mass culling of all animals that have either been exposed
to the disease or even all animals within a particular distance of confirmed cases. Compensation schemes may not only be a matter of justice and solidarity; they can also serve a
pragmatic role in maintaining surveillance. Few farmers may be willing to report disease in
their livestock early on in an outbreak, if they face having their animals culled without
compensation. Compensation schemes may thus support surveillance programmes, which
are essential to any timely and effective response to an outbreak of infectious diseases.
From an ethical point of view, the most appalling public health measures are quarantine
and isolation of persons. Both concepts are used for the same sort of measures, but it makes
sense to distinguish them in a way that is morally relevant. Gostin defines isolation as the
physical separation and confinement of an individual or group of individuals who are
known to be infected with a contagious disease from non-isolated individuals, to prevent or
1
Such a breach of confidentiality may be justified more often in the practice of individual health care.
One such case would be where a health care provider, for example, a family physician, is caring for a
patient who has a sexually transmitted disease and who refuses to tell their spouse, and where the
physician also has a caring relationship with the spouse. The patients refusal would make it
impossible for the physician to sustain a fiduciary relationship with their other client.
106
Section 2: Issues
limit the transmission of the disease to non-isolated individuals (Gostin, 2000: 210).
Quarantine also involves separation, but this applies to healthy individuals or groups who
may have been exposed to a contagious or possibly contagious disease. Both isolation and
quarantine measures can be applied to large groups. As was illustrated in the historical
examples given above, extreme forms of isolation or quarantine may be similar to putting
individuals in jail, excluding them from public life completely possibly for the rest of their
lives. Even if the means of separations are less severe, the impact on personal life may be
overwhelming, as it combines most of the adverse events that arise in all other public health
measures. Isolation and quarantine effectively make it impossible for individuals to continue their lives as planned, to fulfil their jobs and responsibilities, to earn their living, to see
and care for their loved ones. The separation of persons also has an important symbolic
dimension. Individuals or groups are labelled as dangerous, which could undermine their
sense of being part of a community. Being separated from the community, there is a risk
that isolated and quarantined groups will not have sufficient access to such basic needs as
food or health care. In short, all quarantined and isolated individuals are deprived of at least
some essential sources for well-being. Moreover, quarantine measures may mean that all
suspected persons are held together: this includes persons who are in fact exposed to the
disease and may get ill in the short term, as well as those who are only believed to be
exposed but are in fact not infected. The non-infected persons may be detained with people
who may infect them. In this way, quarantine procedures, while intended to reduce the risks
of contagion within the larger population, may actually increase the risk for (at least part of)
the quarantined population.
It is clear that isolation and quarantine procedures can have extremely adverse implications for individuals and that procedures should be applied with due care. In the last few
decades, much work has been done to develop procedures and regulations in such a way
that risks for quarantined persons are minimized, and that they at least have rights to due
process that protect citizens from arbitrary detention. However, even if these measures are
applied with due care, they remain morally problematic and require a strong moral
justification. Before turning to this issue, let me briefly discuss a last category of measures
of infectious disease control.
Mandatory medical treatment

Surveillance, routine testing and quarantine procedures may only have limited effects if
infected patients are not treated against their disease. Medical treatment of infectious
diseases is not only beneficial (if not necessary) for the patient himself, it will also be
essential to prevent further infections of others. Hence, the availability of, and compliance
with, antibiotic or antiviral treatment is not only an issue of individual health care but of
public health. One of the most frequent contexts for compulsory treatment is tuberculosis
control. Antibiotic treatment for tuberculosis may take up to 6 months in normal cases,
and 24 months in cases of multi-drug resistant forms of tuberculosis. Completion of
therapy is not only necessary to result in a cure, and to take away the risk of infection,
but also to prevent the development of drug resistant forms of tuberculosis. However, many
patients may decide to stop treatment as soon as they feel better. Adherence to therapy may
be particularly difficult for persons who belong to marginalized groups such as the
homeless or drug abusers. Hence, various strategies have been developed to promote
adherence to therapy. Apart from voluntary schemes, there are also compulsory options.
107
A common approach to ensure adherence to treatment is directly observed therapy (DOT):

patients are monitored to ensure they take their drugs, at least for a certain period. DOT as
such may not be compulsory, although it does impose restrictions upon a patients freedom.
Yet if this approach is not successful, for example, because a patient does not come to the
physician to receive treatment, public health officials may consider compulsory detention
for the duration of the treatment (Lerner, 1999). The combination of detention with
compulsory treatment is a step further than just isolation of patients, as it involves a
violation of bodily integrity and the right to refuse treatment rights that are central in
modern health care practices in Western societies. However, the right to refuse treatment
and the rules of informed consent may be primarily relevant for medical decisions about
patient care, where the benefits and harms of treatment and non-treatment primarily
concern the patient herself. If patients with infectious diseases neglect the treatment they
need, this could have harmful implications for others, and that may be reason to overrule
the requirements of informed consent.
Just as in the case of isolation and quarantine, it is reasonable to install requirements of
due process, in order to protect patients from unnecessary restriction upon their freedom.
Diseases like tuberculosis are often most prevalent among the marginalized in society, and
detention (sometimes even in gaol) may reflect social prejudice rather than just public
health concerns. Public health officials should not simply assume that patients who belong
to certain risk groups (for example, intravenous drug abusers) can only be treated if they are
detained.
Well-being and freedom

All these measures for infectious disease control involve a tension between the importance
of protecting public health, and the liberty and interests of individual people (patients,
infected persons, possibly infected persons) whose freedom may be curtailed in the name of
public health. Let me review in more detail the main values at stake, well-being and
freedom, before discussing possible ways to justify compulsory measures.
First of all, most interventions, especially when compulsory, curtail the freedom to
move, to travel, to meet persons one wants to meet, etc. Although such liberties are of
intrinsic value themselves, restrictions will especially be problematic where they deprive
people of certain important sources of well-being. For example, restrictions upon travelling may effectively make it impossible for people to go to work, to earn their living, or to
sustain their business. The financial burdens of such forms of control on individuals and
companies may be immense. The losses due to public health constraints, however, go far
beyond economical and financial burdens. Quarantine measures and travel restrictions
may prevent people from seeing their loved ones, or caring for their children, parents or
friends. Many people who are less independent may face significant risks to their health if
caregivers cannot come and help them. The possibly infected people held in quarantine
with others may run a significantly higher risk of becoming a victim of the disease because
they are held together with others who are in fact infectious. Moreover, if certain groups
of people are set aside because they are considered to be a (potential) risk to the rest of
the community, this may have an enormous impact on how they are viewed by that
community. From a moral point of view these groups are innocent; from a medical point
of view they may be a source of infection hence potentially dangerous. In the process
of stigmatization that goes with outbreaks of contagious diseases, medical and moral
108
Section 2: Issues
judgments may get easily confused. As a result, diagnosed patients or suspected persons
may be systematically avoided and have their interests neglected. In short, while public
health measures, from screening to quarantine, are aimed at protecting the health and
welfare of the public, such measures inevitably will have negative if not detrimental effects
on the well-being of certain sub-groups.
Second, even apart from the impact on human well-being, the constraints on liberty and
individual rights are morally controversial as such. Measures of infectious disease control
may force people to behave in specific ways, hence curtailing their freedom of choice and
movement. Surveillance and notification policies may be considered a violation of rights to
privacy and confidentiality; and compulsory vaccinations and treatment violate moral
rights to bodily integrity. A central idea behind these values is that individual persons are
capable of reasoning, making choices and determining the course of their lives and that
these capacities are grounds for respecting persons and the choices they make. In a liberal
society where individual rights are considered of utmost importance, the possibilities to
curtail or overrule those rights should therefore be limited. One should treat persons with
respect: that is, treat them as ends in themselves, and not merely as means to or obstructing
factors for the realization of other ends however worthwhile aims such as the protection
of health might be. They are autonomous beings, that is, beings with practical reason, who
can and do set their own ends. Rules of informed consent as common as they are in
individual health care are often seen as especially important in protecting autonomy. The
ideal route for infectious diseases control is therefore to inform persons about necessary
precautions (to have themselves screened, to isolate themselves, to accept treatment, etc.)
and to trust that they will act accordingly. However, the circumstances of infectious diseases
control, and especially outbreak management, are far from ideal. Infectious diseases may
raise panic, and people may distrust government institutions, and therefore refuse to
cooperate. Moreover, while measures of infectious disease control are aiming at protecting
the health of the many, they will often impose risks on individual persons. Hence, for these
individuals it could be most rational to refuse to cooperate, and to avoid tests, quarantine or
vaccination. In such circumstances, compulsory measures may be inevitable to prevent
spread of disease. Which moral grounds could provide justification for such infringements
on individual rights and freedom?
The harm principle

Even ethical theories that see individual rights and especially liberty rights as the strongest
possible moral claims as trumps vis--vis other moral considerations such as the common
good will necessarily acknowledge limits to those rights. If people use their freedom to
harm others or otherwise impose their will on another person, then the freedom of the
aggressor is undermining the freedom of others. Any liberal framework that endorses
freedom or autonomy as an important value will therefore accept that someones freedom
must be curtailed if this is necessary to protect the freedom of other persons. Ideally every
individual enjoys maximum freedom that is still consistent with the equal liberties of her
fellow citizens. According to John Stuart Mill (1859: 22) the only purpose for which
someones freedom can be constrained is to protect other individuals. One should not
compel persons to act in specific ways for their own good, but it is justified to limit their
freedom in order to prevent evil to third parties. This idea is normally referred to as the
harm principle:
109
HP1: It is justified to restrict the liberty of person A in order to prevent A from causing harm
to person B.2
Moreover, according to Mill, this is the only justification of liberty-limiting actions and
policies. Even the strongest libertarian will agree with liberty-limiting interventions that
protect against harm, and HP1 is therefore relatively uncontroversial. Moreover, the
application of the principle to protection against the harms of infection may seem rather
straightforward. The principle and its application however do raise important questions.
First, the concept of causing harm is not as clear as one might hope certainly not in
relation to the harmful nature of contagion. Second, the harm principle as understood in
HP1 can be given a very strong justification as it is in a sense a necessary condition to
freedom as such, but this is a rather narrow interpretation of Mills harm principle.
Although HP1 may apply to certain measures of infectious diseases control, many forms
of infection (hence policies to prevent infection) fall outside the scope of HP1. Many
authors in fact endorse a broader interpretation of the harm principle, but such an
expanded principle HP2 (see below) may require other sources of justification.
Causing harm
A paradigmatic case of harm, where the harm principle would obviously apply, is one where
someone deliberately intends to infect another with a dangerous disease. A recent example
is the Groningen case in the Netherlands. In 2007, three people in Groningen were arrested
because they had injected others with HIV-infected blood. The culprits all HIV seropositive themselves contacted their victims through internet chat boxes, and invited them to
homosexual sex parties at home. During these meetings they first intoxicated their victims,
and then administered intravenous injections with HIV-infected blood. They also injected
blood in each others veins, just for the kick as they confessed after their arrest.
Although persons with HIV can be treated relatively well (at least in high income
countries), the infection does severely undermine their health. Adequate medical treatment
involves a complex regimen of drugs that negatively affect their quality of life. A deliberate
and malevolent act to infect someone else with HIV is a clear case of harm. State interventions to prevent such harms and to punish evildoers are justified on the basis of the harm
principle.
The Groningen case may be a paradigmatic case of inflicting harm, but it is not a
paradigm case of infectious disease control. Public health measures to control contagious
diseases do not focus on active malevolent behaviour, but on more common means of
infection. For example, even though persons may be aware that they are a carrier of a virus
or other infectious agent, they may simply forget to take precautions, or otherwise act
negligently and spread disease. The story of Typhoid Mary Mallon is a case in point. At
some point she was told she carried a disease that could be easily transmitted if she worked
in a kitchen, but she decided to take up her job after all, hence infecting more persons. As
discussed before, public health measures that demand people isolate themselves, or refrain
from certain activities, or otherwise take precautions, may be detrimental to their own wellbeing, and this may well motivate them to be non-compliant or negligent. The reason why
such measures as compulsory isolation are used is because such behaviour, even though
2
It is not necessary for the principle that B is a specific person. It could be a non-assignable member of
the public.
110
Section 2: Issues
normally not malevolent, can be dangerous to others. If there is a clear risk that infected
persons will harm others, even though the harm is caused by negligent rather than
malevolent behaviour, then the harm principle may justify compulsory measures.
So far we have focused on cases where people have negligently or malevolently imposed
risk on others. Again, these cases cover only part of all current possibilities for compulsory
measures. Infectious diseases may spread where no one is yet aware of the infection.
A disease could have an incubation period where an asymptomatic person is already
infectious. Or persons may have certain symptoms, but not realize that they are infectious.
Many policies including compulsory measures aim to prevent spread of disease in such
contexts: mandatory screening at borders, routine HIV-tests for pregnant persons in risk
groups, prohibition of mass events, closing of schools, etc. Can the concept of harm and the
scope of the harm principle be stretched much further, and justify such measures as well? It
is often assumed in public health law and ethics that the principle can be as broad as this
(Gostin, 2003). Yet this depends partly on how the principle and its core concept, harm, are
understood.
Is it correct to assume that Sally, who does not know she is a hepatitis B carrier, and who
has unprotected sex with Richard, is causing harm to Richard? Is Henry causing harm to his
colleagues if he goes to work, unaware that he is a vector for a new and dangerous influenza
virus? Paul got very ill and was admitted to a hospital, and now it appears that he has SARS
and that he infected two nurses who cared for him. Did Paul harm the nurses?
Maybe Sally is causing harm if we assume that she and Richard should have taken
precautions against sexually transmitted diseases anyway. If that is true, this may reveal that
harming others and even causing harm are normative concepts, that only apply when
certain moral norms are transgressed.3 Joel Feinberg (1984: 36) indeed defines harm as a
wrongful setback to the interests of another person, where wrongful refers to transgression
of a moral norm (Feinberg,4: 36). Certainly Henry can not be considered as harming his
colleagues, if harm would involve the transgression of a moral norm: he just goes to work
and is perfectly justified in doing so. Paul did not do anything, and even though actually
his admittance to the hospital did appear to be a threat, there is no sense in which Paul
could have been responsible for it. This seems to imply that, if we emphasize the moral
nature of causing harm, HP1 only justifies compulsory measures towards persons who are
in some sense responsible for the harm that they produce.5 Compulsory policies towards
persons who are unaware that they pose a threat to the health of others might not be
justified on this interpretation of the harm principle.
The example may provoke the reaction that Richard can not be said to have been harmed if he
agreed to having unprotected sex (volenti non fit injuria), but this only affirms the claim that harm is
a moral concept.
However, Feinberg is specifically focusing on the role of the harm principle in the context of criminal
law, and this may be a reason for thinking that the relevance of his concept of harm in the
context of infectious disease control is limited.
Note that this is not just a result of the definition of harm as such, but also in our understanding of
someone causing harm. Did Henry cause harm to his colleagues, or was the harm caused by the
virus? Statements of causal responsibility may seem purely empirical, but when they concern actions
or omissions of persons, they will inevitably involve evaluative statements that presuppose some
background of norms (Yoder, 2002).
111
Expanding the scope of the harm principle

However, the harm principle could also be interpreted in a less restricted way. In
discussions of the harm principle it is unclear how broad its scope exactly is. Does the
principle only justify compulsion or punishment of person A who is him/herself causing
harm to person B? Mill suggests that the principle may also cover cases where As
omission to act (for example, to rescue) would result in harms for B. In other words,
As actions need not be the source of the harm. If we accept this, in fact we have adopted
an alternative principle:
HP2: It is justified to restrict the liberty of person A in order to prevent harm to others.
(Brink, 2007)
This principle neither requires causal responsibility nor moral responsibility of A for the
harms that are to be prevented. Hence, on this principle also compulsory measures can be
justified towards persons who are unaware that they pose a risk: for example, compulsory
screening of travellers at airports or mandatory HIV tests for pregnant women. A policy by
which all incoming travellers are screened for specific contagious diseases will lower the risk
that outbreaks will occur within a country. Measures that temporarily prohibit the
gathering of large groups of people will inhibit the spread of infection across a population.
This principle could even support compulsory policies towards persons who are known to
be not infected at all, for example vaccination policies of specific groups in order to prepare
for a bioterrorist attack.
As attractive as this interpretation of the harm principle may seem for effective
infectious disease control, it does raise a problem at least for many libertarians.
If it is irrelevant what or who caused the harm that is to be prevented, almost any
compulsory policy that may promote or protect the health of others could be justified.
For example, the state could openly force some persons into quarantine in order to
convince the public that there is a serious threat to public health and that the state
means business, thereby stimulating the public to comply all the better (Wilkinson,
2007). Parents could be required to have their children vaccinated against seasonal
influenza, because that may reduce mortality among the elderly (Reichert et al., 2001).
Citizens could be required to pay for vaccination and treatment of persons who cannot
afford such forms of health care, etc. This is not to suggest that such policies would
necessarily be wrong. However, these policies, and the revised version of the harm
principle, may have little to do with the initial attractiveness of the harm principle
in a liberal framework. The harm principle is often presented as a justification for
compulsion and punishment that even the notorious libertarian could accept. This is
because the principle (at least HP1) is still consistent with allowing every person a
maximum liberty.
The revised version of the harm principle, HP2, however also justifies constraints on
freedom of citizens if, for example, that would protect the health of persons who are
relatively frail anyway. This would raise a discussion about whether such interventions
really aim to prevent harm to those persons, or just confer benefits to them (Brink, 2007).
Libertarians would complain that the revised version of the harm principle is not just aimed
at the protection of liberty, but it is already trading off liberty against health and well-being.
Such a trade-off however may well be acceptable if not essential in many other
normative theories. Let us briefly focus on two such theories.
112
Section 2: Issues
Consequentialist and egalitarian approaches to preventing harm

Contrary to what libertarians assume, it makes sense to hold liberty to be one value
among others, and to reject the idea that freedom from intervention should always be
given priority over other values, such as welfare or equality. Public health measures aim
to protect people from diseases that are detrimental to their health and thereby undermine their opportunities to live a reasonably good life. Consequentialists will argue
that it may be justified to constrain peoples freedom if this would promote the general
good. Infectious diseases may be detrimental to many different sources of well-being;
consequently, effective prevention may promote well-being in various ways. Adequate
protection against contagious diseases first of all promotes the good by preventing
illness, suffering and premature death. Moreover, public health measures, such as
surveillance, screening, testing and vaccination, and adequate structures to respond to
disease outbreaks will also protect and facilitate peoples societal life: to meet others, to
go work, to receive education and to care for the elderly or the weak. If such measures
require compulsion, this will be justified if the consequences are optimal. From a
consequentialist perspective, the causal routes of the harms are rather irrelevant, and
there are good reasons to prefer HP2 over HP1. Moreover, consequentialists may also
acknowledge that compulsion can help to avoid a collective action problem. Suppose
there is an actual threat of a smallpox outbreak, and it will be possible to contain the
outbreak only if sufficient individuals will be vaccinated. The vaccine however is not
completely safe. In such a case, individuals may forego vaccination and hope that they
will benefit from the vaccination of other persons. However, if many people share this
attitude, the public will not be protected at all, and hence all will be worse off. In order
to avoid such problems, the state could compel the relevant groups to accept vaccination, for the sake of protection of all.
Liberal egalitarians, like consequentialists, may also support HP2; though partly for
different reasons. They see equality and fairness, next to autonomy, as fundamental values.
Measures aimed at the prevention of infection can be considered as measures to reduce the
risks of infection (hence the risk that harm will occur) within a population. Egalitarians will
be concerned with a fair distribution of risks to health, but also with a fair distribution of
the burdens of preventive measures. Health inequalities that are unjust, for example because
they are caused by inequitable income distributions, need to be reduced. Infectious diseases
may hit the rich as severely as the poor, but often there are clear correlations between social
deprivation and increased prevalence of infectious diseases. Well-off groups will also have
more possibilities to protect themselves against disease: they have better access to relevant
information; they can afford special vaccinations, safe drinking water and hygienic living
conditions; they may avoid places where infectious diseases are more common, decide to
leave the city (or even their job) in case of an outbreak, and can afford therapeutic care
when they get ill, etc. If it is left up to individual citizens to protect themselves, then
arguably many people in socio-economically worse-off groups will be left unprotected.
Egalitarians therefore support government interventions to protect the health of the public
which go beyond the narrow version of the harm principle. Such interventions will at least
aim at universal access to basic health care, paid from public resources, which includes basic
preventive care and public health services (Daniels, 2001). Compulsory measures to reduce
the risks of infectious diseases may also benefit the worse-off groups, and ensure that the
burdens of protection are distributed equitably.
113
Although consequentialists and egalitarians may both support the expanded version of
the harm principle (HP2), there will also be differences. HP2 fits within a general liberal
framework which the consequentialist may not accept. HP2 still focuses on harm to others
and therefore does not cover paternalistic interventions that restrict someones freedom in
order to prevent harm to him or herself. For the consequentialist, all harms, including selfimposed risks, could be a reason for intervention. Moreover, consequentialists may argue
that in the context of infection it will often be difficult to distinguish acts that are purely
self-regarding. Most contagious diseases do not stick to persons who accepted the risk of
being infected; they will spread from them to others as well.
Liberal egalitarians might respond that at least in some contexts, individuals have the
possibility and responsibility to protect themselves against infection. Sexually transmitted diseases are a good example. Compulsory measures to prevent infection might be
justified where individuals cannot be reasonably expected to protect themselves. If
individuals can easily choose to have protected sex, public health authorities should be
reluctant to intervene and compel them to take precautions. For that matter, compulsory
measures would require interventions in peoples private spheres interventions which,
if not impossible, would at least be highly impractical. Such pragmatic considerations
will also be relevant to the consequentialist. Moreover, many consequentialists may aim
at promoting well-being and freedom, and for them HP2 will be a reasonable trade-off
between both values.
Conditions for justified compulsion in public health

Let me briefly summarize the argument so far. The narrow version of the harm principle
(HP1) may offer justification for compulsory interventions that prevent people from
deliberately or negligently infecting others, for example, infected persons who resist isolation or treatment. This justification is relatively uncontroversial as it can be based upon
the value of freedom itself. For purposes of effective infectious disease control however, the
scope of HP1 is rather limited. Quarantining healthy persons in order to reduce the risk of
contagion, or mandatory screening at borders, or compulsory tests for pregnant women, all
involve situations where it is less clear whether the persons compelled are indeed causing
harm to others or even imposing risk on others. Such measures may fall within the scope
of the expanded version of the harm principle (HP2), because here the causal routes of harm
and the moral connotations of our understanding of causing harm are irrelevant. The
expanded principle can be coherent with several normative theories, notably consequentialism and egalitarianism. However, the scope of the principle may easily become too broad:
the principle could support sacrificing innocent bystanders, if that would help in some way
to reduce even minor risks within a population. Therefore, a solid justification for compulsory measures in terms of the prevention of harm should take into account several further
constraints. Just the fact that a certain compulsory intervention may prevent harm is not
enough to consider the intervention morally right.
Magnitude of harm
A first issue that will be relevant in any justification of compulsory measures is that the
harm or risk to be averted should be significant and realistic. Chicken pox and the
common cold are contagious, but, apart from special circumstances, do not create clear
risks to individuals or the population. Symptoms of chicken pox may be very annoying
114
Section 2: Issues
for patients but it is questionable whether one should consider these as harms at all. Many
other contagious diseases however have irreversible effects on the well-being of patients,
or are otherwise mutilating or lethal. Therapeutic possibilities are limited which further
decreases the prognosis for infected patients, as in the case of extremely resistant forms of
tuberculosis. The magnitude of the harms that can be prevented does not only depend on
how severe a disease is for a patient, but also on the mode and speed of transmission.
Infections that lead to severe disease but that also spread rapidly within a population pose
a serious threat to the public health at large. Such a disease, if not averted, will affect many
individuals, and the ravages will not only be visible in each patient, but also on a
population level.
Effectiveness and evidence

A necessary condition for any compulsory public health measure to be justified is that it
should be effective in preventing infection and reducing the risks of disease. This condition
may seem rather obvious: the rationale for controlling measures is indeed to control
and prevent disease, and interventions that do not work make no sense. However, it is
often difficult to evaluate and assess the effects of public health strategies. Evidence is often
historical or anecdotal, supported by common sense. Randomized trials of policies like
surveillance, isolation, or border control are rare if not impossible. Moreover, evaluation of
effectiveness should not just focus on theoretical effects in ideal circumstances, but also on
the feasibility of measures in times of crisis. Especially where measures are compulsory,
people may try to find ways to avoid them. If persons who are tested positive for disease are
isolated from their families, they may decide not to see a doctor if they have symptoms.
Finally, judgments of effectiveness of planned interventions against future diseases, such as
pandemic influenza, are complicated by many uncertainties about the characteristics of the
virus and its effects (infection rate, transmissibility, mortality, risk groups, etc.).
These problems make clear that assessment of effectiveness is often not a simple matter
of appealing to scientific evidence; it requires judgment on the basis of incomplete evidence.
Proportionality and the least restrictive alternative

A further requirement for controlling measures is that their impact on the liberties and
well-being of individuals should be in the right proportion to the magnitude of harm that
can be prevented. Major threats to public health for example, a bioterrorist attack with
smallpox may require interventions like compulsory isolation and vaccination, yet such
measures may be effective but not reasonable in order to contain an outbreak of less
severe diseases, like rubella or measles.
Reasonable approaches to protect public health will also endorse the principle of the
least restrictive alternative: measures may sometimes justifiably impose constraints on civil
liberties, but such constraints should not be greater than strictly necessary. Mandatory HIV
tests for all persons who visit a hospital may not be justified if routine (opting-out) tests are
known to be equally effective. Detention and forced treatment of tuberculosis is wrong if
directly observed treatment without detention is also possible and effective. In many cases
however, it is less clear what the least restrictive alternative would be. Especially where there
is much uncertainty about the risks of harm or about the effectiveness of various options,
any intervention could be considered to be the least restrictive means necessary. One
interpretation of the principle is that compulsory interventions should only be imposed
115
where less restrictive measures have been tried, and have failed (Annas, 2002). This may be
applicable to treatment of tuberculosis patients, where detention of a patient could be an
option after more voluntary approaches to enhance compliance to therapy have been tried
and failed. However, in emergency situations and public health crises there may be few
possibilities to try voluntary approaches first. The least restrictive alternative therefore
makes sense as a general principle that urges public health authorities to compare options
and reflect on which measures are really necessary, but it is not a criterion that can be
simply applied.
Infectious disease control and public trust

Infectious disease control involves a range of measures that are necessary for the protection
of public health, but which may interfere with the interests and rights of individual persons.
Moreover, for most policies to be effective it is important that members of the public
cooperate, even if that might not be in their best interests. But why would individuals
voluntarily cooperate if they run the risk that their privacy is violated (case reporting;
contact tracing), that they may not be allowed to enter or leave a country (airport
screening); or that they will lose their job because they cannot go to work (isolation)?
One answer could be that they have moral obligations to take precautions against infection
and act in the interest of public health (Verweij, 2005). Yet even if citizens acknowledge that
they have such a duty, they may not be motivated to act according to public health
regulations if they do not fully trust that such regulations actually are for the common
good, that policies are applied fairly and that adverse effects on well being are minimized.
Hence, it is essential for effective infectious disease control that the public will trust public
health authorities and have confidence in the interventions that those authorities impose.
If there is a lack of trust in public health authorities, even compulsory policies will be futile.
Conversely, if public health professionals and authorities can and do publicly justify
regulations that are deemed necessary especially those regulations that are compulsory
in nature they provide reasons for the public to trust such interventions.
Compulsory interventions should also be applied in a fair and non-discriminatory way;
procedural due process for persons who are separated is then important. Public trust may
be further strengthened if adverse effects of controlling public health measures are countervailed where possible (stigmatization, social isolation, health risks, economic losses, etc.),
and if persons can expect reasonable compensation for economic losses they have as a result
of such interventions.6 Unfortunately, the possibilities for due process and full compensation schemes may be limited in large-scale public health emergencies, such as a bioterrorist attack or influenza pandemic. Emergency situations where many people are
quarantined may not allow time for personal hearings and legal representation for each
individual. And if a large part of the population faces economic losses due to public health
measures, it may be impossible to compensate everyone. One might hope that citizens will
accept the limitations of due process and compensation in a large-scale crisis if they have
confidence in the public health measures.
6
In fact, trust cannot be easily created or promoted. Meijboom et al. (2006: 4324) argue that,
ultimately, only trustworthiness strengthens trust. Yet if public health authorities publicly justify the
measures they take, and if they are accountable for the implications for citizens (including fair
compensation schemes, etc.), they do become more trustworthy.
116
Section 2: Issues
Compulsory screening for dangerous infectious diseases, contact tracing, case reporting,
isolation and quarantine, vaccination or treatment all these measures can deeply interfere
with peoples well-being and freedom, and they require a strong moral justification. There
should be sufficient evidence that a coercive measure is necessary to prevent significant
harm to others. Paradoxically, the more this measure and its justification is endorsed by the
public, the less force and compulsion may be required to protect public health.
Acknowledgement
I have had some very illuminating discussions with Franck Meijboom, Angus Dawson and
Frans Brom that have helped me in developing the argument in this chapter.
References
Annas, G. J. (2002) Bioterrorism, public health,
and civil liberties. New England Journal of
Medicine, 346(17): 133742.
Biraben, J. (1976) Les hommes et la peste en
France et dans les pays Europens et
mditerranes. Tome II: Les hommes face la
peste. Paris: Mouton.
Brink, D. (2007) Mills moral and political
philosophy. In The Stanford Encyclopedia of
Philosophy. Available at: http://plato.
stanford.edu/entries/mill-moral-political/
Brody, S. N. (1974) The Disease of the Soul:
Leprosy in Medieval Literature. Ithaca, NY:
Cornell University Press.
Daniels, N. (2001) Justice, health, and health care.
American Journal of Bioethics, 1(2): 216.
Feinberg, J. (1984) Harm to Others. The Moral
Limits of the Criminal Law, Part I. New York:
Gostin, L. O. (2000) Public Health Law. Power,
Duty, Restraint. Berkeley, CA: University of
California Press.
Gostin, L. O. (2003) When terrorism threatens
health: how far are limitations on personal
and economic liberties justified? Florida Law
Review, 55: 1105.
Gostin, L.O. (2006) Public health strategies for
pandemic influenza: ethics and the law.
Journal of American Medical Association,
295: 17004.
Lerner, B. H. (1999) Catching patients:
tuberculosis and detention in the 1990s.
Chest, 115: 23641.
Meijboom, F. L. B., Visak, T. and Brom, F. W. A.
(2006) Public health strategies for pandemic
influenza: ethics and the law. Journal of

Agricultural and Environmental Ethics, 19(5):
42742.
Mill, J. S. (1859) On Liberty. London: John
Parker and Son.
Moote, A. L. and Moote, D. C. (2004) The Great
Plague: The Story of Londons Most Deadly
Year. Baltimore, MD: The Johns Hopkins
University Press.
Noordegraaf, L. and Valk, G. (1988) De Gave
Gods. De pest in Holland vanaf de late
middeleeuwen. Bergen: Octavo.
Porter, R. (1997) The Greatest Benefit to
Mankind. A Medical History of Humanity
from Antiquity to Present. London: Harper
Collins.
Reichert, T. A., Sugaya, N., Fedson, D. S.,
Glezen, W. P., Simonsen, L. and Tashiro, M.
(2001) The Japanese experience with
vaccinating schoolchildren against influenza.
New England Journal of Medicine, 344, 12:
88996.
Rennie, S. and Behets, F. (2006) Desperately
seeking targets: the ethics of routine HIV
testing in low-income countries. Bulletin
of the World Health Organization,
84(1): 527.
Rothstein, M. A. and Talbott, M. K. (2007)
Encouraging compliance with quarantine: a
proposal to provide job security and income
replacement. American Journal of Public
Health, 97(Suppl 1): S4956.
Rouech, B. (1953) Eleven Blue Men and Other
Narratives of Medical Detection. Boston, MA:
Little, Brown and Company.
Verma, G., Upshur, R., Rea, E. and Benatar, S. R.
(2004) Critical reflections on evidence, ethics
and effectiveness in the management of
tuberculosis: public health and global

perspectives. BMC Medical Ethics, 5: 2.
Verweij, M. F. (2005) Obligatory precautions
against infection, Bioethics 19(4): 32335.
Verweij, M. F. and Dawson, A. (2007) The
A. Dawson and M. F. Verweij. Oxford:
117
Wald, P. (1997) Cultures and carriers. Typhoid

Mary and the science of social control.
Social Text, 15: 181214.
Wilkinson, T. M. (2007) Contagious disease
and self-defence. Res Publica, 13(4):
33959.
Yoder, S. D. (2002) Individual responsibility for
health: decision, not discovery. Hastings
Center Report, 32(2): 2231.
Section 2
Chapter
Issues
Population screening
Ainsley J. Newson
Introduction
At first glance, the idea of examining a group of asymptomatic individuals for signs of
disease or illness would appear laudable, even essential. Who would not want to try to
prevent or provide early treatment for serious conditions such as cancer or HIV/AIDS? The
screening tests involved are often minor, can significantly reduce morbidity and can even be
life-saving. The myriad population screening programmes now in existence are underpinned by the rationale that screening can help optimize population health.
Yet issues such as impinging, however slightly, on peoples liberty when offering a
screening test mean that the introduction of this sort of health monitoring in populations
should not be undertaken without careful factual and ethical analysis. Offering an unproven
or ill-chosen test to an individual can have far-reaching implications for their future health
care and well-being. Population screening may also conflict with valued ethical principles in
health care, such as autonomy and consent. Ethical considerations arise with implications
for health professionals, recipients and policy makers.
Conversely, the principles and frameworks for ethical analysis in a public health context
(although methodologically uncertain) arguably differ from traditional analyses in medical
ethics. It may be short-sighted to prioritize individual interests over those of society, as
there are opportunities for rich ethical analysis in assessing the relative values of individual
and collective interests in public health. Focusing on preventive medicine in asymptomatic
populations raises the prospect of a collectivist ethic, in which communal life and the links
that sustain it must be recognized (Jennings, 2007). Public health ethics also forces us to
address difficult questions, such as the acceptability of reducing particular diseases in the
community and the mechanisms by which this may be achieved.
This chapter reflects these broader tensions in public health ethics by highlighting and
analysing some of the ethical considerations that should be borne in mind for population
screening. It is divided into three sections. The first briefly describes what we mean by
screening and reviews the criteria for introducing screening programmes. In the second
section, examples of screening programmes are presented and briefly described. The
majority of the chapter is dedicated to the third section, in which the ethical issues arising
from screening are presented and critiqued, with reference to the examples presented
previously. It emerges that population screening is a complex field with many overlapping
ethical tensions, with several key issues remaining unresolved.
118
Chapter 7: Population screening
119
What do we mean by screening? When should a screening

programme be introduced?
Definitions of screening
Population screening is a diverse aspect of health care, offered in numerous formal (and
sometimes informal) programmes. Broadly, the term screening means the application of a
straightforward but perhaps preliminary test for a designated health problem in an asymptomatic population, to find the few who are at increased risk. This population will often
share a characteristic, but this may not be related to the condition being screened for.
Individuals who return a screen positive result will then be selected for further
investigation (Riis, 1985; Shickle and Chadwick, 1994; Wilfond and Thomson, 2000;
Press and Ariail, 2004).
Screening has two main purposes: primary and secondary prevention. Primary prevention is the detection of a risk of future disease in a completely asymptomatic individual,
such as identifying a pregnant woman as a carrier of sickle cell anaemia. Secondary
prevention involves detecting a disease in its very early stages in order to offer treatment
to maximize prognosis, such as in mammography screening for breast cancer. Both
purposes aim to reduce disease incidence in a population (Riis, 1985; Shickle and
Chadwick, 1994).
The National Screening Committee (NSC) in the UK offers the following definition of
screening:
Screening is a process of identifying apparently healthy people who may be at increased risk of a
disease or condition. They can then be offered information, further tests and appropriate treatment to
reduce their risk and/or any complications arising from the disease or condition.
(National Screening Committee, 2011)
This reveals some important ethical aspects of population screening programmes. First,
screening usually involves a recommendation of testing to a previously unaware population.
The suggestion to test comes proactively from a central organization and is not commissioned following an individuals request (Clarke, 2007).
Second, consequentialist/collectivist reasoning is predominant when assessing the
ethics of population screening, in that a benefitharm calculation is often used to justify
screening programmes (Juth and Munthe, 2007). Screening aims to benefit the health of
the population, as opposed to individuals. While individuals may benefit from screening,
to keep costs down initial screens may not have the same degree of sensitivity and
specificity that clinical tests may have. This is considered an acceptable trade-off as
screening tests are generally low risk (Berlin, 1999; cited in Wilson, 2000), although they
will not help every screened individual and may in fact cause harm. Such harm is a
particularly significant issue if the term is interpreted broadly to include psycho-social
factors such as anxiety and practical considerations such as time allocated to attending for
screening.
Screening is distinct from clinical testing, although the boundaries are easily blurred
and the terms are used interchangeably in the literature. Screening tests are generally fairly
straightforward to administer and interpret and have a low unit cost. Testing is generally
offered to individuals who are already at increased risk of a specific disease or condition.
A diagnostic test may be more complicated, expensive or risky to administer than a
120
Section 2: Issues
screening test; but it will also be more accurate. The purpose of testing is diagnostic: Does
this person have this disease? and involves a definitive test. Screening instead asks: Is this
person at higher identifiable risk for this disease? and results may be probabilistic (Nuffield
Council on Bioethics, 1993; Gostin, 2000; Wilfond and Thomson, 2000; Wilson, 2000; Press
and Ariail, 2004).
This chapter is limited to discussions of population screening formal programmes
implemented by an organization to promote public health objectives. Some forms of
screening, for example whole body magnetic resonance imaging or prostate specific
antigen screening, will not be addressed as they are not yet offered to the public via a
formal screening programme.
Criteria for establishing screening programmes

Formal screening programmes are unlikely to obtain funding or other relevant support
unless they are shown to uphold defined criteria. One of the earliest and best-known lists of
criteria to be met before a screening programme can be established was published by
Wilson and Jungner in 1968. These criteria are presented in Box 7.1.
Whether any hierarchy exists between these principles is unclear, as it will be very
difficult for any one screening programme to meet all of them (Shickle and Chadwick,
1994). However, Wilson and Jungner did emphasize that the second principle (treatment)
was particularly important.
The relevance of these principles remains today, although they have necessarily
been adapted to meet changing health care contexts and priorities. For example the
UK NSC has published its own criteria, all of which should be met before a screening
programme is introduced (National Screening Committee, 2003). The 22 criteria
incorporate Wilson and Jungners first seven principles, while the eighth and ninth
are re-worked to provide additional requirements for diagnostic testing and clinical
management. Further NSC criteria require other primary prevention interventions and
treatments to have been implemented, that no screening programme be introduced without
quality randomized controlled trials, that the programme provides value for money and
that the programme be managed and monitored. Yet whether these are adhered to in
practice is sometimes questionable, as illustrated by the critique of then UK Prime
Box 7.1 Wilson and Jungners principles for screening
1. The condition sought should be an important health problem
2. There should be an accepted treatment for patients with recognized disease
3. Facilities for diagnosis and treatment should be available
4. There should be a recognizable latent or early symptomatic stage
5. There should be a suitable test or examination
6. The test should be acceptable to the population
7. The natural history of the condition, including development from latent to declared
disease, should be adequately understood
8. There should be an agreed policy on whom to treat as patients
9. The cost of case-finding (including diagnosis and treatment of patients diagnosed) should
be economically balanced in relation to possible expenditure on medical care as a whole
10. Case-finding should be a continuing process and not a once and for all project. (Wilson
and Jungner, 1968: 267)
121
Minister Gordon Browns proposals for screening programmes for diabetes, cardiovascular
and renal disease (Tudor Hart, 2008; White, 2008).
The NSC criteria also explicitly account for ethical issues. Wilson and Jungners sixth
principle is supplemented by criteria requiring the complete screening programme to be
clinically, socially and ethically acceptable to both health professionals and the public (NSC
criterion 14). Criterion 15 requires benefits of screening programmes to outweigh physical
and psychological harm. Informed choice (discussed further below) is included as the
justification for providing good evidence-based information to participants.
The notion of a suitable test for a screening programme (Wilson and Jungners fifth
principle) merits further discussion. Any screening test should provide an accurate basis for
determining which participants require additional tests. The terms sensitivity and
specificity are relevant here: sensitivity requires an acceptable percentage of affected
individuals in the population to be positively identified by the test, while specificity denotes
that a percentage of unaffected individuals need to be excluded from further testing by the
initial screen. Sensitivity and specificity are necessarily offset in screening as optimizing
both is impossible. Increasing sensitivity will raise false positives and increasing specificity
leads to more false negatives (Shickle and Chadwick, 1994). Yet finding the right threshold
is vital: a person incorrectly identified as potentially affected may suffer continuing anxiety,
while parents of an ill child whose diagnosis was missed in an initial screening test may
experience difficulty, delay and frustration in obtaining a diagnosis at a later stage.
Meeting screening criteria is often difficult. A condition being screened for may be an
important health problem, yet it could be rare in the population. The rarer a disease is, the
more specific a screening test needs to be, and therefore more expensive. The screening test
will have to provide a trade-off between specificity and cost. Not all proposals for screening
will receive funding, as a programme may not represent an appropriate allocation of
financial and other resources comparable to other health interventions. Significant harm
can also be caused if the threshold for sensitivity is set too low. In the early days of newborn
screening for phenylketonuria (described below), some babies were misdiagnosed and
erroneously treated. Tragically, this caused brain damage in some children who would have
otherwise been of normal functioning and some children died (Fost, 1999).
Common screening programmes

Screening programmes are available to populations at virtually all stages of the lifespan.
These include: the use of mammography in post-menopausal women to aid the detection
of breast cancer, the use of cervical smear tests in women to detect the early stages of
cervical cancer, faecal occult testing to screen for early signs of bowel cancer and genetic
screening.
Genetic screening
From an ethical perspective, an interesting collection of screening programmes are those
involving genetic conditions. The identification of more and more genes implicated in
human disease, together with the decreasing cost and complexity of molecular methods of
diagnosis, has led to a steadily increasing interest in using genetic testing on a populationwide basis to promote the publics health (Presidents Commission, 1983; Khoury et al.,
2000; British Medical Association, 2005). Genetic screening programmes predict or diagnose genetic disease or determine genetic risk for future offspring. They can be used in the
122
Section 2: Issues
Box 7.2 Genetic screening in practice screening for -thalassaemia in Cyprus

b-thalassaemia causes abnormalities in haemoglobin structure. It requires lifelong blood
transfusions, removal of excess iron from the blood and sometimes splenectomy. These
treatments are expensive.
If a couple each have a copy of the mutated gene, they have a 25% chance of having an
affected child. In Cyprus, 1 in 7 people carried the mutated gene and 1 in 158 children
were expected to be born with the disease, placing a demand on health resources.
In the 1970s, a national policy was introduced to reduce new cases of thalassaemia,
incorporating community involvement, public education, service improvement and initially a prevention campaign to discourage carriers from marrying (later replaced by a
prenatal diagnosis campaign).
The programme has led to a significant proportion of the population receiving screening
and a 90% reduction in the number of children born with b-thalassaemia. (Angastiniotis
et al., 1986)
general population or in recognized sub-groups known to be at risk for a designated

condition (European Society of Human Genetics, 2003). For example, antenatal carrier
screening in a couple can lead to a foetus being tested for thalassaemia (see Box 7.2) or
adults may be screened for familial hypercholesterolaemia. As many genetic conditions are
rare and the sensitivity of genetic tests varies, a genetic screening programme will usually
only be developed when the disease has a significant prevalence and when the presence of a
gene change (mutation) strongly predicts disease manifestation. Future genetic screening
programmes may instead take advantage of whole genome sequencing technologies.
The purpose of genetic screening to prevent disease can be categorized slightly differently
to the primary and secondary categories for screening discussed above. Genetic screening can
also be said to focus on either phenotypic or genotypic prevention (Press and Ariail, 2004).
Phenotypic prevention reflects primary and secondary prevention in that the purpose is
to prevent the manifestation of a chosen genetic condition. Genotypic prevention is a form of
primary prevention unique to genetic screening: the purpose is to prevent intergenerational
transmission of a genetic disease in conjunction with promoting informed reproductive
decision-making by couples. The outcome of genotypic prevention may prevent the existence
of an individual and so gives rise to ethical issues such as making value judgements about
the quality of life that those with genetic conditions must live with. The example of screening
for thalassaemia in Cyprus, discussed in Box 7.2, also raises the issue of whether it is ever
appropriate to explicitly set the reduction of a disease as the aim of a screening programme.
These and other ethical issues in genetic screening are discussed below.
A consideration of any genetic screening programme will indicate that it may often be
difficult, if not impossible, to meet those criteria set down by Wilson and Jungner or the
NSC. This problem was recognized some time ago and alternate criteria have been
suggested. Genetic screening should therefore:
1. Contribute to improving the health of persons who suffer from genetic disorders; and/or
2. Allow carriers for a given abnormal gene to make informed choices regarding
reproduction; and/or
3. Move towards alleviating the anxieties of families and communities faced with the
prospect of serious genetic disease. (Lapp et al., 1972, cited by Nuffield Council on
Bioethics, 1993: 17)
123
Antenatal screening
Given the necessity of their contact with health care providers, pregnant women are
principal recipients of offers of screening. In the past 30 years a large number of screening
programmes have been introduced in this group, with the overall goal of optimizing the
health of newborn babies. A secondary goal is to allow women to make informed choices
about continuing a pregnancy when a foetus is at risk of developing a serious health
problem. Tests related to this goal include ultrasound and maternal serum screening to
measure biochemical markers in the pregnant womans blood; both of which provide a risk
profile for several diseases including Downs syndrome. Pregnant women may also be
screened to determine their carrier status for conditions such as cystic fibrosis. Screening
tests are offered at different stages of pregnancy and some women will receive more tests
than others; relative to variables such as maternal age and ethnicity. Women obtaining
higher-risk results from these screening tests will then be offered further invasive
investigations.
Not all screening tests in pregnancy carry the implication of a potential termination of
pregnancy. Since 1999, all pregnant women in the UK have been offered screening for the
presence of HIV. The purpose of such testing is twofold: to provide affected women with
appropriate treatment to optimize their long-term health; and to prevent maternalfoetal
transmission of the virus. Other tests include anaemia, hepatitis B and Rhesus blood group status.
Newborn screening
Programmes to screen the health of newborn babies have been in existence for nearly 50
years (Guthrie, 1986; Fost, 1999; Kayton, 2007). Tests are performed on a small dried blood
sample obtained during the first week of life. Screening aims to detect the presence of those
congenital diseases where early intervention is beneficial, including phenylketonuria (PKU)
and hearing.
Wilson and Jungners criteria for screening are utilized in newborn screening, although
more specific requirements have also been advocated. Clayton, for example, suggests the
following factors:
The disease has a devastating outcome;
Treatment is highly effective in averting this outcome, but only if it is started early;
Affected children cannot be detected on the basis of symptoms in time to start effective
treatment; and
Screening reliably detects most affected children. (Clayton, 2004: 1005)
Newborn screening is not without ethical concerns (Kerruish and Robertson, 2005). As
programmes have developed and expanded, the degree to which the above criteria are still
met is contentious (Wilfond, 1995). While screening for PKU easily satisfies them, they
are arguably less robustly adhered to in screening for conditions such as Duchenne
muscular dystrophy as little can be done to avert the onset of disease. Yet ancillary
benefits can be identified, for example offering parents the chance to avoid subsequent
diagnostic uncertainty, the opportunity for psychological adjustment and the ability to
request prenatal diagnosis in future pregnancies.
Consent, or its absence, is also contentious reflecting the broader tension in public
health ethics. In most US states newborn screening for a large number of diseases, including
mitochondrial disorders, is mandated. Parents in 46 states are technically able to refuse
124
Section 2: Issues
screening (Kayton, 2007), although whether this is actively communicated to parents is

uncertain (Clarke, 1997). If they do refuse, the child may be removed for testing without
their permission (Bratton, 2007) a practice rejected elsewhere (Laurie, 2002). In the UK,
parents are invited to make an informed choice about testing and refusals of testing are
respected but not condoned.
As newborn screening expands, the criteria for testing will be stretched further.
A current issue involves reporting carrier status results, for example after screening for
cystic fibrosis. This approach may be criticized as inconsistent with policies for genetic
testing in children, which generally stipulate that children should only be subjected to
genetic testing when a direct health benefit can be shown. However, the finding of carrier
status occurs as an artefact of the screening test and forms part of the health record, so is
therefore legally accessible by parents. Empirical evidence suggests that parents are not
adversely affected by receiving this information (Parsons et al., 2003).
Further complexity in newborn screening is anticipated. As genes involved with multifactorial diseases such as diabetes are identified, and as whole genome sequencing improves,
the question of screening will arise. Will providing parents with information about their
childs susceptibility to common diseases be a valuable tool for optimizing childrens health?
Or, will parents perceive their children as vulnerable? Early evidence suggests parents will
not worry excessively (Gustafsson Stolt et al., 2003; Kerruish et al., 2007), yet ongoing
monitoring will be important.
Screening in adults
Most of us can expect to be invited to participate in a series of screening programmes
throughout our lives. These will include screening for transmissible viruses, such as HIV,
and acquired conditions, such as cancer.
Modern screening programmes emphasize prevention and early detection. At various
points in our lives depending on our age, gender and ethnicity we will be invited to
undergo screening for HIV, bowel cancer, bone density, cervical cancer, breast cancer,
diabetes, high blood pressure or coronary artery disease, among others (Muir-Gray and
Pearce-Smith, 2007). All of these programmes are subject to costbenefit calculations and
give rise to ethical issues, discussed further below.
Although the diseases forming the focus for screening in the twenty-first century tend to
be acquired or inherited instead of infectious, the SARS epidemic of 2003 illustrates that
infectious disease can still have significant implications for public health. And in the
developing world, the malaria pandemic is a significant cause of morbidity and mortality.
The persistent threat of infectious disease gives rise to the question of whether and when the
state can acceptably intervene to mandate screening of individuals for infectious disease,
and if disease is detected what measures can be taken to restrict a persons liberty.
Ethical issues arising across screening programmes

