The Journal of Socio-Economics 34 (2005) 135150

Behavioral economics research and the

foundations of economics

Vernon L. Smith

University of Anchorage Alaska, USA

Abstract
Five propositions on which economists and psychologists including behavioral economists are in
agreement are presented, leading to a discussion about two kinds of rationality. After some comments
on methodology and on concepts of fairness, I will discuss the question of wealth maximization versus
the economics of survival, and their different implications for behavior.
I want to begin with ve propositions on which economists and psychologists including behavioral
economists are in agreement.
This will lead me to talk about two kinds of rationality.
After some comments on methodology and on concepts of fairness, I will close on the question of
wealth maximization versus the economics of survival, and their different implications for behavior.
2004 Published by Elsevier Inc.
Keywords: Behavioral economics; Economist; Rationality
1. Propositions
Curiously, the image of economists and psychologists as protagonists obscures their
underlying agreement on foundations. Both rely upon the same underlying interpretation
of economic rationality as follows:

This paper is derived from Vernon Smiths Herbert Simon Memorial Lecture, presented at the Society of
Behavioral Economics/International Association for Research in Economic Psychology joint meeting at Drexel
University, Philadelphia, June 16, 2004.

Present address: Interdisciplinary Center for Economics Science, George Mason University, 4400 University
Drive, MSN 1B2 Fairfax, VA 22030, USA. Tel.: +1 703 993 4850; fax: +1 703 993 4851.
E-mail address: [email protected].
1053-5357/$ see front matter 2004 Published by Elsevier Inc.
doi:10.1016/j.socec.2004.09.003
136 V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150
(1) To the extent that markets are rational, as in controlled S &Dexperiments or irrational,
as in experimental asset bubble markets, this derives directly from the rationality or
irrationality of agents.
Thus, even a . . . monopolist . . . has to have a full general equilibrium model of the
economy (Arrow, 1987, p. 207). The default explanation for market rationality is assumed
to derive entirely from individual rationality. Markets cannot be rational if agents are not
fully rational in the particular sense in which theorists have modeled it, and this proposition
has been uncritically accepted by psychologists and behavioral economists.
(2) Individual rationality is a self-aware, calculating process of maximization.
Here is a particularly clear statement of decision as deliberate intentional action: In-
centives do not operate by magic: they work by focusing attention and by prolonged de-
liberation (Tversky and Kahenman, 1987, p. 90). But this is an inference based on how
the theorist conceptualizes and models decision problems, not the subject. The inference
ignores the demonstrated capacity of motivated subjects to nd equilibrium outcomes by
repeat interaction in market experiments without cognitive awareness of this capacity, and
their capability of achieving better outcomes in two-person anonymous trust interactions
than if they applied traditional game-theoretic principles.
(3) Predominantly both economists and psychologists are reluctant to allow that nave
and unsophisticated agents can achieve socially optimal ends without a comprehensive
understanding of the whole, as well as their individual parts, implemented by deliberate
action.
People are not supposed to achieve desirable outcomes, and not be able to tell you about
it. There is no agic; no room for the Gode and Sundar zero intelligence traders who in
simple environments reveal that at least some of the rationality in markets is encapsulated
in the institution; no room for the results of hundreds of single and multiple commodity
market supply and demand experiments in which subjects with private information on values
and costs achieve unintended equilibrium predicted outcomes. The phenomenon, not being
constructively explicable with the tools of standard theory, is in effect denied: if subjects
achieve rational outcomes it is because in effect they think about decision problems as we
economists do.
(4) Consequently, psychologists test the rationality of individual decisions largely by asking
for subject responses to choice problems to discover how they reason.
Rather than challenge this interpretation, economists, subject to the identical
visionhow do agents consciously think about economic choice?are critical of the
questionresponse survey methods used in cognitive psychology: the stakes are zero or
too low, and it is claimed without citing the experimental evidence on stakes, that the sub-
jects are too unsophisticated, inexperienced or untrained to allow a serious researcher to
nd out how real agents really think. Many psychologists appear to nd irrationality ev-
erywhere, and many economists appear to see the ndings as everywhere irrelevant. To
both, how agents think indeed exhausts the core of investigation.