The sometimes divergent approaches of public health ethics and the more traditional
theories of health care ethics provide a key contention in ethical debates over population
screening (Mann, 1997; Hodge, 2004). In traditionally conceived health care ethics (bioethics), significant emphasis is placed on individual rights and responsibilities. Principles such
as autonomy, avoidance of harm and promotion of justice are emphasized. Although public
125
health ethics has not abandoned individualism (indeed, the promotion of voluntariness in
screening is increasingly prevalent, albeit not universally condoned), the focus for the
promotion of health through population screening is more collective.
This juxtaposition of emphasis is reflected in some of the key ethical questions for
population screening. When to introduce screening, whether it should be mandatory or
voluntary and determining how to manage information gained are all important questions,
reflecting some key themes in public health ethics (Hodge, 2004).
Defining success in screening

Ultimately, for the state to support a screening programme and for that programme to be
ethically justifiable, that programme needs to be successful. Gostin (2000: 397), for example,
suggests that: Screening programs can be justified only if they effect a positive outcome that
would not have occurred without the screening.
In the context of genetic screening, Burris and Gostin (2004) have suggested that for a
screening test to be ethical there must be a reasonable probability that it will reduce the
incidence of a genetic condition or mortality from that condition in society or in a
population facing a specific threat.
Preventing disease in the population cannot be all there is to success in screening
however, as prevention per se could come at a great cost to individuals within that
population. One reason for ethical discomfort in screening is that much success is defined
in terms of economics; specifically the chance to reallocate those costs that would otherwise
have been spent on treating a case of disease. In an adult population this is not such a
problem; most people would prefer not to develop a disease were it possible to avoid this
outcome. But in antenatal screening, often the only way to prevent a case of disease is to
terminate a pregnancy. The concern over defining success of screening in terms of cost
savings is that it could lead to accusations of calculating the value of a life in purely
economic terms or accusations of eugenic practice.
Labelling antenatal screening programmes as eugenic could be used as an indefeasible
rejection of this technology. However, the argumentation behind this critique is unclear
(Chadwick et al., 1998). In this context, eugenics is taken to mean a deliberate use of
screening to improve the quality of the human gene pool, leading to socially relevant
changes. Chadwick et al. argue in response that maintaining a status quo can in itself be
eugenic and that genetic screening is unlikely to cause suffering to future individuals
(Chadwick et al., 1998). If we continue to respect the lessons learned from past eugenic
mistakes (Wikler, 1999), screening is not per se problematic.
In the context of screening for Downs syndrome, published programme objectives are
also notable for their absence of emphasis on economic considerations. Nor do they aim
for a reduction in the incidence of the condition in the population. The main aim is to
ensure access to a uniform screening programme, with a further aim to provide women
with the necessary information to exercise an informed choice (National Screening
Committee, 2011). This approach is laudable, but has been subject previously to criticism:
Programmes that claim to justify their existence by offering choice may find it difficult to put this into
practice if the outcome measures they use are, in practice, those relating to uptake of screening,
terminations of pregnancy and the birth incidence of affected infants. . . .
(Clarke, 2007: 429)
126
Section 2: Issues
In defining success in screening, the focus should be broad. Screening should not be seen
as making a judgement about those with a disability. Outcomes should not only consider
those identified as being at higher risk, but those who receive screen-negative results. But as
more screening programmes are implemented, it will remain important to continue to
reflect on the ethical assumptions behind the aims of a given screening programme, and to
question whether it is appropriate to set the reduction of a disease as a goal. While uptake
by a population may be taken as tacit acceptance of the goals of a screening programme,
Clarkes above criticism suggests that participants in screening may not always base a
decision to screen on complete information.
Informed consent and informed choice

Generally, for any diagnostic test in health care, no individual should be tested without full
informed consent from that person or an appropriate surrogate. This upholds modern ethical
and legal doctrines of bodily integrity and autonomy. But in population screening, where
collectivist values are also relevant, should fully informed consent always be required?
To some, the need for informed consent to screening is equivalent to that for any
diagnostic test. Though not endorsing this position, Gostin (2002) points out that screening
can be intrusive; if imposed without consent it can invade personal autonomy and bodily
integrity. For some, screening should therefore be preceded by provision of appropriate
information, the recipient should be competent to make a decision and should decide
voluntarily. The inherent power imbalance in professionalpatient relationships could also
undermine informed consent, as the mere offer of a screening test may be taken to imply
that having the test is a good thing and may lead to a compliance effect on this ground
alone (Shickle and Chadwick, 1994; European Society of Human Genetics, 2003).
But, encouraging fully informed consent may be resource-intensive to satisfy in practice.
This begs the question that even if it is theoretically possible to obtain fully informed
consent to screening, is it desirable? Informed consent will significantly raise the costs of
screening programmes and so it may not always be justifiable, at least for some diseases
where the value implications are less significant (for example, screening interventions that
can be life-saving and are minimal risk). Also, if resource constraints are affecting information provision, participants may not be aware of the condition being screened for or may
struggle to appreciate the (minimal) risks involved raising the question of whether it is
indeed wise to require informed consent at all times. Fostering autonomy could also
manifest in a shifting of the burden of decision making to screening participants (Chadwick
et al., 1998). It may be that the costs of obtaining fully informed consent can be traded off
against other benefits to the population from screening.
These potentially competing concerns necessitate an assessment of priorities in
screening in the developed world: high uptake or an informed decision (Raffle, 2001).
Screening will not be effective in the absence of high uptake; if potential participants are put
off by complex information, public health goals will not be achieved. Further, as screening
targets a healthy population, the ethical assumptions of screening are different to those
found in clinical practice. Effective screening will save lives or improve quality of life via
early diagnosis.
Conversely, persuading people to receive screening tests (such as a cervical smear test or
an antenatal screen for foetal abnormality) may be unethical without fully informed
consent. Even if it harms a programmes success, open and frank communication about
127
some screening programmes is vital to make participants equal partners in screening

(Pfeffer, 2004). Making an informed decision should therefore be included as a measure
of success in some screening programmes, in addition to measuring the numbers who take
up an offer of screening (Raffle, 1997; Foster and Anderson, 1998; Marshall and Adab,
2003). Such an approach may not, however, be appropriate for all screening programmes
informed consent may on occasion be traded off in light of significant health benefits to the
population.
HIV screening is one domain where concerns over informed consent have been raised.
Testing for HIV is now offered on a routine opt out basis in numerous domains, including
at sexual health clinics and in antenatal screening. HIV is now a manageable condition and
peoples best interests are served by knowledge of infection status. Testing is strongly
recommended and a high uptake is expected. In the antenatal context this has been
criticized as undermining a pregnant womans rights to bodily integrity and free choice
(Bennett, 2004; Bennett, 2007) and empirical evidence suggests that only a minority of
women have enough of an understanding of the screening test to meet standards of
informed consent and that women are more likely to participate in screening if they spend
less time discussing it (Sherr et al., 2006; de Zulueta and Boulton, 2007).
Bennett recognizes that pregnant women should know their HIV status, not least
because mechanisms to prevent maternal-foetal transmission of the virus are so effective.
Yet with this knowledge comes the expectation of compliance to a complex regime of
treatments and interventions that can have serious side effects. To this end, she proposes
that antenatal screening for HIV operate on a policy of informed choice, to both promote
screening and ensure womens trust (Bennett, 2004).
Informed choice for screening marks a recent departure from emphasizing screening
as a public health activity designed to reduce the community burden of disease. The concept
has been adopted in many domains of screening and is explicitly recognized in the NSC
criteria (National Screening Committee, 2003) and can be defined as being . . . based on
relevant knowledge, consistent with the decision-makers values and behaviorally implemented (Marteau et al., 2001: 100). Applying this to screening:
[A]n informed choice to undergo a screening test occurs when an individual has a positive attitude
towards undergoing a test, has relevant knowledge about the test and undergoes it. An informed
choice to decline a test occurs when an individual holds a negative attitude towards undergoing a test,
has relevant knowledge about the test and does not undergo it.
(Marteau et al., 2001: 100)
Compared with informed choice, consent involves a more active process of decision making
and almost invariably involves contact with a health professional (Jepson et al., 2005: 193).
It will be more time-consuming and resource intensive. Informed choice may offer a
pragmatic and ethical alternative to fully informed consent. Yet informed choice is not a
one-way exchange of information: any choice that is either based on poor quality information or is inconsistent with the decision makers values will be uninformed and liable to
ethical critique.
Care also needs to be taken regarding the nature of the information to be provided to
facilitate informed choice, insofar as informed choice is appropriate to a particular screening
programme. Unsurprisingly, the lay public process information differently to health professionals or policy makers and potential barriers must be identified. Pre-screen information
should, in general, cover the purpose of screening, the likelihood of false positive or negative
128
Section 2: Issues
results, the implications of testing and plans for follow-up. But conveying these concepts
is often difficult and may not assist in making an informed choice. Participants in
screening may also want to receive information about the diseases being screened for
and to hear narrative accounts from those who have already participated in the programme (Jepson et al., 2007).
Informed choice may offer an appropriate trade-off between high uptake and informed
decisions, although a decrease in uptake of screening is possible. Appropriately implemented
it can offer increased feelings of control, a reduction in anxiety and an improvement in
behavioural change (Marteau and Kinmonth, 2002). How this is put into practice will vary
with the participant population, the clinical significance of the condition being screened for,
the risks associated with the test and resource or time constraints. Quality control and
practitioner training are also important (Jepson et al., 2005).
When should consent be waived? Mandatory and routine screening

We have seen that informed choice can be an appropriate tool for the ethical delivery of
screening programmes. Yet screening is not always voluntary: it can also be administered as
mandatory or routine, with different or absent standards of informed choice. Although
much routine screening is largely uncontroversial as patients, we assume that contact with
health services will at some point lead to screening (Wynia, 2006) these departures from
rigorous informed choice have been subject to criticism.
Newborn screening is effectively mandatory in most US states, as discussed above. This
process is contested (Paul, 1999; Press and Clayton, 2000). Some have argued that if
screening is important for a childs future health, then parents should not have the right
to refuse (Faden et al., 1982a, 1982b). Spending valuable resources on seeking informed
consent could also be inequitable. Further, as virtually all parents agree to the screening test,
the extra anxiety they may suffer through being informed could outweigh any autonomybased considerations around informing them.
The increasing involvement of patients in health care can be cited as reason to reject
mandated screening, although whether this translates from a medical to a public health
context is contentious. That said, the ability to make an informed choice about this practice
helps promote trust in the health professions and allows parents to exercise their right to
be accorded respect in making decisions for their child (Clarke, 1997). It also offers a
valuable opportunity to educate families about the value of this screening (Friedman Ross,
2002) and will minimize harm and promote parental cooperation in the event that an
abnormality is detected. Even if a parent refuses screening, the chances a child will be
harmed are small but are nonetheless real. It may be that even if mandated screening is
unacceptable, parents who refuse newborn screening may be legitimately encouraged to
have the test by health professionals (Newson, 2006).
An emerging problem for informed choice and newborn screening is the rapid expansion of these programmes. Should information be provided about the myriad of conditions
to be tested for, and how? A balance needs to be struck between empowering parents and
ensuring viability of the screening programme. While mandatory screening seems unjustified in the absence of a significant risk to the child, a pragmatic approach may have to be
adopted. Although this de-emphasizes fully informed consent, some degree of informed
choice can still be used. Some interventions in medicine may indeed be collective in nature
and as such are unsuitable for individual decision-making (Nijsingh, 2007).
129
Mandating screening of competent adults is more difficult to justify, as patients have a

general right to refuse medical treatment even if it harms their interests. Despite this, in
some countries mandatory screening for diseases such as tuberculosis is carried out at some
sampling points, such as admission to hospital. But again, should a screening test being
implemented to further public health goals be assessed using the same ethical principles as
health care? Screening may also be more acceptably mandated if an infectious disease
outbreak occurs, for example SARS. Grounds for justification of mandatory screening are
that screening tests are unlikely to be invasive and the outcome can have significant impact
on personal and public health. This demonstrates the Millian harm principle, in that
individual autonomy can be restricted in order to prevent harm to others (Bayer et al.,
1993). The right to privacy is usurped by the need to prevent the spread of disease.
The mandating of any screening programme should, however, only be carried out once
high standards of safety and efficacy are demonstrated. Reversal is also difficult postimplementation (Fost, 1999).
Pregnant women are an ideal group for screening from a public health perspective. They
are necessarily in contact with health services and screening could improve health outcomes
for the next generation. For these reasons a large number of antenatal screening programmes are now routinely offered. These programmes aim to protect the health of
pregnant women and their foetuses.
One aspect of antenatal screening focuses on the foetus future health. This is
unique in public health in that if an abnormality is detected it could lead to the
prevention of a beings existence by terminating a pregnancy. The active promotion
of these screening tests in light of their potentially far-reaching implications has led
some to criticize their routinization as an implicit yet strong recommendation in favour
of testing by health professionals. This may lead to a downturn in informed choice and
may be at odds with individually held values surrounding disability. Empirical studies
have also suggested that not all women make an informed choice when having this
screening, particularly those who accept the tests (Green, 2004; van den Berg et al.,
2005, 2006).
While antenatal screening tests should be made available to women who desire them, it
needs to be experienced by women as a considered choice and not something that merely
happens to them while they are pregnant (Clarke, 2007). So-called institutional directiveness towards abnormality screening in pregnancy can be avoided by providing sound and
unbiased information to pregnant women, ensuring they understand the potentially significant results the test may provide.
Screening and the right to remain in ignorance

Just as we have a right to obtain knowledge to inform our health care, our right to make
autonomous decisions also implies we should have the right to refuse at least some public
health screening tests. In the genetics literature for example, the right not to know ones
genetic status has received significant attention (Raikka, 1998; Bennett, 2001; Andorno,
2004). Broadly, the right to genetic ignorance has two aspects: the right not to know ones
mutation status, and the right not to know that one is at risk at all. The former of these is
most relevant to the screening context given that screening by definition involves an at-risk
population and so respecting the second sense of the right would be incompatible with
education or recruitment policies. That said, individuals who are identified as being at risk
130
Section 2: Issues
of a condition that will have implications for their relatives should exercise due caution
when considering an unsolicited disclosure of risk to other members of their family.
Broadening out the right not to know from the genetic context may not always be
appropriate. Generally, the right not to know that one is at risk of disease is straightforward
to respect, although some have argued it precludes full autonomy and may cause us to fail
to fulfil our obligations to others (Rhodes, 1998; Harris and Keywood, 2001). It also
undermines the collectivist values that underscore ethical analysis in public health. Physically restraining a person to obtain a sample for screening would be unacceptable in all but
extreme cases, such as testing for a highly infectious disease when significant concerns arise
about third-party harm.
Potential for stigma and discrimination
The term stigmatization is used to describe the negative labelling a person can experience
following an event, in this context screening. The Nuffield Council on Bioethics has defined
stigma as the branding, marking or discrediting . . . of a particular characteristic (1993:
para 8.8; cited by Chadwick et al., 1998). Stigma can be experienced as a self-regarding trait,
or something placed upon a person by others. Self-regarding stigma involves feelings of
negative perception following a change in personal status, including the receipt of a test
result. Discrimination is related to stigmatization, but is not the same. Simply, discrimination involves treating like people differently. It leads to differential treatment in light of a
persons status. Any screening programme that led to stigma or discrimination would be
problematic and steps would be needed to minimize this occurring.
Concerns over stigma can arise when a specific population is targeted for screening.
Perhaps the most notorious example arose from the well-intentioned yet ill-fated sickle cell
anaemia screening programmes in the USA in the 1970s. During this decade, many US
states mandated screening in African-Americans to identify carriers of this recessively
inherited genetic condition, with a view to providing health and reproductive advice and
counselling. However, this was not the only effect. Due to misinformation about the nature
of carrier status, those found to have one copy of the mutated gene (who were healthy)
were discriminated against by employers and insurers who believed carriers would become
unwell. Moreover, the African-American population themselves indicated little interest in
testing and were not explicitly consulted during policy development. Additionally, sickle
cell became widely known as a black disease (Kenen and Schmidt, 1987; Fost, 1999;
Wilfond and Thomson, 2000: 64; Bayer et al., 2006: 352).
In the 1980s, screening for HIV also carried concerns over stigma, with potential access
to results by third parties causing significant concerns. In 1990, Almond and Ulanowsky
observed that:
[T]he social and financial implications of a positive diagnosis are serious. Health and life insurance,
mortgages and house purchase may all be affected. There may be problems for some types of
employment . . . .
(1990: 16)
They also show that HIV illustrates how branding can occur:
AIDS . . . has been particularly associated . . . with certain clearly defined groups gay and bisexual
males, IV drug users, and haemophiliacs . . . .
(1990: 16)
131
While treatment for HIV has dramatically improved over the past two decades, with a
consequent positive impact for morbidity and quality of life, people with HIV arguably do
still suffer some degree of stigma. Wynia (2006) suggests this could actually be due to the
exceptional treatment of HIV in screening, given the requirements for in-depth counselling
and pre-test advice. Regardless of its mechanism, the potential for branding or otherwise
derogatively marking people who have HIV remains a problem and must be taken into
account by screening programmes.
A more recent, yet less widely discussed example of potential stigma arising from
screening in target populations is the use of genetic screening for a panel of diseases known
to have a higher prevalence in the Ashkenazi Jewish population (ACOG Committee on
Genetics, 2004; Leib, et al., 2005). This includes TaySachs disease, Gaucher Disease, Canavan Disease and Fanconi anaemia. A significant difference between this programme and
earlier genetic screening programmes is the degree of consultation with, and involvement of,
the target population. Yet problems with stigma may still arise, in two ways. First, these
conditions may be (and in some places, have been) labelled as Jewish diseases, when they in
fact arise across most populations. Second, members of the Jewish population may be labelled
as particularly susceptible to disease. Despite this, the Jewish populations experiences with
screening have been positive and this group exhibits a strong desire for screening.
Discrimination following testing has been discussed at length in the genetic screening
literature. The main concern has been access to results by third parties; notably insurers or
employers. This has been discussed at length elsewhere, but can be summarized as
concerns that:
Individuals will be denied life or other insurance, or charged a disproportionately high
premium, after undergoing a screening test;
Genetic screening will become a prerequisite to employment or obtaining insurance;
Individuals or groups will be excluded from some workplaces due to a genetic risk they
pose, whether or not that risk is related to the particular role; and
Workplaces will decrease their safety standards; instead only hiring employees with lowrisk screening results. (Nuffield Council on Bioethics, 1993: chs 6, 7)
Difficulty in accessing insurance can have devastating consequences, especially in those
jurisdictions lacking a national health service. However, companies providing insurance
products are commercial entities and cannot be expected to act as charities. Adverse
selection, in which consumers disproportionately increase their insurance to counter a
known (but undisclosed) genetic risk could have significant commercial implications and
would raise premiums for all. Therefore, all contracts of insurance should be made on a
level playing field, with full disclosure on both sides.
Potential discrimination in access to insurance has been taken seriously by policymakers in several countries. The UK, for example, currently has an industry concordat
and moratorium on the use of results of predictive genetic tests by insurers (Department of
Health, 2005). This gives a high level of consumer protection and as such a legislative
approach is not deemed to be required. In the USA, however, the Genetic Information
Nondiscrimination Act 2008 does provide legal protections for individuals in relation to
employment and health insurance.
In the employment context, it needs to be recognized that employers can already
legitimately request medical information in some circumstances. Further, testing may
132
Section 2: Issues
not necessarily have to exclude employees screening could be used to indicate

whether any additional risk-avoidance strategies are required. Mass screening
programmes of employees are not only introduced to exclude people and maximize
profits; they may also lead to a reduction in occupational disease (Nuffield Council on
Bioethics, 1993).
Despite this, employers should not act to provide screening without due cause. Indeed,
one US employer has been sued for coercing employees to take non-indicated tests
(Philipkoski, 2001). The Nuffield Council on Bioethics has recommended that screening
is used only where a connection between a specific disease that seriously endangers health
and a workplace has been shown; and for which nothing can be done (Nuffield Council on
Bioethics, 1993: 64). Others have recommended, in addition to standard requirements for
privacy and predictive value, that the genetic factor should be positively correlated with
adequate job performance and that a plan is in place for the use of any data gained (Lapp,
1983; Murray, 1983; Schulte and DeBord, 2000).
Overall it appears that, with some minor exceptions, concerns about discrimination and
stigma following screening remain unsubstantiated. Additionally, as Western populations
become more mixed, some governments have opted to broaden out screening programmes
initially targeted at specific populations; the recent change in practice in the National Health
Service antenatal and newborn screening programmes for sickle cell anaemia and
thalassaemia in the UK being one example (National Health Service, 2006). With an
emphasis on policy monitoring and broad education, concerns over stigma and discrimination can be mitigated. Routinization of screening for other diseases such as HIV could also
help reduce stigma.
Provision of pre-test counselling

For genetic screening in particular, until recently most screening programmes incorporated a counselling element, to try to ensure participants make an informed decision that
reflected their values. Indeed, the provision of counselling (before or after screening) has
been recommended by several authorities in the field (see, for example, Clarke, 1995;
European Society for Human Genetics, 2003). More recently, the growing number of
genetic tests that can detect conditions for which treatment is readily available and which
have little reproductive implications has led some to question whether such tests are really
exceptional when compared to standard screening tests in medicine. Thus the requirement for pre-test counselling could be relaxed and values more associated with public
health (namely high uptake) could be emphasized. The inherited condition familial
hypercholesterolaemia (FH), for example, can cause morbidly high cholesterol if
untreated, can be effectively screened for (using either genetic or clinical means) and a
diagnosis has no reproductive implications. Once a case is identified in an individual, the
care team will cascade out to that persons relatives, in an attempt to identify some of the
thousands of undiagnosed individuals with this condition. Once diagnosed, cholesterol
can be controlled via diet, exercise and statin drugs if needed. For conditions such as
FH, counselling should arguably not be required. If an individuals best interests are
served by knowledge of mutation status, a good intervention is available, the test has no
reproductive implications and stigma is unlikely, then many of the reasons why pre-test
counselling is offered are mitigated. A pre-test counselling process could instead cause
potential participants to be more concerned about what is really a straightforward test,
133
increasing refusal rates. Waiving the need for counselling will not be appropriate in all
circumstances, but adopting a one size fits all approach to counselling in population
screening programmes is equally unwise.
Screening and the potential for psychological harm

A primary ethical principle underpinning screening is that the benefits of early diagnosis
need to be counterbalanced against potential harms. A hallmark of all screening
programmes is that they tend to be offered to otherwise healthy populations. This can have
the effect of introducing concepts of illness and disease to groups which otherwise had no
cause for concern regarding their health. This may cause harm, for example, problems
caused by misleading results leading to unnecessary further testing.
The aspects of screening programmes most likely to give rise to concerns about
psychological harm are the potential for false positive (identifying disease in a healthy
person) or false negative (failing to identify the disease of interest) results, caused by the
necessarily imperfect specificity and sensitivity of screening tests. These rates are not
insignificant: screening tests for some cancers have a 1020% false negative rate.
Both false negative and false positive results can give rise to harm. A false negative result
can potentially mean that a screening recipient will go on to develop the condition or
disease being screened for, although this risk can in some circumstances be mitigated if the
person participates in regular screening and seeks health care when unusual symptoms
develop. A false positive result can give rise to serious feelings of morbidity and anxiety,
even some months after the initial test and despite continuous reassurance (Brett et al.,
1998; cited in Wilson, 2000).
In the context of newborn screening, around 1% of parents will receive a false positive
diagnosis, received at an important point of parentbaby bonding. The effect of this
has been the subject of many empirical studies exploring possible harms such as parental
anxiety or hyper-vigilance about the childs health. While early studies described the
problem of PKU anxiety syndrome (Rothenberg and Sills, 1968), more recent data
suggests that parents display no adverse behavioural attitudes towards their children
(Pollitt et al., 1997; Green et al., 2004). A minority do have lingering concerns about their
childrens health (Pollitt et al., 1997; Green et al., 2004), yet this harm from a false positive
diagnosis can be minimized if good quality information is provided during follow-up
(Green et al., 2004).
False negative results can also have repercussions for individuals, as the example of
cervical screening illustrates. Litigation in response to false negative results for breast and
cervical cancer has occurred in several countries, including the UK and Australia (Wilson,
2000). This has, in some cases, led to a drop in the number of people presenting for
screening, an increase in referrals for invasive tests in patients who have slightly abnormal
results and a demand for more accurate yet vastly more expensive screening tests (Stanley,
1996; Wilson, 2000). Wilson comments that the enthusiasm for screening promoted by
health agencies has provided recipients and the legal fraternity with a false sense of security
as to accuracy.
The possibility of facing further testing following screening can also cause harm. Combined
nuchal and maternal serum screening, for example, is a minimally invasive way for some
pregnant women to reduce their background risk of foetal abnormalities. However, the test is
inherently uncertain and presents results probabilistically. This can cause anxiety or can establish
134
Section 2: Issues
women on a path to invasive testing that they may not have contemplated at the outset. As
this test is only capable of obtaining meaningful results after 12 weeks gestation women who
do later discover their foetus carries an abnormality will face a decision about termination of
pregnancy later than they may feel comfortable with. They may also experience their
pregnancy as tentative until these tests are conducted (Katz Rothman, 1993). Some argue
that the test has also medicalized pregnancy for the many younger women who previously
experienced pregnancy without screening interventions (Lippman, 1991).
In offering routine antenatal screening, careful counselling and information provision is
required to ensure that women make informed decisions. Yet ensuring this is achieved
within the myriad of other tests, interventions and information pregnant women receive is
no easy task.
In light of these concerns, Shickle and Chadwick (1994) argue that a full riskbenefit
calculation must always be performed before a screening programme commences and that
health authorities must be prepared to accept the fact that this may mean that a screening
programme should on occasion not be implemented.
Access to screening programmes justice and distribution

Determining how to allocate scarce resources is inherent to any domain of health care; and
public health is no exception. Appropriate use of funding in public health is imperative to its
success, as this will satiate funders and help ensure public confidence in screening programmes. Thus any screening programme, and certainly those funded from public resources,
will need to be targeted in some way, otherwise the benefits of screening will be undermined
by the inefficiency that inevitably arises when too broad a population is screened.
A consequence of such targeting in screening is that some who may benefit from screening
will be excluded, or may not even be made aware of the programmes existence.
Screening programmes will almost always be competing for resources from a finite pool
of funding. While equity in access to screening regardless of finances is ethically laudable, it
ignores the reality of resource allocation. Distributing funds between groups is often
difficult, as is comparing the value of different interventions. How can we justly compare
the benefits of cervical screening with newborn hearing tests? Both of these appear to be
important health problems, but prioritizing this importance may be difficult and could
involve the use of potentially problematic factors such as ethnicity, socio-economic status or
gender (Chadwick et al., 1998).
Any screening programme should aim to use funds primarily to increase the marginal
health of the population which will require a sound demonstration of the screening tests
efficacy and efficiency. This involves a consideration of more than mere cost-effectiveness.
Population screening for prostate cancer is not currently provided using public funds in the
UK because evidence for benefits of screening is not strong enough to outweigh the
potential harm of false diagnosis of this condition (National Health Service Executive,
1997).The psycho-social impact of any screening test is also relevant. Burris and Gostin
(2004) point out that in the context of genetic screening a test result will be experienced by a
person within a complex set of social links and relationships. Not only are test recipients at
risk of stigma or other kinds of psychological harm, they may suffer economic burdens
through increased health care costs or denial of access to other resources.
A further issue affecting the provision of screening programmes is access to relevant
treatments and/or preventive measures should a person be deemed to be at high risk of
135
disease following screening. A programme should not be instigated until other basic
necessities are in place, such as availability and affordability of the relevant treatment or
preventive tool, or control of the relevant environment (European Society for Human
Genetics, 2003).
In establishing or extending any screening programme, its impact on other populations
who may benefit from screening but whose needs are less visible should also be assessed.
Screening programmes will be problematic if the outcome deliberately or inadvertently
excludes those who may benefit or if the screening process causes harm. For these reasons,
the equity and effectiveness (broadly construed) of all programmes should also be monitored once they are established.
Involvement of the private sector

The risk of some individuals or groups being excluded from a screening programme will
also depend on how the programme is provided through the public sector, via a charity, in
the course of a research project or on a fee-for-service basis. If publicly funded, cost
constraints will inevitably determine the definition and boundaries of the target population.
Under privately funded service provision, cost may instead become the primary barrier to
access, while the programme itself may be provided to a wider cross-section of participants.
The potential for competing interests between the financial motivation of commercial
providers and the avoidance of harm for participants in screening has not gone unnoticed,
particularly in discussions of genetic screening (Chadwick et al., 2001). Many companies
now offer genetic testing direct to consumers, usually via the Internet. Although these
services are subject to some concerns, most fall outside the scope of this chapter in that they
are not explicitly screening programmes. Concerns may arise, however, if a screening
programme is deemed important enough to be established, yet is then provided by a
commercial entity.
Concerns underpinning private provision of screening echo those arising in other areas
of health care. These include:
Denying justice to those who are already disadvantaged.
People who have less financial means may not be able to pay for commercial screening
services. But members of lower socio-economic groups may also stand to gain huge
benefits from screening.
The value to public health will be lost.
If screening becomes accessible only to those who can pay for it, then the potential for
benefit from screening programmes may be reduced. This will then reduce the efficiency
of the programme, which may undermine it being offered at all.
The introduction of a screening test before validity is certain.
Commercial entities will be keen to recoup their costs in developing a test and so may
offer it in the private market before the recommended evaluations have been carried
out. If a test proves not to be beneficial then people would have wasted their money
investing in it.
Screening will exploit the worried well.
Private entities offering screening are very likely to heavily market their tests in order to
ensure maximum customer uptake. Those who do take the test may feel more in control
of their health; yet they are also at risk of taking an ineffective or inappropriate test.
136
Section 2: Issues
Unbiased counselling may be compromised.

Evidence suggests that people who receive a higher risk result after screening do suffer a
degree of anxiety, albeit one that resolves in the longer term. However, this does not
mean it should not be addressed when it occurs. How will private or commercial entities
handle pre-test counselling? Will it be possible to offer independent counselling, or will
this become confused with the need to sell the test? Further, will distressed patients be
pushed back into the public sector to have their problems addressed?
Confounding factors if screening is offered by the charity or research sectors.
If screening is offered by charities, it will be dependent upon funding and donations.
These can at times become unstable, which could affect continuity of service provision.
If screening is offered in the course of a research project, the nature and duration of that
programme will be constrained by the aims and timeline of the particular project.
(Clarke, 1995; R. E. Ashcroft, pers. comm., 2005)
In his thoughtful book, Jim Thornton (1999) argues in favour of a limited programme of
private screening, discussing equity, health needs and free markets. Regarding screening in
adult populations, he argues that consumers are the best persons to judge their own health.
Some screening programmes are ineffective while others may cause actual harm. For these
reasons, adult screening should be provided by the private sector. Some screening programmes may also cause harm. Children and those yet to be born, however, are not capable
of advocating their best interests and so the State must protect them via quality programmes. And if public screening is being requested, the burden of proof for justification
lies with those advocating for it.
If private commercial entities are to be involved in screening, then this should be
entered into with caution and with transparent and rigorous regulation, including oversight
of marketing and advertising. Steps need to be taken to ensure appropriate linking in with a
participants primary care provider and, in the case of genetic screening, to help the
proband identify relatives who may also wish to be screened. Factors such as marketing
and pre-screen information should be monitored and the need for an oversight authority
should be considered (Chadwick et al., 2001).
Emerging issues
Population screening is not a static programme. As more and more screening is performed
for a wider array of conditions or clauses, further ethical issues are sure to arise. For example,
the vast number of biological samples collected during screening can pose problems around
consent and control of future use. Many samples are used again, for example in research or
public health monitoring. This necessitates a balancing of the right to control access to
biological data with the practicality and cost of obtaining fully informed consent for these
kinds of uses. Legislation and policy is also relevant here and in some jurisdictions is complex.
In the field of genetics, genes contributing to complex human diseases and conditions,
such as coronary artery disease, diabetes and cancer, are frequently identified. DNA
sequencing technology is also constantly improving; becoming faster and less expensive. Personalized genomics, whereby we will all be able to obtain a complete sequence
of our genome with an indication of disease risk is already commercially available
and is receiving attention from policy makers (Fink and Collins, 2000). The use of
DNA chips (microarrays) also enables the detection of several hundred gene changes
137
at once. The possible implications of such testing for geneticization of health, our selfunderstanding and discrimination and stigma requires more attention.
Screening programmes also tend to be offered by health systems in developed countries,
leading to further discrepancies in global health worldwide. The equitable and ethical
introduction of health screening worldwide is a significant challenge for the twenty-first
century (Ballantyne, 2006).
Conclusion
Population screening for public health offers myriad complexities; drawing together considerations of ethics, economics, effectiveness and equity. Ethical issues in population
screening tend to be assessed using a utilitarian framework applied consistently with ethical
standards in clinical practice, although some have suggested that as screening programmes
are essentially large-scale experiments a research ethics framework is more appropriate
(Skrabanek, 1990; cited by Jepson et al., 2007). Appropriately analysing these overlapping
considerations and frameworks requires knowledge of each of these domains, and the
literature is illustrative of the collaborations and cross-considerations currently taking place
between academics, health professionals and policy makers.
This chapter has offered an overview of the many debates ongoing in this fascinating area of health care. Ethical issues arise at all stages of screening programmes,
from establishing a case for a new screening test, to recruitment, testing and monitoring.
A balancing of individual and community values is required, with a key theme of
prioritizing benefit over harm to benefit the population and promote and protect
health. Screening is not, however, without risk and to some extent there remains a
question mark over whether risks are always accurately safeguarded against, or what risks
it is acceptable to trade-off to promote population health. Additionally, the issue
of evidence to support screening can be contestable, particularly if screening is time
consuming or invasive.
Acknowledgements
The author would like to thank Dr Angus Dawson, Professor Robyn Martin and Professor
Richard Ashcroft for helpful discussions and assistance during the preparation of this
chapter.
References
ACOG Committee on Genetics (2004) Prenatal
and preconceptional carrier screening for
genetic diseases in individuals of Eastern
European Jewish descent. Obstetrician and
Gynecologist, 104: 4258.
Almond, B. and Ulanowsky, C. (1990) HIV and
pregnancy. Hastings Center Report, 20(2):
1621.
Andorno, R. (2004) The right not to know: an
autonomy based approach. Journal of
Medical Ethics, 30: 43540.
Angastiniotis, M., Kyriakidou, S. and

Hadjiminas, M. (1986) How thalassaemia
was controlled in Cyprus. World Health
Forum, 7: 2917.
Ballantyne, A. (2006) Medical Genetic Services in
Developing Countries: The Ethical, Legal and
Social Implications of Genetic Testing and
Screening. Geneva: World Health
Organization.
Bayer, R., Gostin, L. O., Jennings, B. and
Steinbock B. (eds) (2006) Public Health
Ethics: Theory, Policy and Practice. New
York: Oxford University Press.
138
Section 2: Issues
Bayer, R., Dubler, N. N. and Landesman, S.

(1993) The dual epidemics of tuberculosis
and AIDS: ethical and policy issues in
screening and treatment. American Journal of
Public Health, 83: 64954.
Bennett, R. (2001) Antenatal genetic testing and
the right to remain in ignorance. Theoretical
Medicine and Bioethics, 22: 46171.
Bennett, R. (2004) Routine antenatal HIV
testing and its implications for informed
consent. In Ethics and Midwifery, ed. L. Frith
and H. Draper, 2nd edn. Edinburgh: Books
for Midwives, pp. 7490.
Bennett, R. (2007) Routine antenatal HIV testing
and informed consent: an unworkable
marriage? Journal of Medical Ethics, 33: 4468.
Berlin, L. (1999) Malpractice issues in radiology:
screening versus diagnostic mammography.
American Journal of Roentgenology, 173: 37.
Bratton, A. J. (2007) Deputies seize baby to test
blood against parents will. Iowa State Daily, 22
October.
Brett, J., Austoker, J. and Ong, G. (1998) Do
women who undergo further investigation
for breast screening suffer adverse
psychological consequences? Journal of
Public Health Medicine, 20: 396403.
British Medical Association (2005) Population
Screening and Genetic Testing: A Briefing on
Current Programmes and Technologies.
London: BMA. Available at:
http://www.bma.org.uk/images/
ScreeningBriefing2_tcm41-20993.pdf
(accessed: 04/04/11).
Burris, S. and Gostin, L. O. (2004) Genetic
screening from a public health perspective:
three ethical principles. In A Companion to
Genethics, ed. J. Burley and J. Harris. Oxford:
Blackwell, pp. 45564.
Chadwick, R., ten Have, H., Husted, J., et al.
(1998) Genetic screening and ethics:
European perspectives. Journal of Medicine
and Philosophy, 23: 25573.
Chadwick, R., ten Have, H., Hoedemaekers, R.,
et al. (2001) Euroscreen 2: towards
community policy on insurance,
commercialization and public
awareness. Journal of Medicine and
Clarke, A. (1995) Population screening for

genetic susceptibility to disease. British
Clarke, A. J. (1997) Newborn screening. In
Genetics, Society and Clinical Practice, ed.
P. S. Harper and A. J. Clarke. Oxford: Bios
Scientific Publishers, pp. 10717.
Clarke, A. (2007) Genetic counselling. In
Principles of Health Care Ethics, 2nd edn, ed.
R. E. Ashcroft, A. Dawson, H. Draper and
J. R. McMillan. Chichester: John Wiley,
pp. 42734.
Clayton, E. W. (2004) Genetic testing and
screening: newborn genetic screening. In
Encyclopedia of Bioethics, 3rd edn, ed. S. G.
Post. New York: Macmillan Reference,
pp. 10047.
Department of Health (2005) Concordat and
Moratorium on Genetics and Insurance.
http://www.dh.gov.uk/en/
Publicationsandstatistics/Publications/
PublicationsPolicyAndGuidance/
DH_4105905 (accessed: 27/11/07).
de Zulueta, P. and Boulton, M. (2007) Routine
antenatal HIV testing: the responses and
perceptions of pregnant women and the
viability of informed consent. A qualitative
study. Journal of Medical Ethics, 33: 32936.
European Society of Human Genetics, Public
and Professional Policy Committee
(2003) Population genetic screening
programmes: technical, social and ethical
issues. European Journal of Human Genetics,
11(Suppl. 2): S5S7.
Faden, R., Chwalow, A. J., Holtzman, N. A. and
Horn, S. D. (1982a) A survey to evaluate
parental consent as public policy for neonatal
screening. American Journal of Public Health,
72: 134752.
Faden, R., Holtzman, N. A. and Chwalow, A. J.
(1982b) Parental rights, child welfare and
public health: the case of PKU screening.
American Journal of Public Health, 72:
1396400.
Fink, L. and Collins, F. S. (2000) The human
genome project: evolving status and
emerging opportunities for disease
prevention. In Genetics and Public Health in
the 21st Century, ed. M. J. Khoury, W. Burke

and E. J. Thomson. New York: Oxford
Fost, N. (1999) Ethical implications of screening
asymptomatic individuals. In New Ethics for
the Publics Health, ed. B. Steinbock and D. E.
Beauchamp. New York: Oxford University
Press, pp. 34452.
Foster, P. and Anderson, C. M. (1998) Reaching
targets in the national cervical screening
programme: are current practices unethical?
Friedman Ross, L. (2002) Genetic testing of
children: who should consent? In
A Companion to Genethics, ed. J. Harris and
J. Burley. Oxford: Blackwell, pp. 11426.
Gostin, L. O. (2000) Public Health Law: Power,
Duty, Restraint. Berkeley, CA: University of
California Press.
Gostin, L. O. (ed.) (2002) Public Health Law and
Ethics: A Reader. Berkeley, CA: University of
California Press.
Green, J. M., Hewison, J., Bekker, H. L., Bryant,
L. D. and Cuckle, H. S. (2004) Psychosocial
aspects of genetic screening of pregnant
women and newborns: a systematic review.
Health Technology Assessment, 8(33):
1124.
Gustaffson Stolt, U., Ludvigsson, J., Liss, P.-E.
and Svensson, T. (2003) Bioethical theory
and practice in genetic screening for Type 1
diabetes. Medicine, Health Care and
Guthrie, R. (1986) Newborn screening: past,
present and future. In Genetic Disease:
Screening and Management, ed. P. Carter
and A. M. Wiley. New York: Alan Liss,
pp. 31939.
Harris, J. and Keywood, K. (2001) Ignorance,
information and autonomy. Theoretical
Medicine and Bioethics, 22: 41536.
Hodge, J. G. (2004) Genetic testing and
screening: public health context. In
Encyclopedia of Bioethics, 3rd edn, ed. S. G.
Post. New York: Macmillan Reference,
pp. 101620.
Jennings, B. (2007) Community in public health
ethics. In Principles of Health Care Ethics, 2nd
139
edn, ed. R. E. Ashcroft, A. Dawson,

H. Draper and J. R. McMillan. Chichester:
John Wiley, pp. 5438.
Jepson, R. G., Hewison, J., Thompson, A. G. and
Weller D. (2005) How should we measure
informed choice? The case of cancer screening.
Jepson, R. G., Hewsin, J., Thompson, A. and
Weller D. (2007) Patient perspectives on
information and choice in cancer screening: a
qualitative study in the UK. Social Science
and Medicine, 65: 8909.
Juth, N. and Munthe, C. (2007) Screening:
ethical aspects. In Principles of Health Care
Ethics, 2nd edn, ed. R. E. Ashcroft, A.
Dawson, H. Draper and J. R. McMillan.
Chichester: John Wiley, pp. 60715.
Katz Rothman, B. (1993) The Tentative
Pregnancy: How Amniocentesis Changes the
Experience of Motherhood. New York: W.W.
Norton.
Kayton, A. (2007) Newborn screening: a literature
review. Neonatal Network, 26: 8595.
Kenen, R. and Schmidt, R. (1987) Social
implications of screening programs for
carrier status: genetic diseases in the 1970s
and AIDS in the 1980s. In Dominant Issues in
Medical Sociology, ed. H. D. Schwartz. New
York: Random House, pp. 14554.
Kerruish, N. J. and Robertson, S. P. (2005)
Newborn screening: new developments, new
dilemmas. Journal of Medical Ethics, 31: 3938.
Kerruish, N. J., Campbell-Stokes, P. L., Gray, A.,
Merriman, T. R., Robertson, S. P. and Taylor,
B. J. (2007) Maternal psychological reaction
to newborn genetic screening for type 1
diabetes. Pediatrics, 120(2): 324.
Khoury, M. J., Burke, W. and Thomson, E. J.
(2000) Genetics and public health: a
framework for the integration of human
genetics into public health practice. In
Genetics and Public Health in the 21st
Century, ed. M. J. Khoury, W. Burke and E. J.
Thomson. New York: Oxford University
Press, pp. 324.
Lapp, M., Gustafson, J. and Roblin, R. (1972)
Ethical and social issues in screening for
genetic disease. New England Journal of
Medicine, 286: 112932.
140
Section 2: Issues
Lapp, M. (1983) Ethical issues in testing for

differential sensitivity to occupational
hazards. Journal of Occupational Medicine,
25: 797808.
Laurie, G. (2002) Better to hesitate at the
threshold of compulsion: PKU testing and
the concept of family autonomy in Eire.
National Screening Committee (2003) Criteria

for Appraising the Viability, Effectiveness and
Appropriateness of a Screening Programme.
http://www.screening.nhs/uk/criteria
(accessed: 04/04/11).
Leib, J. R., Gollust, S. E., Hull, S. C. and Wilfond,

B. S. (2005) Carrier screening panels for
Ashkenazi Jews: is more better? Genetics in
Medicine, 7: 18590.
National Screening Committee (2011) Aims and

Objectives of the NHS Downs Syndrome
Screening Programme. London: Department
of Health. Available at: http://fetalanomaly.
screening.nhs.uk/cms.php?folder=2452
(accessed: 04/04/11).
Lippman, A. (1991) Prenatal genetic testing and

screening: constructing needs and
reinforcing inequities. American Journal of
Law and Medicine, 17: 1550.
National Screening Committee (2011) What Is

Screening? London: Department of Health.
Available at: http://www.screening.nhs.uk/
screening (accessed: 5/04/11).
Mann, J. M. (1997) Medicine and public health,

ethics and human rights. Hastings Center
Report, 27(3): 613.
Newson, A. (2006) Should parental refusals

of newborn screening be respected?
Cambridge Quarterly of Healthcare Ethics,
15: 13546.
Marshall, T. and Adab, P. (2003) Informed

consent for breast screening: what should we
tell women? Journal of Medical Screening, 10:
226.
Marteau, T. M., Dormandy, E. and Michie, S.
(2001) A measure of informed choice. Health
Expectations, 4: 99108.
Marteau, T. M. and Kinmonth, A. L. (2002)
Screening for cardiovascular risk: public
health imperative or matter for individual
informed choice? British Medical Journal,
325: 7880.
Muir-Gray, J. A. and Pearce-Smith, N. (2007)
Screening Specialist Library. NHS National
Electronic Library for Health. Available at:
http://www.library.nhs.uk/screening/
(accessed: 22/11/07).
Murray, T. H. (1983) Genetic screening in the
workplace: ethical issues. Journal of
Occupational Medicine, 25: 4514.
National Health Service (2006) NHS Sickle Cell
and Thalassaemia Screening Programme.
London: NHS Antenatal and Newborn
Screening Programmes. Available at:
http://sctscreening.nhs.uk/ (accessed: 04/04/11).
National Health Service Executive (1997)
Population Screening for Prostate Cancer.
London: Department of Health. Prostate
Cancer EL (97)12. Leeds: Department of
Health.
Nijsingh, N. (2007) informed consent and the

expansion of newborn screening. In Ethics,
Prevention, and Public Health, ed. A. Dawson
Press, pp. 198212.
Nuffield Council on Bioethics (1993) Genetic
Screening: Ethical Issues. London: Nuffield
Council on Bioethics.
Parsons, E. P., Clarke, A. J. and Bradley, D. M.
(2003) Implications of carrier identification
in newborn screening for cystic fibrosis.
Archives of Disease in Childhood: Fetal and
Neonatal Edition, 88: F467F471.
Paul, D. (1999) Contesting consent: the
challenge to compulsory neonatal screening
for PKU. Perspectives in Biology and
Medicine, 42: 20719.
Philipkoski, N. (2001) Genetic testing case
settled. Wired Magazine. Available at: http://
www.wired.com/science/discoveries/news/
2001/04/42971 (accessed: 27/11/07).
Pfeffer, N. (2004) If you think youve got a
lump, theyll screen you. Informed consent,
health promotion, and breast cancer. Journal
of Medical Ethics, 30: 22730.
Pollitt, R. J., Green, A., McCabe, C. J., et al.
(1997) Neonatal screening for inborn errors
of metabolism: cost, yield and outcome.
Health Technology Assessment, 1(7).
Presidents Commission for the Study of Ethical

Problems in Medicine and Biomedical and
Behavioral Research (1983) Screening and
Counseling for Genetic Conditions: A Report
on the Ethical, Social and Legal Implications
of Genetic Screening, Counseling and
Education Programs. Washington, DC: US
Government Printing Office. Available at:
http://www.bioethics.gov/reports/
past_commissions/geneticscreening.pdf
(accessed: 04/04/11).
Press, N. and Clayton, E. W. (2000) Genetics
and public health: Informed consent beyond
the clinical encounter. In Genetics and Public
Health in the 21st Century: Using Genetic
Information to Improve Health and Prevent
Disease, ed. M. J. Khoury, W. Burke and E. J.
Press, pp. 50526.
Press, N. and Ariail, K. (2004) Genetic testing
and screening: reproductive genetic testing.
In Encyclopedia of Bioethics, 3rd edn, ed. S. G.
Post. New York: Macmillan Reference, pp.
9961004.
Raffle, A. E. (1997) Informed participation in
screening is essential (letter) British Medical
Journal, 315: 1762.
Raffle, A. E. (2001) Information about
screening: is it to achieve high uptake or to
ensure informed choice? Health Expectations,
4: 9298.
Raikka, J. (1998) Freedom and a right (not) to
know. Bioethics, 12: 4963.
Rhodes, R. (1998) Genetic links, family ties and
social bonds: rights and responsibilities in the
face of genetic knowledge. Journal of
Medicine and Philosophy, 23: 1030.
Riis, P. (1985) Ethical issues in mass screening
procedures. In Ethical Issues in Preventive
Medicine, ed. S. Doxiadis. Dordrecht:
Martinus Nijhoff Publishers, pp. 8489.
Rothenberg, M. B. and Sills, E. M. (1968)
Iatrogenesis: the PKU anxiety syndrome.
Journal of the American Academy for Child
Psychiatry, 7: 68992.
Schulte, P. A. and DeBord, D. G. (2000)
Public health assessment of genetic
information in the occupational setting. In
Genetics and Public Health in the 21st
Century, ed. M. J. Khoury, W. Burke and E. J.
141

Press, pp. 20319.
Sherr, L., Fox, Z., Lipton, M., Whyte, P., Jones,
P., Harrison, U. and the Camden and
Islington Steering Group (2006) Sustaining
HIV testing in pregnancy: evaluation of
routine offer of HIV testing in three
London hospitals over 2 years. AIDS Care,
18: 1838.
Shickle, D. and Chadwick, R. (1994) The ethics
of screening: is screeningitis an incurable
disease? Journal of Medical Ethics, 20: 1218.
Skrabanek, P. (1990) Why is preventive
medicine exempted from ethics constraints?
Stanley, F. (1996) Litigation or science: whats
driving medical decision-making? Australian
Law Reform Commission. Available at:
http://www.austlii.edu.au/au/other/alrc/
publications/reform/reform69/
ALRCR69LITIGATIONORSCIENCE.html
Thornton, J. (1999) Should Health Screening be
Private? London: Institute of Economic
Affairs.
Tudor Hart, J. (2008) A few lessons in screening
for Gordon Brown. British Medical Journal,
336: 123.
van den Berg, M., Timmermans, D. R., ten Kate,
L. P., van Vugt, J. M. and van der Wal, G.
(2005) Are pregnant women making
informed choices about prenatal screening?
Genetics in Medicine, 7: 3328.
van den Berg, M., Timmermans, D. R., ten Kate,
L. P., van Vugt, J. M. and van der Wal, G.
(2006) Informed decision making in the
context of prenatal screening. Patient
Education and Counseling, 63: 11017.
White, C. (2008) Prime minister promises raft of
screening tests. British Medical Journal, 336:
623.
Wikler, D. (1999) Can we learn from eugenics?
Wilfond, B. S. (1995) Screening policy for cystic
fibrosis: the role of evidence. Hastings Center
Report, 25(3, Suppl.): S21S23.
Wilfond, B. S. and Thompson, E. J. (2000)
Models of public health genetic policy
development. In Genetics and Public Health
142
Section 2: Issues
in the 21st Century: Using Genetic

Information to Improve Health and Prevent
Disease, ed. M. J. Khoury, W. Burke and E. J.
Thomson. New York: Oxford, pp. 6182.
Wilson, J. M. G. and Jungner, G. (1968)
Principles and practice of screening for
disease. In Public Health Paper Number 34.
Geneva: World Health Organization.
Wilson, R. M. (2000) Screening for breast and

cervical cancer as a common cause for
litigation. British Medical Journal,
320: 13523.
Wynia, M. K. (2006) Routine screening:
informed consent, stigma and the waning of
HIV exceptionalism. American Journal of
Bioethics, 6: 58.
Section 2
Chapter
Issues
Vaccination ethics
Angus Dawson
Introduction
Ever since it began as an intervention designed to protect against smallpox, vaccination has
been controversial. However, the passion that people bring to debates about vaccination is
not always supported by a fair review of the evidence and issues. In this chapter I will
outline and discuss a few of the key arguments about this important area of public health.
I begin with some clarifications about the limits of this chapter. First, vaccination at its
broadest can be taken to involve some form of artificial stimulation of the immune system
as a response to actual or potential bacterial or virological infection. Vaccination might be
either preventive (given prior to potential infection) or therapeutic (given in response to
infection). This chapter is deliberately termed vaccination ethics as I will restrict my
discussion to priming of the immune system before contact with any disease. This means
we can exclude from this chapter discussion of other forms of immunization such as the
giving of immunoglobulin after possible exposure to, or after infection with, a disease. This
is not because such techniques are unimportant, but because both the use of immunoglobulin and therapeutic vaccination might be thought of as primarily clinical interventions
rather than public health activities. This chapter will concentrate on preventive vaccination
as this is the core controversial issue. Second, the classic image of vaccination consists of an
injection (into the muscle or under the skin). However, in some cases the relevant material
is given orally (and absorbed through the digestive tract) or through a nasal spray, or in the
future it might be delivered in some other way (such as through the consumption of
fortified foods). Lastly, most preventive vaccination is carried out in childhood. This is
because this is the time when the immune system can be primed to the greatest advantage,
but also it is a time of greatest threat to the individual from many diseases. However, it
should not be forgotten that many other vaccinations are carried out with adolescents and
adults. Such vaccination might be for a number of reasons: boosters for childhood vaccinations, because older individuals might be at threat of disease due to travel or because they
are held to be at increased risk for some medical reason (for example, medical conditions
related to immune suppression or general old age) or some expected lifestyle change (for
example, human papilloma virus vaccination prior to sexual activity). I will focus on
childhood preventive vaccination for the rest of this chapter unless I specify otherwise.
I outline three key arguments in the following sections: the nature of our obligation not
to bring about harm to others; the idea of best interests; and a discussion of harms and
benefits. I end with a short discussion of the possible grounds for compulsory vaccination.
143
144
Section 2: Issues
Harm to others
Much contemporary bioethics assumes a broadly liberal set of background commitments.
How might this approach be related to vaccination? Traditional liberals (such as John Stuart
Mill [1859] and Joel Feinberg [1973]) commonly make a distinction between actions likely
to cause harm only to self and those likely to result in harm-to-others. They hold that this
makes a vital difference to the legitimacy of interfering in someones freedom of action. On
this approach third parties have fewer justifiable reasons to intervene on harm-to-self
grounds. While health care professionals may have a duty to warn about a risk of harm
or provide relevant information, any attempt to bring about a vaccination for someones
own good will be met with the charge of paternalism.
What is paternalism? Paternalism can be defined as acting (or not acting) with the
intention of reducing harm or bringing about greater good for the particular individual
affected by the action (or omission). On this definition it is left open whether a paternalistic action is morally justifiable: there are two separate judgments. Is it paternalism? If it
is, is it morally justified? However, many liberals believe that we can use a distinction
between hard/strong or soft/weak paternalism to settle the issue of justifiability. On this
view it is argued that strong paternalism is where an action will overrule the action or
decision of a competent individual. (In the liberals view this is usually held to be
unjustifiable.) In contrast, weak paternalism is where an action is performed on behalf
of an incompetent individual (for example, young child; adult with serious learning
difficulties; adult with dementia, etc.). (In the liberals view this is usually held to be
justifiable.)
Routine childhood vaccinations are usually carried out on very young children, so, if we
are talking about paternalism at all, we are talking about weak paternalism. The issue under
discussion will not be whether someone should make decisions on behalf of children just a
few months old, but who should do so. As we are talking about incompetent individuals
here, we can set aside the harm-to-self argument and concentrate upon harm-to-others
considerations. With young children, harm-to-self considerations end up being, at best,
only a component of a judgment of best interests (as we will see below). In the case of young
children, any argument about harm-to-others considerations related to vaccination will
focus on the potential harm (to third parties) as a result of the parents decision not to
vaccinate their child.
However, note that harm-to-others considerations are likely to be an important consideration in relation to decisions about the vaccination of competent adults as well as
children. For example, an adult traveller might knowingly put others at risk of harm if
they refuse to be vaccinated for a contagious disease. A related issue is whether certain
individuals or groups have special obligations to protect others from harm. For example, in
many parts of the world, health care workers are required to be vaccinated for conditions
such as hepatitis B as a means of reducing the risk of being infected themselves but also of
passing on infections to others. There has been a recent lively discussion about whether
there is an obligation upon workers in care homes to be vaccinated against influenza
(van Delden et al., 2008).
As suggested above, while liberal political philosophy generally leaves it to the individual to make decisions about their own lives, there is an important exception: where your
actions might result in harm-to-others. (Such compulsory powers of detention can be seen
in most jurisdictions and in the International Health Regulations as a means to ensure
Chapter 8: Vaccination ethics
145
that individuals are protected from a potential source of infectious disease, particularly if
treatment is refused.) We can construct a harm-to-others argument related to vaccinations
as follows:
1. Contagious diseases that might result in (more than trivial) harm can be passed on to
others through non-intentional action.
2. This could be prevented through vaccination of any potential source individual in
advance (where a relevant vaccine exists).
3. We have a general moral obligation not to cause harm to others through our own
actions and inactions.
4. Given 1 and 2, an individual can reduce the risk of causing (non-trivial) harm to others
through vaccination for (serious) contagious disease.
Conclusion: given 3 and 4, we are morally obligated to have vaccinations for (serious)
contagious diseases (where available).
Of course there are many issues to discuss in exploring such an argument (Dawson,
2007). However, support for this view comes from the fact that, as mentioned earlier, a
decision not to vaccinate (against contagious diseases) does not just put the non-vaccinated
individuals health at risk. A failure to vaccinate is not like a failure to consent to a blood
transfusion. In the latter case it is only the individual themselves that is harmed, whereas in
the vaccination case others may be harmed as a result of an individuals choice or a parents
decision not to vaccinate their child. In other words, where there are serious public health
issues at stake it is possible to argue that we are under a moral obligation to be vaccinated or
ensure our children are vaccinated on the grounds of potential harm to third parties
(Dawson, 2007).
It should be noted that harm-to-others arguments are not paternalistic, because the
reason for intervention is nothing to do with the good or potential harm relating to the
particular individual we are concerned about. The justification for action is the potential
harm to third parties flowing from the parents decision about their child. This argument is
potentially powerful in the vaccination case, because of the highly contagious nature of
many vaccine-preventable childhood diseases. However, some will object to such an
argument, because they will say that the presumption in favour of parental autonomy
should not be overturned on these grounds in relation to vaccination, perhaps because the
risk of harm resulting from a particular instance of non-vaccination will be too remote.
However, even if this is true, the harm-to-others argument will become stronger the more
imminent and the greater the threat of harm resulting from the decision.
The best interests argument

Liberal democracies, rightly, value individual liberty. As a result competent adults are
usually considered to be the appropriate persons to make decisions about their own health
care treatment. This freedom is usually extended to parents in relation to decisions about
their childrens health. Such a position could be supported on a number of different
theoretical and practical grounds. For example, it is sometimes argued that a parent might
be considered to know better than anyone else what is in an individual childs best interests,
because they can reasonably be expected to know that particular child better than anyone
else. They are also likely to know what will benefit their child, and be aware of the societal
and cultural context within which the child will be raised (and so, what is likely to make that
146
Section 2: Issues
childs life go well). Parents might also argue that such decisional authority is part of what it
is to be a parent. While this latter claim may consist of an assertion of ownership or control,
it need not do so; for example, parents might, alternatively, link the idea to the responsibilities they have as a parent. Ultimately, however such a view is justified, there seems a
reasonable claim at its root, as it will be the parents that have to deal with any negative or
positive consequences that result from any decision made. Another possible argument in
support of parental authority might be based on the idea that the family is a private
institution, and that it is inappropriate for the state to intervene in decision making relevant
to the family, except in exceptional circumstances. Each of these claims has different merits
and problems. Luckily, we do not need to debate them here. We can just accept that liberal
philosophy, however supported, will normally give parents decisional authority in relation
to their children.
How does this relate to vaccinations? Parents generally make decisions about childhood
vaccinations. Is it appropriate for parents to refuse vaccination on behalf of their child? The
child has no say in the matter, and by the time they are competent, damage might have been
caused to the child as a result of the parents decision. While most will agree that the
presumption in favour of parental authority can be overturned in cases where it is a matter
of potential life or death, or serious and significant harm, is there a role for the state to step
in to protect the child from parental decisions and provide vaccinations? Certainly such a
power is used in other health and social contexts. Let us focus on pre-school vaccinations. If
an older child is held to be competent, then we can treat them as we do adults. An argument
can be constructed in favour of vaccinations on the grounds of best interests as follows:
1. Medical decisions about incompetent patients should be made on the basis of what is in
their best interests (where prior wishes are unknown or non-existent).
2. Pre-school infants are incompetent (and have no prior wishes).
3. Therefore, decisions about the medical care of infants should be made on the basis of
what is in their best interests.
4. Best interests in relation to infants should be determined by seeking to balance the
potential harms and benefits of possible actions and inactions.
5. Where the parents make a decision about an infants care which is likely to result in
substantial risk of significant harm to that infant then third parties (such as the state)
have an obligation to intervene to protect the infant from the consequences of that
decision.
6. Given 4, what is in the best interests of infants in relation to vaccinations is to be decided
by seeking to balance the harms and benefits associated with vaccination versus nonvaccination.
7. Given 3, 5 and 6, where it is in an infants best interests to be vaccinated (or not
vaccinated) and the parents decide the other way then the state (or other legitimate third
parties) have an obligation to ensure that the infant is protected from the consequences
of such a decision.
Conclusion: parental decision making about childhood vaccinations can be overruled
legitimately in at least some cases.
This argument requires a lot of explanation and discussion over issues such as what
constitutes best interests. All I suggest here is that it might be possible to construct such an
argument assuming that parents do not by definition decide what is in their childs best
147
interests: that is, parents can in fact be error about what is in their childs best interests
(Dawson, 2005). On this view best interests are decided on the basis of an overall welfare
judgment, and on at least some occasions other parties may step in to ensure that children
are protected from the consequences of their parents decision making.
A possible objection is that a judgment about best interests is always made in relation to
an individual child. In circumstances where herd protection exists in a population, it looks
as though a judgment about best interests may favour non-vaccination (assuming there is
any possibility of harm resulting from that vaccination). Some might argue that there is a
potential issue of justice here and that such parental decision making is essentially free
riding upon the actions of others (Cullity, 1995). There is a sense in which this is true, but it
is not clear that objections based on free-riding are sufficient to impose an obligation to
vaccinate in such circumstances (Dawson, 2007). In addition, it should not be forgotten that
there will be individual benefits from vaccination as well (for example, the child might
travel to another population where herd protection does not exist) and there will always be
strong pragmatic reasons in favour of vaccination even where herd protection exists
(Dawson, 2005, 2007).
Harms and benefits

Arguments related to the balancing of harms and benefits are important to vaccination
policies. You dont need to be committed to consequentialism to accept this, and it is
important to see that this is not the only possible moral argument about vaccination. The
vital and difficult issue to decide is: what are the relevant harms and benefits? In this section
I will begin by suggesting some relevant considerations, and then argue that in vaccination
policy (like in many preventive programmes) the focus cannot just be on harms and benefits
in relation to particular individuals but that any such judgment needs to take into account the
consequences for populations not just individuals. This complicates the ethical discussion.
Vaccination brings potential benefit to the individual receiving the vaccination
because they are less likely to develop that particular disease if they come into contact
with it. However, there is also an important benefit to society if sufficient members of
that population are vaccinated to create herd protection (Paul, 2004). Herd protection
means that all members of the community are at reduced risk of attack by infectious
disease. This is because if such a disease enters the population it is far less likely to
become an epidemic or pandemic as any diseased individual is unlikely to pass on the
infection if the surrounding individuals have been vaccinated prior to contact. In
addition, any unvaccinated individuals in the population are better protected, as they
are less likely to come into contact with an infected individual. (At least some unvaccinated individuals are not at risk due to their own decisions. For example, neonates might
not be old enough to be vaccinated; the ill and those with compromised immunity might
be unvaccinated for sound medical reasons; vaccination might fail or be insufficient to
give immunity; scheduled vaccinations might have been missed due to population
movement, etc.) Herd protection offers all these groups their only vaccine-related
protection against the risk of disease.
The supporters of vaccination will argue that few other medical interventions have
had such a positive impact upon the worlds health (CDC, 2006). Smallpox could not
have been eradicated without vaccination and poliomyelitis (despite some problems) is
148
Section 2: Issues
close to eradication as a result of global efforts and a sustained vaccination programme.

Much effort has gone into the preparation of a vaccine in response to the latest influenza
pandemic as a direct means of preventing the spread of such a disease. If routine
vaccination for all recommended diseases were available to everyone across the world
the impact on global health would be hugely significant (WHO, 2002). Despite this,
some authors have expressed scepticism about the scale of the contribution of vaccination to the dramatic reduction in childhood mortality from infectious disease since the
mid-nineteenth century (Keown, 1976). They suggest that improvements in nutrition,
water quality and sanitation have had more impact upon the disease-related mortality
figures. I would argue that these factors have certainly contributed to this fall, as a
healthy child is better able to fight infection. However, Keowns thesis almost certainly
underestimates the impact of vaccination in preventing or reducing the force of waves of
disease. The (additional) impact of vaccination can be seen by looking at infection rates
for such diseases as diphtheria in the UK where the number of infections (and deaths)
plummeted after routine vaccination was introduced in the 1940s and 1950s (Salisbury
and Begg, 1996: 68).
Ironically, in the developed world it is the very success of vaccination in maintaining
unprecedented low rates of many infectious diseases that has proved to be part of the reason
why vaccination has become such a controversial issue at times. Few adults in the developed
world have any experience of previously very common infectious diseases (for example,
diphtheria, measles, pertussis), and so it is easier to downplay or ignore the risks of
such diseases and overemphasize any potential risks from a vaccine. Of course, vaccinations
can cause harm. Such harms range from inflammation and pain at the site of injection
through to anaphylaxis and death. However, any adverse events are rare, and serious
adverse events are very rare indeed for most vaccines. Occasionally, the public can lose
confidence in a particular vaccine as happened in the UK during the 1970s with pertussis
and MMR since 1998. In some cases, with particular vaccines, the risks of vaccination can
be higher than would be acceptable for a routine vaccination programme and this can lead
to poor uptake in target populations (for example, recent smallpox vaccination among first
responders in the USA was held to be disappointing. See Yih et al. [2003]). Media reports,
rumours and misunderstandings can feed these concerns. However, vaccines are subject to
at least the same rigorous development standards as any other medicinal products. In
practice, though, risks tend to be much lower because of the need to administer vaccines
to very large asymptomatic populations. We can distinguish between the perception of risks
and the statistical reality of risks but the former matter as vaccination programmes need to
be acceptable to the target population (Verweij and Dawson, 2004).
Different vaccines have different adverse event profiles depending upon how they are
made. Vaccines can be manufactured using live but weakened (attenuated) pathogens, dead
pathogens, part pathogens, inactivated toxins, or recombinant techniques (a kind of genetic
engineering). In some cases there is a choice of vaccines for the same disease (as with polio)
and they will have different modes of operation and side-effects (Paul and Dawson, 2005).
In other cases, the choice of a particular strain might make a difference to the likelihood of
side-effect as some strains will be more virulent than others (see Kretzschmar et al., 2006,
on smallpox vaccines). In addition, the vaccine will not just contain an active ingredient
designed to induce immunity but also other things including adjuvants (ingredients to
stimulate an immune reaction) as well as preservatives. Sometimes these elements can be
the reason for concern, such as in the recent discussion in the USA about the role of
149
mercury as a preservative in vaccines (Institute of Medicine, 2004). Vaccine development

involves a constant process of refinement and improvement with the aim of trying to reduce
any potential risks to a minimum.
There are some other possible harm-related objections to vaccination that I will leave
to one side. For example, it might be claimed that there is no risk of harm involved in
common childhood diseases, or that the risk of harm from vaccines is disproportionate to
the threat from such diseases. Such arguments tend to be over-generalized or involve
dubious empirical claims. Much of the popular anti-vaccination material available in the
public domain is based on no or poor evidence, with little attempt to consider the issues
fairly (Sorell, 2007). Objections to vaccinations as such will not be considered here
(although I do not think they make any sense, as any harmbenefit judgment can only be
focused on a consideration of an individual vaccine and disease).
As mentioned earlier, it is important that in considering harms and benefits in
relation to vaccination we do not merely focus on how these issues affect individuals
but also how they affect the whole population. This is vital in relation to vaccination
because of the issue of herd protection. Herd protection is a good example of a benefit
that exists at the population level. However, this population benefit is not an instance of
a mere aggregation of individual benefits. Where herd protection exists, the benefit
extends much further than the sub-group of the total population that have been
vaccinated. One way of conceptualizing this population-benefit is in terms of seeing it
as a public good. A public good is a good that cannot be created by any individual alone:
it takes collective efforts. It cannot be broken down into individual goods and distributed among the members of a population. All benefit in a population, if it exists. None
can enjoy it, unless all benefit. There are various problems associated with the creation
and maintenance of public goods, but the importance for this section is that such a
benefit as herd protection cannot be easily entered into any simple harm/benefits
calculation in relation to individuals: the population level is relevant to such deliberations (Dawson, 2007).
It is important to see that this appeal to population benefits does not mean that
individuals need to be sacrificed for the good of the population, just that the relevant
benefits to the population as a whole are a relevant consideration in any deliberations about
harms and benefits. While it can be argued in relation to at least some preventive public
health policies that risks are run by individuals, while benefits accrue to the population, this
is not true of vaccination policies (Dawson, 2004). Vaccinated individuals do benefit from
their participation in the programme; it is just that they will gain an extra benefit if herd
protection exists. Where herd protection is an explicit aim of a vaccination programme and
vaccination is strongly encouraged, it is very important to consider the question of compensation for any possible vaccine-related harm: even the perception of injustice, of unnecessarily sacrificing individuals for the good of the population, should be avoided if possible
(Paul and Dawson, 2005).
In addition to seeing herd protection as a public good, we might also see it as a
common good. Common goods are those goods that are created and maintained
through social interaction in a joint or common project: they are about what we do
together. They often include the kinds of things that are related to the background
conditions for a flourishing life. Common goods are the product of shared social norms
and meanings, and they are not maintained through potential enforcement (unlike at
least some public goods, that may involve state pressure to ensure participation).
150
Section 2: Issues
Common goods in relation to vaccination can perhaps be best seen at certain historical
periods. For example, the pursuit of a vaccination for polio in the USA during the 1940s
and 1950s was largely driven by a collective sense of threat and widespread support
through small charitable donations to groups such as March of Dimes. The discourse of
common goods is not often used these days, for a number of reasons. Talk of common
goods does not fit well with the general focus in todays society upon individual choices
and rights, but this does not mean that such a common or joint sense of activity does
not continue to exist in relation to vaccination. We can see the collective reduction of
risk to a society as a whole, through the creation and promotion of herd protection, as
being a clear example of this (despite the fact that this is not the way it is talked about).
Vaccination, where herd protection is a plausible aim, provides a way for us to think
about responding to common threats to our society and our way of life, and how we as
a group can respond and protect. When I choose to vaccinate my children, I do more
than protect them in an individual sense, as I am also contributing to a collective
project. If I choose not to vaccinate them, I am not only putting them at increased risk
of harm as individuals, but also failing to contribute to the maintenance of a common
good. Of course, it may be objected that parents will make the decision based upon what
is in their childs interests. This is fair enough, but it is worth drawing peoples attention
to the fact that, once again, we can see that choices about vaccination do not merely
have consequences for the direct individuals involved, but for the whole of society
(a society that all such individuals belong to). Our connectedness to others becomes
more and more invisible, and we are starting to lose the language of such social
connections. However, invoking the idea of common goods, and promoting a sense of
public health ethics that contains such a notion, will, arguably, result in a richer and
more socially based view of ethics in general. Such a view would allow us to invoke other
positive, non-individual values such as solidarity, reciprocity and social justice. The idea
of common goods can also be used, in turn, to offer further support for other moral
arguments in favour of vaccination, such as those based on both harm-to-others and
best interests. A society with a declining sense of the values we share is contributing to
an increased possibility of harm, and one with a commitment to common goods is
clearly in the best interests of a child. In conclusion, this suggests that such values ought
to at least be on the table, and we should not just assume that the only relevant values
are those related to individuals, or are those that have come to the fore in medical ethics
over the last forty years (see Dawson, Chapter 1).
Compulsion?
Such arguments based on shared or common values are not going to be acceptable to
everyone. However, one of the interesting things about vaccination is that we can
produce many different arguments about it. Even if someone wants to remain within
the framework of values derived from medical ethics and its dominant liberal discourse,
certain arguments in favour of vaccination remain powerful. For example, on the latter
approach, appealing to both the idea of harm-to-others and best interests might provide
some justification for a moral obligation to vaccinate a child in at least some circumstances. Let us assume this is the case does it follow that compulsion is therefore
justified? It always seems slightly odd to me that many people consider this to be the most
important ethical issue in relation to vaccination. It is, of course, possible to have legally
151
compulsory vaccination, but in many cases this is not necessary. Compulsion may be
tempting for legislators for a number of reasons, such as where public health structures
are weak; there is a lack of trust or little feeling of social solidarity; or where there is a
social or political culture committed to individual freedom as the primary value of
public policy. However, in many parts of the world, such measures are not necessary,
and vaccination rates remain sufficient for herd protection to exist. Rather than seeing
the justifiability (or not) of compulsion as the central issue in vaccination ethics, we can
almost take the fact that this is an issue for public policy as a sign that something has
gone wrong with the sense of values in such a population. A society without what we
might call active social values (that is, social values that have purchase within discourse
about public policy) is a failing society. A society that agrees upon the states central role
as being one of coercing free citizens into acting in various ways (including accepting
their moral obligations) is to accept an impoverished model of social relations: one
where making sense of public health becomes very difficult, if not impossible. There
is an odd paradox here. A society that privileges the value of individual liberty above
all others ends up needing compulsion to ensure reasonable consideration for other
valuable considerations, such as freedom from potential harm, whereas a society that
actively embraces a range of values may have little need for compulsion at all. As a
result, to frame the debate in terms of being in favour or against compulsion is already
to miss the point.
Whatever a reader thinks of this last point, it is important to see that the issue of legal
compulsion is different from that of the existence of a moral obligation. An argument for
compulsion requires a further step to move from our moral condemnation of a parent, to
an act of interference in the family to ensure a certain end. However, in certain circumstances the latter might be justified, for example where the risk of harm was great enough,
or perhaps where a court orders vaccinations after a parental dispute (Dawson, 2005). Of
course, compulsion in the everyday sense implies the use of force or legal sanctions such
as fines. However, related activities might count as indirect compulsion and these can
cover a range of cases from the requirement to have vaccinations before enrolment in
school, to a presumption in favour of vaccination with little possibility to opt out and little
attempt to offer the relevant information for an informed consent. Of course, direct
compulsion will generally be an action of last resort. In many cases there will be strong
arguments against, because if it is not proportionate it might actually result in a decline in
vaccination uptake. However, some people will be opposed to compulsion on principle, so
in the rest of this section I will consider some categorical objections to compulsory
childhood vaccination.
First, it might be argued that it can never be appropriate to overrule parents
decisions about vaccinations, because vaccination is a preventive rather than a therapeutic measure, and in most cases the diseases are trivial and low risk. Invoking the
powers of the state to interfere in the family in such a case, it could be argued, is just
inappropriate. However, it is not clear that this argument will succeed as any attempt to
draw a morally significant difference between an action based upon the fact it is either
preventive or therapeutic is potentially problematic. It can be argued that it should be a
balance between all relevant harms and benefits that matters (Dawson, 2004). In
addition, any risk of disease must be balanced against the risk of vaccination. As
previously suggested, any risks of vaccination related to common diseases are low.
While the risk of contracting such a disease in the developed world is also usually
152
Section 2: Issues
low, if contracted, the potential impact of many of these diseases should not be
underestimated. (Neither should it be forgotten that the calculations of such risks and
benefits might vary widely depending upon the background societal conditions. In many
parts of the developing world childhood diseases are endemic and millions of children
die each year from vaccine-preventable diseases.)
Second, it might be argued that parents can invoke a right to refuse this medical
intervention. Presumably such a claim would have to involve some kind of justification
for seeing this particular right as taking precedence over other rights, including a right of
the child to be protected from potential harm. I do not have space to consider rights
discourse in any detail here, but it is worth pointing out that the parents refusal in this case
is not governing their own care, but that of their child. This may make a significant
difference to the case. Even if such a right to refuse treatment is held to be central to the
debate about vaccinations, it is only likely to block any best interests argument, leaving the
harm-to-others argument untouched.
Third, parents might use another form of appeal to rights. In this case, they may argue
that the child has a right to bodily integrity, and that this will be transgressed in the case of
compulsory vaccination. Once again, the important thing about rights is that there are
many such rights and, when they are invoked, it is a requirement of their defender to
produce an argument not just to say why we have such a right, but explain why that
particular right is supposed to take precedence in our moral deliberations. While such an
argument can, no doubt, be produced, it can surely be contested. Once again, while such a
right might, at most, have some claim in relation to arguments about best interests, it is not
clear how it might deflect any harm-to-others argument.
While all of these objections may be important, as the risk of harm-to-others from the
relevant disease grows, it becomes more and more difficult to hold that any such rights take
precedence. Even in relation to best interests, we might doubt whether these arguments are
decisive. This is because any deliberations about best interests mean that no parent has an
absolute right to do as they want with a child. Parents have a great deal of leeway about how
they choose to bring up their children, but there are serious constraints upon what will
count as being legitimately in the best interests of any child.
Conclusion
These different arguments will work in different ways depending on the nature of
the particular disease, vaccine and potential recipient(s). Determining the most relevant
and effective vaccination policy is not easy. However, it will certainly involve a consideration of the risks of harm and benefits from vaccination and non-vaccination, as well
as more theoretical arguments about harm-to-others and best interests. Considering
the ethical issues related to vaccination requires new thinking because the focus of traditional bioethics has been on the individual. This is unhelpful as it misses the fact that
vaccination is not just about individuals and their choices but population health as well.
Acknowledgement
This chapter is a revised version of a chapter, used with permission, that first appeared as:
Dawson, A. (2007) Vaccination ethics. In Principles of Health Care Ethics, 2nd edn, ed.
R. Ashcroft, A. Dawson, H. Draper and J. McMillan. Chichester: Wiley.
References
CDC (2006) Vaccine preventable deaths and the
global immunization vision and strategy,
20062015. MMWR, 12 May, 55(18): 51115.
Available at: http://www.cdc.gov/mmwr/
preview/mmwrhtml/mm5518a4.htm
(accessed: 1/9/06).
Cullity, G. (1995) Moral free riding. Philosophy
and Public Affairs, 24(1): 334.
Dare, T. (1998) Mass immunizations
programmes: some philosophical issues.
Bioethics, 12(2): 12549.
Dawson, A. (2004) Vaccination and the
prevention problem. Bioethics, 18(6): 51530.
Dawson, A. (2005) The best interests argument
and childhood vaccinations. Bioethics, 19(2):
188205.
Dawson, A. (2007) Herd protection as a public
good: vaccination and our obligations to
others. In Ethics, Prevention and Public
Health, ed. A. Dawson and M. Verweij.
153
Mill, J. S. (1859) On Liberty. Reprinted 1974.

Harmondsworth: Penguin.
Paul, Y. (2004) Letter: Herd immunity and herd
protection. Vaccine, 22: 3012.
Paul, Y. and Dawson, A. (2005) Some ethical
issues arising from polio eradication
programmes in India. Bioethics, 19(4):
393406.
Salisbury, D. and Begg, N. (eds) (1996)
Immunisation Against Infectious Diseases.
London: HMSO.
Selgelid, M., Battin, M. and Smith, C. B. (2006)
Ethics and Infectious Disease. Oxford:
Blackwell.
Sorell, T. (2007) Parental choice and
expert knowledge in the debate
about MMR and autism. In Ethics,
Press.
Feinberg, J. (1973) Social Philosophy. Engelwood

Cliffs, NJ: Prentice-Hall.
van Delden J. J. M., Ashcroft R., Dawson A.,

Marckmann, G., Upshur R. and Verweij,
M. F. (2008) The ethics of mandatory
vaccination against influenza for health care
workers. Vaccine, 26(44): 55626.
Institute of Medicine (2004) Immunization

Safety Review: Vaccines and Autism.
Washington, DC: National Academies Press.
Verweij, M. and Dawson, A. (2004) Ethical

principles for collective immunisation
programmes. Vaccine, 22: 31226.
Kretzschmar, M., Wallinga, J., Teunis, P., Xing, S.

and Mikolajczyk, R. (2006) Frequency of
adverse events after vaccination with different
Vaccinia strains. PLoS Medicine, 3(8).
World Health Organization (2002) State of the

Worlds Vaccines and Immunization. Geneva:
WHO.
McKeown, T. (1976) The Role of Medicine:

Dream, Mirage, or Nemesis. London: The
Nuffield Provincial Hospitals Trust.
Yih, W. K., Lieu, T. A., Rgo, V. H., et al. (2003)

Attitudes of healthcare workers in U.S.
hospitals regarding smallpox vaccination.
BMC Public Health, 3: 20.
Section 2
Chapter
Issues
Environment, ethics and public

health
The climate change dilemma
Anthony Kessel and Carolyn Stephens
In this chapter we explore the relationship between the environment, ethics and public
health (Kessel, 2006). To do this we take the theme of climate change and public health
(with climate representing the natural environment), and we examine the relationship
between our climatic environment and public health ethics. We illustrate the changing
relationship between public health and the environment, and examine the place of ethics in
the relationship between the environment and public health. In particular, we critique the
status of utilitarianism as a moral foundation for public health. The importance of environmental justice, environmental philosophy and development ethics to the future of public
health are discussed.
Background: definitions and methods

As an introduction to this chapter it is important to refer to some difficulties around
definitions, and also to briefly describe the methods used. To philosophers definitions are
often key to their debates. For one philosopher in particular, Ludwig Wittgenstein, issues to
do with definitions are at the root of all philosophical problems (Elliott, 2001). A look at the
words used in the title of this chapter readily illustrates the challenges of writing on this
topic. To Socrates, ethics was about how we ought to live, and why. More contemporarily,
however, ethics has been divided into normative ethics (systematic examination of organized theories) and speculative (or reflective) ethics (Honer et al., 2006). In this chapter we
look at moral theories as well as adopting a more speculative approach.
The environment, likewise, has different meanings to different people in different
contexts, all of which are relevant to health debates. There is the physical (sometimes
called the built) environment, for instance buildings and urban infrastructure. The social
environment, in contrast, tends to refer to demographic factors in populations (age
structures, ethnic groups, inequalities), and the natural environment usually relates to
aspects such as air, water, green spaces (for example, parks, forests), and biodiversity. In
this chapter we are largely discussing the natural environment. Within the public health
literature the term environmental health has tellingly been used as will be returned to
later in this chapter largely to capture the impacts of man-made contamination of the
environment on human health (for example, chemical hazards, toxins), rather than to
describe the intrinsic health of the natural environment per se (Environmental Health
Commission, 1997; Jukes, 1999).
154
Chapter 9: Environment, ethics and public health
155
What is meant by public health is also open to debate. In 1988, Donald Acheson led an
inquiry into the future of public health in the UK and drawing on Winslows 1920
definition described public health as the art and science of preventing disease, promoting
health, and prolonging life through organized efforts of society (Department of Health,
1988: 1); the Institute of Medicine in the USA had similarly drawn on Winslow in a report
of the same year (Institute of Medicine, 1988). The health historian Dorothy Porter has, in a
related vein, labelled public health as collective action in relation to the health of populations (Porter, 1999: 4).
In 2002, the Faculty of Public Health Medicine (UK) was still using Achesons definition, and added that public health is . . . concerned primarily with health and disease in
populations, complementing, for example, medical and nursing concerns for the health of
individual patients and its chief responsibilities are monitoring the health of a population,
the identification of its health needs, the fostering of policies which promote health, and the
evaluation of health services (FPHM, n.d.). What this definition illustrates is the perceived
difference between professional public health (practitioners trained in the speciality) and
non-professional public health (anyone working to improve population health). There are
other qualitative distinctions that characterize the various interpretations of public health,
some of which relate to the different histories of public health in different countries, but in
this chapter we predominantly use Porters definition, except when referring specifically to
the professional component.
This is an inter-disciplinary chapter. The over-arching academic framework is ethical
theory and philosophy, but discussions about history, public health science and policy are
integrated within the analysis (Baum, 1995; Green and Thorogood, 2004).
Climate change and public health

In the approach taken to dealing with climate change, the debate about our environment has provided the opportunity to rethink the relationship between mankind
and nature, and the moral dimensions of public health theory and practice. Often
referred to synonymously as greenhouse warming, climate change presents an instance
of the health effects of western lifestyles being borne by those at a distance in time and
place. Unlike those affected by, say, passive smoking, those affected by climate change
have little or no connection with the perpetrators, yet are left with the consequences.
And this raises fundamental questions about the geographical, temporal and moral
boundaries of public (health) responsibilities, as well as the place of utilitarianism in
public health theory.
The scientific background to the process of climate change is well described in the
literature, and will not be repeated here (McMichael, 1995; IPCC, 1996, 2001; Hamilton,
1999). The consequences of climate change to human health are similarly well articulated
and most conveniently grouped into direct and indirect effects (McMichael, 1996). The
direct effects result from increased exposure to thermal extremes (heat waves and severe
cold) and associated climatic changes (deaths in vulnerable groups, domestic violence, civil
disturbances and riots, floods, storms, cyclones, hurricanes and bushfires, injuries, psychological disorders and infectious diseases). Indirect effects stem from disturbances to complex ecological systems (changes in the ranges and activity of vectors and infective parasites
such as those responsible for malaria, dengue fever, trypanosomiasis and the viral encephalitides), and changed food productivity will result in malnutrition, hunger, impaired
156
Section 2: Issues
child development and growth, with increased morbidity and mortality (Epstein, 1995;
McMichael and Haines, 1997).
Climate change and public health philosophy

From the perspective of public health philosophy, what is fascinating about climate change
is that it throws open three new aetiological dimensions to population health and disease.
First, the causes of greenhouse warming and the resultant climate change and its health
effects are anthropogenic. Excluding lifestyle diseases which an individual predisposes
him- or herself to through personal activity1 there are plenty of examples of illnesses
created by human activity, such as occupational cancers or the passive smoking example
mentioned earlier. But in these situations only the populations that create the environmental hazard experience the consequences.
What is different about climate change is that certain communities (and the individuals
within them), through their adopted activities, will affect the health of other communities
that may well not have taken up such activities. And this opens up interesting, hitherto
unexplored questions about personal responsibility, and also about the relationship between
public responsibility and how this is expressed through policies such as those concerning
public health. In other words, how do such responsibilities fit into the public health
philosophy and practice of the perpetrating communities?
Second, the health effects of climate change are, to a substantial degree, likely to
impact at a large geographical distance from their source. Aside from the equity
issues relating to the differential impact and ability to mitigate or adapt accordingly
which are looked at later in this chapter it is difficult to think of any other
example2 in which the activities of one community could so connectedly affect the
health of a population afar. War is perhaps the closest parallel. Related to this point,
the third new aetiological dimension that climate change throws up is that the health
impacts of current (and past) activities will likely be the burden of generations to
come. Once again, it is difficult to recall any similar example in the history of public
health. So the question arises again of how do these spatial and chronological dimensions fit into the public health philosophy and practice of the perpetrating
communities?
Like much public policy, public health is informed heavily by one moral theory:
utilitarianism. Yet utilitarianism is problematic and seems out of touch with the worlds
current problems. As a guide for both personal and public morality, traditional utilitarianism appears anachronistic. Indeed, the roots of all the new dimensions of health
effects of climate change outlined above, can be tracked against the deficiencies of
utilitarian theory.
So the next section looks specifically at utilitarianism, its moral limitations and the
relevance of these to climate change and public health philosophy. After that, the first major
challenging moral framework for public health is considered, one based on John Rawlss
vision of social justice.
1
2
Personal choice, however, such as the ability to stop smoking, may be affected by factors such as
employment status and social support, both of which are linked to deprivation.
Apart from other economic activities related to human lifestyles.
157
Utilitarianism, climate change and public health