V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150 137
(5) Finally, psychologists and economic theorists have identied rationality almost entirely
as expected utility (including expected prot, or wealth) maximization.
But is this the concept of rationality that best explains either behavior or what agents seek
as optimal for themselves? In my closing remarks I will discuss a particularly appealing
alternative: the theory of economic survival developed by Roy Radner and his coworkers.
1.1. Opinion surveys
In point of fact opinion surveys can provide important insights. Survey ndings can be
tested more rigorously with reward motivated choices and some have predictive content,
e.g., the choice asymmetry between losses and gains in wealth. But sometimes what people
actually do completely contradicts what they say, and sometimes you cannot nd out by
asking because the agents themselves do not know what they will do or are doing. Here are
some examples:
(1) Comparisons of risk preferences under low and high monetary stakes have shown that
actual reward levels do have a statistically signicant effect on decision, but that the
qualitative conclusions from hypothetical choice response surveys are not refuted by
studies using very high stakesthe accumulated payoffs average three times subjects
normal monthly living expenses (Kachelmeier and Shehata, 1992; also see Binswanger,
1980 for similar ndings). These ndings have been cited defensively with the argu-
ment that payoffs do not change the conclusions from hypothetical choice studies.
Yes, and the scholars in question do not claim otherwise, but until their work, we did
not know; now we do; it could have gone the other way for all anyone could say; in
fact observed outcomes are effected, so indeed payoffs do make a difference, but not
enough to rescue expected utility theory from the qualitative results in previous contra-
dictions. In other contexts, payoff levels have sometimes, and sometimes not, created
important shifts in observations (Holt and Laury, 2002). The correct conclusion is that
you better do some testing using incentive rewards, because it can matter. The wrong
conclusion is that the results from unmotivated choice surveys generalize to motivated
choice.
(2) Consider the double auction in classroom demonstration experiments: in debriengs
afterwards students deny that there is any kind of quantitative model that could predict
their market price and exchange volume, or that they were able to maximize their
prots; but a participant with a sealed envelope containing the predictions provided in
advance, opens it to reveal that this consensus is false. The dispersed private value/cost
information is aggregated into prices that are at the equilibrium and each agent is
maximizing his or her prot given the behavior of all others. Here there is indeed a
kind of magic, but only, I think, in not being well-understood or modeled at the game-
theoretic level of individual choice. In this sense our bounded rationality as economic
theorists is far more constraining on economic science, than the bounded rationality of
privately informed agents is constraining on their ability to maximize the gains from
exchange in markets.
(3) In asset trading, participant survey responses reect the disparity between their infor-
mation on fundamental value and their puzzling experience of a price bubble and crash
138 V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150
generated on the long path to the rational expectations equilibrium in a stochastically
stationary environment (Schwartz and Ang, 1989).
(4) Opinion polls administered to the Iowa Electronic Market traders show the same judg-
ment biases that psychologists and political scientists nd in public opinion polls, but
these biases do not interfere where the market is able to predict the popular vote out-
comes (Forsythe et al, 1992).
(5) In preference reversal survey experiments subjects report many inconsistent choices:
gamble A is preferred to B but a subject will sell A for less than B. Arbitraging the
subjects cash motivated choices quickly reduces these inconsistencies (Chu and Chu,
1990, p. 906), and it has been shown that the inconsistencies are unbiased random
errors under some, but not all, conditions (Cox and Grether, 1996; also see Soper
and Gigiolotti, 1993, where choice intransitivity is studied directly and the errors are
found to be random). But reducing inconsistency through experimental treatments,
while of interest, does not mean that the subjects will generalize that learning to other
contexts.
In these examples it is in private information environments, where the market is aggre-
gating information far beyond the reach of what each individual knows, understands and is
able to comprehend, that the solicited opinions are so far off the mark. The surveys yield
no useful understanding because the subjects have none to relate. In the complete informa-
tion asset market, subjects are aware of its fundamental value structure, and come to have
common expectations through an experiential process of repetition; i.e. initial common in-
formation is not sufcient to induce common expectations. They trade myopically and their
expressed bafement (prices rise without cause) reects this myopia. These observations
suggest that much insight might be obtained from the systematic comparison of the condi-
tions under which survey results are robustly informative and the conditions under which
they are not.