Utilitarianism falls into the consequentialist class of moral theories, in which the rightness
or wrongness of an action, or rule, is determined by the consequences of that action or rule.
Despite relentless ongoing criticism utilitarianism has proved a remarkably tenacious moral
theory, the corner-stone to liberal democracy, and both its persistence in and significance to
western political philosophy inevitably tie utilitarianism to ethical issues in public health.
Utilitarianism became applied politically in the eighteenth century, and is most famously associated with Jeremy Bentham (17481832) and, a little later, John Stuart
Mill (180673). But utilitarianism had antecedents, and the main tenets of the theory
were laid down earlier by philosophers such as John Locke (16321704) and David Hume
(171176).
Bentham was a lawyer and was most interested in the relevance of his ideas to
legislation. This element connected him strongly to one of his followers, Edwin Chadwick,
because of a shared belief in improving the lot of those worst off through reform. However,
Benthams concept of equality was strikingly at odds with that of certain successors, such as
Marx and Engels, who provided a very different explanation for the historical processes
determining how inequalities arose and what should be done to redress them. To Bentham,
equality formed the basis of a calculus3 in which each individual was of equal value (Russell,
1991).
Benthams theory was founded on two linked principles, the principle of association and
the principle of utility. The principle of association was a deterministic account of linked
mental occurrences, akin to the modern conditioned reflex but without the physiology.
The principle of utility, or the greatest-happiness principle is, however, what Bentham is
best known for, and rests on the premise that what is good is pleasure, and what is bad is
pain. Bentham came to this position through the belief, articulated in his 1789 Introduction
to the Principles of Morals and Legislation, that human beings are subject to, and slaves to,
two poles of sensation (Scruton, 1996). Bentham took the leap of designating happiness as
the moral goal, and his principle of utility . . . approves or disapproves of every action
whatsoever, according to the tendency which it appears to . . . augment or diminish the
happiness of the party whose interest is in question; or what is the same thing in other
words, to promote or to oppose that happiness (Bentham, quoted in Rachels, 1993: 91).
Extrapolated from the individual to the larger, social domain, the principle of utility
states that . . . the greatest happiness of all those whose interest is in question . . . [is] . . . the
only right and proper and universally desirable end of human conduct (Bentham, quoted
in Scruton, 1996: 224). So, one set of affairs is better than another if there is a greater
balance of pleasure over pain, or a smaller balance of pain over pleasure. Empiricism was
thus brought firmly into the foreground, as the right action could in theory at least be
determined by summing up individual experiences of these two sensations. This process of
quantification was Benthams felicific calculus,4 in which the audience to be considered
was all those affected by the action, each counting equally.
In his 1863 book Utilitarianism, John Stuart Mill, like Bentham, extended moral
consideration to the whole of sentient creation but, differently, made qualitative
3
4
The calculus referred to a calculation, rather than the modern understanding as a particular method
in mathematics.
Also known as the optimific, or hedonic, calculus.
158
Section 2: Issues
distinctions between pleasures (Harris, 1997: 128). Although critiques of modern utilitarianism, and their relevance to public health, will be looked at a little later, it is
necessary to point out here a serious problem with the theorys early forms. That
is, utilitarianism as depicted by both Bentham and Mill, makes a conceptual leap of
inferring from what is in the world to what ought to be. Bertrand Russell (1991: 744)
is straightforwardly damning:
John Stuart Mill, in his Utilitarianism, offers an argument which is so fallacious that it is hard to
understand how he can have thought it valid. He says: Pleasure is the only thing desired; therefore
pleasure is the only thing desirable. He argues that the only things visible are things seen . . . and
similarly the only things desirable are things desired. He does not notice that a thing is visible if it
can be seen, but desirable if it ought to be desired. Thus desirable is a word presupposing an
ethical theory; we cannot infer what is desirable from what is desired.
With this seemingly mistaken leap in mind, it is perhaps surprising that utilitarianism
has endured. Yet few would demur that utilitarianism strongly underpins much of
contemporary moral and political thinking and action. Good national policies are
judged to be those that increase overall wealth, the modern euphemism for the greatest
happiness, and good public health policies are judged to be those that demonstrably
improve population health.
However, the pursuit of happiness has become something closer to a taken-forgranted, a lifelong endeavour shaped by society, unquestionably accepted and followed.
And the inevitably elusive chase finds happiness disguised as, inter alia, healthism, obsession with risk aversion and consumerism (Forde, 1998; Porter, 1999). Not surprisingly
therefore, faced with the significant lifestyle changes that would be required to offset climate
change and its global health effects, most individuals do not really want to sacrifice or
undermine pursuit of their own happiness-oriented goals, despite superficial environmental
soundings to the contrary.
Picking up on this, the philosopher Alasdair MacIntyre has put the blame for todays
moral ambivalence squarely on the shoulders of utilitarianism, and the selfishness it has
engendered. He (1989: 243) argues that . . . the individualism of modern society and the
increasingly rapid and disruptive rate of social change brings about a situation in which for
increasing numbers there is no overall shape to the moral life but only a set of apparently
arbitrary principles inherited from a variety of sources. Further, MacIntyre (1989: 2378)
holds the early utilitarians directly responsible for todays woes, and emphatically questions
the price to be paid:
The concept of happiness is, however, morally dangerous in another way; for we are by now well
aware of the malleability of human beings, of the fact that they can be conditioned in a variety of
ways into the acceptance of, and satisfaction with, almost anything. That men are happy with their
lot never entails that their lot is what it ought to be. For the question can always be raised of how
great the price is that is being paid for the happiness.
Critiques of utilitarianism, and relevance to climate change

The first criticism is that utilitarianism, in its classical or present economic form, necessitates the enumeration and summation of utilities in some shape or form. And utilitarianism
then uses the results of this process as the moral basis to guide actions or policies. In its
early form an obvious difficulty was how to quantify happiness, along with the problem
159
outlined earlier of whether happiness is an appropriate moral goal in the first place.
Preference- and welfare-based utilitarianism circumvent the latter issue, but do not get
around the issue of quantification.
In fact, modern versions of utilitarianism do precisely the opposite. They are reliant,
perhaps more than ever, on empirically obtained information as the basis for acting. They
place, metaphorically, all the moral eggs in the basket of a positivist conception of science.
In a classic book containing essays for and against utilitarianism, the philosopher Bernard
Williams describes contemptibly the appeal of utilitarianism in that it picks up little of the
worlds moral luggage, preferring instead to place huge demands on information because
even insuperable technical difficulty is preferable to moral unclarity, no doubt because it is
less alarming (Williams, 1991: 137).
This moral side-step may be economically and politically advantageous, in the short
term at least, but it raises almost insuperable problems for climate change. There may be a
general consensus now on the scientific proof that climate change is actually happening, but
there is no agreement about what should be done about it (Menne and Bertollini, 2005;
DEFRA, 2006).
More subtly, however, centralizing utilitarianism in the climate change debate raises the
important point that some elements are more amenable to scientific enquiry and analysis
than others. The environment, for example, is excluded from investigation. After all, how
do you place a utility function on the value individuals may, or may not, place on retaining
a beautiful area of wilderness, or an unpolluted atmosphere? It is clearly easier to calculate
the economic costs of climate change and the mitigation strategies to prevent it, than to
reliably quantify the health impacts or environmental utilities, so creating a bias in areas of
consideration; and this does not even touch the question of how to compare different
utilities. This is a fundamental issue recognized by Williams (1991: 148):
For to exercise utilitarian methods on things which at least seem to respond to them is not
merely to provide a benefit in some areas which one cannot provide in all. It is, at least very often,
to provide those things with prestige, to give them an unjustifiably large role in the decision,
and to dismiss to a greater distance those things which do not respond to the same methods. Just
as in the natural sciences, scientific questions get asked in those areas where experimental
techniques exist for answering them, so in the very different matter of political and social
decision weight will be put on those considerations which respected intellectual techniques can
seem, or at least promise, to handle.
The second criticism of utilitarianism, and its framing of climate change policies, relates to
proximity. As has been described, classical and modern versions of utilitarianism involve
quantification and summation of individual utilities, whether happiness, preferences or
interests. But who should be included in the arithmetic? Bentham and Mill predicated that
the pleasures and pain of all affected by the action, the audience, should be considered,
including to a lesser degree non-human animals.
Although circumstances in the nineteenth and twentieth centuries were more contained
than today by the technological allowances of the time, the focus on individuals (the
audience members) close in space and time applies to both eras. This is because, as before,
the utilitarian calculus favours consideration of issues around which there is greater
certainty. The philosopher Robert Goodin (1997: 247) highlights that utilitarians may want
to include the utilities of all those affected by an action in any given calculation, but in
practice it is unlikely:
160
Section 2: Issues
. . . utilitarians can go on to say, perfectly properly, that as a purely pragmatic matter their calculations
will often lead us to show some apparent favouritism toward those near and dear to us. It is easier
to know what people nearby need, and how best we can help; . . . Those are purely contingent,
pragmatic considerations, to be sure. In the ideal world, they may be absent. But in the real world,
they are powerfully present.
This creates special problems for policies relating to climate change. At the national level,
and at the local level within countries, policy makers usually take into account the effects of
their policies on individuals within their immediate boundaries. Climate change would
appear to open up the borders by demanding that those from afar are considered too. But it
is hard at present to know how to incorporate such requirements, and it remains difficult to
believe that such tough decisions will be made by politicians with national, party, and their
own interests at heart. The limited concessions to date in the high profile international
climate change meetings affirm the somewhat bleak outlook.
So far, the focus of this second criticism has related to geographical proximity. But
utilitarianism also has a temporal bias. The utilitarian philosopher J. J. C. Smart (1991)
argues that it is impossible to envisage the total future situation because it stretches to
infinity. According to Smart it is unnecessary in practice to consider very distant consequences, as these in the end approximate rapidly to zero like the furthermost ripples on a
pond after a stone has been dropped into it. He (1991: 334) defends this presentism:
The necessity for the ripples in the pond postulate comes from the fact that usually we do not know
whether remote consequences will be good or bad. Therefore we cannot know what to do unless
we can assume that remote consequences can be left out of the account.
This issue is particularly acute for climate change, and policies related to it, as the environmental, financial and health effects will not only occur in the future, but in the distant
future. Economists have a general way of dealing with this phenomenon called discounting, an analytical tool to compare economic effects that occur at different points in time.5
But there are different discount rates available and the choice of discount rate is of crucial
technical importance for analyses of climate change policy, because the time horizon is
extremely long, and mitigation costs tend to come much earlier than the benefits of
avoiding damages (IPPC, 1996: 8).
There has been extensive, unresolved debate about discounting in assessment of climate
change policies, a debate which reminds us that facts alone cannot provide moral judgements.
The recent IPCC publication (2001: 97) emphasizes that uncertainty regarding the discount
rate relates not to calculation of its effects, which is mathematically precise, but to a value
judgement about the appropriateness of the present generation valuing services for future
generations. Environmental philosophers have pointed out that any form of discounting
devalues the environment, and the benefits that the environment holds for future generations.
The final criticism of utilitarianism relates to equity. The summation and averaging of
utilitarian calculations insufficiently recognizes the importance of how utilities are distributed within the population under consideration. Whether the utility is health or wealth,
there is scant difference between a population in which a small number have a lot of (good)
The basic premise behind discounting is that a million pounds to me now is of more value than a
million pounds in a years time.
161
health and the remainder have poor health, and a population in which everyone is
reasonably healthy. And this does not sit comfortably with our common-sense morality,
as Williams (1991: 1423) states:
In this light, utilitarianism does emerge as absurdly primitive, and it is much too late in the day to be
told that questions of equitable or inequitable distribution do not matter because utilitarianism
has no satisfactory way of making them matter. On the criterion of maximising average utility, there
is nothing to choose between any two states of society which involve the same number of people
sharing in the same aggregate amount of utility, even if one of them is relatively evenly distributed,
while in the other a very small number have a very good deal of it; and it is just silly to say that in
fact there is nothing to choose here.
So, if climate change illustrates that utilitarianism is a limited moral determinant of public
health policies, an alternative is needed. And here, recent developments in the climate
change debate suggest an alternative might be emerging.
Social justice and climate change

There is a huge literature on justice stretching back as far as the Greeks. Aristotle, for
instance, in the Nicomachean Ethics considers just actions, and likens the characteristic of
being just to the other excellences or virtues of character. For Aristotle justice is a mean,
injustice represents the extremes and the just man6 recognizes how to determine an
individuals appropriate share (Urmson, 1998).
In contemporary times, however, social justice has come to embody aspects of the last
part of Aristotles definition, fairness and proportionality. In contrast with legal and
retributive justice, social justice is about the distribution of societys benefits and burdens
and the socio-political mechanisms that enable such distribution to occur (Daniels, 2000).
This climate change dilemma has certainly extended the boundaries of moral debate in
areas of public, and public health, policy making. Because the causes and effects of climate
change are differentially distributed, the reasonableness of basing decisions purely on
utilitarian economic thinking has been questioned.
As a result, there has been a flurry of academic work looking at equity considerations in
the climate change debate. One way of determining how to distribute the costs of climate
change mitigation and adaptation policies would be not to try to falsely distribute such
costs at all, but to allow the market to decide. But libertarian, or market utilitarian,
approaches would likely lead to rich countries not valuing or wanting to pay for such
policies and poor countries being unable to afford them. Unless there was some kind of
catastrophic threat from climate change, poor countries might well be left to simply deal
with the consequences.
An alternative framework would be contractarian, also sometimes called administrative utilitarian. In this approach, the limits of using total sum or average utility as a
sufficient determinant of policy are acknowledged, and efforts are made to incorporate
We are using man here, rather than person, to represent Aristotles depiction, which focused
predominantly on men.
162
Section 2: Issues
additional dimensions to economic calculations to allow for more informed and,

apparently fairer, distribution.
This framework has been argued is more egalitarian and has drawn strongly on the
concept of social justice. In fact, one persons theory has been stressed within this dimension
of the debate, John Rawls, whose name can be found scattered among the articles, discussion
papers, and policy-related documents on equity issues in climate change. Rawlss (1999)
A Theory of Justice was first published in 1971 and 40 years later the impact remains
remarkable. There has been much discussion of Rawlss theory, and a number of expositions,
and it is not appropriate to detail these here. However, a very brief recap is necessary as a
prelude to the following sections.
Rawls thinks of justice as fairness. He starts from the premise that utilitarianism is an
inadequate, inappropriate, and ultimately unjust moral or politico-economic tool for
making distributive decisions in society. For Rawls (1999: 3), justice denies that . . . the
loss of freedom for some is made right by a greater good shared by others. And justice
does not allow . . . that the sacrifices imposed on a few are outweighed by the larger sum
of advantages enjoyed by many. Instead Rawls (1999: 5) defines justice as . . . a
characteristic set of principles for assigning basic rights and duties and for determining
what they take to be the proper distribution of the benefits and burdens of social
cooperation.
Rawls makes two other key assertions. First, he argues that peoples perceptions of
entitlement and so too of justice or fairness are inevitably shaped by their own
backgrounds, interests and social organizations. While Rawls accepts that human beings
naturally have certain interests for instance striving for basic primary goals most
interests are not of this nature and any agreed notion of justice needs to be reached
before the undue influence of unnatural interests. Second, he predicates that any social
advantages obtained through chance by birthright or natural endowment are
essentially unfair.
Putting these together Rawls sets out to establish the principles of justice for the
basic structure of society that would be agreed by individuals in an original (or abstract
pre-existence) state. Taking the form of a social contract these principles are those that
. . . free and rational persons concerned to further their own interests would accept in
an initial position of equality as defining the fundamental terms of their association
(1999: 10). This initial, or original, position corresponds to the state of nature in the
traditional theory of the social contract. Rawls purports that the original position is . . .
the appropriate initial status quo, and the fundamental agreements reached in it are fair.
This, he (1999: 11) continues, . . . explains the propriety of the name justice as
fairness: it conveys the idea that the principles of justice are agreed to in an initial
situation that is fair.
Fairness issues in the climate change debate

Armed with the basics of Rawlss theory of justice, it is possible now to return to the climate
change debate. The starting-point for the distributive concerns in climate change are three
related questions: who is responsible for the problem; who will suffer (most) from the
problem, and how; and who will bear the costs of abatement? The four fairness issues in
climate change policy that correspond to these questions have been expressed as follows
(Reichart, 1998; Rayner et al., 1999):
163
1. What is a fair allocation of the costs of preventing the global warming that is still
avoidable?
2. What is a fair allocation of the costs of coping with the social consequences of the global
warming that will not, in fact, be avoided?
3. What background allocation of wealth would allow international bargaining (about
the first two points) to be a fair process?
4. What is a fair allocation of greenhouse gases over the long-term and during transition
to the long-term allocation?
In trying to address the fairness issues in climate change, debate has actually focused on
an administrative utilitarian (or contractarian) approach, drawing in, to a degree, some
Rawlsian ideas of social justice. The IPCC, for example, distinguishes two categories of
equity as significant to climate change analyses: procedural equity and consequentialist
equity. The former is largely about making policy, focusing on the criteria and methods for
implementing fair procedures for design of, and participation in, the decision-making
processes, as well as respect for legal rights. It is about inclusion, fairness and openness at
all stages in the policy making processes and corresponds to item four on the list above.
Consequentialist equity, in contrast, and is about the outcomes of climate change (and
policies addressing climate change): justice and fairness in respect of the impacts of climate
change, and justice and fairness in respect of abatement, in other words the distribution of
burdens and allocation of benefits associated with reducing greenhouse gas emissions and
managing climate change. Consequentialist equity has been further divided into intragenerational equity (although actions by individuals in contributing to greenhouse gas emissions may affect anyone, impacts reflect vulnerability and are borne differentially by social
groups or countries depending on their geography, economic development and so forth)
and intergenerational equity (costs of abatement may be borne now but benefits may not be
realized well into the future). Consequentialist equity takes on board the widely heralded
precautionary principle, which dictates that when there is serious doubt about likely
environmental impacts and consequences, decisions should be made that err on the side
of safety (Hayry, 2003).
Climate change and climate justice

The intellectual and theoretical developments described in the previous section have been
mirrored in two parallel, connected sets of processes in the climate change debate: developments in international policy around managing climate change; and growth in the campaigning efforts of pressure groups. There is not scope here to look at how international
policy developments have captured Rawlss ideas, although these are described elsewhere
(Kessel, 2006). Instead, climate justice will be explored here.
In parallel to policy developments and negotiations (and sometimes providing evidence
for them) there has been a groundswell in independent think tanks, non-profit making
organizations, and other new bodies established to press for fair and generally more
aggressive policy targets relating to climate change. A number of these have expressed
their opinions and activities in terms of global justice, and their mix of consciencedriven academics and pressure-group campaigners has provided both the intellectual
base and the energy needed to drive activities forward. There is the feel of a throwback
to the lobbying efforts of the first half of the twentieth century to clean the skies of
air pollution.
164
Section 2: Issues
The Global Commons Institute (GCI), for instance, was set up in 1990 in London, and has
been encouraging awareness of its solution to climate change called Contraction and Convergence. Put forward as the suggested international framework for the arrest of greenhouse gas
emissions Contraction and Convergence has argued that economic growth can continue at
current (business as usual) rates only provided large efficiency gains are made and nearly all
energy comes from renewable sources (Global Commons Institute, n.d.; Meyer, 2000).
Another group, the cleverly named US-based EcoEquity, is committed to advancing
equal rights to global commons resources, in particular the principle of equal per capita
rights to the atmosphere. Lamenting both US rejection of the Kyoto Protocol and also the
Byrd-Hagel resolution,7 EcoEquity (n.d.) has argued that fairness . . . cannot and will not
mean that the rich go on as before, and that a climate treaty will have to embody a . . .
fairness that is acceptable in China as well as the United States. EcoEquity has intended to
deepen and clarify the meaning of climate justice through drawing together academics and
non-government organizations into the global justice movement: What will we be doing?
has been asked or the EcoEquity (n.d.) website Working to bring the many threads now
being spun around climate justice together into a stronger web, one that can support a
broader political strategy.
There are other individual groups or organizations, but a powerful coalition of groups
including CorpWatch, Friends of the Earth International, OilWatch Africa and the World
Rainforest Movement gathered as the International Climate Justice Network at the final
preparatory negotiations for the Earth Summit in Bali in June 2002. The coalition
developed a set of principles aimed at putting a human face on climate change. The Bali
Principles of Climate Justice first outline the nature of the problem (caused primarily by
the rich, and felt disproportionately by small island states, coastal peoples, women, the poor
and others; violating human rights), then state 27 core principles of the international
movement for Climate Justice (CorpWatch, n.d.).
Within the context of this chapter climate change dilemmas as a theme to explore the
relationships between the environment, ethics and public health these principles illustrate
that Rawlss social justice has provided an alternative moral framework to utilitarianism for
public health. The climate change debate has spawned a range of academic, policy and
pressure group writings reflecting ideas articulated by John Rawls. Connected to this, the
climate change debate has also become an arena for expression and discussion of the
perceived reasons for many of the worlds ills: the impact of industrialization and of
modern western lifestyles, global poverty, and the conceptual imperialism of economics
(Victor, 1999; Anthanasiou and Baer, 2002; Brown, 2002).
Discussion: environmental ethics, environmental justice

and public health
Professional public health in the UK emerged at a time of great social and philosophical
change and its conceptual framework was immediately linked to the utilitarian thinking of
the period. Around the turn of the twentieth century, as the bacteriological model of disease
pathology became widely accepted, medical and public health theory became increasingly
7
A campaign prior to the Kyoto negotiations of 1997 led to 95 US senators demanding developing
countries also take on firm reduction commitments, so challenging the UNFCC principle that
developed countries take the lead in reducing emissions.
165
underpinned by science and epidemiology. Huge strides have since been made in western
medico-science but, in parallel, both medicine and public health have become more
distanced from the environment.
The climate change dilemma illustrates the limitations of public health theory, founded
on utilitarianism, to cope with modern challenges such as climate change. And at a less
global level, public health practitioners, through the twentieth century, have found it
increasingly hard to engage in real issues about the health of the environment. Such
developments have not occurred in isolation.
Separation of public health and the environment

The origins of todays environmental problems, and the relevance of this to public health,
can be traced through the inter-connected paths of progress in scientific medicine, public
health and political philosophy over the past few hundred years.
Before the era of Descartes, the relationship between mankind and nature was more
integrated. In the centuries of Hobbes and Locke, however, this outlook changed dramatically. The development of mechanistic philosophy and the progression of science somewhat
removed humans from the natural environment, which itself was mechanistically objectified. The era of individualism had begun, with justification of self-interested behaviour and
an emphasis on individual and private rights. Personal morality no longer had a special
relationship to the state, whose role became that of partner in a dispassionate arrangement
that primarily provided an environment suitable for promotion of the individual (Cohen,
2001). Mary Midgley (2001: 159) captures this well:
Since the Renaissance, this kind of contraction has in any case been happening in political philosophy
in the West. Political thinkers of the Enlightenment systematically shrank morality by making it
essentially a civic affair a matter of mutual bargaining between prudent citizens within a limited
society. Contract thinking sought to abolish the idea of duties towards anyone or anything outside
that society . . . But this move had unintended side-effects. It now makes it quite hard for us to
make sense of our responsibility towards humans outside our own society, and almost impossible
to explain our responsibilities towards non-human nature.
The seventeenth-century philosopher John Lockes emphasis on individual rights and property rights illustrate how the era also proclaimed mankinds dominion over nature. The
natural environment was articulated in inert, demarcated terms, largely devoid of value, and
humans would be morally justified in manipulating it any way necessary to further legitimate
personal interests. This tied in with ownership, rather than stewardship, of nature and began
to set in stone an image of the natural environment detached and there for human needs
which has only relatively recently been challenged by environmentalists (Passmore, 1974).
In fact, despite some romantic inclinations, this image of nature was reinforced
during the eighteenth and nineteenth centuries as utilitarian political philosophy took hold
(Rousseau, 1762). As discussed earlier, utilitarianism has (indirectly) reinforced moral
justification for individual pursuit of that which gives pleasure, with maximizing human
happiness as the overall goal. Manipulating nature to meet these ends has ethical validation
and modern welfare economics the corner-stone of liberal democracies is grounded in
these ideals. Yet utilitarianism focuses proximally, both in terms of the audience within its
calculation (failure to include impacts on those at a distance) and with regard to time the
difficulty of incorporating the needs and desires of future generations. And there is little or
no accounting for the intrinsic worth of nature.
166
Section 2: Issues
Utilitarianism emerged politically at a time of corresponding changes in science,

medicine and biology. The connection of human health with nature through miasmatic
theories of disease was gradually replaced at the end of the nineteenth century by bacteriological explanations, which catalysed the reductionism of medical science. And at that time
Darwin and his colleagues were providing a vision of nature that placed self-interested
behaviour at its very core, the driver for change, integrally related to adaptation to, and
manipulation of, the environment (Dawkins, 1976).
Through the twentieth century individuals have often been seen as disconnected from
other individuals and the natural world, with purpose, values and goals narrowly defined.
Midgley (2001: 69) again captures this well:
It is the social atomism that lies at the heart of individualism the idea that human beings are
essentially separate items who only come together for contingent reasons of convenience [original
emphasis]. This is the idea expressed by saying that the state is a logical construction out of its
members, or that really there is no such thing as society. A social contract based on calculations of
self-interest is then supposed to account for the strange fact that such things as human societies
do actually exist.
These developments have clearly impacted on how public health has progressed. Public
health theory and practice has become distanced from the environment. Within contemporary public health the term environmental health has come to signify how degraded
aspects of the environment affect human health, rather than reflect the true health of the
environment per se.
There has, however, been a recent counter-vision, in the form of environmental ethics,
which has provided a different way of understanding the world. A look at environmental
ethics, and the related areas of development ethics and environmental justice, will enable a
synthesis of the implications for public health.
Environmental ethics
Although environmental ethics has blossomed as an academic activity over the last two
decades, its main tenets can be traced back to earlier this century. And, though these
fundamentals have been subject to considerable theoretical and philosophical debate, they
have also become inescapably linked to socio-political ideologies and movements (Light and
Holmes, 2003).
It is difficult to place the various philosophical perspectives on the environment into a
bag labelled environmental ethics, as they differ in many important aspects, but what they
share is a fundamental questioning of the value, or values, ascribed to nature. Yet even here
there are different approaches, or ways in, to examining this core. One such approach, a sort
of starting point in environmental ethics, is to distinguish between anthropocentric
(human-centred) and non-anthropocentric ethics. This is seen as a good place to begin
because an oft shared belief in environmental ethics is that the roots of todays environmental problems lie in the moral favouritism given to human interests; this is in itself linked
to developments in science and political philosophy discussed in the previous section. The
moral favouritism, the anthropocentric ethical framework, is then disapproved of in
different ways and for different reasons.
In trying to summarily address what an environmental ethic is, Robert Elliot captures
this overview and presents five sub-divisions. The first, human-centred ethics, has modern
utilitarianism as an exemplar, in which facts are needed to calculate the happiness yielded
167
by options, but only humans are treated as morally considerable, that is, are included in the
calculus. An animal-centred ethic treats individual animals as morally considerable, but
may allow ranking to account for different interests and capacities. Treating equal interests
equally and unequal ones unequally, for example, would accommodate human ranking
above animals based on a different capacity for rational autonomous action. A life-centred
ethic, on the other hand, counts all living things as morally considerable, not just humans
or non-human animals. However, while some would ascribe equal moral considerability to
all life, such as the biotic egalitarianism of Norwegian philosopher Arne Naess, others
allow differentiation, for instance by complexity. This may favour, for example, the
biosphere over humans and leads to a special kind of life-centred ethic termed ecological
holism, which grants moral considerability to wholes, such as large ecosystems or the
biosphere: individuals or species are only important in relation to these wholes. The final
environmental ethic, called rights for rocks by Elliot (1997), extends moral considerability
to all as an everything ethic.
Underlying these divisions, or different perspectives, is the question of what
makes something worthy of moral considerability worthy of consideration when
judging the morality of action (Goodpaster, 1978). Humans are morally considerable
because they have interests that can be promoted or harmed, based on their human
capacities for rational thought and action, and sentience. However, not only is sentience
shared by some animals (which could extend moral consideration to them), but moral
considerability could lie elsewhere, in some other intrinsically valuable property, for
instance complexity or even beauty. This in turn would shift moral considerability to
non-sentient animals, plants,8 ecosystems or the wilderness, and could include nonliving9 entities.
The different perspectives within environmental ethics lie along a spectrum, which
stretches from humans to animals, plants, all living and non-living things, incorporating
different concepts of what matters morally. Des Jardins, for example divides the spectrum
up a little differently, but it still incorporates the same elements. His grouping are:
biocentric ethics, which is centred around (all) life and has correlative duties10; ecological
ethics, which focuses on ecological communities and embraces ethical holism; the land
ethic, articulated first by Aldo Leopold in 1949, which embraces living things, ecosystems
and the land (Callicott, 1979); deep ecology, especially that of Naess, which emphasizes the
deep roots of environmental crises, the radical cure needed (personal and cultural transformation) and forcibly expresses its distinction from shallow anthropocentric environmentalism; and social ecology and ecofeminism, which explore how social structures serve
the interests and power of certain groups, reflected in (and reinforced by) domination over
nature (Des Jardins, 1997).
However the continuum within environmental ethics is separated out, academically
or theoretically, a common thread is the difficulty, or failure, to ascribe inherent value
8
9
10
The difference between having interests and goals has been stressed by philosophers. A plant may
grow towards light or a tree may wither and die, but neither the plant nor the tree, arguably, has
attitudes towards these happenings.
The distinction between living and non-living is often neither biologically nor philosophically
clear. For instance, a rock may be considered non-living or inert, but what about soil?
These are non-maleficence (to any organism), non-interference, fidelity (to not betray or deceive
wild animals) and restitutive justice (to restore balance if harm done).
168
Section 2: Issues
to non-human nature, whether that is other animals, vegetation or alternative concepts

of what might exist. The anthropocentric nature of western ethics gives, at best,
instrumental value to anything non-human; in other words wombats, wild flowers
and the wilderness are of value only by way of serving human interests as pets, for
rambling or as potential new medicines. This has arisen because of entwined developments in science, medicine, and moral and political philosophy over several hundred
years. It may be fair to reflect that dominion over beasts was heralded back in Aristotelian times, but the contemporary situation is rather different in terms of the success
and value placed on liberal individualism, materialism and the socio-political structures
enshrining these ideologies. The present situation is also vastly different in terms of the
depth of environmental crises affecting the planet, of which greenhouse warming is just
one example. The Australian philosopher Peter Singer (1999: 285), despite holding
sentience alone as morally considerable, is sure of the seriousness of the problem, and
the extent of change needed:
Now we face a new threat to our survival. The proliferation of human beings, coupled with the byproducts of economic growth, is just as capable as the old threats of wiping out our society and
every other society as well. No ethic has yet developed to cope with this threat. Some ethical principles
we do have are exactly the opposite of what we need. The problem is that . . . ethical principles
change slowly and the time we have left to develop a new environmental ethic is short. Such an ethic
would regard every action that is harmful to the environment as ethically dubious, and those that
are unnecessarily harmful as plainly wrong.
Singer then outlines his environmental ethic as including consideration of all sentient
creatures now and well into the future, aesthetic appreciation of wild places and nature,
rejection of materialistic ideals, promotion of frugality and reassessment of extravagance.
He (1997) has espoused these views elsewhere in further detail. For Singer, and many
others, the connection between environmental ethics and environmental activism (or
environmentalism) is strong.
Environmental justice and development ethics

Environmental ethics, as a field within ethical theory, has not developed in a vacuum. Two
complementary concepts have emerged in the same recent period: environmental justice
and development ethics. We do not have time to explore these concepts fully but will outline
them briefly in the context of discussion of climate change and environmental ethics
because of their relevance to the theme.
Environmental justice originated in protests in the 1980s by community groups in the
USA against the repeated siting of polluting factories and waste sites in predominantly black
neighbourhoods and indigenous peoples reservations. Civil rights protestors highlighted
the disproportionate burden of negative environmental impacts these caused for the most
vulnerable sectors of society (Stephens and Bullock, 2000).
In 1994, the issue reached the White House when then President Clinton issued
Executive Order 128298: Federal Actions to Address Environmental Justice in Minority
Populations and Low-Income Populations. This order reinforced the Civil Rights Act 1964
by requiring federal regulatory agencies to make environmental justice a part of all they do
(EPA, n.d.).
Environmental justice is generally defined in normative terms, specifying a set of
conditions or expectations which should be aspired to, sought after or demanded. Two
169
definitions provide examples. The US Environmental Protection Agency (n.d.) defines

environmental justice as:
. . . the fair treatment and meaningful involvement of all people regardless of race, color, national
origin, or income with respect to the development, implementation, and enforcement of
environmental laws, regulations, and policies. Fair treatment means that no group of people,
including a racial, ethnic, or a socioeconomic group, should bear a disproportionate share of the
negative environmental consequences resulting from industrial, municipal, and commercial
operations or the execution of federal, state, local, and tribal programs and policies. Meaningful
involvement means that: (1) potentially affected community residents have an appropriate
opportunity to participate in decisions about a proposed activity that will affect their environment
and/or health; (2) the publics contribution can influence the regulatory agencys decision; (3) the
concerns of all participants involved will be considered in the decision making process; and (4) the
decision makers seek out and facilitate the involvement of those potentially affected.
Environmental justice has also been conceived in terms of rights and responsibilities. For
example, Stephens and Bullock (2000) assert that environmental justice means:
. . . that everyone should have the right and be able to live in a healthy environment, with access to
enough environmental resources for a healthy life
and
. . . that responsibilities are on this current generation to ensure a healthy environment exists for
future generations, and on countries, organizations and individuals in this generation to ensure that
development does not create environmental problems or distribute environmental resources in ways
which damage other peoples health.
This idea of rights and responsibilities has a strong modern resonance with climate
justice discourse, and indeed the two movements have begun to come together. The
notion of responsibility also resonates with Singers concepts of ethics in the age of
individualism. And to an extent, this notion has become part of modern consciousness
in wealthy countries and is articulated through the growing trend towards organic and
locally sourced foods, recycling of waste and more sustainable building design. None of
these actions may have a direct benefit on the individuals doing them, but the notion of
responsibility towards the planet, other species and the future seems to motivate
people nonetheless.
A final stream of complementarity is with the emerging field of development ethics. The
central concept of development ethics is a reflection on where we are going as a species: the
clearest articulation of this is from the International Development Ethics Association
(IDEA) formed in 1984. They argue that: . . . international development ethics is ethical
reflection on the ends and means of local, national and global development (IDEA, n.d.).
We do not have the space to expand their ideas fully, but their initial declaration outlines
the framework (IDEA, 1989):
the absolute respect for the dignity of the human person, regardless of gender, ethnic
group, social class, religion, age or nationality;
the necessity of peace based on a practice of justice that gives to the great majorities
access to goods and eliminates the conditions of their misery;
the affirmation of freedom, understood as self-determination, self-management, and
participation of peoples in local, national and international decision processes;
170
Section 2: Issues
the recognition of a new relation of human beings with nature, facilitating responsible
use, respectful of biological cycles and the equilibrium of ecosystems especially
those of tropical forests and in solidarity with future generations; and
the stimulus to construct a rationality suited to exploited peoples, one that accords
with their cultural traditions, their thought, their interests, and their needs and that
involves a new valuing of self-esteem based on their being subjects rather than objects
of development.
There are strong themes of justice and respect in this statement. It also articulates a form of
procedural justice in its emphasis on freedom and participation.
Both environmental justice and development ethics bring the field of environmental
ethics into the more activist realm of human rights, and overall development policy. Linked
to climate change, development ethics would argue that our whole development trajectory
is flawed built on wasteful materialism and pursuit of human happiness articulated
through ever increasing material wealth. Utilitarianism would see no problem with this
but both environment justice activists and development ethicists would argue that, with the
changes in our environment brought about by human material development, the entire
human development trajectory needs to be reviewed.
Conclusion
To have purchase, attempts to seriously tackle problems such as climate change
need simultaneously to address the roots that have bred liberal individualism, dominant
utilitarian-based political philosophy, materialism and social atomism.
Environmental ethics reminds us that the roots of current environmental ideas lie deep.
Several hundred years of separation in western thought of mind from matter, subject from
object, values from facts, has resulted in the dominance of scientific reductionism over
holism, and the devaluing of nature. Connected developments in moral and political
philosophy have ingrained utilitarianism and liberal individualism, justifying self-interested
behaviour and leading to social atomism. The depth of the problem means meaningful
solutions need to be radical.
Within public health, the environment has predominantly been viewed rather than
inherent value of instrumental, as concerned with the human health consequences of
environmental damage. The climate change dilemma is illustrative of connections between
environmental ethics and human welfare, and reflective of the limitations of utilitarianism
as the underpinning moral theory of public health. The alternative of Social and environmental justice is presented.
Over the last two centuries the means to health have been thought to be through improved
material well-being initially clean water, sanitation, safe housing. While even these
conditions are still lacking for a shocking majority of the worlds population, for a minority
the definition of material well-being is not only a selfish level of material consumption, but
deeply damaging to the planet and its future. While environmental matters such as chemical
hazards and even outdoor air pollution are undeniably important, they really only attend to
the superficial end of the spectrum, representing shallow environmentalism.
For the future of the publics health a more substantial change in attitudes is required:
the discipline of public health needs to embrace the ideals of environmental ethics and
social justice, and incorporate the idea of development ethics with its emphasis on critical
reflection on our very development model.
References
Athanasiou, T. and Baer, P. (2002) Dead Heat:
Global Justice and Global Warming.
New York: Seven Stories Press.
Baum, F. (1995) Researching public health:
behind the qualitative-quantitative
methodological debate. Social Science and
Medicine, 40: 45968.
Brown, D. (2002) Ethical Problems with
the United States Response to Global
Warming. Blue Ridge Summit: Rowman and
Littlefield.
Callicott, J. B. (1979) Elements of an
environmental ethic: moral considerability
and the biotic community. Environmental
Ethics, 1: 7181.
171
Environmental Health Commission (1997)

Agendas for Change. London: Chadwick
House Groups.
Environmental Protection Agency (EPA)
(n.d.) Environmental justice. Available at:
http://www.epa.gov/compliance/
environmentaljustice (accessed:
17/5/07).
Epstein, P. (1995) Emerging diseases and
ecosystem health: new threats to public
health. American Journal of Public Health, 85:
16872.
Faculty of Public Health Medicine (FPHM)
(n.d.) A career in public health. Available at:
http://www.fphm.org.uk/CAREERS/Careers.
htm (accessed: 9/8/02).
Cohen, M. (2001) Political Philosophy: From

Plato to Mao. London: Pluto.
Forde, O. H. (1998) Is imposing risk awareness

cultural imperialism? Social Science and
Medicine, 47(9): 11559.
CorpWatch. (n.d.) Bali Principles of

Climate Change. Available at: http://www.
corpwatch.org/article.php?id=378
Global Commons Institute (n.d.) Basic

climate scenarios. Available at:
http://www.gci.org.uk/scenarios.html
Daniels, N. (2000) Accountability for

reasonableness. British Medical Journal, 321:
13001.
Goodin, R. E. (1997) Utility and the good. In

A Companion to Ethics, ed. P. Singer. Oxford:
Blackwell.
Dawkins, R. (1976) The Selfish Gene. Oxford:

Goodpaster, K. (1978) On being morally

considerable. Journal of Philosophy,
75: 30825.
Department for Environment, Food and Rural

Affairs (DEFRA) (2006) Tomorrows Climate:
Todays Challenge (Climate Change UK
Programme 2006). London: HMSO.
Green, J. and Thorogood, N. (2004)

Qualitative Methods for Health Research.
London: SAGE.
Department of Health (1988) Public Health in

England: Committee of Inquiry into the
Future Development of the Public Health
Function (Acheson Report). London: HMSO.
Hamilton, C. (1999) Justice, the market and

climate change. In Global Ethics and
Environment, ed. N. Low. London:
Routledge. pp. 90105.
Des Jardins, J. R. (1997) Environmental Ethics:

An Introduction to Environmental
Philosophy. Belmont, CA: Wadsworth.
Harris, C. E., Jr (1997) Applying Moral Theories.

Belmont, CA: Wadsworth.
EcoEquity. (n.d.) About EcoEquity. Available at:

http://www.ecoequity.org/about.html
Hayry, M. (2003) European values in

bioethics; why, what and how to be used?
Theoretical Medicine and Bioethics, 24(3):
199214.
Elliott, C. (ed.) (2001) Slow Cures and Bad

Philosophers: Essays on Wittgenstein,
Medicine and Bioethics. London: Duke
University Press.
Honer, S. M., Hunt, T. C., Okholm, D. L. and

Safford, J. L. (2006) Invitation to Philosophy:
Issues and Options, 10th edn. Belmont, CA:
Wadsworth.
Elliot, R. (1997) Environmental ethics. In

A Companion to Ethics, ed. P. Singer. Oxford:
Blackwell, pp. 28493.
Institute of Medicine (1988) The Future of

Public Health. Washington, DC: National
Academy Press.
172
Section 2: Issues
Intergovernmental Committee on Climate

Change (IPCC) (1996) Climate Change 1995:
Economic and Social Dimensions of Climate
Change. Contributions of Working Group III
to the Second Intergovernmental Panel on
Climate Change. Cambridge: Cambridge
University Press.
Intergovernmental Committee on Climate
Change (IPCC) (2001) Climate Change 2001:
Impacts, Adaptation, and Vulnerability.
Cambridge: Cambridge University Press
(for IPCC).
Meyer, A. (2000) Contraction and Convergence:

the Global Solution to Climate Change.
Dartington: Green Books.
Midgley, M. (2001) Science and Poetry. London:
Routledge.
Passmore, J. (1974) Mans Responsibility for
Nature. London: Duckworth.
Porter, D. (1999) Health, Civilization and
the State: A History of Public Health from
Ancient to Modern Times. London:
Routledge.
Rachels, J. (1993) The Elements of Moral
Philosophy. New York: McGraw-Hill.
International Development Ethics Association

(IDEA) (n.d.) What is development ethics?
Available at: http://www.development-ethics.
org/default.asp?cid=5010 (accessed: 17/5/07).
Rawls, J. (1999) A Theory of Justice, rev. edn.

International Development Ethics Association

(IDEA) (1989) Mrida Declaration July 7,
1989, Mrida, Yucatan, Mexico. Available at:
http://www.development-ethics.org/default.
asp?cid=5011 (accessed: 18/5/07).
Rayner, S., Malone, E. L. and Thompson, M.

(1999) Equity issues and integrated
assessment. In Fair Weather: Equity Concerns
in Climate Change, ed. F. L. Toth. London:
Earthscan, pp. 1143.
Jukes, G. (1999) Environmental health

perspectives. In Perspectives in Public Health,
ed. S. Griffiths and D. J. Hunter. Oxford:
Radcliffe Medical Press.
Kessel, A. (2006) Air, the Environment and
Public Health. Cambridge: Cambridge
University Press.
Reichart, J. E. (1998) The tragedy of the

commons revisited: a game theoretic
analysis of consumption. In The Business of
Consumption: Environmental Ethics and the
Global Economy, ed. L. Westra and P. H.
Werhane. Oxford: Rowman and Littlefield,
pp. 4766.
Light, A. and Holmes, R., III (eds) (2003)

Environmental Ethics: An Anthology. Oxford:
Blackwell.
Rousseau, J. J. (1762/1959) The Social Contract

and Discourses, Trans. G. D. H. Cole.
New York: Dutton.
MacIntyre, A. (1989) A Short History of Ethics.

London: Routledge.
Russell, B. (1991) History of Western Philosophy.

London: Routledge.
McMichael, A. J. (1995) Planetary Overload:

Global Environmental Change and the Health
of the Human Species. Cambridge:
Scruton, R. (1996) A Short History of Modern

Philosophy. London: Routledge.
McMichael, A. J. (1996) Global environmental

change and human health. Paper presented at
seminar on Global Changes and Human
Health. Royal Swedish Academy of Sciences,
Stockholm, 29 May.
McMichael, A. J. and Haines, A. (1997)
Global climate change: the potential
effects on health. British Medical Journal,
315: 8059.
Menne, B. and Bertollini, R. (2005) Health and
climate change: a call for action. British
Singer, P. (1997) How Are We to Live? Ethics in

an Age of Self-interest. Oxford: Oxford
University Press.
Singer, P. (1999) Practical Ethics. Cambridge:
Smart, J. J. C. (1991) An outline of a system of
utilitarian ethics. In Utilitarianism: For and
Against. Cambridge: Cambridge University
Press, pp. 374.
Stephens, C. and Bullock, S. (2000) Environmental
justice: an issue for the health of the children of
Europe and the World. In The Environment and
Childrens Health in Europe, ed. G. Tamburlini.
Rome: World Health Organization.
Urmson, J. O. (1998) Aristotles Ethics. Oxford:

Blackwell.
Victor, D. G. (1999) The regulation of
greenhouse gases: does fairness matter? In
Fair Weather: Equity Concerns in Climate
173
Change, ed. F. L. Toth. London: Earthscan,

pp. 193206.
Williams, B. (1991) A critique of utilitarianism.
In Utilitarianism: For and Against.
Section 2
Chapter
10
Issues
Public health research ethics

Is non-exploitation the new principle
for population-based research ethics?
John McMillan
Introduction
The ethical principles that should govern public health research must accommodate public
healths focus upon the health of populations, as well as the different research methodologies that it tends to use. Whereas biomedical research ethics has tended to emphasize the
centrality of consent, research upon populations often raises additional ethical issues and
consent may often be insufficient or inappropriate. The Belmont principles of respect for
persons, justice and beneficence are intended to cover all biomedical and behavioural
research involving human subjects, so should be applicable to public health research.
However, because of the population-based focus of public health research work, the scope
and application of these principles need to be reconsidered. An additional consideration is
whether once their scope and application has been reconfigured, the three Belmont principles are suited to the ethics of public health research. Non-exploitation is a moral concept
that has been mooted as an additional principle for biomedical research. Given that a
principle of non-exploitation can be readily applied to populations and some unethical
public research can be described as exploitative, this would appear at first to be a promising
principle for public health research. This chapter will consider the merit of these claims and
conclude that the Belmont principles can be interpreted so that they imply that exploitative
public health research is wrong.
What is public health research?

Before considering the appropriate normative framework for public health research it is
important to characterize this kind of research as clearly as is possible. In The meaning of
public in public health Verweij and Dawson (2007) note that Public Health is a
contested concept and that there are many competing definitions of this concept. While
this is likely to be the case a working definition is useful for characterizing the domain
under discussion.
The Institute of Medicine gave the following influential definition in 1988: Public
Health is what we, as a society, do collectively to assure the conditions in which people can
be healthy. While this definition is influential and inclusive it gives little guidance to the
methodology of public health and given that this is important for characterizing public
health research, a sharper definition is helpful for my purposes.
174
Chapter 10: Public health research ethics
175
Childress et al. (2002: 170) say:

Public health is primarily concerned with the health of the entire population, rather than the health of
individuals. Its features include an emphasis on the promotion of health and the prevention of disease
and disability; the collection and use of epidemiological data, population surveillance, and other
forms of empirical quantitative assessment; a recognition of the multidimensional nature of the
determinants of health; and a focus on the complex interactions of many factors biological,
behavioral, social, and environmental in developing effective interventions.
Like the Institute of Medicine these authors emphasize the importance of a population level
view about the aetiology and prevention of disease, but they also point to the core activities
and methodology that characterize public health. Given that health promotion and the
prevention of disease and disability are some of the central activities of public health, this
helps to characterize the domain of public health research. The last part of their definition
emphasizes the complicated explanatory level that public health often operates at. Clearly,
the explanatory complexity of public health influences the study methodologies adopted
and in turn will have implications for the moral framework that is best suited to public
health research ethics. While there is a broad array of research methodologies that can be
considered public health research, epidemiology, population surveillance and other kinds of
empirical quantitative assessment are the uncontroversial core of public health research.
While there are clear examples of public health research there are some interesting cases where
it is less obvious whether the systematic collection of information counts as public health research.
As is the case in the social sciences there is still a significant debate within public health about
the relative merits of qualitative and quantitative methodologies. It has been claimed that
Methodological problems remain at the forefront of public health research because the ideology
underpinning public health is often in flux (Research in Health and Behavioral Change, 1989).
This is significant for thinking through the ethics of public health research. Those who
advocate qualitative methods also insist that the ethical issues that need to be thought
through for this kind of research are very different from those for quantitative methodologies. This is partly because qualitative research has a radically different account of what
constitutes validity and the implications that this has for what is considered good study
design. There are also some interesting differences in that some qualitative researchers see
themselves as engaged in what is essentially, an empowering experience for the people
taking part in the research (Punch, 1994). Certainly, it does seem the case that for many
qualitative studies, the likelihood of physical harm is radically different from clinical trials
and often a participant can exercise their right not to participate by clamming up and
refusing to speak to the researcher. While there are some reasons for thinking that ethical
scrutiny is less important for qualitative studies the very permissive situation does seems to
be changing somewhat at least in the UK: an important change is the introduction of a
requirement for all research funded by the Economic and Social Research Council (the major
social science funding body in the UK) to be ethically reviewed (ESRC, 2006). The ethical
principles that provide an adequate framework for public health research also provide
an adequate framework for qualitative research. However to argue this thesis out fully would
require another chapter because it would require a full exploration of qualitative methods.1
For an introduction to the ethical considerations that are relevant to qualitative research see Gauld
and McMillan (1999).
176
Section 2: Issues
The scope of this chapter is limited to the issue of whether the conventional principles of
research ethics can accommodate cases of exploitative public health research.
Another important ambiguity is whether we can determine whether collecting public
health information in a systematic way is research or audit (collecting health information
for the purposes of measuring care against already existing standards). This distinction
matters because in many places there is an obligation upon researchers to gain approval
from a research ethics committee if they are conducting research, which usually entails a
significant amount of extra work and time. Audit can be considered an essential component
of checking the quality of services that are provided and therefore is usually not reviewed by
a research ethics committee. Wade (2005) has discussed some of the difficult boundary
cases that are reasonably common in biomedical research. One common way in which this
boundary can become blurred is when data is collected in a systematic way for the purposes
of audit, which when analysed unexpectedly generates findings that are generalizable and
likely to be of interest to a broader audience. In this kind of case what started out as audit
and was not scrutinized has morphed into research that probably should have been
subjected to ethical scrutiny. While these tricky boundary cases are reasonably common
in biomedical research they are likely to be even more prevalent for public health, given the
research methodologies commonly used. Because public health tends to function at a
population level it can be hard to distinguish data collection that is part of routine public
health from public health research, because they use common methods.
MacQueen and Buehler (2004) describe a number of possible public health scenarios
where it would be unclear whether whats being done is research or a public health
intervention. Their suggestion is that we should focus upon the primary intent of the
researcher/physician to distinguish whether this is a public health intervention or research
(as we might to distinguish audit from research). In cases where the methods used for a
public health intervention are very similar to those used for public health research then
there seems little choice but to make this determination on the basis of what it is that is
intended, although this will not avoid the problem of unexpected generalizable findings.
Codes and principles

Up until the 1980s there were relatively few codes of research ethics. The Declaration of
Helsinki (2008), the Nuremberg Code (Annas and Grodin, 1995) and the CIOMS (2002)
guidelines were and still are important moral touchstones for human subject researchers.
While these codes lay down rules that have been very significant in shaping the conduct of
human subject research, they are limited in a number of important respects. All codes of
ethics contain a set of moral rules that are thought to be important guides to conducting
research in an ethical way. Inevitably codes of ethics have to strike a difficult, and perhaps
impossible, balance between being too restrictive and setting conditions that are unlikely to
be met and being overly permissive or vague with the result that the code has little
normative force. For example, the Nuremberg Code says that all research should have the
voluntary consent of the research participant which has the virtue of being a clear and
prescriptive rule but seems to make it impossible to do research upon those who are unable
to give consent, even when this research would clearly be in their interests or could not have
any impact upon the interests of that person (para. 1). The Declaration of Helsinki requires
researchers to ensure that research participants have access post trial to study medication if
trial medication is shown to be effective for participants (para. 30). This rule is likely to have
177
the consequence that some important research in the developing world cannot be conducted because of the difficulty that researchers would face in ensuring that the infrastructure for the continuation of the same kind of care is in place.
While the moral rules of research ethics play an important role they need to be
supplemented by moral principles or other foundational values that can:
justify exceptions to moral rules;
justify the refinement of moral rules;
determine the scope and applicability of a moral rule;
provide guidance in cases which are not covered by an appropriate rule.
Even though rule based codes of ethics are important we often need to move to a deeper
level of moral analysis which can provide better justified moral reasons, instead of just
relying upon the more precise prescriptions that tend to be expressed by rules.
In addition to the difficulties inherent in applying moral rules to the complex and
novel moral scenarios that are common in research ethics, the large number of codes of
research ethics makes it difficult to know whose rules should be followed. Since the 1980s,
when Nuremberg, Helsinki and CIOMS were the most prominent codes, the amount of
guidance on research ethics has increased hugely. Codes of research ethics have proliferated to the extent that it would be unusual for any professional body whose members
conduct research to be without its own code of ethics. To a significant extent, codes are
parasitic so there is a degree of consistency between them. However, there are also many
examples where different professional bodies express rules about the same thing in
significantly different ways. For example the Royal College of Physicians (1996: SS 7.3)
Guidelines claim that the level of risk that is appropriate for healthy volunteer research
should not exceed that which would be presented to people travelling by car but less than
that which would be presented by travelling by pedal or motorcycle. While this rule has
the virtue of clarity this degree of precision means that it is likely to contradict the rules of
other professional bodies who also specify the level of risk that is appropriate. Making a
judgement about which rule is applicable in a particular case can be made by moving to a
level of analysis where the fundamental values underpinning these rules can be weighed
and applied to the rule and case.
As I will argue in the following sections, the Belmont principles can be interpreted so as
to specify the fundamental values that underpin public health research. If they are carefully
applied and interpreted in exploitative cases they can explain the wrongness of these cases.
The Belmont principles

The Belmont Report (National Commission for the Protection of Human Subjects of
Biomedical and Behavioural Research, 1979) attempts to articulate the core moral values
that should be implicit in all ethical research involving human subjects. The Reports
authors hoped to provide an analytic framework for considering the ethics of all human
subject research. The idea is that by laying out the fundamental moral principles of
research, argument about the appropriate rules and specific cases can proceed with a
common basis in a set of shared moral commitments. The Belmont principles are:
Respect for persons
Beneficence
Justice
178
Section 2: Issues
The Report gives an account of what these principles mean and examples of some of the
moral rules and practices that can be justified by reference to them.
The Report (1979: B1) interprets respect for persons as implying:
. . . at least two ethical convictions: first, that individuals should be treated as autonomous agents, and
second, that persons with diminished autonomy are entitled to protection. The principle of respect
for persons thus divides into two separate moral requirements: the requirement to acknowledge
autonomy and the requirement to protect those with diminished autonomy.
This way of explicating respect for persons emphasizes the centrality of autonomous
individuals and their right to self determination. This account leads naturally to requirements such as informed consent and the right to have ones sensitive medical information
protected which are fundamental practices and rules for public health research ethics.
Belmonts authors are clear that these obligations are not the only obligations that follow
from respect for persons which is consistent with their intention to provide a framework for
moral analysis rather than a set of moral rules that will cover all cases. Nonetheless, their
brief account of respect for persons does not mention some other aspects of this principle
and this has led some to suggest that other principles need to be added to the Belmont three.
In the following sections I will show that concerns about the relevance of non-exploitation
to public health research are implied by a more developed account of respect for persons.
While Beauchamp and Childress (1994; Beauchamp, 2007) think that it is necessary to
have four principles of biomedical ethics, the Belmont Report (rightly in my view) includes
non-maleficence under its principle of beneficence.
Persons are treated in an ethical manner not only by respecting their decisions and protecting them
from harm, but also by making efforts to secure their well-being. Such treatment falls under the
principle of beneficence. The term beneficence is often understood to cover acts of kindness or
charity that go beyond strict obligation. In this document, beneficence is understood in a stronger
sense, as an obligation. Two general rules have been formulated as complementary expressions of
beneficent actions in this sense: (1) do not harm and (2) maximize possible benefits and minimize
possible harms.
(1979: B2)
As I will show, the obligation to benefit research participants is the subject of a significant
debate and some have argued that the principle of beneficence is not a core foundational
value for research ethics. The Reports authors could have done more to show why beneficence is a core value, particularly research that is combined with professional care.
The Belmont Report (1979: B3) describes the question of justice as being primarily
about
Who ought to receive the benefits of research and bear its burdens? This is a question of justice, in the
sense of fairness in distribution or what is deserved. An injustice occurs when some benefit to
which a person is entitled is denied without good reason or when some burden is imposed unduly.
Another way of conceiving the principle of justice is that equals ought to be treated equally.
The Report conceives of justice as being primarily a matter of justice in distribution. In the
context of public health research ethics this will imply a number of things, including that
the benefits and burdens of research should be distributed equally and account should be
taken of any particular vulnerabilities that may lead to inequity. This might mean that
public health research that is conducted upon those who have very poor access to care, puts
179
a group at significant risk or cannot benefit members of that research group, distributes the
benefits and risks of research in an unfair way. Applying the principle of justice to public
health research might imply that care needs to be taken over the recruitment of groups with
a particular vulnerability.
While the Reports authors are right to emphasize this as a core element of justice there
are other ways in which the concept of justice can be relevant to public health research ethics.
In A Theory of Justice, Rawls (1972) defends the idea of justice as fairness, which includes
not only a principle of distributive justice but also a liberty principle which requires us to
maximize the liberties of agents within a political community. This is significant because (as
I will show) focusing upon this aspect of justice provides a way to value non-exploitation
without adding another principle to the Belmont three.2 A broader conception of justice is
important for the ethics of public health more generally and a great deal more could be said
about this. For my present purposes it will be sufficient to show interpreting the principle of
justice in a fuller sense can accommodate worries such as exploitation.
All of the Belmont principles can be explicated by reference to different and often
divergent moral theories. As is the case for justice, there are theoretical accounts of
what respect for persons amounts to,3 many of which emphasize different things and some
of which might imply contradictory things. Defenders of principlist approaches can insist
that this is one of the strengths of focusing upon principles rather than particular moral
theories. The theoretical debates about, for instance, justice, are contested and complicated.
When we need analytic tools for arguing about the ethics of applied topics such as public
health research ethics we should aim at reaching a sound answer to an applied question
rather than settling a theoretical debate about moral theory. Nonetheless, fleshing out moral
principles in the way that I think we should, is consistent with the spirit of the Belmont
Report and other authors who defend a principles based approach to research ethics.
Moving towards consistency between moral judgements or rules and the principles of
research ethics by a careful process of minor revision and elaboration can be considered
an instance of Rawlsian reflective equilibrium.4 The coherence of principles, rules and
judgements is important for providing a consistent account of the morality relevant to
public health research ethics. Inevitably this will require elaboration upon the scope and
meaning of principles as well as the clarification, reformulation and possible rejection of the
rules and judgements about research ethics.
Before considering how the Belmont principles can apply to public health research it is
important to see the reasons why principles are useful moral tools for research ethics but
perhaps are less obviously so in clinical medicine. Beauchamp and Childress (1994)
defend the use of four principles of biomedical ethics for all areas of biomedicine. There
are some good reasons for supposing that principles are more appropriately applied to the
ethics of research than they are to clinical medicine more generally. Research very often
has an international dimension that raises important ethical issues. Some of the most
significant recent debates about the ethical standards of research have focused on whether
2
3
4
Others have applied Rawls to health care. In Just Health Care (1985) Norman Daniels argues
that Rawlss principle of justice should be applied when considering the just distribution of health
care within a nation state.
For a useful introduction to the ways in which autonomy (which is one reading of what respect for
persons amounts to) can be a value in a health care setting, see Stoljar (1997) and Dworkin (1988).
See Rawls (1972: 51) and Van Willigenburg (2007).
180
Section 2: Issues
research in different countries should follow exactly the same moral rules or whether
there are principled reasons for permitting different rules in different parts of the world.
The most prominent recent example of this is the standard of care debate that followed
the publication of Lurie and Wolfes (1997) article on AZT trials conducted in the
developing world and was continued in articles published in the American Journal of
Public Health (Annas and Grodin, 1998; Bayer, 1998; Karim, 1998). One of the major
arguments in support of a principles-based approach is that it provides a common
morality that can be interpreted and applied so as to take account of the particulars
relevant to different contexts. In the debate over the AZT trials the extent and way in
which our obligations under the principle of beneficence should be discharged in the
developing world was one of the principal bones of contention. Principles are particularly
useful for debates of this kind because they provide moral standards that are both
universal and have the flexibility necessary to accommodate the particulars of different
contexts. The Nuremburg Code was adopted during the trial of Nazi physicians so that
the Nuremberg judges had a universal code which could be used for assessing the wrongs
committed in the name of research.
Clinical medicine is different, at least to some extent. The prominence given to autonomy by Beauchamp and Childress is not emphasized to the same extent in all other parts of
the world. Clinical medicine in some countries still operates on a predominantly paternalist
model. Of course, it might be that these countries should move towards a model that
embraces patient autonomy. However, this seems unjustifiably imperialistic in a way that it
does not in the case of research ethics. It might be that there are a number of different
reasons for this intuition, but one that is particularly important is the different ends of
clinical medicine and research. The fundamental objective of clinical medicine is patient
welfare. While it is right that research should also have the welfare of research participants
as a core moral commitment, the fundamental objective of research is the production of
knowledge. When clinical medicine fails to promote a patients welfare, clinical medicine
has, in this case, failed. When research fails to promote a research participants welfare, it
might still produce useful knowledge. A physician who wants to be a successful physician
fails to succeed when she does not promote a patients welfare. In his classic article Ethics
and clinical research (1966) Beecher described the incentives for young researchers to take
ethical short cuts because of the importance for their careers of generating research funding
and publications. Given that the incentives, as well as the success conditions for research,
can differ significantly from what promotes the patients welfare or autonomy this means
that there is a greater need for articulating the core values that are relevant to research So
long as clinical medicine succeeds by promoting the patients welfare (and by this I mean
welfare in an extensive sense comprising more than just the patients medical interests) then
failing to emphasize autonomy to the same extent seems more defensible, if not less bad.
The Belmont principles and population-level research

While the Belmont principles should be applied to public health research, they need to be
reconfigured so as to accommodate the population-level emphasis of public health. In
Ethical principles for the conduct of human subject research, Larry Gostin (1991: 191)
shows how the Belmont principles can be applied to public health research:
Ethical principles help support autonomy and self determination, protect the vulnerable, and
promote the welfare and equality of human beings. But traditional ethics focuses primarily upon
181
individual rights and duties and does not always see individuals as part of wider social orders and
communities. A person dominated medical ethic is insufficient for the task of setting moral and
human rights boundaries around the conduct of research on populations.
Given that public health research interventions usually operate at a population level,
principles that emphasize only the interests of individual actors are unlikely to capture
whats ethically salient about public health research. The Belmont discussion of respect for
persons emphasizes the centrality of consent without mentioning the other ways in which
persons can fail to be respected. The Belmont discussion of beneficence emphasizes the
importance of not causing harm to research participants and working towards an appropriate risk/benefit ratio. Given that public health research is more concerned with the health
of populations, how can respect for persons and beneficence be made relevant to public
health research?
Gostin describes a number of the problems that can result when respect for persons is
taken to imply only informed consent. These include the problems that can result when
there is a significant degree of mistrust between researchers and participants as well as the
problems that can occur in countries where decision making is made on a communal basis.
Gostin suggests that when applying respect for persons to public health research in addition
to inferring informed consent we should also consider:
. . . the interrelated concepts of individual consent, consensus should be regarded as additional ethical
obligations that are to be examined within the context both of international human rights norms
and local cultural and ethical beliefs.
(Gostin, 1991: 194)
Belmont emphasizes the importance of adopting appropriate measures when autonomy is

compromised. In public health research this way of cashing out respect for persons implies
that groups that might be considered vulnerable (for example, prisoners or sex workers)
should also be considered more carefully so that individuals are respected as persons.
Gostin describes a number of ways in which the Belmont principles can be interpreted
so as to provide rules that are relevant to public health research. Rather than evaluating all
of these it is more relevant to examine the cases which are hard for the Belmont principles
to accommodate that have led some to propose that Belmont needs to be revised.
Exploitation and unethical public health research

Many instances where the ethics of public health research has gone awry can be reasonably
described as instances where researchers have exploited their research subjects. Researchers
involved in the Tuskegee syphilis study failed to provide research participants with penicillin when it was clear that this would be an effective treatment for their condition (Caplan,
1992; Northington Gamble, 1997). Instead they were given heavy metals therapy and other
treatments that were unlikely to provide them with any benefit. While it would be hard to
defend what was done, the Tuskegee study did produce useful results on the natural
progression of syphilis in African American males. This seems a case where the wrong that
has been committed is one that is naturally described as exploitation and the actions of the
researchers as exploitative.
A more recent and perhaps more arguable example is a Baltimore public health study
on lead abatement therapy that was considered in the judgement Grimes v. Kennedy
Krieger Institute Inc 2001. The study ran from 1993 until 1995 and was part of a broader
182
Section 2: Issues
public health initiative to improve the quality of housing for those in lower socioeconomic groups. The research aimed to identify the least expensive, effective level of lead
abatement therapy.
Identifying the minimal effective level of lead abatement was considered important because of the
need to preserve availability of low-rent urban housing that might be abandoned by landlords if they
had to pay for the expensive repairs needed to eliminate lead using standard abatement methods . . .
1068 Baltimore rental properties were classified into 5 groups. Three groups of housing received less
than full lead abatement (a different level for each group); housing in the 2 control groups either had
previously undergone full lead abatement or had been constructed without lead paint. Over a 2-year
period, the researchers were to measure and compare lead dust levels collected in the housing with
lead levels in blood samples drawn from children living in those homes. Informed consent was to be
obtained, and parents were to be notified of their childrens blood levels and the results of lead dust
collection in their homes. Blood levels would provide evidence regarding the effectiveness of a
particular level of abatement.
(Mastroianni and Kahn, 2002: 1074)
The researchers might argue that the children benefited by being in the trial because for any
of the arms they might be randomized to they would be likely to be better off than they
otherwise would have been. If they were randomized to one of the two arms that had full
abatement therapy they would be better off because there was no lead in those houses. If
they were randomized to one of the houses with the partial lead abatement therapy they
would be likely to be better off because they would have avoided the risk of being in a house
with no lead abatement therapy. Even though the children in these three groups might have
been better off if they had been in one of the two groups that provided full abatement
therapy, the fact that they were in the study has made them better off than they otherwise
would have been. Given that parental consent had been obtained, the researchers might
argue that they had benefited the children as well as respecting them as persons via the
process of gaining parental consent.
This moral argument not withstanding there is something wrong about allowing poor
children to live in houses that have higher levels of lead contamination than is thought
healthy to find out whether in fact living with that level of contamination is unhealthy. The
court in Grimes v. Kennedy agreed and made comparisons with the Tuskegee syphilis study
while describing the research as exploitative (Mastroianni and Kahn, 2002: 1075).
More needs to be said about what exploitation is before reaching a conclusion about its
viability as a principle but on the face of it, it does seem to be a specific kind of wrong that
can occur even when there is consent and research participants have not been made worse
off over all.
Non-exploitation as a new principle for research ethics

In an influential article, Miller and Brody (2003) raise important critical questions about the
normative basis of research ethics; questions that they think have important implications
for the ethical conduct of research. They think that the authors of the Belmont Report made
a mistake when they included beneficence as a principle of research ethics. According to
Miller and Brody the obligation to benefit is not an obligation that should befall researchers
and conflates the ethics of clinical care with the ethics that should be relevant to research.
They (2003: 20) claim that: An ethical framework that provides normative guidance about a
practice should accurately characterize the practice.
183
In other words, research involving human subjects does not ordinarily involve a
commitment to benefit research participants so it is inappropriate to impose this ethical
norm when it is not part of the practice. There are a number of reasons why this is not a
particularly good argument. First, it does not follow from the fact that a practice exists that
normative guidance must accurately characterize it there are plenty of immoral practices
that need to be reformed. Second, bioethics has been criticized in the past for being merely a
legitimating exercise, that is, whatever new technology or research practice emerges ethicists
will end up providing a rationale for that activity. Insisting that ethical principles must be
derived from practice runs the risk of bioethics uncritically accepting the practice when the
appropriate question is whether the norms are right. A third problem is that the rules that
the Belmont authors derive from beneficence are generally accepted as norms for research:
minimizing harm and maximizing benefit are hardly controversial rules.
While Miller and Brody think that this is an important argument against an obligation
to benefit research participants, it is not their main objection. They (2003: 20) think that
applying the principle of beneficence to research involves a misguided attempt to bridge the
ethics of research and therapy:
In view of the nature and purpose of clinical research, the principles of beneficence and nonmaleficence applicable to clinical research lack the therapeutic meaning that guides their application to
medical care . . . and as treatment conducted by physicians who retain fidelity to the principles of
beneficence and therapeutic non-maleficence that govern the ethics of clinical medicine. The doctrine
of clinical equipoise has emerged as the bridge between medical care and scientific experimentation.
Miller and Brody seek to completely overhaul the rules and principles of research ethics.
Instead of beneficence and the requirements such as clinical equipoise they think that we
should adopt a set of rules derived from a principle of non-exploitation. They endorse some
of the rules proposed by Emanuel et al. (2000) and suggest that at least some of these rules
derive from a principle of non-exploitation. While many instances of unethical research do
intuitively seem to be instances of exploitation it is important to consider whether nonexploitation is a viable principle for research ethics in general and public health research
ethics in particular.
There is something intuitively right about the idea that exploitation plays an
important role in identifying unethical research. The wrongs of the Tuskegee syphilis
study, Willowbrook (Pappworth, 1968) and other instances of unethical public health
research in the twentieth century were intuitively instances where research subjects were
exploited. The therapeutic misconception is notoriously difficult to avoid in clinical
research (Appelbaum et al., 1987). While it is more likely that public health physicians/
researchers can do more to avoid this, it still seems to be at the root of the problem for
the examples of unethical public research discussed thus far. A useful initial way of
characterizing exploitative research is that it uses an inequality so that persons can be used
as research subjects. Beauchamp and Childress (1994: 441) note:
. . . because investigators and subjects are unequal in knowledge and vulnerability especially if sick
patients are involved public policy and review committees must act to prevent potentially
exploitative contracts. . . .
While the power and knowledge differential between researchers and subjects provide good
reasons for supposing that exploitation is a key problem for research it is also important to
be able to give a coherent account of what it is that exploitation amounts to.
184
Section 2: Issues
Wertheimer and exploitation

Arguably the most important recent theory of exploitation is that produced by the political
philosopher, Wertheimer (1996). He sets himself the task of giving an account of what the
wrong making features of an exploitative transaction are. Rather than merely giving an
analysis of what the concept means he is concerned with producing an account that brings
out the distinctive normative features of exploitative arrangements. This kind of account is
what is required in order for exploitation to play a useful role in public health research
ethics. Unless a reasonably coherent account can be given of what exploitation is and why it
is bad then it will tend to function merely as a rhetorical device or term of art. While there
are a number of theoretical explications of each of the Belmont principles all of them can be
characterized in a way that makes their normative force clear and this seems like a
reasonable prerequisite for any moral principle.
Wertheimers account is the most plausible non-Marxist account of this kind and this
reason alone is sufficient for using his account in this context. A second, perhaps more
important, reason for using Wertheimers analysis is that those who have suggested a
principle of non-exploitation for research ethics have referred to his account and Wertheimer
(2007) has recently spelled out some of the implications of his theory for research ethics. As
I will show, Wertheimers account is not as obviously useful for research ethics as many think.
Wertheimer distinguishes a number of meanings of exploitation: harmful exploitation,
wrongful use exploitation and non-consensual exploitation. He argues that a fourth sense:
mutually advantageous, consensual, exploitation is a moral concept that is distinct from
other moral considerations such as respect for persons.
Harmful exploitation
The Tuskegee syphilis study resulted in large numbers of African Americans not being
provided treatment for syphilis so that the natural progression of the disease could be
observed (Caplan, 1992). It seems reasonable to describe the African Americans who
suffered from untreated syphilis at a time when penicillin was available as being exploited
in the most appalling way.
The problem is that for Wertheimer this kind of wrong is not unambiguously identified
as being an instance of exploitation. His project is to discover the distinctive wrongness of
exploitation. While the Tuskegee study is an appalling wrong, its wrongness can just as
readily be explained in terms of the harm caused to these people and the failure to provide
treatment when it was available. In other words this wrong making aspect of the situation is
well covered by the principle of beneficence.
Wrongful use exploitation

Many commentators including Miller and Weijer (2009) take exploitation to mean what can be
called wrongful use exploitation. The idea here is that transaction X between agents A and B is
exploitative if and only if X involves A and/or B being treated as a mere means. In other words,
wrongful use exploitation violates the Kantian injunction against the instrumental use of
persons. Clearly exploiting research subjects in this way is a very serious wrong and would
provide a prima facie reason for criticizing any trial that did this. Most instances of wrongful
use will be instances of harmful exploitation but there will be some cases where it might be
plausible to describe a person as being used inappropriately even though they have not been
made worse off or harmed, so it is worth treating this as a distinct kind of case.
185
However, if exploitation is to play a distinctive role as a key principle of research it

should do moral work that is not already provided by extant moral principles. Wrongful
useful exploitation is a paradigmatic violation of the importance of respect for persons.
While the Belmont report mentions consent as one of the implications of respect for
persons there is no reason why the principle does not also imply the wrongfulness of using
human beings instrumentally and neglecting their humanity. Wrongful use exploitation is
clearly morally significant and a serious wrong but it is a paradigmatic implication of
respect for persons.
Non-consensual exploitation
One common objection to allowing payment for research participation is that it will induce
participants to take part in research. More needs to be said on the nature of what constitutes
an inducement, but a number of important commentators, Faden and Beauchamp (1986)
being two notable examples, think that inducements are problematic because they compromise the validity of consent. They give the example of a woman who agrees to take part in risky
and unpleasant research because of her desperate need and the large amount of money that
she will receive for participation. Faden and Beauchamp are overly paternalistic in claiming
that this will compromise her ability to consent and make this arrangement non-consensual.
It might be that she makes a cool assessment of what is in her interests on the basis of good
information about the study and the usefulness of the money that it will provide. (For more
on this see Wilkinson and Moore, 1997.) Nonetheless let us assume for arguments sake that
this offer is what Wertheimer would call seductive and results in her making a serious error
in judgment about what is in her interests. We might be tempted to say that the researchers
have exploited this womans vulnerability with the result that she takes part in research that
she would otherwise have refused. This would be an instance of what Wertheimer calls nonconsensual exploitation and while this clearly is morally problematic its not unambiguously
an instance of exploitation. This is because this kind of case is well covered by respect for
persons and the importance of informed consent. While exploitation might be a convenient
label for this kind of case it is not of novel normative significance.
Mutually advantageous, consensual, exploitation

Having put non-consensual, harmful and wrongful use cases to one side Wertheimer gives a
series of examples that he thinks bring out the distinctive wrongness of exploitation. The
specific instances that he thinks are revealing are what he calls instances of mutually
advantageous, consensual, exploitation (MACE). In other words these are unfair transactions or arrangements where an individual has given valid consent and will not be harmed
by the arrangement, yet is exploited nonetheless.
It is an open question whether there is anything wrong with this kind of exploitation. As
Feinberg (1990) suggests, if a person gives valid consent to an arrangement that will do
them no harm then there is a serious question about whether or not they have any grounds
for complaint. Nonetheless Wertheimer (1996: 22) does generate a number of plausible
instances of this kind of exploitation that show that there is at least something wrong with
transactions of this kind.
An unexpected blizzard hits an area and people rush to the hardware store to buy a shovel. The
hardware store owner sees the opportunity to make an abnormal profit and raises the price of the
186
Section 2: Issues
shovel from $15 to $30. If B agrees to pay $30 for the shovel, because the shovel is worth more than
$30 to B under the circumstances, then the transaction is clearly Pareto superior. Both parties gain.
But B feels exploited because B gains less (or pays more) than B thinks reasonable. A similar structure
applies to some of the other cases of alleged exploitation . . . profits from the AIDS drug AZT,
surrogate motherhood contracts, organ sales. We need not deny that B benefits from these
transactions, all things considered.
This idea of one party contributing more than they should or perhaps benefiting less than
they should is very close to the heart of what was thought by many to be wrong about the
AZT trials conducted in the developing world. Those who argued for the acceptability of
these trials often pointed to the fact that no treatment was available for HIV/AIDS in many
of these countries and that research subjects were therefore not made worse off and in many
cases benefited from these trials. Nonetheless it is hard to dispute that it is unfair that
subjects in these trials did not receive therapy that they would have if they lived in North
America or Europe. Of course it is another argument to say that these trials should not have
been conducted on the basis of this unfairness or that those who designed or conducted the
trials had an obligation to provide this benefit.
The Baltimore lead abatement study also had this kind of moral problem. While questions
were raised about whether parents were informed about all of the risks, the primary moral
objection appears to have been the fact that although taking part in this research can be argued
to have benefited them, their participation was obtained because a vulnerability was exploited.
In other words the participation of the Baltimore subjects was mutually advantageous and
possibly consensual but exploitative because of the lack of options for these subjects.
So far so good. These examples do suggest that MACE is relevant to at least some
important examples in public health research ethics. Although Wertheimer aims to give an
account of MACE he does not think that it always necessitates interference. So while in the
snow shovel example B might have grounds for feeling exploited by A it does not follow that
anything should be done to stop A exploiting B in this way. This is not in and of itself a
reason to dismiss exploitation as a moral principle because all principles should be
considered prima facie. However in more recent work Wertheimer has suggested that
MACE as a concept wont help us determine whether or not an exploitative arrangement
is wrong (2007: 253).
Problems for MACE

While Wertheimers account is very influential he now concedes that his ideal market
account of exploitation fails. Arneson has demonstrated that Wertheimer does not succeed
in identifying the appropriate moral comparator for unfair transactions.
Wertheimer argues that at least for an important range of cases, the baseline that fixes fair terms of
interaction deviation from which qualifies as substantially unfair is the competitive market price for a
good or service. If a perfectly competitive market actually exists, transactions under these conditions
are not exploitative. If a perfectly competitive market does not exist, but we can envisage these
conditions obtaining, then this hypothetical perfectly competitive price sets an appropriate baseline
for determining whether actual terms of trade are substantially unfair.
(2001: 890)
Arneson is referring to Wertheimers (1996: 232) substantive account of what would make a
MACE transaction fair and this says:
187
When a market is perfectly competitive, no one is able to take special advantage of particular
defects in the other partys decision-making capacity or special vulnerabilities in the other partys
situation.
In a public health research context this might imply that research participation is not
exploitation when neither researcher nor participant has any vulnerability that either part
could take advantage of. Its worth thinking of exploitation as being possible from both
directions: while it does not seem very likely it is possible that a participant could exploit a
researcher if the researcher were particularly desperate for their participation. Participants
in the Baltimore lead abatement study might have been in a fairly competitive research
participation market place if they had not been disadvantaged by their lack of a reasonable
alternative for housing.
Arneson (2001: 890) points out that there are problems with this account because:
. . . suppose there is a perfectly competitive market for child care and house cleaning services in
the city where I live and I am hiring John to care for my children. Nothing constrains me from
offering him an above market wage. The market wage may be abysmally low.
A second counter example:

A perfectly competitive market for doctors services may obtain, but this situation does not prevent
Dr Ann from reflecting on what she ought to do and deciding to work at a medical clinic that
serves poor people and pays her a below-market price for her services.
(Arneson, 2001: 891)
The point of these counter examples is that perfect competition does not have the normative significance that Wertheimer thinks and that perfect competition does not provide a
standard for a just price. Wertheimers account of MACE and the problems with it might
appear to have drifted a little way away from the present concerns about public health
research ethics. Nonetheless the important point here is that the most viable theoretical
account of exploitation has problems that its author thinks decisive. Wertheimer does not
think that this means that he has misidentified MACE but instead he has not succeeded in
giving us an account of to what extent these cases are wrong. In other words consensual,
mutually advantageous exploitation is a wrong without a viable theoretical underpinning.
This matters because it is reasonable to expect that a moral consideration that is to play
the role of a principle should have at least one plausible theory that underpins it. One of the
motivations behind Beauchamp and Childresss insistence that there should be four principles is that there are a number of plausible theoretical accounts of autonomy, beneficence
and justice. A principle of non-exploitation is either redundant or applies only awkwardly
in a limited number of cases.
MACE, justice and public health research ethics

A precondition for a principle being a plausible moral principle is that it has grounding in
at least one reasonable theory and this does not appear to be the case for Wertheimers
analysis of exploitation. The other senses of exploitation, harmful, wrongful use and nonconsensual are important values for public health research ethics but they are all implied by
the Belmont principles.
Nonetheless MACE does seem to pick out the wrongness of the Baltimore lead abatement
study and the AZT trials (on the assumption that they were wrong). Wertheimers account of
188
Section 2: Issues
MACE might be cashed out by coming up with a normative baseline that is more appropriate
for public health research ethics. Even if Arneson had not shown the competitive market
place to be a mistake its a comparator that is tricky to apply to research ethics, especially
given prevailing views about payment for research participation. What seems more relevant
is what the exploiter should have done morally. Dr Anns below market rate for her services
reflects a moral determination about what is just, given her obligations. This kind of
comparison is useful when thinking through the Baltimore and AZT examples. Our intuitions about what a just researcher would do in that situation seem highly relevant to
judgements about whether or not they have exploited their research subjects.
When the Belmont report discusses justice it describes it primarily in terms of distributive justice while not mentioning the importance of liberty as discussed by Rawls. One way
of teasing out the appropriate normative baseline for MACE cases is in terms of an unfair
set of liberties. If we consider the AZT trials and Baltimore lead abatement study wrong
because they involve MACE then the appropriate normative baseline might be the appropriate liberties for these participants. The participants in the AZT trials would not have
been exploited if their decision to take part was one that occurred when they could access
appropriate care under similar conditions to those in other countries. The participants in
the Baltimore lead abatement study would not have been exploited if they had agreed to
participate when they were able to access reasonable housing.
A normative baseline of this kind does seem to be consistent with the wrongness of
research arrangements of this kind. However, a normative baseline of this kind will not
help those who want non-exploitation to be a principle of research ethics. If the
normative baseline in MACE cases involves an appeal to justice then these cases are
reasonably covered by the principle of justice. If research participants take part in public
health research because their liberty to improve their situation in other ways is seriously
limited then this is a case of injustice, even if we might also describe that transaction
as MACE.
References
Annas, G. and Grodin, M. (1995) The Nazi
Doctors and the Nuremberg Code: Human
Rights in Human Experimentation. Oxford:
Annas G. and Grodin, M. (1998) Human rights
and maternal child HIV transmission
prevention trials in Africa. American Journal
of Public Health, 88: 5603.
Appelbaum, P. S., Roth, L. H., Lidz, C. W.,
Benson, P. and Winslade, W. (1987) False
hopes and best data: consent to research and
the therapeutic misconception. Hastings
Arneson, R. (2001) Exploitation. Mind, 110:
88891.
Bayer, R. (1998) Research to prevent maternalfetal transmission of HIV: racist exploitation
or exploitation of racism? American Journal
of Public Health, 88: 56770.
Beauchamp, T. (2007) The four principles

approach to healthcare ethics. In The
Principles of Healthcare Ethics, 2nd edn, ed.
R. Ashcroft, A. Dawson, H. Draper and
J. McMillan. Chichester: John Wiley.
Beauchamp, T. and Childress, J. (1994) The
Principles of Biomedical Ethics. Oxford:
Beecher, H. (1966) Ethics and clinical research.
New England Journal of Medicine, 274:
135460.
Caplan, A. (1992) Twenty years after. The legacy
of the Tuskegee Syphilis Study. When Evil
Intrudes, Hastings Center Report, 22: 2932.
Childress, J. F., Faden, R. R., Gaare, R. D., et al.
(2002) Public health ethics: mapping the
terrain. Journal of Law and Medical Ethics,
30: 1718.
Council for International Organizations of
Medical Sciences (CIOMS) in collaboration
with the World Health Organization (WHO)

(2002) CIOMS International Ethical
Guidelines for Biomedical Research Involving
Human Subjects. Geneva: CIOMS.
Daniels, N. (1985) Just Health Care. Cambridge:
Dworkin, G. (1988) The Theory and Practice of
Autonomy. Cambridge: Cambridge
University Press.
Economic and Social Research Council (ESRC)
(2006) Research Ethics Framework. Swindon:
ESRC.
Emmanuel, E., Wendler, D. and Grady, C.
(2000) What makes clinical research ethical?
Journal of American Medical Association, 283
(20): 2701.
Faden, R. and Beauchamp, T. (1986) A History
and Theory of Informed Consent. Oxford:
189
MacQueen, K. and Buehler, J. (2004) Ethics,

practice and research in public health.
American Journal of Public Health, 94(6): 928.
Mastroianni, A. and Kahn, J. (2002) Risk and
responsibility: ethics, Grimes v Kennedy
Krieger, and public health research involving
children. American Journal of Public Health,
92: 10736.
Miller, F. and Brody, B. (2003) A critique of
clinical equipoise: therapeutic misconception
of the ethics of clinical trials. Hastings Center
Report, 33(3): 1928.
Miller, P. and Weijer, C. (2009) The trust based
obligations of physicians to patients in
clinical research. In The Limits of Consent:
A Socio-legal Approach to Human Subject
Research in Medicine, ed. O. Corrigan,
K. Liddell, J. McMillan, M. Richards and
C. Weijer. Oxford: Oxford University Press.
Gauld, R. and McMillan, J. (1999) Ethics

committees and qualitative health research in
New Zealand. New Zealand Medical Journal,
112: 1957.
National Commission for the Protection of

Human Subjects of Biomedical and
Behavioural Research (1979) The Belmont
Report: Ethical Principles and Guidelines for
the Protection of Human Subjects of Research.
Bethesda, MD: Office of Protection from
Research Risks, US Public Health Service.
Gostin, L. (1991) Ethical principles for the

conduct of human subject research:
population-based research and ethics. Law,
Medicine and Health Care, 19: 191201.
Northington Gamble, V. (1997) Under the

shadow of Tuskegee: African Americans and
health care. American Journal of Public
Health, 87: 17738.
Hope, T. and McMillan, J. (2004) Challenge

studies of human volunteers: ethical issues.
Pappworth, M. (1968) Human Guinea Pigs:

Experimentation on Man. London:
Beacon Press.
Institute of Medicine (1988) The Future of Public

Health. Washington, DC: National Academy
Press.
Punch, M. (1994) Politics and ethics in

qualitative research. In The Sage Handbook
of Qualitative Research, ed. N. K. Denzin and
S. Yvonna. Lincoln. Newbury Park,
CA: SAGE.
Feinberg, J. (1990) Harmless Wrongdoing.

Karim, S. (1998) Placebo controls in HIV

perinatal transmission trials: a South
Africans viewpoint. American Journal of
Public Health, 88: 5646.
Lurie, P. and Wolfe, S. (1997) Unethical trials of
interventions to reduce perinatal
transmission of the human
immunodeficiency virus in developing
countries. New England Journal of Medicine,
337: 115961.
McMillan, J. and Conlon, C. (2002) Developing
world research and the Nuffield Councils
Report. Journal of Medical Ethics, 30: 2046.
Rawls, J. (1972) A Theory of Justice. Oxford:

Research in Health and Behavioral Change
(1989) Changing the Public Health.
Chichester: John Wiley.
Royal College of Physicians (1996) Guidelines on
the Practice of Ethics Committees in Medical
Research Involving Human Subjects, 3rd edn.
London: Royal College of Physicians.
Stoljar, N. (2007) Theories of autonomy. In The
Principles of Healthcare Ethics, 2nd edn, ed.
190
Section 2: Issues
R. Ashcroft, A. Dawson, H. Draper and

J. McMillan. Chichester: John Wiley.
Wertheimer, A. (1996) Exploitation. Princeton,

NJ: Princeton University Press.
Van Willigenburg, T. (2007) Reflective

equilibrium as a method in health care
ethics. In The Principles of Healthcare Ethics,
2nd edn, ed. R. Ashcroft, A. Dawson,
H. Draper and J. McMillan. Chichester:
John Wiley.
Wertheimer, A. (2007) Exploitation in health

care. In The Principles of Healthcare Ethics,
2nd edn, ed. R. Ashcroft, A. Dawson,
H. Draper and J. McMillan. Chichester:
John Wiley.

Ethics, Prevention and Public Health, ed.
University Press.
Wade, D. (2005) Ethics, audit and research:
all shades of grey. British Medical Journal,
330: 473.
Wilkinson, M. and Moore, A. (1997)

Inducement in research. Bioethics, 11(5):
37389.
World Medical Association (2008) Declaration
of Helsinki. Ethical Principles for Medical
Research Involving Human Subjects. 59th
WMA General Assembly, Seoul, Korea,
October 2008. Available at: http://www.wma.
net/e/policy/b3.htm (accessed: 8/12/08).
Section 2
Chapter
11
Issues
Equity and population health

Toward a broader bioethics agenda
Norman Daniels
Introduction
In its early decades, bioethics concentrated on two important problem areas: (1) the dyadic,
very special relationships that hold between doctors and patients and researchers and
subjects; and (2) Promethean challenges, the powers and responsibilities that come with
new knowledge and technologies in medicine and the life sciences, including those that bear
on extending and terminating life. The dyadic relationships yield important goods, impose
significant risks, are rife with inequalities in power and authority, and yet are bound by
complex rights and obligations. They provide a rich field for ethics to explore. The
Promethean challenges are the favourites of the media: how god-like can we become in
our relations with people, with animals, and with our environment without losing our
moral footing? They attract serious inquiries about how to use knowledge and technology
responsibly for the individual and collective good. Unfortunately, they also form the
frontline trenches for contemporary culture wars.
Bioethics focus on the largely non-institutional examination of these dyadic relations
and the emergence of exotic technologies means other important issues concerning
population health and its equitable distribution were not addressed. (There are exceptions
to this generalization.) The doctorpatient relationship and the researchersubject relationship do have a bearing on population health, since medicine and medical research
have impact on the health of individuals and populations. By not examining the broader
institutional settings and policies that mediate population health, bioethics has sometimes
been myopic, not seeing the context in which these relationships operate. Similarly, the
focus on exotic technologies may blind bioethics to the broader determinants of health
and thus to factors that have more bearing on a larger good both domestically and
globally.
To motivate a broader bioethics agenda, I shall focus on equity in three problem areas:
(1) health inequalities between different social groups and the policies needed to reduce
them; (2) intergenerational equity in the context of rapid societal aging; and (3) issues of
equity raised by international health inequalities and the international institutions and
policies that have influence on them. Each area has both domestic and international
implications.
There are good reasons for pursuing this broader agenda. The agenda aligns bioethics
with the goal of more effectively promoting a fundamental good, improved population
191
192
Section 2: Issues
health, especially for those who enjoy less of it domestically and internationally. It
focuses bioethics on the pursuit of justice. Justice obliges us to pursue equity and
fairness in the promotion of health, but policy needs the guidance of ethics in determining what this means. These population issues provide the relevant institutional
context in which we should think about the role of new technologies and the dyadic
relationships I mentioned earlier. Finally, for bioethics to play this role, it must draw on,
and train its practitioners in, a wider range of philosophical skills and social science
disciplines.
What must we do to pursue equity in health?