2. Two kinds of rationality
I think the agreement between economists and psychologists on the constitution of ratio-
nality in the above propositions is wrong. My views have been shaped by the discoveries of
experimental economists in both impersonal market exchange and personal social exchange.
I want to argue that the key to understanding our data is to be found in recognizing two
distinct conceptions and sources of rational behavior: constructivist and ecological. This
distinction grew out of my experience in the laboratory in which time and again subjects
produced results that were not part of their conscious intentions.
2.1. Constructivist rationality
The rst concept of a rational order derives from the standard socioeconomic science
models (SSSM) going back to the seventeenth century. The SSSM is an example of what
Hayek has called constructivism, which stems particularly from Descartes and others,
who believed and argued that all worthwhile social institutions were and should be created
V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150 139
by conscious deductive processes of human reason. In the 19th century Bentham and John
Stuart Mill were among the leading constructivists.
As Hayek put it, the constructivist view of rationality . . . gives us a sense of unlimited
power to realize our wishes . . . (and) . . . holds that human institutions will serve human
purposes only if they have been deliberately designed for these purposes . . . (moreover, an
institutions existence) . . . is evidence of its having been created for a purpose, and always
that we should so re-design society and its institutions that all our actions will be guided
by known purposes . . . This view is rooted . . . in a . . . propensity of primitive thought to
interpret all regularity . . . as the result of the design of a thinking mind . . . (Hayek, 1973,
pp. 89).
In economics the SSSM leads to rational predictive models of decision that motivate
research hypotheses that experimentalists have been testing in the laboratory since mid-
twentieth century. Although the test results tend to be conrming in impersonal market
exchange, the results are famously and recalcitrantly mixed in personal exchange, no-
tably in a great variety of two person extensive form games where some half of the peo-
ple attempt and frequently succeed when risking cooperation, even when anonymously
paired.
Behavioral economics has made a cottage industry of showing that the SSSM assump-
tions seemto apply almost nowhere to decision making. This is because the methodology of
behavioral economists unabashedly champions a deliberate search for identifying the ways
in which behavior differs from the standard model . . . (Mullainathan and Thaler, 2001).
This search must necessarily succeed for it includes search in the tails of distributions, and
this inappropriate methodology explains why there is so much variation in the degree of
similarity in the ndings of experimental and behavioral economists.
Thus, there is agreement between EE and BE that in certain standard contexts the SSSM
fails to predict the high levels of cooperation in ultimatum and dictator games. But in
exploring and trying to understand such behavior EEhas discovered conditions under which
cooperation in these games are greatly reduced or become inconsequential. For example,
ultimatum game proposals vary from 44% to 28% depending upon whether the task is
formulated as a market exchange, and whether a pre game exercise is used to earn the rst
mover position.
Similarly, in dictator games the percentage of the pie given by dictators falls from 23%
to 10.5% when conducted double blind, and further to only 3% when dictators must rst
earn their stakes of $40 (essentially, the $40 becomes their own money before it can be
givena condition believed by some behavioral economists to be important). Alternatively
the percentage given rises from 23% to 52% as in the two stage investment trust game, with
gains from exchangethe amounts sent are tripled and any amount can be returned. We
now understand that giving can be very self-interested as in the SSSM if social distance is
great and subjects have to earn their stakes perhaps creating an endowment effect (Smith,
2004, Tables 4 and 5).
The results from some of these games have motivated constructivist extensions of game
theory based on other-regarding, in addition to own-regarding, preferences (e.g. Bolton,
1991; Rabin, 1993), but these models perform poorly in competition with exchange models
when applied to wider classes of games and are inconsistent with the above very large
variation in context, stakes and anonymity.
140 V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150
Moreover there now exists a large data set from trust games in which subjects achieve
cooperative outcomes in single play, and make more money than if they applied game-
theoretic principles, contrary to both the predictions of game theory and the behavioral
economics search for irrational behavior.