Health egalitarians and health maximizers
I take health to mean normal functioning, that is, the absence of pathology, mental or
physical.1 This biomedical account of health is clearly narrower than the widely quoted
WHO (1946) definition: Health is a state of complete physical, mental and social well-being
and not merely the absence of disease or infirmity. The WHO conception erroneously
expands health to include nearly all of well-being, so it may no longer function as a limit
notion. People who actually measure population health, such as epidemiologists, concentrate on departures from normal functioning. As we shall see, understanding health as
normal functioning is quite compatible with taking a broad view of the determinants of
health revealed by the social determinants literature.
This characterization of health has implications for what counts as pursuing equity in
health.2 Every society has some healthy and some unhealthy people. One natural way to
understand the goal of equity in health the goal of health egalitarians is to say that we
should aim, ultimately, to make all people healthy, that is, help them to function normally
over a normal lifespan.3 Pursuing equality means levelling up bringing all those in less
than full health to the status of the healthy.4
Surprisingly, the ultimate aim of health maximizers is identical to that of health
egalitarians: We maximize population health if all people function normally over a normal
This characterization is neutral between a value-free, stochastic account of normal functioning, such
as Boorses (1975, 1997), and a modestly normative, etiological (or evolutionary) account, such as
Wakefields (1992) account of mental disorders as harmful dysfunctions. Neither account views
pathology simply as an unwanted condition without providing a clear, objectively ascribable view of
what makes it a dysfunction at some level within the organism.
I side-step a debate about whether to measure all individual variation in health, later factoring in
information about demographic sub-groups (Gakidou et al., 2000), or to focus on health
inequalities across important sub-groups, for example, by class or race or ethnicity (Braveman et al.,
2002).
Rawlss social contract situation involves the simplifying assumption that all people are fully
functional over a normal lifespan. We might take this to be an egalitarian default position (Rawls,
1971).
My health egalitarian behaves like Parfits (1995) prioritarian: one would not level down the
better health of some to make them more equal with those in worse health (blind the sighted to
equalize health states with the blind) if there were no reasonable off-setting gain to those who are
in worse health. Doing so would frustrate the ultimate egalitarian goal of making all fully normal
over a normal lifespan.
Chapter 11: Equity and population health
193
lifespan. Health is clearly different from income (and possibly well-being). There is no
natural stopping point for income the rich can always be richer but health is a limit
concept. Being completely healthy is being completely healthy. Despite convergence
on ultimate aims, health egalitarians and health maximizers generally support different
strategies or policies for achieving their common ultimate aim of producing a completely
healthy population.5 A maximizing strategy or policy will seek the highest achievable
aggregate measure for resources invested, regardless of how the health is distributed.
Someone concerned with equity in health will put important constraints on how the health
is distributed.
Unsolved rationing problems

A family of unsolved distributive problems has been discussed by people in the social
sciences and bioethics (Kamm, 1987, 1993; Brock, 1988; Daniels, 1993, 1994). In these
problems, maximizing strategies are pitted against fairness or equity considerations. For
example, when we select an intervention because it has the best cost-effectiveness ratio, we
are maximizing health benefits at the margin regardless of how the benefits are distributed. Consider three ways this maximizing strategy conflicts with concerns about equity.
First, it gives no priority to those who are worse off or in greater need. Most people want
to give some priority to those who are worse off even if they do not want to give them
complete priority. How much priority should we give? That is the priorities problem.
Second, cost-effectiveness analysis (CEA) allows us to aggregate minor benefits to larger
numbers so that they outweigh significant benefits to fewer people. Yet, even though most
people accept some forms of aggregation, they reject unrestricted aggregation, refusing to
allow, for example, life-saving treatments to a few to be outweighed by very minor
benefits to very large numbers. The aggregation problem asks when should we aggregate?
Third, CEA doggedly pursues best outcomes while denying fair chances for some benefit
to those with worse outcomes. Yet most people reject a strict maximizing strategy,
preferring to give people fair chances at some benefit. How should we balance
best outcomes against fair chances? This is the best outcomes/fair chances problem.
We have considerable trouble agreeing on what the appropriate middle ground is in each
of these problems.
Reducing health inequalities and unsolved distributive problems

The same distributive problems arise when we think about eliminating health inequalities.
Five of the eight internationally negotiated Millennium Development Goals (MDG) are
inequality reducing, aiming at poverty reduction or providing primary education to those
who lack it. The three health targets, however, are stated in terms of reducing population
aggregates, for example, mortality of children under age 5. Gwatkin (2002) models two extreme
approaches to these aggregate goals. A maximizing approach aims at rapid achievement of
the target goal by directing resources to those who are already better off but easier to reach
with strategies for improvement. An egalitarian approach aims to help those who are worst
off first, then the next worst off and so on. Programme incentives and the geo-politics
5
I set aside ethical and conceptual problems that arise in the construction of summary measures of
population health, which allow us to aggregate across various health conditions of different
seriousness and length.
194
Section 2: Issues
surrounding the MDG programme mean that the maximizing strategy is more likely to
be implemented, since funders want rapid results, although it actually increases inequality in the population.
The unsolved distributive problems are raised in contexts where it is not morally
problematic why those who are worse off are so they are just sicker than the others for
whatever reason. In the MDG problem, as in the concern about inter-group health inequalities more generally, the baseline distribution is itself morally problematic because there is
some social responsibility for creating the basic health inequality. Racial disparities in the
USA, ethnic disparities in the UK and gender inequalities in the prevalence of HIV/AIDS in
sub-Saharan Africa may be clear examples. The injustice of the existing baseline may give
extra weight to the concern that we minimize inequalities, giving impetus to efforts to draw
attention to race disparities in health in the USA and to stronger efforts to reduce class
disparities in the UK, Sweden and elsewhere in Europe. Indeed, the WHO Commission on
the Social Determinants of Health encourages broad attention to health disparities and their
origins in the inequitable distribution of various other goods.
How much should this consideration of the injustice of the baseline outweigh our
concern that we are not achieving best outcomes in the aggregate? Some may object that
if we single out some groups as more deserving because they were wronged, then we are
abandoning the principle that we ought to focus only on need in medical contexts. We
should not, then, give priority to fixing the broken leg of the mugging victim over the skier.
When the Chinese decided to give priority in access to antiretroviral treatments to victims
of infected blood, they were criticized for stigmatizing as less deserving those people
infected in other ways.
Moral disagreement about these conflicting concerns will be sharp. There will be
disagreements about who is really responsible for the baseline and efforts to explain its
injustice away (perhaps in the form of victim-blaming). The unsolved distributive problems
are thus made even more difficult. Bioethics has barely risen to the challenge of solving
them when they have morally neutral baselines. It must also address the added challenge
posed by inequitable baselines.
Social determinants and health inequalities
Most Americans, and I suppose most British, who are asked, What does it take to assure
people of equity in health? will respond with what they take to be a truism, Give people
equal access to appropriate medical care, for example through a universal coverage
insurance scheme. At best, this apparent truism is but a small part of the answer; at worst,
it is misleading in important ways.
Equal access to medical services does not by itself assure equity if we have made the
wrong trade-offs in our health system between equity and the maximization of aggregate
health benefits. Just as important, we cannot produce equity in health simply by distributing
medical or even public health resources equitably. Health inequalities have more complex
origins. We know from the longitudinal Whitehall studies of British civil servants of
different employment ranks, for example, as well as from other studies, that health inequalities in a population may not decrease, and may increase, even with universal coverage
(Marmot and Wilkinson, 1999). The Whitehall study involves a study population that
suffers no poverty and has had basic education. Our health is thus affected not simply by
the ease with which we can see a doctor or be treated in a hospital, and not simply by the
195
reduction in risks that come from traditional public health measures though these factors
surely matter but also by our social position and the underlying inequality of our society.6
If we accept as otherwise just the inequalities we allow in our society, but these
inequalities contribute to health inequalities, then should we view these health inequalities
as themselves just? Or are significant health inequalities across groups always grounds for
altering the distribution of other goods? Our answer may depend on the kinds of other
inequalities that we see producing health inequalities (Daniels et al., 1999, 2000). Turn from
class, for the moment, to race.
American data reveals a significant but complex independent effect of race or racism
on health. African Americans have worse health than whites at every income and educational level. Institutional and overt racism must be included as further social determinants
of health. For example, the increasing de facto residential segregation that we see in
America contributes significantly to these inequalities. The complex pattern by race and
ethnicity of key behavioural risk factors (diet, tobacco, alcohol, and substance use and
abuse, violence) contributes to, but does not account for, race and ethnic inequalities in
health. In addition, medical treatment patterns differ by race, a result, perhaps, of conscious
and unconscious stereotyping. A society that has a legacy of caste structure and exclusion
creates psycho-social stresses in many institutional settings (schools, the workplace, shopping malls) that are implicated in health inequalities. Similar issues affect many immigrant
ethnic minorities in European countries.
Race inequalities seem to be the easy case. What about the socio-economic status (SES)
induced inequalities we began with? We live in societies that tolerate and even encourage
some significant degree of inequality as incentives, as justifiable desert, as an expression of
diversity. Should we count as unfair or unjust health inequalities that result from other
social inequalities that we think acceptable or justifiable?
In earlier work (Daniels et al., 1999), I argued that Rawlss principles of justice as
fairness quite unexpectedly capture what the social epidemiological literature picks out as
the key social determinants of health ranging from effective political participation rights
to education and early childhood training to significant restrictions on income and wealth
inequalities to supports for the social basis of self-esteem. Conformance with them would
flatten socio-economic gradients of health more than any we see around us. Social justice,
that is, fair terms of social cooperation developed in abstraction from thinking about health,
is good for our aggregate health and leads to a more equitable distribution of it.
This conclusion is portrayed in the following argument:
1. To maximize population health requires making all people healthy. Making all people
healthy achieves complete equity in health. (This is a conceptual point.)
2. There is no social justice without equity in health. (This is a widely held normative
belief.)
3. There can be no equity in health without social justice. (This is an empirical and
causal claim that depends on what we know about the social determinants of health,
combined with the hypothesis that distributing them in accord with Rawlsian
principles of justice flattens health inequalities [Daniels et al., 1999]).
6
Because these additional social factors so crucially affect population health and its distribution, the
distinction that labels health as a natural rather than as a social good, as in Rawls (1971) or Nagel
(1977), is less clear.
196
Section 2: Issues
4. Therefore, achieving the best level of population health requires (causally) that we
pursue social justice more broadly.7
If social justice is important to population health and its fair distribution, then the policies
aimed at equity in health must be intersectoral in scope. All socially controllable factors that
affect the distribution of health become the concern of those pursuing equity in health. In a
striking way, this perspective challenges one version of the view that we should treat health
as a separate sphere focusing on health benefits without thinking about the contributions
health makes across spheres.
We live in a non-ideal world that does not comply with Rawlsian principles of justice.
We face important questions left unanswered by ideal theory. Many health improving
interventions we may undertake have the property that they increase the levels of health
of all parts of the population even as they may increase health inequalities (Mechanic, 2002).
For example, black infant mortality rates (IMRs) were 64% higher than whites in 1954 but
were 130% higher in 1998, even though white rates dropped by 20.8 per 1000 and black
IMR dropped by 30.1 per thousand. Mechanic (2002) concludes from this and other cases
that it is reasonable to accept increasing health inequalities that result from policies that
improve population health as long as the health of all groups is being improved.
Suppose, however, that we have two interventions (whatever sector, whatever novel
technology) that both raise the health of all groups. If intervention A does less for those who
are worse off than B but much more for those much better off, then both satisfy Mechanics
criterion. We may have strong views about whether to pursue A or B, depending on further
facts about the magnitude of the effects or other facts about the sizes of the groups and thus
the total impact of the programs. If society is responsible for causing the initial inequality
through unfair policies, it may have special obligations to give more weight to equity than
maximization and to considering the speed at which it rectifies the effects of past injustice.
The complexity of inequality itself

Policy choices about reducing health disparities are especially complex because they are at
the interface of claims about injustice and standard distributive problems about which
reasonable people disagree. Unfortunately, another source of complexity derives from what
Temkin (1993) has identified as the complexity of inequality itself.
Temkin describes situations in which two or more groups of individuals differ in their
levels of well-being. He then asks the normative, not descriptive, question, which has the
worse inequality? Specifically, someone who is worse off has a complaint about the unfairness of the inequality. The strength of that complaint depends on whether we compare
those who are worst off (1) with those who are best off, (2) with all those better off than they
7
Suppose we flatten SES gradients of health as much as the principles of justice would seem to require
us to but there remain socio-economic inequalities that induce some health inequalities. Are these
residual health inequalities just? Or must we eliminate all social or economic inequalities that
contribute to health inequalities? Some might interpret the priority Rawls gives to opportunity as
requiring this response. Then, Rawlss theory becomes more egalitarian than was supposed.
Alternatively, we might come to understand the mechanisms through which health inequalities are
produced by other inequalities and intervene to reduce them without having to reduce otherwise
justifiable inequalities. On another reading, Rawlss theory may not specifically answer this question
about residual health inequalities (see Daniels et al., 1999).
197
are or (3) with the average to determine the magnitude of their complaint. To determine
when one inequality is worse than another, we must not only assess the strength of each
complaint, but we must aggregate those complaints within each situation. There are three
approaches to aggregation: a maximin egalitarian view, an additive view, or a weighted
additive view about how to sum up complaints. The nine combinations of these bases for
judging inequalities better or worse yield divergent judgments about cases, including ones
with multiple groups and ones involving welfare transfers among groups. All nine
approaches, for example, prefer to make the worst off individual or group better before
adding comparable benefits to any of the other individuals or (equal-sized) groups, but they
differ on judgments about other cases.
Temkin argues that none of these nine combinations can be dismissed outright as
inconsistent or otherwise completely implausible. Consequently, Temkins egalitarian must
accept the fact that reasonable people will often disagree about when one situation is worse
with regard to inequality than another. Since egalitarians will give more weight to reducing
worse inequalities than ones that are not as bad, they will have reasonable disagreements
about which inequalities to give priority to reducing. This explanation for the disagreement
in judgments about when one inequality is worse than another may underlie some disagreements about how much priority to give worst off individuals or groups.
Broadening the bioethics agenda

Bioethics is not the right field to find the relevant policy levers to reduce health inequalities.
That is a task of social epidemiologists and other social scientists. But bioethics should
provide guidance, in light of the complexity we have discussed, to the policy decisions that
involve different ways of trading off equity against maximization. There are two key
dimensions to that body of work.
First, there is purely normative work trying to find consensus on principles that might
guide us in the range of cases posed by our policy options, including those that arise in
developing and disseminating new technologies. These bioethics agenda items bear on this
normative work:
1. Advance the existing ethical work on the unsolved distributive problems.
2. Clarify when a health inequality is unjust.
3. Explain how that injustice affects the unsolved distributive problems.
4. Clarify what counts as a reasonable rate of progress toward reducing health inequalities.
5. Test the implications of (1)(4) in the context of actual policy choices about reducing
health disparities, including those that involve the dissemination of new technologies.
Second, bioethics must consider what to do when we cannot achieve consensus on principles that can resolve the disagreements we encounter in agenda items (1)(5). All these
problems must be solved in ways that are perceived to be fair and legitimate in real time.
Where we lack consensus on distributive principles, we must rely on procedural justice to
give us fair and legitimate resolution to moral disagreements. In effect, procedural justice
must supplement principled approaches to problems wherever the principles we can agree
on are too indeterminate or coarse-grained to resolve disputes.
Sabin and I developed an approach we call accountability for reasonableness and have
used it to examine medical resource allocation in managed care contexts in the USA
(Daniels and Sabin, 2002). The approach appeals to key elements of deliberative democratic
198
Section 2: Issues
theory to characterize the features of fair process in a range of decision-making contexts

and institutions. I have used it, for example, to address issues of equity in scaling up
antiretroviral treatments in high prevalence countries in the context of the WHO 3 by 5
programme (Daniels, 2005). It is also being used to improve the legitimacy of decisionmaking about coverage of treatments within the catastrophic insurance scheme of the
Mexican Seguro Popular. It is cited as a framework for ethical deliberation about the
implications of cost-effectiveness analysis in a recent IOM report on regulatory contexts
(Miller et al., 2006). Others are using it to improve decision making in publicly managed
systems in Canada, Norway, Sweden, New Zealand and the UK (where NICEs Citizens
Council derives some support from our approach). Still, there are many problems developing appropriate versions of this approach at the various institutional levels where policy
regarding health inequalities is made and implemented.
My expanded agenda calls for bioethics to:
6. Develop the general account of fair process so that reasonable people who disagree can
view policies as fair and legitimate.
7. Apply the account to the various institutional contexts at which they must be addressed.
Equity between age groups and birth cohorts in the context of

societal aging
Societal aging, especially in developing countries, will emerge as a major public health
problem of the twenty-first century. Societal aging intersects with and complicates two
under-analysed problems of intergenerational equity. Though I have earlier written about
the problems of equity between age groups and equity between birth cohorts, I underestimated the difficulty of integrating solutions to these problems in the face of persistent
societal aging.
Societal aging is greeted as a crisis in many recent book titles (which refer to an age
quake, age wave or generational storm to note a few (Peterson, 1999; Wallace, 2001;
Kotlikoff and Burns, 2004)), even though it is a result of the success, not the failure, of
widely pursued health and family planning policies aimed at reducing mortality and fertility
rates. It is accentuated when some birth cohort is much larger than the one following it as
with the American post-war baby boom or the Chinese cohort that enjoyed dropping
mortality rates that preceded the one child policy. Societal aging is a global phenomenon
that has broad impacts on social structure and not just health.
In the USA, Kotlikiff and Burns (2004) observe that walkers replace strollers.
By 2030, nearly 20% of the US population will be 65 or over, whereas only 4% were in
1900. By 2040, the number of Americans over 80 (26.2 million) will exceed the number of
pre-school children (25 million) (Peterson, 1999). European countries, including the UK,
have already reached zero population growth. In Italy, the fertility rate (1.2) is well below
the level at which a population can reproduce itself (2.1), and the working age population
is already shrinking (as it also is in Japan) (CSIS Commission on Global Aging, 2002).
Americas fertility rate of 2.1 helps insulate it from the more extreme aging Italy faces.
The UN predicts that Italy will have a median age of 54 by 2050, second only to Spain. But
Italy is not alone. All the European G7 countries are below the replacement level in
fertility rates. By 2050, half of Continental Europe will be 49 or older, and well before that,
by 2030, one of every two adults in developed countries will have reached retirement age.
199
The UN projects that the ratio of working-age adults to elderly in the developed world will
drop from 4.5 : 1 today to 2.2 : 1 in 2050.
While the proportion of the elderly in developed countries is due to double over the next
50 years, from 15% to 27%, it is due to triple in East Asia, from 6% to 20%. By 2050, there
will be 332 million Chinese aged 65 years or over, equivalent to the worlds elderly
population in 1990 (Jackson and Howe, 1999). The two billion people over age 60 who
will live in our aging world by 2050 will mostly be living in developing countries.
This rapid societal aging in developing countries takes place without the wealth and the
sophisticated economic institutions that exist in developed countries. As one commentator
noted, China will grow old before it grows rich, unlike the developed countries. And China
is not alone. The rate of increase in the number of older people between 1990 and 2025 is
projected to be eight times higher in developing countries, such as Colombia, Malaysia,
Kenya, Thailand and Ghana than it is in the UK and Sweden (International Center for
Longevity, 2006). By 2050, the transitional economies of Eastern Europe will have populations with 28% elderly, and Latin America will have over 17%, well over the current US rate
(CSIS Commission on Global Aging, 2002).
Two effects of societal aging

Societal aging dramatically changes the profile of needs in a country, creating new and
intensified forms of competition between age groups for scarce resources. It also reduces
societys ability to sustain measures for meeting those needs, sharpening competition
between birth cohorts. Together these effects bring questions about intergenerational equity
to the fore that may have not been noticed under different demographic conditions.
We all know, for example, that the rapid growth of those over 75 Bernice Neugartens
(1982) old-old, those elderly who face especially increased disability and dependency will
bring with it increased burdens for the management of chronic disease and long-term care.
Despite the presence in some developed countries of publicly funded long-term care, most
such care is provided by family members, so the burden of societal aging will increasingly
fall on adult working children, usually women. Yet, nearly one-quarter of all the elderly in
the USA in 1989 had no children, and another 20% had only one child. With more women
in the workforce, the problem of providing family care is intensified, since women have
traditionally been the primary care givers. Pressures will increase to provide costly public
programmes at the same time that the working age population is shrinking.
In developing countries, the problem is not the sustainability of the kinds of publicly
supported social and medical services provided for the elderly in developed countries, but
the sustainability of informal, social structures of support, such as the traditional patterns of
care that involve aged parents living with adult children, as in Japan and China. China, for
example, must face the specific consequence of the success of its very strict population
policy: one child for urban couples, two for rural ones. Like the USA, China will have many
elderly with no children and even more elderly who have only one child to support them.
The Chinese refer to this as the 124 problem, one child must care for two parents and
four grandparents. In 1996, the Chinese government made it a legal requirement that adult
children support their elderly parents, obviously anticipating that traditional filial obligations would be strained to the breaking point by the new demographic realities. The law is
not going to solve the problem. Nor would such a law work in other developing countries in
which rapid aging, extensive urbanization and industrialization, and a lack of existing
health care and income support systems of the elderly collide with traditional family values.
200
Section 2: Issues
The increase in medical needs and the shift in the profile of those needs with societal
aging is much broader than the problem of long-term care for frail elderly people. With
societal aging there are increases in the prevalence of cardiovascular disease, chronic
pulmonary disease, diabetes, arthritis and cancer, as well as Alzheimers disease and other
dementias. The increased cost of treating the greater prevalence of these illnesses imposes
great strains on resources and intensified competition for them in developed countries. The
problem will be even worse in developing country health care systems that have barely
begun to gear up to meet the needs posed by chronic diseases. In poorly funded systems,
beefing up medical services for the chronic illnesses of middle and older age means diverting
resources from primary care and preventive care for the whole population.
Age groups and birth cohorts: two distributive problems
Age groups and birth cohorts are easy to confuse, for the term generation is ambiguous
between them. But they are different. Birth cohorts age, but age groups do not. At any given
moment, an age group consists of a birth cohort; over time, it consists of a succession of
birth cohorts.
The age group problem raises these questions: how do we treat age groups fairly within
distributive schemes, such as health care systems? What is a just allocation of resources to
each stage of life, given that needs vary as we age? When is a distributive scheme age-biased
in an unfair way? Is age itself a morally permissible criterion for limiting access to new
technologies?
Medicare in the USA recently approved three very expensive technologies: left ventricular assist devices (LVAD) as destination therapy for patients ineligible for heart transplants
but suffering from advanced congestive heart failure; lung volume reduction surgery
(LVRS) for select patient groups with chronic obstructive pulmonary disease; and implantable defibrillators (Gillick, 2004). Only the last one fell within any usual cost-effectiveness
threshold. No consideration of opportunity costs entered the deliberation. Since Medicare is
a system only for the elderly, unlike universal coverage systems in other countries, equity
issues in allocation over the lifespan were impossible to raise. We could, for example,
produce more health for both the young and the old were proper screening and treatment
for high blood pressure implemented instead. The billions spent on these technologies are
arguably better spent on other forms of care for the elderly themselves.
How should we think about health care resource allocation across age groups? The key
to thinking about this age group problem is the observation that we all age, though we do
not change race or sex (Daniels, 1988). Treating people differently at different ages,
provided we do so systematically over the lifespan, creates no inequalities across persons.
Treating people differently by race or class or gender creates inequalities that are always in
need of justification. Indeed, treating ourselves differently at different stages of life can
make our lives go better over all we invest in our youth, at some sacrifice of immediate
revenues and pleasures, in order to be rewarded more later in life. We take from ourselves
in our working years in order to make our later, retired years go better.
Prudent allocation over the stages of life should be our guide to fair treatment among
age groups (even if prudence is not a general guide to justice). This prudential lifespan
account must be properly qualified by assuming we already enjoy just distributions across
persons and that we will live with the results over our whole lifespan (Daniels, 1988).
Specifically, we should allocate health care so that it promotes the age-relative fair share of
opportunities (or capabilities).
201
Rationing by age on this view is permissible under some scarcity conditions because it
would not be imprudent to so allocate. This argument does not rely on specific, contested
intuitions about the fairness of age rationing (as do Williams [1997] claim that the old have
had their fair innings or Frances Kamms [1993] claim that the young need more years
more than the old). It relies only on the general prudential allocation model I used. Since
reasonable people disagree about the acceptability of this model and about specific issues,
such as age rationing, we will need fair procedures of the sort I noted before to resolve
disputes about priority setting among age groups. Properly used, a transfer scheme based
on prudential allocation or on some other view of fair outcomes that emerges from fair
process would solve the age group problem.
A solution to the age group problem must also be compatible with solutions to the birth
cohort problem. Imagine that over time different birth cohorts pass through a scheme that
solves the age group problem to our satisfaction. These cohorts are each treated fairly,
I proposed two decades ago, if they have comparable benefit ratios as they age through the
schemes. New technologies that were not available for the elderly when they were young but
will be available over the lifespan of those now young pose a special problem of inter-cohort
equity.
This approach to the birth cohort problem can adjust for temporary demographic shifts,
such as those produced by the US baby boom cohort. It is less clear how it can be modified
to accommodate persistent population decline. That is the new challenge we find in global
aging.
The bioethics agenda I am proposing must:
8. Address the distributive issues raised by the age group problem, including the impact of
new technologies on resource allocation over the lifespan; assess the permissibility of
rationing by age; consider age-bias in health systems, such as the inadequacy of longterm care in the USA and elsewhere, and age bias in our methodologies, such as cost
effectiveness analysis; and consider fair process where reasonable people disagree about
these issues.
9. Address how persistent societal aging affects the complex problem of treating cohorts
equitably while at the same time not undermining proper solutions to the age group
problem.8
International equity and health

Earlier, I suggested that health inequalities are unjust if they arise from an unjust distribution (as specified by Rawlss principles of justice as fairness) of the socially controllable
factors that affect population health and its distribution. Judged from this ideal perspective,
there are indeed many health inequities by race and ethnicity, by class and caste, and by
gender around the world. Health inequity is pervasive globally.
This account is unfortunately silent about important questions of international justice.
When are inequalities between different societies unjust? What do better off societies owe as
a matter of justice (not charity), by way of improving the health of the population in less
8
Privatization strategies do not solve the problem they just represent one conclusion about what
such equity requires and they do so without allowing us to use a scheme that addresses the age group
problem at the same time. In addition, privatization is not even a starter for lifespan health systems
they way it is for income support.
202
Section 2: Issues
healthy societies? Suppose countries A and B each do the best they can to distribute the
socially controllable factors affecting health fairly, and there are no glaring sub-group
inequities. Nevertheless, health outcomes are unequal between A and B because A has
many more resources to devote to population health than B. Is the resulting international
inequality in health a matter of justice? Suppose B fails to protect its population health as
best it can, leading again to population health worse than that of A. Is the resulting health
inequality a matter of international justice?
Recasting the problem as an issue of a human right to health and health care does not
improve the situation for two reasons. First, the obligation to secure a right to health for a
population falls primarily on each state for its own population. Although international
human rights agreements and proclamations also posit international obligations to assist
other states in securing human rights, the international obligations cannot become primary
in the human right to health and health care. External forces cannot assure population
health across national boundaries in the way they might intervene to prevent violation of
some other rights, even when they can afford assistance.
Second, even when a right to health is secured in different states, health inequalities
between them may exist. Since conditions do not always permit everything to be done to
secure a right in one country that may be feasible in another, the right to health and health
care is viewed as progressively realizable.9 Reasonable people may disagree about how best
to satisfy this right, given the trade-offs priority setting in health involves. Consequently
some inequalities may fall within the range of reasonable efforts at progressive realizability. In addition, because of their unequal resources, different states may achieve unequal
health outcomes while still securing a right to health and health care for their populations.
Appealing to human rights does not tell us if these inequalities are unjust and remains silent
on what obligations better off states have to address these inequalities.
Though nearly all people recognize some international humanitarian obligations of
individuals and states to assist those facing disease and premature death, wherever they
are, there is substantial philosophical disagreement, even among egalitarian liberals, about
whether there are also international obligations of justice to reduce these inequalities and to
better protect the rights to health of those whose societies fail to protect them as much as
they might. Nagel, who affirms these humanitarian obligations, argues that socio-economic
justice, which presumably includes the just distribution of health, applies only when people
stand in the specific relation to each other that is characterized by a state. Specifically,
concerns about equality are raised within states by the dual nature of individuals both as
coerced subjects and as agents in whose name coercive laws are made (Nagel, 2005). Rawls
(1999) also did not include international obligations to assure a right to health on the list of
human rights that liberal and decent societies have international obligations of justice to
protect.
This statist view encounters a strong counter-intuition. Life expectancy in Swaziland is
half that in Japan (40 versus 80 [WHO, n.d.]). A child unfortunate enough to be born in
Angola has 73 times as great a chance of dying before age five as a child born in Norway
9
The UN Commission on Human Rights (2003) Urges States to take steps, individually and through
international assistance and cooperation, especially economic and technical, to the maximum of
their available resources, with a view to achieving progressively the full realization of the right of
everyone to the enjoyment of the highest attainable standard of physical and mental health by all
appropriate means, including particularly the adoption of legislative measures.
203
(UNICEF, 2000). A mother giving birth in southern sub-Saharan Africa has a 100 times
as great a chance of dying from her labour as one birthing in an industrialized country
(WHO, 2007). Many of us think there is something not just unfortunate and deserving of
humanitarian assistance, but something unfair about the gross inequality.
Those who claim the gross health inequalities are unjust have quite different, incompatible ways of justifying that view. For example, those who believe that any disadvantage
that people suffer through no fault or choice of their own is unjust would assert that the
disadvantage facing the Angolan child is therefore unjust. The underlying principle of
justice is applied to individuals wherever they are in the cosmos and regardless of what
specific relationships they stand in to others contrary to the RawlsNagel account,
which applies principles of justice to the basic structure of a shared society. The disadvantage of the Angolan child might also be thought unjust by those who, like Rawls or
Nagel, think principles of justice are relational and apply only to a basic social structure
that people share, but who, unlike Rawls or Nagel, believe we already live in a world in
which international agencies and rule-making bodies constitute a robust global basic
structure that is appropriately seen as the subject of international justice developed
perhaps through a social contract involving representatives of relevant groups globally
(Beitz, 2000). Fair terms of cooperation involving that structure would, some argue,
reject arrangements that failed to make children in low income countries as well off as
they could be. Clearly, there may be more agreement about some specific judgments of
injustice than there is on the justification for those judgments or on broader theoretical
issues.
I shall examine briefly two ways of trying to break the stalemate between statist and
cosmopolitan perspectives. One approach aims for a minimalist (albeit cosmopolitan)
strategy that focuses on an international obligation of justice to avoid harming people by
causing deficits in the satisfaction of their human rights (Pogge, 2002, 2005a). It is a
minimalist view in the sense that people may agree on negative duties not to harm even if
they disagree about positive duties to aid. This approach handles some international
health issues better than others. A more promising (relational justice) approach, which
I can only briefly illustrate, requires that we work out a more intermediary conception
of justice appropriate to evolving international institutions and rule-making bodies,
leaving it open just how central issues of equality would be in such a context (Cohen
and Sabel, 2006).
Harms to health: a minimalist strategy
If wealthy countries engage in a practice or policy or impose an institutional order that

foreseeably makes the health of those in poorer countries worse than it would otherwise be,
specifically, making it harder than it would otherwise be to realize a human right to health
or health care, then, Pogge (2005a) argues, it is harming that population by creating this
deficit in human rights. Since this harm is defined relative to an internationally recognized
standard of justice, the protection of human rights, Pogge argues that imposing the harm is
unjust. Moreover, if there is a foreseeable alternative institutional order that would reasonably avoid the deficit in human rights, there is an international obligation of justice to
produce the rights-promoting alternative.
There remains some unclarity about how the baseline against which harm is measured is
specified. When is there a deficit in a human right to health? Whenever a country fails to
204
Section 2: Issues
meet the levels of health provided, say, by Japan, which has the highest life expectancy? Or
is there some other, unspecified standard? Consider two examples.
The brain drain of health personnel

The brain drain of health personnel from low-income to OECD countries may exemplify
Pogges concerns. The situation is dire. Over 60% of doctors trained in Ghana in the 1980s
emigrated overseas (WHO, 2003). In 2002, in Ghana 47% of doctors posts were unfilled and
57% of registered nursing positions were unfilled. Some 7000 expatriate South African
nurses work in OECD countries, while there are 32,000 nursing vacancies in the public
sector in South Africa (Alkire and Chen Alkire, 2004). Whereas there are 188 physicians per
100,000 population in the USA, there are only 1 or 2 per 100,000 in large parts of Africa.
The brain drain does not cause the whole of the inequality in health workers, but it
significantly contributes to it.
International efforts to reduce poverty, lower mortality rates and treat HIV/AIDS
patients the Millennium Development Goals (MDG) agreed upon in 2000 are all
threatened by the loss of health personnel in sub-Saharan Africa. An editorial in the Bulletin
of WHO points out that the MDG goals of reducing mortality rates for infants, mothers and
children under five cannot be achieved without a million additional skilled health workers
in the region (Chen and Hanvoravonchai, 2005). The global effort to scale up antiretroviral
treatments poses a grave threat to fragile health systems, for its influx of funds hardly a
bad thing in itself may drain skilled personnel away from primary care systems that
already are greatly understaffed.
What about causes? There is both a push from poor working conditions and opportunities in low-income countries and a pull from more attractive conditions elsewhere. Is
this simply the market at work, backed by a right to migrate?
Pogges argument about an international institutional order has more specific grip than
the vague appeal to a market. When economic conditions worsened in various developing
countries in the 1980s, international lenders, such as the World Bank and International
Monetary Fund (IMF) insisted that countries severely cut back publicly funded health
systems as well as take other steps to reduce deficit spending. In Cameroon, for example,
in the 1990s, their measures included a suspension of health worker recruitment, mandatory retirement at 50 or 55 years, suspension of promotions and reduction of benefits. The
health sector budget shrank from 4.8% in 1993 to 2.4% in 1999, even while the private
health sector grew (Liese et al., 2004). As a result, public sector health workers migrated
to the private sector and others joined the international brain drain. The international
institutional order increased the push.
The pull attracting health workers to OECD countries is also not just diffuse economic
demand. Targeted recruiting by developed countries is so intensive that it has stripped
whole nursing classes away from some universities in the south. In 2000, the Labour
government in the UK set a target of adding 20,000 nurses to the NHS by 2004. It achieved
the goal by 2002. The UK absorbed 13,000 foreign nurses and 4000 doctors in 2002 alone.
Recruitment from EU countries was flat (many of these countries also face shortages due to
aging populations), but immigration from developing countries continued despite an effort
to frame a policy of ethical recruitment (Deeming, 2004). Arguably, even if there were a
diffuse economic pull, in the absence of active recruiting, the harm would be much less.
The remedy for this harm is not a prohibition on migration, which is protected
by various human rights. The UK has recently announced a tougher code to restrict
205
recruitment from 150 developing countries. In addition it has initiated a US$100m

contribution to the Malawi health system aimed at creating better conditions for retaining
health personnel there. The UK has thus taken two steps that are intended to reduce both
the push and the pull behind the brain drain. Other countries have not followed suit.
International property rights and access to drugs

The minimalist strategy becomes harder to apply in a clear way to other international health
issues. The problem of international property rights and the incentives they create goes
beyond the issue of access to existing drugs, such as the antiretroviral cocktails that were the
focus of attention in recent years. Big Pharma has long been criticized for a research and
development bias against drugs needed in developing country markets. Indeed, it has
responded to existing incentives by concentrating on blockbuster drugs for wealthier
markets, including many me too drugs that marginally improve effectiveness or reduce
side effects slightly. Funding the research needed to develop a vaccine against malaria, for
example, has fallen to private foundations.
Do intellectual property rights and the incentive structures they support create a
foreseeable deficit in the right to health that can be reasonably avoided? Pogge (2005b)
argues that they do. Nevertheless, many drugs developed by Big Pharma under existing
property right protections have filtered into widespread use as generics on essential drug
formularies in developing countries. Health outcomes in those countries are much better
than they would be in the absence of such drugs. Since many of these drugs would not have
been produced in the absence of some form of property right protections, people are not
worse off than they would be in a completely free market with no temporary monopolies on
products.
Arguably, however, different property right protections and different incentive schemes
would make people in these poor countries with poor markets better off than they currently
are. Which schemes ought we to select? Pogge (2005b) proposes that we revise incentives
for drug development by establishing a tax-based fund in developed countries that would
reward drug companies in proportion to the impact of their products on the global burden
of disease. For example, drugs that meet needs in poor countries with very high burdens of
disease would pay more, even if the drugs are disseminated at a cost close to the marginal
cost of production. The tax, he admits, would be hard to establish, but it would be offset in
rich countries by lower drug prices. The programme could be limited to essential drugs
leaving existing incentives in place for other drug products. Even so, the tax and thus the
incentives could vary considerably, presumably with consequences of different magnitude
for the global burden of disease. How do we pick which alternative to use as a baseline
against which a deficit in the right to health is specified? Pogge does not tell us.
Leaving aside the problem of vagueness, Pogges proposal cannot be justified by
appealing to the no harm principle alone. The proposed incentive fund would better help
to realize human rights to health, as Pogge argues, but not optimally helping is not the
same as harming and so the justification has shifted. There may well be good reasons for
an account of international justice to consider the interests of those affected by current
property right protections more carefully than those agreements now do but that takes us
into more contested terrain than the minimalist strategy.
International harming is complex in several ways. The harms are often not deliberately
imposed, and sometimes benefits were arguably intended. The harms are often mixed with
benefits. In any case, great care must be taken to describe the baseline against which harm is
206
Section 2: Issues
measured. Such a complex story about motivations, intentions and effects might seem to
weaken the straightforward appeal of the minimalist strategy, but the complexity does not
undermine the view that we have obligations of justice to avoid harming health.
In any case, international harming is only one of three causes of international health
inequalities. International health inequalities are also the result of (1) domestic failures to
promote population health adequately or fairly and (2) differences across countries in levels
of human and natural resources and in the natural conditions that contribute to the
prevalence of disease (such as malaria). Thus the minimalist strategy fails to address many
inequalities that some believe raise questions of international justice.
The new terrain of global justice: where the action is

Although I noted the strong pull of the cosmopolitan intuition about the unfairness of
international health inequalities, there is also a strong intuition that obligations of justice
arise only when people stand in certain specified relations to each other (relational justice).
If we abandon the idea that such relations can arise only within states, we may find
attractive the view argued for by Cohen and Sabel (2006). They sketch three types of
international relationships that might give rise to obligations of justice going beyond
humanitarian concerns, international agencies charged with distributing a specific good,
cooperative schemes and some kinds of interdependency. Each may give rise to obligations
of justice, such as concerns about inclusion. These may range from an obligation to give
more weight to the interests of those who are worse off if it can be done at little cost to
others, to obligations of equal concern, perhaps yielding far more egalitarian obligations.
I shall illustrate each of these relationships and the obligations they give rise to with
examples focused on key issues of global health.
WHO plausibly illustrates the idea that institutions charged with distributing a particular, important good, such as public health expertise and technology, must show equal
concern in the distribution of that good. The organization would be charged with being
unfair if it ignored the health of some and attended more to the health of others. For
example, this point about showing equal concern arises in other debates about the methodologies WHO employs. We saw that CEA ignores issues of equity in the distribution of
health and health care (Daniels and Sabin, 2002). These criticisms of CEA thus challenge the
unconstrained use of CEA by the WHO whether it is using the methodology to determine
health policy within a specific country or across countries. WHO is constrained by its
mission of improving world health to consider equity in distribution in all contexts in
which it works within and across countries.
Concerns about equity show up in programmatic discussions as well. WHO paid
attention to equity in the distribution of antiretroviral treatments (ARTs) for HIV/AIDS.
I noted earlier WHOs sponsorship of a Commission on the Social Determinants of Health,
with its focus on equity. Both these examples illustrate behavior compatible with and
required by the institutional charge to WHO. Either this is a misguided focus of energy
for WHO, as seems to be implied by Nagels strong statist view, or it is an implication of the
obligation to show equal concern that arises within institutions charged with delivering an
important good whether they operate within states or across them.
Consider now the international bodies that establish rules governing intellectual
property rights, including those that are key to creating temporary monopolies over
new drugs. Such a scheme is consequential in that it increases the level of cooperation
207
in production of an important collective good, research and development of drugs, and it

does so in a way that has normatively relevant consequences (Cohen and Sabel, 2006).
Suppose we conclude that this mutually cooperative scheme generates considerations of
equal concern, or at least that it must be governed by a principle of inclusion. Then we
might view quite favourably Pogges suggestion about structuring drug development
incentives so that they better addressed the global burden of disease. Earlier, I said
Pogges proposal could not be defended on the minimalist grounds that it avoided doing
harm because of the problem of specifying the relevant human rights baseline. Now,
however, we have a new basis on which to defend the justice of Pogge-style incentives.
Such an incentive scheme, supplementing existing property rights or modifying them
appropriately, would greatly enhance the benefits to those who are largely excluded from
benefit for a significant period of time, and it would do so at only modest cost to those
profiting from the endeavour. Minimally, it illustrates what a more inclusive policy
should include; one can build into it even stronger egalitarian considerations, if the
cooperative scheme gives rise to concerns about equality and not simply inclusion.
Exactly what form the policy would take, or the justification for it deriving from the
form of cooperative scheme involved, remains a task for further work. With these issues
worked out, we might then support Pogges incentive schemes as a way of moving some
countries closer to satisfaction of a right to health, connecting the effort to human rights
goals as he does.
Consider again the example of the brain drain of health personnel from low- and
middle-income countries to wealthier ones. Nagel (2005: 130) notes that nations generally
have immunity from the need to justify to outsiders the limits on access to its territory,
though this immunity is not absolute, since the human rights of asylum seekers act as a
constraint. Still, the decisions different countries make about training of health personnel
and about access to their territories have great mutual impact on them. There is an
important interdependency affecting their well-being, specifically, the health of the populations contributing and receiving health personnel. The British decision in 2000 to recruit
30,000 new nurses from developing countries rather than try to train more greatly affected
the fate of people being served by health systems in southern Africa. The under-funding of
salaries for African nurses and doctors, in part a legacy of Structural Reform Programs
imposed by the IMF and World Bank, but clearly continued by local governments, helps
create the push factor driving these workers abroad. Arguably, this relation of interdependence brings into play obligations of inclusion, perhaps those of equal concern, going
beyond in any event humanitarian considerations. In addition to Pogges no harm or
minimalist approach, we thus have available obligations of inclusion requiring us to
consider the interests of all those in the interdependent relationship. These obligations
can be translated into various policy options that address the brain drain: it may be
necessary to restrict the terms of employment in receiving countries of health workers
from vulnerable countries; it may be necessary to seek compensation for lost training costs
of these workers; it may be important to contribute aid to contributing countries aimed at
reducing the push factors; it may be necessary to prohibit active recruitment from vulnerable countries.
We might combine these relationships of interdependence with the relationships and
obligations that arise from cooperative schemes. The International Organization for
Migration, established in 1951 to help resettle displaced persons from the Second World
War, now has 112 member states and 23 observer states. It manages various aspects of
208
Section 2: Issues
migration, providing information and technical advice, and arguably goes beyond its initial
humanitarian mission. Suppose it took on the task of developing a policy that helped to
coordinate or manage the frightening health personnel brain drain. Minimally, it might
seek internationally acceptable standards for managing the flow standards on recruitment,
on compensation, on terms of work. More ambitiously, it might seek actual treaties that
balanced rights to migrate with costs to the contributing countries, countering at least some
of the pull factors and even providing funds that might alleviate some of the push factors
underlying the brain drain. In seeking these, it might work together with the International
Labour Organization (ILO), with the World Trade Organization (WTO), with WHO and
with the United Nations (UN). Such a cooperative endeavor would reflect the common
interest in all countries of having adequate health personnel and thus being able to assure
citizens a right to health and health care as well as the common interest in protecting
human rights to dignified migration.
The fuller development of a plausible account of justice in these intermediary institutions is a task for the expanded bioethics agenda I have been charting:
10. Assess the implications of the obligation not to harm for reducing health inequalities
internationally.
11. Develop an account of justice for the evolving international institutions and rulemaking bodies that have an impact on international health inequalities.
12. Examine Promethean challenges from the perspective of their impact on international
health inequalities and obligations of justice regarding them.
Preparing the field

The broader bioethics agenda I have described poses two distinct and significant challenges to the field. The first challenge is one of training. Many of the problems take us
outside the more familiar domain of ethics and the clinical practice of medicine into the
far less familiar terrain of political philosophy and a wide range of social sciences. The
relevant philosophical literature is less familiar to many in the field and it would have to
be mastered by both those who teach bioethics and those who engage in it. Some training
programs in bioethics already include some of this material, but many would have to
expand their focus.
The second challenge is political. An implication of my earlier argument is that social
justice is good for population health and is essential to its fair distribution. But engaging in
the pursuit of social justice and not simply justice in health care can be divisive in a novel
way. Many people already agree that we have social obligations to give people equal access
to an appropriate array of health care interventions. Perhaps this is because they see the
threat of disease and disability as a part of the struggle of humans against nature. In that
battle, we are united in our vulnerability. We hope that technology will rescue all of us. The
Promethean battle of humans against nature and our own human weaknesses is in a
certain sense unifying.
Many fewer people, however, understand the broad ways in which the distribution of
other goods affect health inequalities, let alone agree about how to distribute those goods.
Indeed, this is a context in which we cannot all easily unite against nature. Rather, there are
divisions of interest and perspective among all of us, including across nations. Shifting a
bioethics agenda to address the causes of health inequality can thus be politically divisive,
both domestically and internationally.
209
Acknowledgements
This article is reprinted (with permission) from the Hastings Center Report (Daniels, 2006).
It was originally adapted from a lecture delivered to the Nuffield Council of Bioethics (May
2005). I also draw on material from my book, Just Health: Meeting Health Needs Fairly (2008).
References
Alkire, S. and Chen Akire, L. (2004) Medical
exceptionalism in international migration:
should doctors and nurses be treated
differently? JLI Working Paper, 73: 110.
Beitz, C. (2000) Rawls law of peoples. Ethics,
110(4): 66996.
Boorse, C. (1975) On the distinction between
disease and illness. Philosophy and Public
Affairs, 5(1): 4968.
Boorse, C. (1997) A rebuttal on health. In What Is
Disease?, ed. J. M. Humber and R. F. Almeder.
Totowa, NJ: Humana Press, pp. 1134.
Braveman, P., Starfield, B. and Geiger, H. J.
(2002) World Health Report 2000: how it
removes equity from the agenda for public
health monitoring and policy. British Medical
Journal, 323: 67881.
Brock, D. (1988) Ethical issues in recipient
selection for organ transplantation. In Organ
Substitution Technology: Ethical, Legal and
Public Policy Issues, ed. D. Matthieu. Boulder,
CO: Westview Press, pp. 8699.
CSIS Commission on Global Aging (2002)
Meeting the Challenge of Global Aging.
Washington, DC: Center for Strategic and
International Studies. Available at: http://
www.csis.org/gai (accessed: 27/8/05).
Chen, L. and Hanvoravonchai, P. (2005) HIV/
AIDS and human resources. Bulletin of the
World Health Organization, 83: 1434.
Daniels, N. (2005) Fair process in patient

selection for antiretroviral treatment for
HIV/AIDS in WHOs 3 by 5 program.
The Lancet, 366: 16971.
Daniels, N. (2006) Equity and population health:
toward a broader bioethics agenda. Hastings
Daniels, N. (2008) Just Health: Meeting Health
Needs Fairly. Cambridge: Cambridge
University Press.
Daniels, N., Kennedy, B. and Kawachi, I. (1999)
Why justice is good for your health: Social
determinants of health inequalities.
Daedalus, 128(4): 21551.
Is Inequality Bad for Our Health? Boston,
MA: Beacon Press.
Daniels, N. and Sabin, J. E. (2002) Setting Limits
Fairly: Can We Learn to Share Medical
Resources? New York: OUP.
Deeming, C. (2004) Policy targets and ethical
tensions: UK nurse recruitment. Social Policy
and Administration, 38(7): 22792.
Gakidou, E. E., Murray, C. J. L. and Frenk, J.
(2000) Defining and measuring health
inequality. Bulletin of the World Health
Organization, 78: 4254.
Gillick, M. (2004) Medicare coverage for
technological innovations time for new
criteria? New England Journal of Medicine,
350(21): 2199203.
Daniels, N. (1993) Rationing fairly: programmatic

considerations. Bioethics, 1: 25571.
Gwatkin, D. R. (2002) Who Would Gain Most

from Efforts to Reach the Millennium
Development Goals for Health? An
Inquiry into the Possibility of Progress that
Fails to Reach the Poor. Health, Nutrition
and Population Discussion paper.
Washington, DC: World Bank.
Available at: http://poverty.worldbank.org/
files/13920-gwatkin1202.pdf (accessed:
27/8/05).
Daniels, N. (1994) Four unsolved rationing

problems. Hastings Center Report, 24(4): 279.
International Centre for Longevity (2006) Aging

on the World Stage. New York: ICL.
Cohen, J. and Sabel, C. (2006) Extra

rempublicam nulla justitia. Philosophy and
Public Affairs, 34(2): 14775.
Daniels, N. (1988) Am I my Parents Keeper?
An Essay on Justice Between the Young
and the Old. New York: Oxford University
Press.
210
Section 2: Issues
Jackson, R. and Howe, N. (1999) Global Aging:

The Challenge of The New Millennium.
Washington, DC: CSIS/Watson Wyatt
Worldwide. Available at: http://www.csis.
org/gai (accessed: 27/8/05).
Kamm, F. (1987) The choice between people,
commonsense morality, and doctors.
Bioethics, 1: 25571.
Kamm, F. (1993) Morality, Mortality, Volume I:
Death and Whom to Save from It. Oxford:
Pogge, T. (2005b) Human right and global

health: a research program. Metaphilosophy,
36: 889909.
Rawls, J. (1971) A Theory of Justice. Cambridge,
MA: Harvard University Press.
Rawls, J. (1999) The Law of Peoples. Cambridge,
Temkin, L. (1993) Inequality. New York: Oxford
University Press.
Kotlikoff, L. J. and Burns, S. (2004) The Coming

Generational Storm. Cambridge, MA: MIT
Press.
UN Commission on Human Rights (2003)

Resolution 2003/28: The Right of Everyone to
the Enjoyment of the Highest Attainable
Standard of Physical and Mental Health.
Geneva: UNHR.
Liese, B., Blanchet, N. and Dussault, G. (2004)

The Human Resource Crisis in Health Services
in Sub-Saharan Africa. Background Paper:
World Development Report 2004. Washington
DC: The World Bank.
UNICEF (2000) The State of the Worlds

Children. Under-Five Mortality Rankings.
Copenhagen: UNICEF. Available at: http://
www.unicef.org/sowc00/stat2.htm (accessed:
12/8/05).
Marmot, M. and Wilkinson, R. G. (1999) Social

Determinants of Health. Oxford: Oxford
University Press.
Wakefield, J. C. (1992) The concept of mental

disorder. On the boundary between
biological facts and social values. American
Psychologist, 47: 37388.
Mechanic, D. (2002) Disadvantage, inequality,

and social policy. Health Affairs, 21(2): 4859.
Miller, W., Robinson, L. S. and Lawrence, R. S.
(eds) (2006) Valuing Health for Regulatory
Cost-Effectiveness Analysis. Washington, DC:
National Academies Press.
Nagel, T. (1997) Justice and nature. Oxford
Journal of Legal Studies, 17(2): 30321.
Nagel, T. (2005) The problem of global justice.
Philosophy and Public Affairs, 33(2): 11347.
Neugarten, B. L. (1982) Age or Need? Public Policies
for Older People. Beverley Hills, CA: SAGE.
Parfit, D. (1995) Equality or priority? Lindley
Lecture at the University of Kansas.
Peterson, P. (1999) Gray Dawn: How The Coming
Age Wave Will Transform America And The
World. New York: Three Rivers Press.
Pogge, T. (2002) World Poverty and Human
Rights: Cosmopolitan Responsibilities and
Reforms. Cambridge: Polity Press.
Pogge, T. (2005a) Severe poverty as a violation
of negative duties. Ethics and International
Affairs, 19(1): 5583.
Wallace, P. (2001) Agequake: Riding the

Demographic Rollercoaster Shaking Business,
Finance and Our World. London: Nicolas
Brealey.
Williams, A. (1997) Intergenerational equity: an
exploration of the fair innings argument.
World Health Organization (n.d.) WHO
Mortality Database: Tables. Available at:
http://www.who.int/healthinfo/morttables/
en/index.html (accessed 23/8/07).
World Health Organization (1946) Preamble to
Constitution of WHO, adopted by
International Health Conference, NY 1922
June 1946; entered into force 7 April 1948,
never amended.
World Health Organization (2003) International
Migration, Health & Human Rights. Geneva:
WHO.
World Health Organization (2007) Maternal
Mortality in 2005: Estimates Developed by
WHO, UNICEF, UNFPA, and the World
Bank. Geneva: WHO.
Section 2
Chapter
12
Issues
Health inequities
James Wilson
Introduction
The infant mortality rate (IMR) in Liberia is 50 times higher than it is in Sweden, while a
child born in Japan has a life expectancy at birth of more than double that of one born in
Zambia (Central Intelligence Agency, 2007).1 And within countries, we see differences that
are nearly as great. For example, if you were in the USA and travelled the short journey
from the poorer parts of Washington to Montgomery County Maryland, you would find
that for each mile travelled life expectancy rises about a year and a half. There is a twentyyear gap between poor blacks at one end of the journey and rich whites at the other
(Marmot, 2004: 2).
There are two types of questions that it is important to ask about inequalities in health
such as these. The first are social scientific questions about the extent of inequalities in
health and the factors which are causally responsible for these inequalities. Examples of
social scientific questions to ask might be: how do infant mortality rates in the UK differ
according to social class? What is the difference in life expectancy between Japanese who
emigrate to the USA and those who remain in Japan? Why do civil servants in higher
ranked jobs tend to live longer than civil servants in lower ranked jobs?
The second type are normative questions about the reasons we have to care about
inequalities in health. Important normative questions to answer are: which inequalities in
health should we care about (all inequalities or merely some of them)? When is an
inequality in health unjust? How should we weigh our concern for equality in health against
other factors such as maximizing the health achievement of community?
This chapter focuses on these normative questions. But I shall first briefly outline some of the
main findings of the social sciences literature in order to put the normative questions in the right
perspective. Much of the relevant literature has focused on the relationship between socioeconomic status (SES) and health achievement, and I shall follow this lead in my summary.
The gradient in health

There is evidence of a socio-economic gradient in health in all countries for which
statistics are available. In other words, as a persons SES increases, so her life expectancy
1
The infant mortality rate in Liberia is 149.73 per 1,000; in Sweden it is 2.76. The life expectancy at
birth in Japan is 82; in Zambia it is 38.4 (Central Intelligence Agency, 2007).
211
212
Section 2: Issues
improves and so also do a range of other important health indicators.2 The gradient is by
no means confined to groups who suffer absolute deprivation. It also occurs within
groups who are in absolute terms, fairly well off. For example the famous Whitehall
studies showed a gradient in health among civil servants, all of whom were, in absolute
terms, comfortably off (Marmot et al., 1978). So it is not just that the poor, as a group
have worse health, but also that, as we go up the social scale, each rung we ascend will
increase life expectancy, and decrease the chance of developing many diseases, such as
stroke or heart disease.3
The socio-economic gradient in health is less steep in some countries than in others
(and also changes in severity in the same countries over time), which suggests that there
must be social factors that can either flatten gradients or make them more vertiginous. And
if the socio-economic gradient in health has social causes, then it seems plausible to think
that it will be in our power to flatten it if we want to.
Does low socio-economic status cause ill health?

The correlation between low SES and ill health is robust. However, correlation is not
causation. It does not follow logically from the fact that people with low SES tend to have
worse health and lower life expectancies that having a low SES causes ill health. For all that
we have so far seen, it might be ill health that causes low SES, or it might be that there is
some further factor that causes both low SES and ill health, while neither low SES nor ill
health cause one another.
Many people think that the direction of causation of health inequalities is important,
because they think that it is morally worse if an inequality in SES causes an inequality
in health, than if an inequality in health causes an inequality in SES. As Daniels et al.
(2004: 63) put it, Many who are untroubled by some kinds of inequality are particularly
troubled by health inequalities. They believe that a socio-economic inequality that otherwise
seems just becomes unjust if it contributes to heath inequalities. Because of this, much of
the social sciences literature on health inequalities has been devoted to establishing that it is
low SES which is responsible for poor health, rather than vice versa.
No one disputes that some of the correlation between SES and health is caused by
the effects of ill health. For example, it is very plausible to think that, as a group, people
2
The term SES is usually used to refer to the relative position of a family or individual on a
hierarchical social structure, based on their access to or control over wealth, prestige and power
(Mueller and Parcel, 1981). However, when defined in this way SES is difficult to measure accurately
and so studies tend to use something easier to measure as a proxy for SES. Popular candidates for
proxy measures include level of education, current income, overall wealth, type of occupation or
some combination of these measures (Shaver, 2007). Each of these measures has their advantages
and disadvantages, and clearly the fact that different studies use different measures creates a degree
of difficulty in comparing the results of different studies examining the relationship between SES and
health (Braveman et al., 2005). However the finding that there is a correlation between SES and
health is sufficiently robust that it holds whichever way we measure SES, so I shall set this problem of
measurement aside here.
There are some (very few) conditions, such as breast cancer in European countries, which show a
reverse socio-economic gradient. (A possible explanation for the reverse economic gradient in
breast cancer is that having children later increases risk of breast cancer, and women with a higher
SES tend to have children later. See Strand et al., 2007.)
Chapter 12: Health inequities
213
who are unable to work because of chronic illness will tend to have a lower income than
people who are able to work. The interesting question is whether all (or even most) of the
socio-economic gradient in health can be explained in this way. And in fact the evidence
shows that the effects of this health selection are fairly small in comparison to the overall
size of the socio-economic gradient.4
Of course, there are other factors which might explain the correlation. One obvious
factor could be that it is inequalities in access to health care which explain the social
gradient in health. However, this does not seem to be a very significant cause of variation
in health, given that we see a significant social gradient in health even in countries such as
the UK, which have a nationalized health system. Another factor could be intelligence: it
might be the case that more intelligent people will tend to do better in their jobs, and also
will tend to take more health preserving behaviours than less intelligent people, so that
intelligence will tend to influence both SES and health. Ultimately, though the question
remains whether we can plausibly account for all the correlation of health and SES without
allowing that low SES causes a significant proportion of the variation in health. And it
seems that we cannot.
In addition, we have a number of possible models which seek to explain how low SES
could cause ill health, by explaining how the social factors associated with lower SES could
have a bad effect on health. Marmot (2004) hypothesizes that people of a lower SES tend to
have less control over their working and living environments, and that this sense of lack of
control leads to stress responses, which predictably cause conditions such as atherosclerosis
and obesity. Wilkinson (1996) argues that it is income inequality that is the key factor which
affects the size of the socio-economic gradient, and that other things being equal, the health
of all members of society tends to be worse in an unequal society. This suggests that goods
such as social capital and social cohesion play a role in governing health states presumably
ultimately via similar causal pathways to those hypothesized by Marmot.
There is very much more that could be said about the social sciences literature, in
particular on the question of how low SES causes ill health. However, our focus is on the
normative questions that arise about health inequalities, and as we shall see below, there is
good reason to think that these questions of causation are only of tangential relevance to the
fundamental normative question of which inequalities in health are unjust.
From social science to political philosophy

Until fairly recently, there was little interaction between social scientists working on the
extent and causes of health inequalities and political philosophers seeking to answer
normative questions about which inequalities are unjust. One reason for this was a failure
(by both sides) to see how materials produced by the other could be relevant to the
questions that they were asking. For instance, it is notable that Rawls, the most influential
political philosopher of the second part of the twentieth century, scarcely addresses health at
4
For example, Chandola et al. (2003) argue that in the Whitehall study, the data show that the effect
of social position on health was over two and a half times greater than the effect of health on
social position. Clearly the size of the causal effect of health on SES will vary from country to country
and from situation to situation, depending on the level of support a society provides for those
unable to work through illness. But nowhere is it plausible to attribute all the socio-economic
gradient in health to health selection.
214
Section 2: Issues
all, and where he does, questions of inequalities in health do not even enter onto his radar.5
Meanwhile, Marmot (2004), who has done more than anyone to raise and to answer the
social scientific questions, reports that it was only fairly recently that he realized that the
normative implications of his work on inequality in health had already been separately
explored at length by political philosophers.6
However, recent times have seen a change. Workers within empirical fields are becoming increasingly aware of the role that ethical thinking and particularly political philosophy
can play, while political philosophers have become increasingly aware of the relevance of
the social scientific literature. One of the major aims of this chapter is help to strengthen
this dialogue.
The concept of a health inequity

Before going any further, we must make an important distinction between health inequalities and health inequities. I shall take health inequalities to be the generic term used to
designate differences, variations, and disparities in the health achievements of individuals
and groups (Kawachi et al. 2002: 647), while I shall take health inequities to be those health
inequalities that are, all things considered, unjust.
We need to make this distinction because it is plausible to think that there are at least some
inequalities in health that are not unjust. And where an inequality is not unjust, it would be
wrong to think that we have a duty to alleviate it or eliminate it.7 To give an example, a recent
study of 1000 major European and North American pop stars 19562005 revealed that they
experience significantly higher mortality (more than 1.7 times) than demographically matched
populations in the USA and UK (Bellis et al., 2007: 896), thus showing a significant health
inequality between pop stars and the ordinary members of the public. But I think that few
would feel that the researchers had thereby uncovered an inequity that the European and US
governments should be in a hurry to address, given that pop stars decreased life expectancy
seems to be due in large part to their choices to adopt high-risk behaviours.
When dealing with concepts such as that of a health inequity, which imply a normative
judgement about those things that fall under them, it is often useful to begin by clarifying
the role that we think the concept should play before going on to specify which things
should fall under the concept. For it is usually much easier to get a consensus on the role
5
There is one passage which is particularly telling, where Rawls allows that health is equally as
important a good as the primary goods which his theory of justice picks out to be distributed fairly.
However he denies that we should treat health as one of the primary goods, on the grounds (1) that
it is natural rather than a social good (meaning that it is much less closely affected by changes in
the basic structure of society than are the genuine primary goods such as rights, liberties and
opportunities, income and wealth, and the social bases of self-respect), and (2) his theory of
justice concerns only the justice of the basic structure of society (1971: 62). While this was a
reasonable view of health to hold in 1971, the mass of literature produced on the social determinants
of health over the past thirty years has clearly demonstrated that (1) is false. Rawls seems never to
have revised his views to take account of the social determinants of health.
As he (2004: 38) puts it, I read for the first time [in 1998], Amartya Sens writings on Inequality . . .
I felt like the man who had discovered he had been talking prose. I was intrigued to discover that
some of the conclusions towards which my colleagues and I had been struggling towards on the basis
of our evidence, were elegantly laid out there.
Of course, it would not necessarily be wrong to alleviate or eliminate it. My point is simply that we
would not be obliged to do so.
215
that a normative concept should play than to get a consensus on what things should fall
under that concept, and once we have this consensus it is easier to understand and to
adjudicate the disputes that will then inevitably arise about which things should fall under
the concept. So the rest of this section aims to clarify the role that the concept of a health
inequity should play.
I shall briefly examine the two most influential definitions of a health inequity, the first
by Whitehead Dahlgren (1992, 2006; Whitehead, 1990), and the second by Kawachi et al.
(2002). I shall argue that neither is adequate, and will suggest a more precise definition.
In a widely cited article, Whitehead (1990: 5) defines health inequities as differences in
health which are unnecessary and avoidable but, in addition are also considered unfair and
unjust.8 Whitehead does not provide much by way of an explanation as to why the
differences should have to be unnecessary and avoidable as well as unjust, and it seems
to me that these additional qualifications are unhelpful. For if we already thought that a
given health inequality was unjust, we would usually be taken to be already making the
claim that it amounted to a health inequity. So I do not think that being also told that the
inequality was unnecessary and avoidable would add anything. Conversely, if we did not
know whether a difference in health was inequitable, discovering that the difference was
unnecessary and avoidable would not provide a reason in and of itself to think that the
difference in question was inequitable. For instance, when competent adults undertake
dangerous sports such as mountaineering, they knowingly increase their risk of death and
injury. This increased risk creates a health inequality between mountaineers and nonmountaineers which is unnecessary and avoidable, but not unjust (and hence not inequitable) (Kawachi et al. 2002: 648). So at the very least, adding the claims that the differences
have to be unnecessary and avoidable adds nothing to the definition of health inequity.
However, things are worse than this. The assumption that health inequities by definition
are unnecessary and avoidable has led Whitehead and Dahlgren and those who have
followed them to make some very implausible claims about what kinds of things could
and could not count as health inequities. Whitehead and Dahlgrens (2006: 2) underlying
thought seems to be that it is only where social processes . . . produce health differences
rather than these being determined biologically, that it makes sense to say that there is a
health inequity, and that hence only inequalities which are caused by social as opposed to
natural factors are even candidate health inequities. Their argument for this claim seems to
be as follows:
Human beings vary in health as they do in every other attribute. We will never be able to achieve a
situation where everyone in the population has the same type and degree of illness and dies after
exactly the same life span. This is not an achievable goal, nor even a desirable one. Thus, that portion
of the health differential attributable to natural biological variation can be considered inevitable
rather than inequitable.
(Whitehead, 1990: 67)
The thought seems to be that inequalities which are due to human activity are avoidable
because they are caused by human action, whereas inequalities caused by nature are not
8
It is unclear why there is a reference to differences which are both unfair and unjust, given that
she does not make any distinction between these two terms, which are in any case close to
synonymous in ordinary usage. I shall assume that we should delete unfair, and simply leave
ourselves with the idea of unjust differences.
216
Section 2: Issues
caused by human action and so we are powerless to do anything about them. There are two
mistakes contained in this: first, it does not follow that just because something has a social
cause that we will be able to successfully stop it by social intervention. (It is by no means
clear that we will be able to stop climate change, for instance.) Second, it does not follow
from the fact that something is caused by nature either that we are powerless to stop it, or
that we should refrain from trying to stop it. (Presumably all of medicine could legitimately
count as trying to stave off what would otherwise be the inevitable operation of nature, as
John Stuart Mill [1874] points out.)
Moreover the conclusion does not follow from the premises. For, even if it would be
unjust and undesirable to attempt to equalize health achievement across a society, it would
not follow that people who are born with worse health have no claim to some form of
rectification on grounds of justice. For example, we may not be in a position to give
someone who is blind sight, and clearly it would be undesirable to blind the rest of the
population to restore equality. But there are other things we can do, such as ensuring that
buildings are arranged for easy navigation by blind people, or ensuring that all official
documents are available in braille, which many people take to be required by justice. Hence
the fact that we cannot alleviate blindness does not mean that we have to say that it is merely
down to nature, and that the disadvantage to the blind person should be considered
inevitable rather than inequitable.
Further, much of the mainstream philosophical writing on justice over the last 30 years
has taken it as axiomatic that undeserved disadvantages due to bad luck (such as being born
blind, or without the use of ones legs, or having a very low life expectancy) do raise issues of
justice, and that undeserved disadvantages of this kind give the affected individuals claim
for rectification under egalitarian justice. This position has now come to be known as luck
egalitarianism. See for example, Dworkin (1981a, 1981b), Arneson (1989) and Cohen
(1989). For critiques of this position, see Anderson (1999) and Scheffler (2003).
This means that Whitehead and Dahlgren must either allow that health inequities
account for only some of the inequalities in health achievement which a just society should
be concerned about, or they must be committed to the claim that luck egalitarianisms
central claim is false, and that in fact it is only inequalities with a social cause which are of
concern for egalitarian justice. Both of these options seem inappropriate for an account of
the concept of a health inequity. Absent any account of why we should treat unjust
inequalities with social causes differently from unjust inequalities with natural causes, it
looks arbitrary to suggest that we should address them separately. And it would seem to be a
mistake to build a controversial claim about the nature of egalitarian justice into our
concept of a health inequity. While such controversial positions might be true, they need
to be argued for as the best accounts of health inequity, rather than simply presupposed as
following from the very concept of a health inequity.
Where does this leave us? I take it that the only part of Whitehead and Dahlgrens
original definition of a health inequity still standing is the idea of an unjust inequality. This
idea forms the kernel of the second major definition of health inequity in the literature,
namely that Health inequity refers to those inequalities in health that are deemed to be
unfair or stemming from some form of injustice (Kawachi et al., 2002: 647). This is, I shall
suggest, much better, but still in need of a little more refinement.
Kawachi et al.s definition suggests that health inequalities which stem from injustice
should count as health inequities. However, it is unclear why we should accept this. For not
every inequality which stems from an injustice is itself unjust. Some inequalities which stem
217
from injustices will be trivial and morally insignificant, while other inequalities might even
benefit those who have been unjustly treated. (For instance, a society which allowed men,
but not women, to smoke would treat women unjustly. However, the health inequalities in
womens favour to which this would no doubt lead would not rightly be considered to be
health inequities.)
So it seems too strong to suggest, as Kawachi et al. seem to, that all inequalities which
stem from injustice are themselves inequitable. Rather, it is only when an inequality which
stems from injustice adversely affects something that we already think we have reason to
care about from the perspective of justice, that we think it is unjust. Hence inequalities in
health that stem from injustice will count as health inequities only if we already have a
reason to think that inequalities in health are something that we ought to care about from
the perspective of justice. And we ought to care about inequalities in health from the
perspective of justice only if such inequalities are either unjust in themselves or contribute
to states of affairs which are unjust in themselves. So if we build into our definition of a
health inequity the claim that inequalities in health which stem from injustice are inequitable, we risk begging the question.
It follows, I take it, that a concern with health equity is purely and simply a concern for
justice in the distribution of health achievement. So a health inequality is a health inequity if
and only if it is an inequality which a just society would seek to counteract. The key
question then is which health inequalities are unjust?
Two dimensions of egalitarian justice

I shall assume that the correct account of justice is in some broad sense egalitarian, and
that a just society should seek to treat all its citizens as equals.9 There is a dispute in
egalitarian justice as to the appropriate scope of obligations of justice: cosmopolitans such
as Pogge (2008) argue that obligations of justice are global, while others such as Rawls
(1999) and Nagel (2005) argue that obligations of socio-economic justice fundamentally
apply only within societies. We shall not enter into this dispute here, but I should
point out that the outcome of this debate will have a major impact on which types of
health inequalities could count as inequities. For if there are no strict obligations of global
socio-economic justice, then health inequalities between nations will not count as health
inequities.
There are two axes along which theories of justice vary. First, there is the what
dimension: what good or goods need to be distributed fairly to each in order to ensure
9
Egalitarianism in this broad sense has come to be central to the way we think about justice. Whereas
in former times the idea that some form of inegalitarianism that for example, took men to be
worth more than women, white more than black citizens, or rich more than poor was expressed and
even defended, the thought that a just society ought to treat all its citizens as equals enjoys
overwhelming support today. In Sens (2004: 22) words, a theory of justice in the contemporary
world would not have any serious plausibility if it did not value equality in some space a space that
would be seen as important in that theory. This is not to say that in the world as we encounter it
women are in fact treated as equals to men; and the poor are in fact treated as equals to the rich, but
rather that where such things occur they are widely recognized to be unjust. Defending this intuition
of fundamental human equality is beyond the scope of this chapter. (I go some way towards
defending it in Wilson [2007a, 2007b]). For the best current overview and defence of the idea of
fundamental human equality, see Waldron (2000).
218
Section 2: Issues
that a society is just? Second, there is the how dimension: how should we distribute those
goods which are the appropriate concern of egalitarian justice? (Hurley, 2007).
Cohen (1989: 906) famously labelled the what question the question of the currency of
egalitarian justice: [w]hat aspects of a persons condition should count in a fundamental
way for egalitarians, and not merely as a cause of or proxy for what we regard as
fundamental? For instance, a crude egalitarian might take financial wealth to be the only
currency of egalitarian justice, while more sophisticated egalitarians argue that the relevant
currency is opportunity for welfare, or access to advantage.10 The key normative question
for us is whether health matters in a fundamental way for egalitarian justice. If it does not,
then talk of health inequities risks being needlessly imprecise and potentially misleading,
given that the relevant inequity will not lie in the maldistribution of health, but rather in the
maldistribution of another, more fundamental good. I take up the question of health and
egalitarian justice in the following section.
The key question posed by the how question is how we should seek to distribute those
goods which are of fundamental concern for egalitarian justice. There are three main
approaches within the broadly egalitarian framework. Strict egalitarians take it that our
goal (insofar as we are concerned with egalitarian justice) should be to equalize the amount
of those goods that are of fundamental importance to justice which each person receives.
Importantly, strict egalitarians believe that it can be legitimate to level down namely to
remove goods which are of fundamental importance from the perspective of justice from
those who are better off, just to make the distribution more equal, even if no individual
persons life is made better by this. Prioritarians take it that it is not an equal distribution of
goods per se which matters, but rather how each individual person is faring relative to how
they might be faring. According to the prioritarian, we should give priority (either absolute
or weighted) to improving the condition of those who are worst off in the distribution of
those goods which are of fundamental importance from the perspective of justice. This
avoids the counterintuitive result of levelling down. Lastly, sufficientarians take it that what
matters from the perspective of justice is that each person have enough of the goods that are
of fundamental importance for justice, and that once this threshold has been reached,
a person has no more claims on justice. As we shall see below, how to apply these answers to
the how question to the domain of health is a complex question.
Is health of fundamental importance for egalitarian justice?

It is implausible to think that health is the only currency of egalitarian justice. For there
seem to be goods which are important to a just society which are neither reducible to fair
distribution of health achievement, nor valued only for their contribution to fair distribution of health achievement. For example, it would seem strange to describe a society which
was rife with racism and discrimination, and prevented women from voting or from
holding political office, but yet where fortuitously everyone had the same level of health
10
Focusing only on financial wealth is a crude view, first because financial wealth is not important in
itself, but rather only important for the goods it gives access to. Second, different people will
have different levels of efficiency in converting money into things that do matter for their own sake.
(For instance, if we gave the same amount of money to all persons, this money would go much less
far for someone who requires expensive equipment to counteract a disability, than it would for
someone who lacked this disability.) Cohen (1989) defends access to advantage, while Arneson
(1989) defends opportunity for welfare, as the appropriate currency of egalitarian justice.
219
achievement as one which was just in an egalitarian sense. Nor is it much more plausible to
claim that health is the most important good that a just society should be aiming at. For this
would seem to turn our commitment to health into a bottomless pit, as there will always be
further interventions we could make that would marginally improve health, which would
have to be bought at the cost of our commitment to goods other than health11 (Dworkin,
2000: 309).
It follows that egalitarian justice must care about more than merely health. There are
two major options. Either there is a single currency of egalitarian justice, and this currency
is not health (and so health is important from the perspective of justice only for the impact
that it has on this currency), or there are multiple (and mutually irreducible) currencies of
egalitarian justice, and health is but one of them.12 In this section, I shall examine three well
worked out versions of these options. First, Dworkins monistic approach, which takes there
to be a single currency of egalitarian justice, which is not health. Second, the pluralistic
capabilities approach of Nussbaum and Sen, according to which the capability to live a
healthy life of a normal length is one of the currencies of justice. And lastly, I will look at
Daniels approach, which aims to argue that while health is not itself a currency of
egalitarian justice, it is so closely related to something that is a currency of egalitarian
justice (namely opportunity), that we ought to treat health as special from the perspective of
justice.
Before getting to this, we must briefly address two issues. First, health is a good which is
much more difficult to redistribute than others we are usually concerned about from the
perspective of justice. Second, there is a question about which health inequalities we should
be focusing on: only those which arise between different socio-economic groups, or also
those which arise between individuals considered separately?
Distributing health
Most of the goods we might be concerned about from the perspective of justice are divisible
and redistributable, and so it is easily possible to remove some of them from those who have
too much, and bestow more on those who have too little. For instance, if we want to combat
an unjust income inequality, we can quite easily redistribute money from the rich to the
poor, by taxing the rich and then giving the resulting money to the poor. Or if we find that
there is an unjust distribution of liberty (as, for example, in a society which allowed men,
but not women to own property), we could change the law so that both men and women
would equally have the liberty to own property.
However, if we were to uncover an unjust distribution of health, it would be rather more
difficult to address it by redistribution, as health is not (in general) directly transferable
11
12
The World Health Organization (WHO) definition of health, namely that health is a state of
complete physical, mental and social well-being and not merely the absence of disease or infirmity
makes it less implausible to think that health so defined could be the most important good for a
just society to focus on. However this is due to the fact the definition of health is simply too
all-encompassing to be useful as a definition of health. (For the debate on the WHO definition of
health, see further Callahan [1973] and Bok [2004]).
There is also a third possibility, which I shall set aside: it might be the case that there are multiple
currencies of egalitarian justice, and that health is not one of them. From the perspective of our
interest in health inequalities, this would be normatively little different from the first option: in both
cases health would be only indirectly relevant to justice.
220
Section 2: Issues
from one person to another.13 So we can only rectify unjust distributions of health
indirectly. One such way would be to ensure a just distribution of the social determinants
of health. While ensuring a just distribution of the social determinants of health would
clearly help to produce a just distribution of health, it would not fully rectify any current
injustice, given that many of the ill health conditions which are caused by an unjust
distribution of the social determinants of health will not be reversible. For example if
someone has atherosclerosis as a result of working for a long time in a stressful environment in which they experienced little control over what they were doing, then allowing
them decent working conditions would not undo their atherosclerosis.14 Another way of
counteracting the effects of unfairly poor health would be by compensating those people
with unfairly poor health with a different and more readily redistributable good (such as
money, free health care or free mobility equipment). However, it is far from clear that ill
health can be fully compensated by being provided with other goods (see Deciding between
these approaches, below).
The upshot seems to be that even if we do think that health is an appropriate currency of
egalitarian justice, it will not be easy, and may in fact be impossible, to bring it about that
there is an equitable distribution of health.
Individual and group inequalities in health

Until fairly recently, researchers who have been interested in health inequities have assumed
that the relevant inequities that we should be worrying about are inequities between groups
rather than individuals, so that they have thought it was a cause for concern if certain
groups (such as African Americans) do worse than other groups (such as white Americans).
However, Murray et al. (1999, 2000) argue that we should also be interested in inequalities
in health between individuals. Unless we do so, they argue, we fail to attend to the
inequalities within these groups, and thus mask part of the inequality present in the
population (Murray et al., 1999: 537).
It seems to me that it is only if we are committed to the view that health is a fundamental
currency of justice and we think that inequalities with natural as opposed to social causes
can be unjust that we ought to be concerned about individual health inequality. (If these
conditions held, it would follow that each individual was owed a fair share of health. It
would then simply be false to say to an individual that we had treated him justly with
respect to health because the social class or race of which he was a member was sufficiently
healthy.)
However, if we do not think that health is a currency of justice, we will find it much
more useful to know for social policy purposes the kinds of questions that social scientific
researchers have tended to concentrate on, namely how inequalities in health correlate with
other variables, such as social class that are of fundamental concern for justice. Knowing
about individual inequalities in health will not tell us very much about what we would need
13
14
Unless, that is, we redistribute healthy organs from one person to another person with less healthy
organs. However this would be such a gross violation of the self-ownership of the people from
whom the organs were taken that it can be safely set aside as a solution to how to distribute health
justly (Segall, 2007: 358).
In addition, as we saw earlier, arguably there are some unjust distributions of health which do not
have a social cause, so these would remain untouched by addressing the social determinants of ill
health.
221
to do to make society more just, given that justice will depend on the distribution of goods
other than health (Hausman et al., 2002; Hausman, 2007). In addition, measurements of
individual health inequalities do not allow us to filter out what is caused by social factors
from what is caused by natural factors, and so they will be a useful measure only if there is
no normative difference between these two types of health inequalities (Asada and
Hedemann, 2002).
Monistic approaches
Any view of justice that takes there to be a single currency of justice will be forced to
conclude that this single fundamental currency is not health. So all monistic approaches to
justice present a standing challenge to the normative significance of health inequalities. We
shall take Dworkins (1981a, 1981b) account as an example of a monistic approach to
egalitarian justice. Dworkin picks out resources as the currency of egalitarian justice. He
uses the concept of resources in a broad way to include not just external resources such as
land and money, but also talents, which are theorized as internal resources. Disability and
ill health are understood as negative internal resources.15
Dworkins view has two main implications for the study of health inequities. First, our
duty would be to ensure a fair share of resources for each person. Health is only one such
resource, and can reasonably be traded against other goods. So the fair distribution of
resources would not require us to take health to be a special case, and to seek to equalize it
separately from other goods.
Second, in Dworkins view (in common with other luck egalitarians) the distinction
between the natural and social is not of normative significance, and so should not be
presupposed in a theory of justice. What matters according to Dworkin is whether someone
can fairly be held responsible for the shortfall in their combined bundle of internal and
external resources. If a person suffers a shortfall for which they cannot be held responsible,
then this requires rectification, regardless of whether the cause of the shortfall is natural or
social. So, for example, on Dworkins view, the natural inequality in life expectancies
between women and men would raise a prima facie case for rectification.16
In addition, Dworkin (along with other luck egalitarians) does not think that justice
requires us to equalize advantages or disadvantages which result from choices that competent adults make. Hence on Dworkins account disadvantages in terms of health or other
resources caused to individuals by their competent choices to drink heavily or to not take
15
16
Dworkin (1981b: 312) denies that ill health and disability give someone a direct claim to
equalization of resources, on the ground that this would amount to a slavery of the talented, and he
argues that justice requires only that each person be given sufficient resources to enable them to
purchase insurance against ill health and disability. However, it is questionable whether Dworkins
argument for this claim is consistent with his broader position, as Cohen-Christofiadis (2004)
argues. For a good overview of the debate on egalitarianism and disability, see Wolff (2008).
There are two reasons why Dworkin would be unlikely to suggest that we take steps to remedy this
gender based health inequality. First, his theory requires us to equalize resources, not health. The
natural inequality in life expectancies favours women, and so if anything will help to counteract the
gender imbalance of resources which is likely to result in a male dominated society. Second, it is
difficult to see how we could systematically favour one sex over another in the distribution of
health care related resources, without violating the core idea of egalitarian justice, namely that each
should be entitled to treatment as an equal (Sen, 2004: 24).
222
Section 2: Issues
exercise do not call for rectification. This would mean that we would need to work out to
what extent (if any) individuals can fairly be held responsible for the types of health
disadvantage which are correlated with, for example low SES.17
Capabilities approaches
Capabilities theorists, of whom Sen (1999, 2004) and Nussbaum (2000) are the pre-eminent
exponents, differ in two fundamental ways from Dworkin. First, they believe that resources
are the wrong distributive space to be working in. They argue that what is of value is people
being able to function in various characteristically human ways, such as using their practical
reason, or playing, or living a healthy life of an ordinary length, rather than the amount of
resources at their disposal. Capabilities theorists argue it would be a mistake for a society to
attempt to provide everyone with the given functioning, as someone may legitimately
choose not to exercise it. (For example, someone may wish to fast for religious reasons,
and at such a time it would be wrong for a society to force them to have the functioning of
being well nourished.) So capabilities theorists argue that we should focus on providing
each person with the capability to function in the valued way, not ensure that they do
actually function in this way.
Second, they argue that we should be pluralists when it comes to justice: there are a
plurality of capabilities which are jointly necessary for a flourishing human life, and we
cannot fully compensate a shortfall in one capability with a superfluity of another. Given
this broad pluralistic framework, the argument for why health is a functioning that matters
fundamentally for egalitarian justice is simple. In Sens (2004: 23) words:
. . . health is among the most important conditions of human life and critically significant constituent
of human capabilities which we have reason to value. Any conception of social justice that accepts
the need for a fair distribution as well as efficient formation of human capabilities cannot ignore the
role of health in human life and the opportunities that persons respectively have to achieve good
health free from escapable illness, avoidable afflictions and premature morality.
Approaches inspired by Rawls

Rawls (1971, 2005) argues that there are two principles of justice which determine the
justice of the basic structure of society. First, and foremost, the liberty principle, which
states that each individual has an equal right to protection by a fully adequate scheme of
basic liberties. Second, a two-part principle governing which social and economic inequalities are acceptable, namely that such inequalities must first be open to all under conditions
of fair equality of opportunity (the opportunity principle), and second the resulting
advantages must be to the advantage of the least advantaged members of society (the
difference principle).
As we have already mentioned, Rawls does not theorize a place for health in his theory
of justice. There are at least three ways that we might account for health within the broad
context of Rawls theory of justice. First, we could do what Peter (2001) does, and argue that
17
Solving the problem of how to determine which actions people can reasonably be held
responsible for is a difficult problem for the luck egalitarian. For an interesting (though
controversial) approach to solving it, see Roemer (1993). For further discussion of personal
responsibility and health inequities, see Wikler (2004).
223
Rawls theory of justice is already adequate as it stands, and that we should understand
health inequities to be those inequalities in health which are the result of unjust social
arrangements.18 Second, we could conceive of health as a primary good, to be distributed
(insofar as this is possible) according to the difference principle.19
Third, we could do what Daniels (1985, 2008; Daniels et al., 2004) does, and account for
health by reference to the opportunity principle. Daniels argues that we should greatly
extend the opportunity principle so that it is not simply a matter of getting a fair
opportunity to compete for jobs and offices (as in Rawls vision), but rather becomes a
matter of guaranteeing to each individual a fair share of the normal range of opportunities
for someone with their talents. Health (argues Daniels) is a necessary condition for
someone being able to access the normal opportunity range for their talents, and so a
commitment to fair equality of opportunity commits us to treating health (and health care)
as special. Treating health and health care as special amounts to focusing on these as goods
to be equalized on their own quite apart from our more general commitments to egalitarian
distribution.
Deciding between these approaches

Daniels approach claims an impressive pedigree; Rawlss theory of justice is by general
consensus the best worked out and most comprehensive theory of justice that we currently
have, and so to provide an account of justice and health which dovetailed neatly with Rawls
would be very desirable.
However, there are two problems which Daniels account struggles to overcome. First, it
is far from clear that the importance of health and health care are correctly explained in
terms of their impact on opportunity (Segall, 2007). Second, Daniels account appears to be
internally inconsistent. Daniels argument for the specialness of health depends on the claim
that health is a condition for the possibility of a normal opportunity range, whereas other
goods (such as wealth) implicitly are not. He then draws the conclusion from this that
whereas other goods can be distributed according to the difference principle, or even
according to the free market, health and health care must be governed by the more stringent
opportunity principle. However, the literature on the social determinants of health
which we reviewed above shows that health is pervasively determined by the distribution
of other goods, such as workplace culture, levels of income equality, the amount of social
capital in a society. Hence it follows that someones share of the normal opportunity range
is significantly determined by the distribution of these other goods. And given that health
18
19
Peter describes this as an indirect account of health inequity. It threatens to share the same
problem we saw with Kawachi et al.s (2002)s definition of a health inequity, namely that it
becomes unclear why we should treat the health inequalities which result from unjust social
arrangements as health inequities. Peter (2001: 164) acknowledges this and adds that an indirect
approach will need some justification for why health is considered a relevant indicator and
such a justification will draw upon the good of health, but does nothing further to explain how we
should relate this conception of the good of health to a Rawlsian approach to justice. It is unclear
how she can ultimately avoid a commitment to either a position analogous to Dworkins (which
says that health matters only for the influence it has on other factors which are directly relevant to
justice), or a position analogous to the capabilities approach (which makes health one of the
currencies of justice).
This seems to be Veatchs (1999) position.
224
Section 2: Issues
does not have the asymmetrical causal role that his argument requires, we cannot argue
from it for the view that health makes special demands of justice because it has an
asymmetrically fundamental causal role in as a condition of opportunity for other goods
(Hurley, 2007: 328).
This, as I see it leaves Daniels with a dilemma. Either he can expand his account of what
he takes to be special, so that all the social determinants of health now count as special too
because of their impact on opportunity (but, this would have results which are unattractive
for egalitarianism in general20) or, he could drop his claim that health is special but this
would be to give up the essence of his position on health and justice.21
It is more difficult to determine whether a monistic account such as Dworkins or a
pluralistic account such as Sen or Nussbaums provides the best way of accounting for the
place of health in a theory of justice. Ideally, we would want a theory of egalitarian justice
that was both fully sensitive to all those feature(s) that make us equals, and all those goods
that need to be distributed fairly if a society is to be a just one (call this the accuracy
requirement), and enabled us to make useful comparative judgements about which of two
situations departs more fully from what justice requires (call this the indexing requirement). The main problem is that the accuracy and the indexing requirements conflict. The
indexing requirement will tend to push us towards a monistic currency of justice. This is
because, if two goods, or two forms of advantage and disadvantage, cannot be compared,
then they cannot be placed on a common scale, and so it will become impossible, in
many cases, to say whether one person is worse off or better off than another (Wolff and
de-Shalit, 2007: 23). And clearly this problem will only get worse as we increase the plurality
of incommensurable goods which we allow as currencies of justice.
However, the accuracy requirement will tend to push us towards acknowledging a
plurality of goods, none of which is fully reducible to the others. Wolff and de-Shalit point
out that if some form of monism about justice were true (for the sake of convenience,
assume that the sole currency of justice was resources in Dworkins sense), it would follow
that any disadvantage relevant to justice could be fully rectified by providing a suitable
amount of this one currency. However, this seems not to be the case, particularly if the
disadvantage to be suffered is an increased risk of early death, or chronic ill health. And so
we seem forced by a concern for accuracy to admit that there are multiple goods which are
of relevance to justice (Wolff and de-Shalit, 2007: 2135).
This, I take it leaves us with a very difficult problem to solve in normative political
philosophy. I would (tentatively) suggest that we should favour accuracy over indexing
here, as there seems little point in starting our accounts of justice from a theory of value we
already know to be inadequate. Although the indexing problem for a plurality of different
goods looks to be very difficult to solve, it is plausible to hope that we may be able to make
20
21
As Segall (2007: 360) explains, the claim that health is special mandates that entitlement to
health care should not be curtailed due to inferior or superior wealth. . . . But while this feature
(working both ways) appears attractive when it comes to medical care, it appears considerably less
attractive with regard to the other social determinants of health. Egalitarians typically do want to
allocate more (social bases of) self-respect to those who have less of other goods (for example
income, looks) and conversely, allocate more income to those who have smaller bundles (compared
to others) of other social (and natural) assets. But treating the social determinants of health as
special prohibits this.
I examine Daniels position in greater depth in Wilson (2009).
225
some progress on it in the future.22 And if we do allow that there is a plurality of goods
which should count as currencies of egalitarian justice, then it seems overwhelmingly
plausible that health should be one, given both its importance for human life in its own
right, and its status as a precondition for many other important functionings.
The how of health equity