An alternative explanation of some of the contradictions to theory is that people may use
ancient social-grown norms of reciprocity
i
(including equity, meaning to each according to
his justly earned desert; i.e., equality of opportunity, not outcome) to achieve cooperative
states superior to individually rational defection outcomes. Technically, the issue can be
posed as one of asking how most productively to model agent types by extending game
theory so that types are an integral part of its predictive content, rather than merely imported
as an ex post technical explanation of experimental results. For example, moves can signal
types, and effect decision, which explains why game-form matters, and why opportunity
costpayoffs available, but foregonecan effect outcomes. These considerations must be
part of the internal structure of the theory such that outcomes become predictions conditional
on the elementary characteristics of players who read each others intentions. If such a
programwere successful, many of the basic results in game theory would become irrelevant
or special cases of the extended theory.
In market experimentswhere cooperation can occur through the coordination function
of prices produced by, but simultaneously resulting from, interaction with individual choice
behaviorthe results are more commonly in accord with standard competitive models
that maximize group welfare. This cannot be claimed as a professional economic victory,
however, because of the complete failure of the SSSM to predict the surprisingly weak
conditions under which the results obtain.
Thus, for tractability, Cartesian rationalism provisionally assumes or requires agents
to possess complete payoff and other informationfar more than could ever be given
to one mind. In economic analysis the resulting exercises are believed to sharpen eco-
nomic thinking, as if-then parables. Yet, these assumptions about the economic envi-
ronment are unlikely to approximate the level of ignorance that has conditioned either
individual behavior, or our evolved institutions as abstract norms or rules indepen-
dent of particular parameters, which have survived as part of the world of experience.
The temptation is to ignore this reality because it is poorly understood, and does not
yield to our familiar but inadequate modeling tools, and to proceed in the implicit be-
lief that our parables capture what is most essential about what we observe. Having
sharpened our understanding on Cartesian complete information parables we carry these
tools into the world for application without all the necessary caveats that reect the
tractability constraints imposed by our bounded professional cognitive capacities as theo-
rists.
In summary, constructivism uses reason to deliberately create rules of action, and design
human socioeconomic institutions that yield outcomes deemed preferable, given particular
circumstances, to those produced by alternative arrangements. Although constructivism is
one of the crowning achievements of the human intellect, it is important to remain sensitive
to the fact that human institutions and most decision making is not guided only or even
primarily by constructivism. Emergent arrangements, even if initially constructivist in form
must have survival properties that take account of opportunity costs and environmental
challenges invisible to our modeling efforts.
V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150 141
2.2. Limitations and distractions of constructivist rationality
Since our theories and thought processes about social systems involve the conscious
and deliberate use of reason, it is necessary to constantly remind ourselves that human
activity is diffused and dominated by unconscious, autonomic, neuropsychological systems
that enable people to function effectively without always calling upon the brains scarcest
resourceattention and reasoning circuitry. This is an important economizing property of
howthe brain works. If it were otherwise, no one could get through the day under the burden
of the self-conscious monitoring and planning of every trivial action in detail. Thus, If we
stopped doing everything for which we do not know the reason, or for which we cannot
provide a justication . . . we would probably soon be dead Hayek (1988, p. 68). No one
can express in thoughts, let alone words, all that he or she knows, and does not know but
might call upon, or need to discover, for some purposive action. Imagine the strain on the
brains resources if at the supermarket a shopper were required to explicitly evaluate his
preferences for every combination of the tens of thousands of grocery items that are feasible
for a given budget.
Such mental processes are enormously opportunity-costly and implicitly our brain
knows, if our conscious, planning, modeling mind does not know, that we must avoid
incurring opportunity costs that are not worth the benet. The challenge of any unfamiliar
action or problemappears rst to trigger a search by the brain to bring to the conscious mind
what one knows that is related to the decision context. Context triggers autobiographic expe-
riential memory, which explains why context surfaces as a nontrivial treatment, particularly
in small group experiments.
We fail utterly to possess natural mechanisms for reminding ourselves of the brains
ofine activities and accomplishments. This important proposition has led Gazzaniga (1998)
to ask why the brain fools the mind into believing it is in control: By the time we think
we know somethingit is part of our conscious experiencethe brain has already done
its work. It is old news to the brain, but fresh to us. Systems built into the brain do their
work automatically and largely outside of our conscious awareness. The brain nishes the
work half a second before the information it processes reaches our consciousness . . . We
are clueless about how all this works and gets effected. We dont plan or articulate these
actions. We simply observe the output . . . The brain begins to cover for this done deal
aspect of its functioning by creating in us the illusion that the events we are experiencing are
happening in real timenot before our conscious experience of deciding to do something
(Gazzaniga, 1998, pp. 6364).
As stated by Benjamin Libet, whose pioneering contributions to the study of brain func-
tion in conscious mental activity dates from the 1960s, . . . experimental evidence does
show that certain stimuli in the sensory pathway, even when inadequate to produce any
conscious experience, can nevertheless be usefully detected by the human subject. The
important inference is, then, that neural activities inadequate to produce any subjective
awareness can nevertheless help to mediate functions without awareness. Indeed much of
our brain activities are of that nature (Libet, 2004, p. 28).
And to Hayek, who had a thorough grasp of these propositions, but without the advantage
afforded by recent neuroscience understanding, what was the fatal conceit? The idea that
the ability to acquire skills stems from reason. The constructivist mind makes a fatal
142 V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150
error, blinding itself to understanding, as we are warned, one should never suppose that
our reason is in the higher critical position and that only those moral rules are valid that
reason endorses (Hayek, 1988, p. 21).
That the brain is capable of off-line subconscious learning is shown by experiments with
amnesiacs who are taught a new task. They learn to perform well, but memory of having
learned the task escapes them (Knowlton et al., 1996).
Also, there are numerous anecdotal reports of people seeking a solution to a problem,
who abandon it unsolved, but awaken in the morning with a solution. Such reports involve
famous composers, poets and many scientists (Mazzarello, 2000). For example, Mendeleyev
reported that the critical rule underlying his periodic table of the elements came in a dream
after unsuccessful efforts to gain insight while awake. Poincar e (1913) reports having given
up on solving a particularly difcult mathematical problem only to have the solution sud-
denly appear to him on a trip to Lyon.
It is known from many experimental studies that neuronal patterns associated with task
stimulus-responses when awake are reactivated subsequently during sleep (Maquet, 2001).
Its like the brain is practicing while asleep. Sleep appears to have a consolidating effect on
memory, but may also restructure previous representations and yield insight. Thus, Wagner
et al. (2004) provide evidence that sleep has a facilitating role in subjects discovery of a
hidden abstract rule that abruptly improved their response speed in a mental task using two
simple explicit rules. More than twice as many subjects experienced this insight following
sleep as after wakefulness independent of time of day.
Perhaps almost all of our individual mental activities and accomplishments are inaccessi-
ble to our conscious awareness. Similarly we are not aware of a great range of socioeconomic
phenomena, such as the productivity of social exchange systems and the external order of
markets that underlie the creation of social and economic wealth.
2.3. Ecological rationality
These considerations lead to the second concept of a rational social order: an ecological
system, designed by no one mind, which emerges out of cultural and biological evolutionary
processes: home grown principles of action, norms, traditions, and morality.
Constructivism, however, can apply reason by rational reconstruction to examine the
behavior of individuals based on their experience and folk knowledge, who are nave
in their ability to apply constructivist tools to the decisions they make; to understand the
emergent order in human cultures; to discover the possible intelligence embodied in the
rules, norms and institutions of our cultural and biological heritage that are created from
human interactions but not by deliberate human design. People follow rules without being
able to articulate them, but they may nevertheless be discoverable.
This is the intellectual heritage of the Scottish philosophers and Hayek, who described
and interpreted the social and economic order they observed, and its ability to achieve
desirable outcomes. The experimental laboratory provides a tool for testing hypotheses
derived from models of emergent order.
David Hume was famously concerned with the limits of reason, the bounds on human
understanding, and with scaling back the exaggerated claims of Cartesian constructivism.
To Hume, rationality was phenomena that reason discovers in emergent institutions. Thus,
V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150 143
the rules of morality . . . are not conclusions of (our) reason (Hume, 1739/1985, II, p.
235).
Adam Smith developed the idea of emergent order for economics. Truth is discovered in
the form of the intelligence embodied in rules and traditions that have formed, inscrutably,
out of the ancient history of human social interactions. This is the antithesis of the anthro-
pocentric belief that if an observed social mechanism is functional, somebody, somewhere,
somehow in the unrecorded past surely must have used reason consciously to create it to
serve its perceived intended purposes. This is the default folk belief in the historic origins
of any legacy that is functional.
But in cultural and biological evolution, order arises from mechanisms for generating
variation to which is applied mechanisms for selection. Reason is good at providing vari-
ation, but poor at selection. Constructivism is a powerful engine for generating variation,
but it is far too limited and inexible in its ability to comprehend and apply all the relevant
facts in order to serve the process of selection, which is better left to ecological processes.
In experimental economics the eighteenth century Scottish tradition is revealed in the
observation of emergent order in numerous studies of existing market institutions such as
the continuous double auction (CDA). To paraphrase Adam Smith, people in these exper-
iments are led to promote group welfare enhancing social ends that are not part of their
intention. This principle is supported by hundreds of experiments whose environments
and institutions (sealed bid, posted offer and others besides CDA) may exceed the ca-
pacity of formal game-theoretic analysis to articulate predictive models. But they do not
exceed the functional capacity of collectives of incompletely informed human decision
makers, whose autonomic mental algorithms coordinate behavior through the rules of the
institutionsocial algorithmsto generate high levels of measured performance.
Both kinds of rationality have inuenced the design and interpretation of experiments
in economics. Thus, if people in certain contexts make choices that contradict our formal
theory of rationality, rather than conclude that they are irrational, some ask why, reexamine
maintained hypotheses including all aspects of the experimentsprocedures, payoffs, con-
text, instructions, etc.and inquire as to what new concepts and experimental designs can
help us to better understand the behavior. What is the subjects perception of the problem
that he/she is trying to solve?
Finally, understanding decision requires knowledge beyond the traditional bounds of
economics, a challenge to which Hume and Smith were not strangers. Thus, for Hayek, an
economist who is only an economist cannot be a good economist. This is manifest in the
recent studies of the neural correlates of strategic interaction (Kevin McCabe calls it neuro-
economics) using fMRI and other brain imaging technologies. This research explores the
neuro-correlates of intentions or mind reading, and other hypotheses about information,
choice, and own versus other payoffs in determining interactive behavior.
2.4. How are the two concepts of a rational order related?
Constructivism takes as given the social structures generated by emergent institutions
that we observe in the world, and proceeds to model them formally. Thus, constructivist
models need not ask why or how an auction institution arose; or what were the ecological
conditions that created it; or why there are so many distinct auction institutions. In some
144 V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150
cases it is the other way around. Thus, revenue equivalence theorems in auction theory show
that the standard auctions generate identical expected outcomes, which, if taken literally,
leave no modeled economic reason for choosing between them. But society chooses between
them in particular applications. It is important to ask how and why, and to avoid dismissing
such learning as irrational because of revenue equivalence.
More generally, using rational theory, one represents an observed socioeconomic situa-
tion with an abstract interactive game tree. Contrarily, the ecological concept of rationality
asks certain prior questions: from whence came the structure captured by the tree? Why
this social practice, from which we can abstract a particular game, and not another? Were
there other practices and associated game trees that lacked survival properties and were
successfully invaded by what we observe? There is a sense in which ecological systems,
whether cultural or biological, must necessarily be, if not already, in the process of becom-
ing rational: they serve the tness needs of those who unintentionally created them through
their interactions.
To understand what isthe surviving tip of the can-do knowledge icebergrequires
understanding of a great deal more that is not. This is because of the rich variety of alter-
natives that society may have tried, but that have failed. Nor is there any assurance that
arrangements t for one economic and social environment may be t for another. In the
laboratory we can not only rationally reconstruct counterfactuals, as in economic history,
but also use experiments to test and examine their properties.
3. Behavioral methodology
As I have indicated researchers in psychology and behavioral economics almost uni-
formly report results contrary to accepted rational theory (Hogarth and Reder, 1987). It
was not always so, but the focus on what are called anomalies, beginning in the 1970s,
converted an emerging discovery enterprise into a search for contradictions between reports
of behavior and standard decision theory. This has been documented by Lopes:
Prior to 1970 or so, most researchers in judgment and decision-making believed that
people are pretty good decisionmakers . . . Since