Egalitarian justice in the broad sense has its basis in the idea that we should treat persons as
equals. Even leaving aside disputes about which goods should count as currencies of justice,
there are different interpretations of how egalitarians in this broad sense should aim to
distribute those goods that are agreed to be currencies of justice. There are three main
positions developed in the literature: strict egalitarianism, which argues that distributive
equality is an end in itself; prioritarianism, which argues that we should give priority in the
distribution of goods to those who are worst off; and sufficientarianism, which argues that
justice requires only that we ensure that each person has a sufficient quantity of the thing
being distributed.23
Strict egalitarianism
Strict egalitarians believe that the best distribution of those goods that are currencies of
justice is an equal one, and that there is a value to equal distributions in and of itself, even
where no individual is made better off by such distributions. If we cared only about equality
in health, then we would prefer a society in which there was very little inequality in health,
but where life expectancies were lower, to one in which everyone lived longer, but there was
much greater inequality in health. For the purposes of simplicity, let us say that we have a
choice of two societies: in society A, everyone (rich or poor) dies at the age of 60, while in
society B, all the poor die at 65, while all the rich die at 90.24 Strict egalitarians argue that,
insofar as we are egalitarians, we should favour society A to society B, as it has a more equal
distribution of health.
However, while it is certainly true that society A has a more equal distribution of health,
it is far from clear that we should choose it over society B, given that moving from
society B to society A would be levelling down. No one is better off in society A than they
are in society B, as both the poor and the rich have a higher life expectancy in society B. And
in fact society A makes both rich and poor worse off than they would otherwise be. Many
take this to be a strong intuitive complaint against strict egalitarianism.
Strict egalitarians (such as Temkin, 1993) make two moves to attempt to ward off the
levelling down objection. First, they argue that equality is only one of the values which we
should deploy when deciding how goods should be distributed all things considered. So
while society A may be better from the perspective of equality, there may nonetheless be
22
23
24
Wolff and de-Shalit (2007: 89107) propose a practical strategy for solving indexing problems in
real world solutions, which looks like a promising avenue for further research.
In this section I shall assume that health is a currency of justice in the relevant sense. I do not take it
that this has been established by the argument in the previous section. I make this assumption
simply because, unless health is a basic currency of justice, there will be little further to be said about
how we should go about distributing health justly.
For the purposes of simplicity I am making the rather unrealistic assumption that there are no
individual health variations within the groups.
226
Section 2: Issues
compelling reasons to favour society B, all things considered. Second, they argue that the
levelling down objection presupposes a person-affecting requirement, namely that one
situation cannot be better than another unless there is a person for whom it is better;
however this requirement is false.
The strict egalitarians first response may well appear to be dodging the issue: what we
are attempting to work out is the overall principle or principles by which to distribute those
goods which are the currencies of justice. To be told that strict egalitarianism is but one
principle that needs to be weighed against as yet to be specified others, is not perhaps as
helpful as one would have liked.
The problem with the strict egalitarians second reply is that, as Hurley (2007) points
out, it is not true that all objections to levelling down presuppose the person-affecting
requirement. So even if the person-affecting requirement were false, it would not follow that
levelling down in health is acceptable. If a good which is a currency of justice is not only
good for the person who has it, but is also impersonally good, then there would be reasons
to object to levelling down in respect of that good that are separate from the person
affecting requirement. Hurley (2007: 3323) argues that health is just such a good:
Health is a distinctive type of flourishing, with a specific natural character and basis . . . It is not just
good for people to be healthy rather than unhealthy; it is also good in itself for there to be healthy
people rather than unhealthy people.
If this is true, then there is an intuitively obvious objection to levelling down in health which
does not rely on the person affecting principle, namely that leveling down wastefully throws
away the higher reaches of good (2007: 332). And so the objection to levelling down seems
to stand, at least when we are distributing health.
Prioritarianism
Thinking about the problem of levelling down has led many to the view that the core
commitments of egalitarian justice in health should not be strict egalitarianism, but rather
to what has come to be known as prioritarianism, namely that we should give priority to
improving the condition of those who are worst off (Parfit, 1997). Rawls difference
principle is an explicitly prioritarian principle of justice, namely that social and economic
inequalities are to be to the greatest benefit of the least advantaged members of society.
Prioritarianism is very appealing as an account of what we owe to one another when
it comes to health. However, while it may be true in most obvious cases that we should
give priority to alleviating the condition of the worst off, giving an absolute priority to
improving the condition of the worst off can have counterintuitive implications. For it may
well be the case that it is much less cost effective to attempt to improve the condition of
those who are the very worst off, than to attempt to improve the conditions of some other
groups who are slightly better off. (This is often the case in health care, where those who are
worst off frequently require very expensive treatments which only succeed [if at all] in
improving their condition marginally.) And so a focus on giving priority to the worst off
will sometimes conflict with considerations of cost-effectiveness and efficiency.25
25
The underlying worry here is similar to the levelling down objection. But where the levelling down
objection concerns cases where no one gains and there is a significant loss to some, this objection
concerns cases where there is a slight benefit to some, but a much more significant loss to others.
227
Prioritarians can address this challenge from efficiency in one of two ways: either they
can deny that considerations of efficiency have any role to play here, and claim that even if
much more good could be done to those who are less badly off, we should still focus our
efforts on those who are worst off. Or they can shift to a position which has come to be
known as weighted prioritarianism, which claims that priority to the worst off is to be given
a high (but not absolute) weighting, so that in certain circumstances priority to the worst off
can legitimately be overridden by the demands of efficiency.
Sufficientarianism
Sufficientarians (Frankfurt, 1987; Crisp, 2003a, 2003b) argue that priority is only to be
given to those who are worse off than a certain threshold, and that if everyone is above this
threshold, benefits to those who are comparatively better off count equally to those who are
comparatively worse off. The policy implications of a sufficientarian approach to health
would clearly depend on what we took the relevant threshold to be for health achievement.
If the threshold was set high (for instance to 90 years of healthy life), then a sufficientarian
approach would not be distinguishable in practice from a prioritarian approach. However,
if the threshold was set low (for instance, to 50 years of moderately healthy life), then socioeconomic conditions which caused someone to die at 55 or 60 would not deprive someone
of a share of healthy life to which they were entitled by justice.
Crisp (2003a, 2003b) argues that we should set the level of sufficiency through the exercise
of compassion from the perspective of an impartial observer. However, this seems not to be a
very complete answer, given that what we reasonably feel compassion for is closely related to
what we feel is to be expected. And indeed, it is hard to see how there could be a nonarbitrary
way of deciding how long a life was long enough. As Hooker (2008: 190) puts it, As long as a
pleasant, intellectually active, socially interactive life is possible, I cannot see why anyones
needs expire at 70 years, or at 100 years, or at 150 years. So it may be that when we are
thinking about the length of life, there is no relevant sufficiency threshold. It is more plausible
to think that there could be an objective standard for when a life of a given length is sufficiently
healthy. But even here we would face the very difficult task of deciding what that level should
be, a problem the prioritarian (who only needs to determine who is worse off) does not face.26
I take it that these difficulties give us some reason to favour prioritarianism over
sufficientarianism when we are distributing health, even if only on pragmatic grounds.
Conclusion
Social scientists working on health inequalities have often thought that it is simple to
identify health inequities. For example, Dahlgren and Whitehead (2006: 3) forthrightly
state that:
In todays Europe, working out what social differences in health are fair and unfair is unnecessary.
Essentially, all systematic differences in health between different socioeconomic groups within a
country can be considered unfair and, therefore, classed as health inequities. There is no biological
26
As Hooker (2008) reminds us, All prioritarianism has to do is determine who is worse off, and this
seems much easier than determining whether ys getting x is something that y needs or merely
something that would benefit y . . . Crisp needs to defend a line; prioritarianism has no line to
defend.
228
Section 2: Issues
reason for their existence, and it is clear that even systematic differences in lifestyles between
socioeconomic groups are to a large extent shaped by structural factors.
This chapter has argued that things are rather less clear cut than this, and that it is a much
more complex matter to determine which inequalities in health should count as inequities.
Which inequalities in health should count as inequities is, we argued, on analysis a matter of
which inequalities in health are unjust.
We saw that there are two different dimensions to theories of egalitarian justice. First,
there is the question of which goods we should take to be the currencies of justice. Here we
argued that, despite the indexing problems it causes, there is reason to favour a pluralistic
account of the currencies of justice over a monistic account, and that if we adopt a
pluralistic theory of justice it is overwhelmingly plausible to think that health should be
one of the currencies. Second, there is the question of how we ought to distribute those
goods that are currencies of justice. Here we suggested that when we are distributing health,
an approach that gives priority to those who are worst off may be preferable both to a strict
egalitarian one, and a sufficientarian one.
Assuming this is correct, it follows that justice requires us above all to concentrate on
improving the condition of the worst off, and that what makes one person worse off than
another will be, among other things, their health state. Flattening socio-economic gradients
in health will be one important way of improving the condition of those who are worst off,
especially as so doing will also require us to distribute goods other than health more
equitably.
References
Anderson, E. (1999) What is the point of
equality? Ethics, 109: 287337.
Arneson, R. (1989) Equality and equal
opportunity for welfare. Philosophical
Studies, 56: 7793.
Asada, Y. and Hedemann, T. (2002) A problem
with the individual approach in the WHO
health inequality measurement. International
Journal for Equity in Health, 1(2). Available
at: http://www.equityhealthj.com/content/1/
1/2 (accessed: 14/02/11).
Bellis, M., Hennell, T., Lushey, C., Hughes, K.,
Tocque, K. and Ashton, J. (2007) Elvis to
Eminem: quantifying the price of fame
through early mortality of European and
North American rock and pop stars. Journal
of Epidemiology and Community Health,
61: 896901.
health research: one size does not fit all.

Journal of the American Medical Association,
294: 287988.
Callahan, D. (1973) The WHO definition of
health. Hastings Center Studies, 1(3): 7787.
Central Intelligence Agency (2007) The World
Factbook 2007. Available at: https://www.cia.
gov/library/publications/the-world-factbook/
(accessed: 18/07/08).
Chandola, T., Bartley, M., Sackera, A.,
Jenkinson, C. and Marmot, M. (2003) Health
selection in the Whitehall II study, UK. Social
Science & Medicine, 56(10): 205972.
Cohen, G. (1989) On the currency of egalitarian
justice. Ethics, 99(4): 90644.
Cohen-Christofiadis, M. (2004) Talent, slavery
and envy in Dworkins equality of resources.
Utilitas, 16(3): 26787.
Crisp, R. (2003a) Equality, priority, and
compassion. Ethics, 113: 74563.
Bok, S. (2004) Rethinking the WHO definition

of health. Harvard Centre for Population and
Development Studies: Working Paper Series,
14: 7.
Crisp, R. (2003b) Egalitarianism and

compassion. Ethics, 113: 119126.
Braveman, P., Cubbin, C., Egerter, S.,

Chideya, S., Marchi, K., Metzlr, M. and
Posner, S. (2005) Socioeconomic status in
Dahlgren, G. and Whitehead, M. (1992) Policies

and Strategies to Promote Equity in Health.
Copenhagen: WHO Regional Office for
Europe. (Document number: EUR/ICP/RPD

414 (2). Available at: http://whqlibdoc.who.
int/euro/-1993/EUR_ICP_ RPD414(2).pdf
(accessed: 18/07/08).
Dahlgren, G. and Whitehead, M. (2006)
Levelling Up (Part 1): A Discussion Paper on
Concepts and Principles for Tackling Social
Inequities in Health. Copenhagen: WHO
Regional Office for Europe. Available at:
http://www.who.int/social_determinants/
resources/leveling_up_part1.pdf (accessed:
18/07/08).
Daniels, N. (1985) Just Health Care. Cambridge:
Daniels, N. (2008) Just Health. Cambridge:
Health and inequality, or why justice is good
for our health. In Public Health, Ethics and
Equity, ed. S. Anand, F. Peter and A. Sen.
Oxford: Oxford University Press, pp. 6392.
Dworkin, R. (1981a) What is equality? Part 1:
equality of welfare. Philosophy and Public
Affairs, 10: 185246. (Reprinted in Dworkin
2000.)
Dworkin, R. (1981b) What is equality? Part 2:
equality of resources. Philosophy and Public
Affairs, 10: 283345. (Reprinted in Dworkin
2000.)
Dworkin, R. (2000) Sovereign Virtue: The Theory
and Practice of Equality. Cambridge,
Frankfurt, H. (1987) Equality as a moral ideal.
Ethics, 98: 2142.
Hausman, D. M. (2007) Whats wrong with
health inequalities? Journal of Political
Philosophy, 15, 1: 4666.
Hausman, D. M., Asada Y. and Hedemann T.
(2002) Health inequalities and why they
matter. Health Care Analysis, 10, 2: 17791.
229
Kawachi, I., Subramanian, S. and AmmeidaFilho, N. (2002) A glossary for health

inequalities. Journal of Epidemiology and
Community Health, 56: 64752.
Marmot, M. (2004) Status Syndrome: How Your
Social Standing Directly Affects Your Health
and Life Expectancy. London: Bloomsbury.
Marmot, M., Rose, G., Shipley M. and
Hamilton, P. (1978) Employment grade and
coronary heart disease in British civil
servants. Journal of Epidemiology and
Community Health, 32: 2449.
Mill, J. (1874) On Nature. (Originally published
in Mills collection Nature, The Utility of
Religion and Theism. Now most conveniently
available at: http://www.lancs.ac.uk/users/
philosophy/texts/mill_on.htm [accessed:
18/07/08].)
Mueller, C. and Parcel, T. (1981) Measures of
socioeconomic status: alternatives and
recommendations. Child Development,
52: 1330.
Murray, C., Gakidou, E. and Frenk, J. (1999)
Health inequalities and social group
differences: what should we measure?
Bulletin of the World Health Organization,
77(7): 53743.
Murray, C., Gakidou, E. and Frenk, J. (2000)
Response to P. Braveman et al. Bulletin of the
World Health Organization, 78(2): 2345.
Nagel, T. (2005) The problem of global justice.
Peter, F. (2001) Health equity and social
justice. Journal of Applied Philosophy,
18(2): 15970.
Nussbaum, M. (2000) Women and Human
Development: The Capabilities Approach.
New York: Cambridge University Press.
Parfit, D. (1997) Equality and priority. Ratio,
10: 20221.
Hooker, B. (2008) Fairness, needs, desert. In The

Legacy of H. L. A. Hart, ed. M. Kramer and
C. Grant. Oxford: Oxford University Press.
Pogge, T. (2008) World Poverty and Human

Rights, 2nd edn. Cambridge: Polity Press.
Hurley, S. (2007) The what and the how of

distributive justice and health. In
Egalitarianism: New Essays on the Nature and
Value of Equality, ed. N. Holtung and
K. Lippert-Rasmussen. Oxford: Oxford
Rawls, J. (1999) The Law of Peoples. Cambridge,

Rawls, J. (1971) A Theory of Justice. Cambridge,

Rawls, J. (2005) Political Liberalism, expanded

edn. New York: Columbia University Press.
230
Section 2: Issues
Roemer, J. (1993) A pragmatic theory of

responsibility for the egalitarian planner.
Scheffler, S. (2003) What is egalitarianism?
Segall, S. (2007) Is health care (still) special? The
Journal of Political Philosophy, 15(3): 34261.
Sen, A. (1999) Development as Freedom. Oxford:
Sen, A. (2004) Why health equity? In Public
Health, Ethics and Equity, ed. S. Anand,
F. Peter and A. Sen. Oxford: Oxford
Shavers, V. (2007) Measurement of
socioeconomic status in health disparities
research. Journal of the National Medical
Association, 99(9): 101323.
Strand, B., Kunst, A., Huisman, M., et al. (2007)
The reversed social gradient: higher breast
cancer mortality in the higher educated
compared to lower educated. A comparison of
11 European populations during the 1990s.
European Journal of Cancer, 43(7): 12007.
Temkin, L. (1993) Inequality. Oxford: Oxford
University Press.
University of St Andrews, May 2000.

Available at: http://weblaw.usc.edu/
academics/assets/docs/waldron.pdf (accessed:
18/07/08).
Whitehead, M. (1990) The Concepts and
Principles of Equity and Health. Document
number: EUR/ICP/RPD 414. Copenhagen:
WHO Regional Office for Europe.
Available at: http://whqlibdoc.who.int/
euro/-1993/EUR_ICP_RPD_414.pdf
(accessed: 18/07/08).
Wikler, D. (2004) Social and personal
responsibility for health. In Public Health,
Ethics and Equity, ed. S. Anand, F. Peter and
A. Sen. Oxford: Oxford University Press,
pp. 10934.
Wilkinson, R. (1996) Unhealthy Societies: The
Afflictions of Inequality. London: Routledge.
Wilson, J. (2007a) Nietzsche and equality. In
Nietzsche and Ethics, ed. G. von Tevenar.
Oxford: Peter Lang, pp. 22140.
Wilson, J. (2007b) Transhumanism and moral
equality. Bioethics, 21(8): 41925.
Wilson, J. (2009) Not so special after all? Daniels
and the social determinants of health. Journal
of Medical Ethics, 35: 36.
Veatch, R. M. (1999) Justice, the basic social

contract and health care. In Contemporary
Issues in Bioethics, 5th edn, ed. T. Beauchamp
and L. Walters. Belmont, CA: Wadsworth,
pp. 36874.
Wolff, J. (2008) Disability among equals. In

Philosophy and Disability, ed. K. Brownlee
and A. Cureton. Oxford: Oxford University
Press.
Waldron, J. (2000) Basic equality. Delivered as

the Sir Malcolm Knox Lecture at the
Wolff, J. and de-Shalit, A. (2007) Disadvantage.

Index
abortion, 535, 57
Abortion Act (1967),
53, 62
disability and abortion,
5365
LIFE (anti abortion charity),
62
adjudication between patients
differing degrees of desert,
90
age groups and birth cohorts,
2001
aggregative dimension of
population health, 21
AIDS, 100
alcohol consumption, 4
allocation of scarce
healthcare resources,
89, 90
altering the behaviour of
smokers, 92
alternative hypothesis, 79
American College of
Obstetricians and
Gynecologists, Committee
on Genetics, 131
analysis of personal health
information without
consent, 2
and value, 713
Angastiniotis, M., 122
anti-scientific world view of
risk, 82
approaches to harm
prevention, 11213
Aristotle, 161
Arneson, Richard., 1867,
216
as an attempt to improve
the human gene pool,
58
Ashcroft, Richard, 85, 136
association between obesity
and socioeconomic
status, 5
authoritarian reduction of
obesity, 60
autonomous choices of the
individual, 88
Belmont Principles, 174,

17781, 184, 187
Belmont Report, 177, 178, 179,
182, 185, 188
benefit to society, 147
benefiting the individual, 21
benefits and burdens
calculations (of harm), 4
Bentham, Jeremy, 1578, 159
Big Pharma, 205
bioethics, 8, 18, 51, 71, 183,
1912
bioethics agenda, broadened,
191, 197, 2089
biotic egalitarianism, 167
British Medical Association
(BMA), 121
Brown, Gordon, 86, 121
Brlde, Bengt, 20, 23, 2930,
345, 38, 39
BSE crisis, 76
Buehler, James, 176
burden of regulation in
proportion to the harm
prevented, 91
ByrdHagel Resolution, 164
causal routes of harms, 112
Centers for Disease Control
and Prevention (US), 147
Central Intelligence Agency
(US), 211
chicken pox, 113
Civil Rights Act (1964, US),
168
classes of outcomes, 75
classical liberal approach to the
regulation of personal
behaviour, 91
climate change
Bali Principles of Climate
Justice, 164
climate change and public
health, 154
climate change debate,
fairness within, 1623
climate justice, 1634
contractrarian framework
for, 161
Earth Summit, Bali (2002),

164
egalitarian framework for,
162
impacts of, 163
international distribution of
burdens, 163
targets relating to, 163
utilitarian perspective,
15761
Clinton, William Bill, 168
collective action (liberalism), 12
Commission on the Social
Determinants of Health
(WHO), 206
common good (disease
prevention), 115
communitarians, 89
compromise of functional
ability, 93
compulsory health
interventions, 107
compulsory interventions, 115
compulsory medical treatment,
103
concept of, 3
confidentiality, 108
consequentialist outcomes,
723
consequentialists
justification of the constraint
of freedom, 112
constructivist accounts of risk,
68
contact tracing, sexually
transmitted diseases, 104
contagious nature of many
vaccine-preventable
childhood diseases, 145
contemporary risk-analysis,
73
correlations between alcohol
and disease, 4
Council for International
Organizations of Medical
Sciences (CIOMS), 176,
177
creation of equal opportunities
for health, 20
231
232
Index
critical level healthcare and the

poverty line, 44
CSIS Commission on Global
Aging, 198
Darwin, Charles, 166
de Zulueta, P., 127
Declaration of Helsinki, 1767
Declaration of Helsinki (2008),
177
defence of government health
policy, 70
definitions of screening', 119
deontological distinctions
between types of risks, 76
deontological intuition, 74
Department for Environmen
Food and Rural Affairs
(DEFRA, UK), 159
Department of Health, 54, 131,
155
Descartes, Ren, 165
desert concept of, 90
diminishing marginal value of
health, 43
disability-adjusted life
expectancy, 26
Disability Discrimination Act
(2004), 62
disability-free life expectancies,
26
Disability Rights Commission,
623
disapproving of the conduct of
smokers, 89
disease, 20
disease and disability
reduced welfare, 62
disease burden, 2, 24
disease control
historic examples, 101
disease prevention
justification of constraints on
civil liberties, 114
the least restrictive
alternative, 114
disease, risk factors, 2
disorders
disorders, definition of, 29
disorders and public health, 33
disorders as internally caused
undesirable conditions,
2930
disorders, definition of, 29
dissemination of precautionary
information, 108
distribution of health-risks in
the population, 67
distributions of health
inequality, 42
distributive justice in
healthcare, 90
DNA sequencing technology,
136
dominance of liberal
assumptions, 11
Downs Syndrome and
abortion, 54
EcoEquity, 164
Economic and Social Research
Council (ESRC), 175
education or recruitment
policies, 129
egalitarian framework for
climate change, 162
egalitarian view of health
inequality, 41
egalitarianism
aboandonment of health
inequalities, 22
Engels, Friedrich, 157
environment, ethics and public
health, 171
environmental ethics, 71
environmental health, 154
environmental justice and
public health, 164
environmental justice and
development ethics,
16870
environmental damage,
human-caused, 74
environmental damage,
natural, 74
environmental ethics and
public health, 168
Environmental Health
Commission, 154
environmental policy, 71
Environmental Protection
Agency, 169
epidemics and pandemics,
reduced risk of, 147
epidemiological testing, 78
epidemiology, ethics of, 83
epistemic rules for policymaking, 80
equity and population health,
209
age groups and birth cohorts,
2001
allocation of healthcare
resources to different age
groups, 2001
burden of disease, 205
cost-effectiveness analysis
(CEA), 193, 201, 206
determinants of health, 192
distributive healthcare
rationing issues, 1934
equity in health, 1923
global migration, 208
global terrain of health
justice, 206
health egalitarians and health
maximizers, 1923
health inequalities
between different social
groups, 191
health inequalities, reduction
of, 1934
international equity and
health, 2013
international health
inequalities, 191, 206
international property rights
and access to drugs, 2056
Mexican Seguro Popular
(health insurance scheme),
198
migration, push and pull,
2045
Millennium Development
Goals (MDG), 193, 204
societal aging as an emerging
public health problem,
198200
working age population,
shrinkage of, 199
eugenics, 5864
abortion and eugenics, 58
antenatal screening labelled
as eugenic, 125
authoritarian eugenics,
5960
composition of future
populations, 64
compulsory sterilisation
programmes, 59
equal value principle, 624
genetic improvement as a
flawed concept, 61
government influence within
reproduction and family
planning, 59
Human Fertilisation and
Embryology Act (1990), 62
Index
human gene pool,

improvement of, 612
laissez-faire eugenics, 59
moral standing of, 601
Nazi eugenic policies, 48, 59,
64
permissible and wrongful
eugenics, distinction
between, 61
preimplantation genetic
diagnosis, 49, 50, 51,
53, 55, 578, 59, 61,
623, 64
prenatal genetic diagnosis, 60
pro-life (abortion/eugenics),
61
to reduce the prevalence of
disease and impairment,
62
eugenics as a strategy for
improving public health,
48
eugenics, immoral features of,
48
European Commission, 71
European Society of Human
Genetics, 122, 126, 132,
135
evolutionary malfunction, 32
expected costs and benefits
(risks precautions),
70
Faculty of Public Health
Medicine (UK), 155
fair innings argument, 26
false negatives, 80
false positives, 80
Federal Actions to Address
Environmental Justice in
Minority Populations and
Low-Income Populations
(US), 168
freedom and individual
autonomy, 108
Frost, Robert, 89
Galton, Francis, 58
General Medical Council
(GMC), 128
Genetic Information
Nondiscrimination Act
(US, 2008), 131
Gillon, Raanon, 61
Global Commons Institute
(GCI), 164
goals of public health, 2045

aggregative considerations,
73
assignment of value to
different life years, 25
Atkinson inequality measure,
42
biological disadvantage, 31
biomedical definitions, 34
central risk factors, 22
connection between disorder
and harm, concept, 30
creation of equal
opportunities for health,
20
critical levels of healthcare,
445
Cummins-function, 32
disparities between life
expectancy and increases
in health, 26
distributions of health
inequality, 42
distributive dimension of
population health, 21
egalitarianism and the
priority view, 43
fair innings argument, 26
good life, definitions of, 38
health, functional
definitions, 345
health, pluralistic definitions,
356
healthy life expectancies,
261, 401, 43, 45
improvement of average
health, 21
improvement of social
conditions, 22
improvement of the health
of the most unhealthy,
43
inequality adjusted
happiness (IAH), 43
inequality reduction, 434
life expectancy, 267
machine faults, 313
measurement of
improvements in health,
27
measurement of personal
health, 34
minimally acceptable health
and life expectancy, 44
person trade-off, 378,
45
233
preference-based health
value index, 36
prevention of disease and
injury, 29
prioritization of the worseoff, 43
priority views of healthcare,
43
prolongment of lives,
246
promotion of physical and
mental health, 33
quality adjusted life years,
28
reduction of inequity, 434
Rosser classification, 36
social health promotion, 22
standard gamble, 36, 38
time trade-off, 36, 38
well-being and public health,
3940
goals of public health,
conception of, 23
good moral reasons for
regulating the behaviour
of individuals (general),
89
good or bad outcomes, 80
government regulation of
hazards to health, 88
Great Plague of London (1665),
1012
harm, 4
justification for
interventions, 113
magnitude of harm, 113
harm and the causation of
harm, judgements
concerning, 82
causal responsibility nor
moral responsibility, 111
expanded (HP2), 113
harm principle, common good,
108
harm principle, expanded
(HP2), 11112, 113
harm, broader notion of, 4
harm, risk of, 69, 144, 145, 149,
150, 151, 152
harms
weighed against the benefits
of prevention, 4
harms of infection, 109
harms to self, the undeserving
patient \r, 89
234
Index
health disorders as machinefaults, 31

health equity between age
groups and birth cohorts,
198201
health inequities, 228
approaches to health
inequalities, 222
avoidable health inequalities,
215
deciding between approaches
to health inequalities, 2235
decreased life expectancies
and high-risk behaviours,
214
distributing health, 21920
egalitarian justice, 217,
21819, 220
health differentials
attributable to biological
variation, 215
infant mortality rates,
international comparisons,
211
life expectancies, differences
in, 211
monastic approaches to
health inequality, 2212
obligations of socioeconomic
justice, 217
prioritarianism, 2267
social determinants of
health, just distribution of,
220
social science to political
philosophy, 214
socioeconomic gradients in
health, 213
socioeconomic status and ill
health, 212
strict egalitarianism, 2256
sufficientarianism, 227
health inequality, complexity
of, 1967
health screening and testing,
112
healthcare conflicts between
individual and
population, 6
healthcare interventions, 4
health-related utility values, 38
Hepatitis B screening, 103
herd, 151
HIV/AIDS, 100, 109, 186
HIV, lower rates of, 16
Hobbes, Thomas, 165
Hockney, David, 857, 889,

93, 96
Hourdequin, Marion, 73
House of Commons Science
and Technology
Committee (2006), 75
human-centred environmental
ethics, 166
Human Fertilisation and
Embryology Act (1990),
62
Human Fertilization and
Embryology Authority
(HFEA), 55, 57, 63
Hume, David, 157
individual and group
inequalities in health,
2201
infectious disease control, 116
antibiotics, 100
compulsory vaccinations,
108
contact-tracing and further
testing, 1045
correlation between social
deprivation and increased
prevalence of infectious
diseases, 112
decreased morbidity and
mortality due to infectious
diseases control, 100
directly observed therapy
(DOT), adherence to
treatment, 107
effectiveness in preventing
infection, 114
Groningen case
(Netherlands, 2007), 109
harm principle, infectious
diseases, 108
Hepatitis B screening, 103
HIV screening, 103, 124, 127
individual responsibility for
infectious disease
protection, 113
infectious disease control
and public trust, 11516
isolation of patients, 102
justified compulsion in
public health, 113
leprosy, 1012
liberty, restriction of, 102
patient confidentiality,
2, 103
quarantine, 102, 1056, 107
reduction of infection risks,

114
risk to the community, 107
screening of travellers, 104
sexually transmitted diseases,
104, 113
smallpox, 100
social distancing to prevent
infectious disease, 101
Spanish Flu (1918), 100
surveillance, screening, and
notification, 1034
tuberculosis control, 106
typhoid, 100
World Health Organization
preparation for an
influenza pandemic, 100
influence of drinking cultures
on individual behaviour, 4
influences on the health of
other communities, 156
Institute of Medicine, 149, 155,
174, 198
Intergovernmental Panel on
Climate Change (IPCC), 163
International Center for
Longevity, 199
International Committee on
Climate Change (IPPC),
160
International Development
Ethics Association
(IDEA), 16970
International Health
Regulations, 144
International Labour
Organization (ILO), 208
International Monetary Fund
(IMF), 204, 207
International Organization for
Migration, 207
interventions
generating anxiety and other
harms, 4
individual focused, 4
involuntarily incurred harms to
third parties, 88
Jardins, Des, 167
Jepson, Joanna, 54
John, Stephen, 67
judgment pluralism, 9
justification for compulsory
interventions, 113
justification of screening
programmes, 119
Index
Kamm, Frances, 193, 201

Kantian injunction, 184
Kyoto Protocol, United States
rejection of, 164
Labour Government (UK),
204
Leopold, Aldo, 167
less severe infectious diseases,
114
levels of health quality, 26
liberal approach
autonomous decisionmaking, 7
choice and responsibility, 8
health care, equality of
access, 12
intervention ladder, 12
liberal neutrality and
tolerance, 9
liberal framework of public
health, 2
moderated liberal positions,
1012
narrow liberal views, 610
non-interference, 1, 67
relationship between law and
ethics, 9
liberty as a value, 108, 112, 145,
151
liberty-limiting actions and
policies, 109
life expectancy, 267
lifestyle choices
outside the control of the
individual, 4
limitations on the idea of
harm, 4
limits of ethical health
promotion, 7
liver cirrhosis in France, 4
Locke, John, 157, 165
machine-fault model, 34
MacIntyre, Alasdair, 158
MacQueen, Kathleen, 176
magnitude estimation, 36
malaria eradicatuion (risk), 67
malevolent behaviour, 109
Mallon, Mary, (Typhoid Mary),
1023, 109
mandatory screening of adults,
129
maximization of expected
utility, 69
McMillan, John, 174
measurement of personal
health, 34
measures combining health
quality and quantity, 28
Mechanic, David, 196
medical ethics, 1
beneficence and
non-maleficence, 6
biomedical ethics, four
principles of, 178
clinical medicine (focus of
contemporary medical
ethics), 3
contemporary medical
ethics, 2, 3, 4, 6, 8
criticisms drawn from public
health, 12
expansion of traditional
frameworks, 8
medical ethics, framework
of, 2
traditional medical ethics, 6
unethical behaviour, 2
Medicare (US), 200
mental disorders, 31, 33
mental or somatic disorders, 29
Midgley Mary, 1656
Mill, John Stuart, 67, 12, 109,
111, 129, 144, 1578, 159,
216
Miller, Franklin, 1823
mitigation and adaptation
policies, 161
mitochondrial disorder
screening, 123
mobile phone masts, 79
mobile phone masts, potential
risk of harm, 77
mobile phones, potential health
threats from location of
masts, 7780
models of scientific inquiry
(risk proc), 81
moral judgement of past or
present smoking, 90
moral standing of, 601
morality of, 58
Naess, Arne, 167
National Commission for the
Protection of Human
Subjects of Biomedical
and Behavioural Research,
177
National Health Service, 52, 53,
56, 64, 901, 132
235
National Screening Committee

(UK), 119, 120, 122, 125,
127
Neugarten, Bernice, 199
newborn screening, 1234
Newson, Ainsley J., 118
nicotine dependency, 87
non-anthropocentric
environmental ethics,
166
non-ascertained risk of harm,
77
non-exploitation as a
principle of research
ethics, 188
non-exploitation in research
ethics, 1824
non-principled evaluation
healthcare distributions,
45
non-standardconsequentialism, 73, 76
normal harms expected sum
total, 80
notion of implicit harm, 6
Nuffield Council on Bioethics,
10, 12, 120, 122, 130,
1312
null hypothesis, 789
obesity, 16
objectivist pluralism, 39
Organisation for Economic
Co-operation and
Development (OECD), 204
pandemic influenza, 8
parents, abortions and
eugenics, 58
paternalism, 5
paternalism, as a positive
approach, 5
paternalistic intervention, 88
paternalistic regulation, 88
paternalistic regulation of
smoking, 88
pathological and nonpathological disorders, 29
patient welfare as a
fundemental principle of
clinical medicine, 180
population screening, 137
access to screening
programmes, 134
antenatal screening, 122, 123,
127, 129
236
Index
population screening (cont.)

availability and affordability
of relevant treatment, 135
to benefit populations, as
opposed individuals, 119
benefits of diagnosis
counterbalanced against
potential harms, 133
collectivist population
screening, 126
conflicts with ethical
principles, 118
criteria for establishing
screening programmes,
1201
denial of life or other
insurance, 131
as distinct from clinical
testing, 119
exclusion from the
workplace, 131
genetic screening, 121, 123
harm potential from false
positive or false negative
results, 133
HIV screening, 103, 124, 127
informed consent and
informed choice, 1268
litigation following false
negative results for breast
and cervical cancer
screening, 133
mandatory and routine
screening, 1289
potential for stigma and
discrimination, 1301
psychological harm,
potential for, 133
the right to remain ignorant
of potential disease, 129
screening in adults, 124
tuberculosis screening, 103
vulnrable groups and the
provision of screening,
135
plague, 100
pluralism, 10
judgement pluralism, 8
pluralism in ethics, 9
population-orientated
approach, 12
Porter, Dorothy, 155, 158
practicality constraints, 77
precaution in practice, 7681
precautionary principle, 6875,
767, 78, 80
precautionary reasoning, 73
Presidents Commission
(1983), 121
prevention and reduction of
harm, 4, 6, 17
prevention of disease through
the termination of
pregnancies, 129
prevention of diseases through
the termination of
pregnancies, 125
prevention of obesity, 5
preventive measuresagainst
infective diseases, 103
privacy and confidentiality,
104
private sector involvement, 135
prolongment of lives, 246
promotion of health as a goal of
public health, 36
promotion of physical and
mental health, 33
property right protections, 205
provokation of complex social,
moral and legal reactions,
85
public goods, 53
public health, 4, 1557
aims of, 3
appropriate ethical theory, 2
individual healthcare, 1
individuals and populations, 3
notion of, 3
primary focus on
populations, 1
theoretical perspective of, 5
uncertainty and precautionary
motivation, 4
public health actions, unethical
activity, 18
public health activities, 20, 29
public health activity, 3, 4
public health crises such as
DDT or thalidomide, 72
public health emergencies, 2,
115
public health ethics,
development of, 1
public health ethics,
relationship to medical
and environmental ethics,
71
public health ethics, substantive
account of, 12
public health interventions
impact of, 103
public health outcomes, 81

public health policies and
activities
ethically dubious, 1
public health policy, 71
public health practice and
policy
concept of harm, 6
public health practice, discord
with moral theory, 2
public health research, without
individual consent, 2
public health, adverse effects of
controlling, 115
public health, ethics of, 83
public health, individuals and
populations, 3
public nature of public health,
12
public understanding of
the probabilities of risk,
75
public health research ethics,
188
AZT trials, 180, 186, 1878
Baltimore lead abatement
study, 181, 186, 1878
beneficence and nonmaleficence, principles
applicable to clinical
research, 183
benefits and risks of research,
179
clinical trials, 175
ethical scrutiny of research,
176
exploitation and unethical
public health research,
1812
exploitative reasearch,
features of, 184
Grimes vs Kennedy Krieger
Institute Inc. (2001), 181,
182
harmful exploitation,
1845
health research and
interventions, distinction
between, 176
immoral research practices,
183
moral rules of research
ethics, 177
mutually advantageous
consensual exploitation,
1858
Index
nonconsensual exploitation,
185
non-maleficence under the
principle of beneficence,
178
Nuremberg Code, 176, 180
patients welfare, promotion
of, 180
public health information
collection, distinction
between research and
audit, 176
qualitative and quantitative
methodologies, 175
reasearch codes and
principles, 1767
research ethics committees,
176
research ethics, overhauling
the principles of, 183
Tuskegee syphilis study, 182,
183, 184
vulnerable groups, research
of, 181
Wertheimer and
exploitation, 184
public health, resetting the
parameters, 18
analysis of personal health
information without
consent, 2
causal and moral
responsibility, 8
clinical medicine (differences
from public health), 3
clinical medicine and public
health ethics, 16
clinical relationships
between doctor and
patient, 8
collective interventions, 3, 5
common goods, 16
common interests, 16
communitarianism, 12
congruent interests, 15
convergent interests, 15
core public health
activities, 1
government decision making
regarding lifestyle choices
of individuals, 5
harm principle, 7, 1089,
113
health outcomes impossible
to achieve by the
individual alone, 5
healthcare conflicts between

individual and population, 6
interventions in health, 3
known population risks
applied to the lives of
individuals, 4
nature of public health
condition, 2
population and
communities, populationlevel interventions, 2
prioritizing populations over
individuals, 2
promotion of healthier
lifestyles, 2
public health emergencies, 2,
115
reduction of risk, 1, 82
relationship between
population health and
individual health, 4
restrictions on liberty, 7
social inequalities as a
population effect, 17
social justice, conceptions of,
1718
theoretical concerns of
public health, 8
unethical healthcare policy, 2
quality adjusted life years, 28
quality of life, 389
quality-adjusted life years, 26
racial influence on health
inequalities, 195
rating scale, 36
Rawls, John, 701, 156, 162,
163, 164, 179, 188, 195,
201, 2023, 213, 217,
2223, 226
rcba, 84
reduce health inequalities,
434
reduce mortality (av health), 21
reduction of morbidity (average
health), 21
reduction of socioeconomic
inequalities, 22
regulation in the interests of
health or pleasure, 86
regulation of private sectorprovided screening, 136
regulation of risk, 74
regulation of the food
industry, 5
237
relationship between economic

wealth or income and life
expectancy, 97
relationship between selective
reproduction and public
health, 48
relativistic pluralism, 9
relevant individual outcomes,
234
relevant value of additional life
years, 24
Repository for Germinal
Choice, 49
restrictions on the liberties of
individuals and
populations, 107
retrospective clinical
judgement, 90
rights to privacy, 108
risk and precaution, 83
balancing risks, 82
consequentialism in riskassessment, 75
creation of risk-free
environments, 67
environmental degradation
and public health, 71
ethics of risk, 689, 81, 83
health-and-safety regulation,
70, 74
identification and
management of risk, 68,
69, 81, 83
laissez-faire risk
management, 77
precautionary approaches to
public health policy, 68, 75
precautionary health-andsafety, 77
probabilities of suffering
harm, 69
public health policy and risk,
67
public perception of risk, 76
reduction of risk, 1, 82
risk, tolerability, 813
scientific and constructivist
views of risk, 756
value chains for comparison
of risk, 70
risk-based health policy, 75
risk-cost-benefit-analysis,
697, 757, 812
risks to individuals or
populations, 113
Rose, Geoffery, 4, 7
238
Index
Rosser classification, 36
Russell, Bertrand, 1578
Royal College of Physicians, 94,
97, 177
SARS, 8, 100, 103, 124, 129
scientific misunderstanding of
risk, 75
scientific standard proof of risk,
80
scientific statistical testing, 78
scope for prevention (harms), 4
screening and the right to
remain in ignorance,
12930
screening for infectious diseases
(compulsory), 116
selective reproduction and
public health, 58
abortion, 535, 57
benefit and harm in selective
production, 52
contraception and sexual
abstinence, 49
disability rights, 53
discouragement of
smoking during
pregnancy, 50
embryo-implantation, 48
ethics of selective
reproduction clinical
practice, 63
gametic essentialism, 51
in vitro fertilisation (IVF), 57
non-health purposes of
sexual reproduction, 52
origin view, 51
policy and regulatory issues
of reproduction and
health, 55
potential harm from selective
reproduction, 52
preimplantation genetic
diagnosis, 49, 50, 51, 53,
55, 578, 59, 61, 623, 64
public health interventions
in pregnancy, 57
reproductive and genetic
technologies, 48
selection to avoid disease, 48
sex selection, 52
size of the future population,
49
teenage pregnancy, 49
welfare gains from selective
reproduction, 52
self-inflicted harms, 87
sensitivity and specificity of,
121
separation of public health and
the environment, 1656
serious handicaps and abortion,
54
smoking
health inequalities, 97
smoking and risk of lung
cancer, 78
smoking and teenagers, 92
smoking as a vice and virtue,
867
smoking as a voluntary, selfinflicted harm to health,
88
smoking as public good, 96
smoking cessation,
consequences of, 87
smoking during pregnancy, 95
smoking in enclosed public
places, harms and public
goods, 93
smoking in public and the
workplace, 95
smoking in the home, 93, 95
smoking, health and ethics, 98
as a public health problem,
85
attempts to control smoking,
85
banning smoking in enclosed
public places, 89
clinical approach to
smoking-related illnesses,
89
compulsory smokingcessation, 95
consumption of and
contributions to healthcare
resources, 91
costs of smoking-related
illness, 90
criminalization of smoking,
95
ethics and smoking, 957
marketing of tobacco
products, 85
non-treatment of smokers,
91
passive smoking, impact of,
91
reduction of smoking in
poorer socioeconomic
groups, 97
regulation of smoking, 86,

88, 89, 914, 96
relationship between income
and smoking, 97
smoking, health information,
23
smoking, moral uncertainty
and moralism, 889
smoking, regulation of, 91
smoking, the morality of,
856
social determinants and health
inequalities, 1946
social health disparities at
national and international
levels
access to medicines, 8
social injustices, 41
social justice, 196
social justice and climate
change, 1612
societal level policy initiatives, 2
socioeconomically worse-off
groups, 112
Socrates, 154
standard consequentialist
outcomes, 73
standard statistical testing, 80,
81
Stephens, Carolyn., 154, 168,
169
stewardship model, 10, 12
Stewart report (2000), 71, 77
stigma, sickle cell anaemia
screening (US), 130
taxation on tobacco products,
90, 91
taxes on tobacco products, 89
technocratic accounts of risk,
68
termination time limits, 53
the precautionary principle,
812
third dimension of population
health, 22
to reduce health inequalities,
20
tolerability of risk, 82
traditional liberal approaches to
public health, 1
traditional medical ethics, 6
traditional medical model
(machine-fault model)
disorders, 29
treatment of obesity, 5
Index
tuberculoisis marginalized in
society (vulnerable
groups), 107
tuberculosis, 8, 107, 114, 129
tuberculosis, contact screening,
104
tuberculosis, detention and
forced treatment of, 114
unequal healthcare distributed,
41
unequal socioeconomic
determinants, influence on
health, 3
United Nations (UN), 71, 199
UNICEF, 203
unintentional transmision of
infectious diseases, 110
unjust inequalities, 44
vaccination, 22
vacination ethics, compulory
vaccination, 1502
value pluralism, 9
vaccination ethics, 153
anti-vaccination
information, quality of,
149
best interests argument,
1457
childhood vaccinations, 144,
151
compulsory vaccinations, 108

contagious nature of many
vaccine-preventable
childhood diseases, 145
diphtheria infection
reduction after the
introduction of
vaccination, 148
endemic childhood diseases
in the developing world,
152
failure or refusal to vaccinate,
145
herd protection, 1, 15, 22,
147, 14950
legal compulsion as distinct
from moral obligation, 151
lost public confidence in
vaccination, 148
MMR (measles, mumps and
rubella), 148
moral arguments about
vaccination, 147
negative or positive
consequences, 146
non-vaccination of
individuals, 147
obligation to protect others,
144
parental decision-making
about childhood
vaccinations, 146
239
paternalistic arguments, 145

potential benefit to the
individual, 147
potential harm to third
parties, 145
reasons for and types of
vaccination, 143
reduction in childhood
mortality from infectious
disease, 148
tetanus vaccination, 15
vaccination as a public good,
149
Verweij, Marcel, 3, 4, 6, 12,
216, 556, 57, 100, 115,
148, 174
viruses of the mind (mental
disorders), 33
Wertheimer, Alan, 184, 185,
1868
WHO Commission on the
Social Determinants of
Health, 194
Wittgenstein, Ludwig, 154
World Bank, 204, 207
World Health Organization
(WHO), 35, 100, 148,
192, 198, 202, 204,
206, 208
World Trade Organization
(WTO), 208

Public Health Ethics Key - Concepts and Issues in Policy and Practice

Uploaded by

Copyright:

Available Formats

Public Health Ethics Key - Concepts and Issues in Policy and Practice

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Public Health Ethics Key - Concepts and Issues in Policy and Practice

Uploaded by

Copyright:

Available Formats

Public Health Ethics

Key Concepts and Issues in Policy and Practice

Public Health Ethics

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town,

Resetting the parameters: public

Environment, ethics and public

Health, disease and the goal of

Selective reproduction, eugenics

10 Public health research ethics: is nonexploitation the new principle for

Risk and precaution

Smoking, health and ethics

Infectious disease control

11 Equity and population health:

Resetting the parameters

insufficient individuals contribute, then this important protection cannot be achieved.

A necessary condition for any adequate account

Chapter 1: Resetting the parameters

improve a populations health through population-level interventions. Third, population

Chapter 1: Resetting the parameters

The inadequacy of liberal medical ethics as a means of thinking

Narrow liberal views

Chapter 1: Resetting the parameters

In addition, a common inference from the liberal non-interference approach to framing

Chapter 1: Resetting the parameters

Moderated liberal positions

Chapter 1: Resetting the parameters

In my comments I will concentrate on general remarks about this kind of approach,

Resetting the parameters: in defence of a substantive

Chapter 1: Resetting the parameters

Substantive notions of the good

to a substantive notion of the good such as human flourishing or a particular notion of

Collective values: interests, goods and harms

Chapter 1: Resetting the parameters

Figure 1.1 Congruent interests. The ellipse

Figure 1.2 Convergent interests. The ellipse

Figure 1.3 Private interests. The circles represent the

Chapter 1: Resetting the parameters

Figure 1.4 Common interests. Our individual

Conceptions of social justice

and public health. Journal of Medical Ethics,

Chapter 1: Resetting the parameters

relationship. In Public Health Ethics &

Martin, R. (2009) The role of the law in public

Dawson, A. and Garrard, E. (2006) In defence of

Mill, J. S. (1859) On Liberty. (Many editions).

Dawson, A. and Verweij, M. (2008) The steward

Nuffield Council on Bioethics. (2007) Public

Feinberg, J. (1984) Harm to Others. Oxford:

Munthe, C. (2008) The goals of public health: an

Nussbaum, M. C. (2000) Women and Human

Gostin, L. O. (2005) Public health

Nys, T. (2008) Paternalism in public health care.

Gostin, L. O. and Stone, L. (2007) The health of

Parfit, D. (1984) Reasons and Persons. Oxford:

Griffin, J. (1986) Well-Being: Its Meaning,

ONeill, O. (2002) Public health or clinical

Paul, C. (2009) The common good argument

Jennings, B. (2007a) Public health and civic

Raz, J. (1986) The Morality of Freedom. Oxford: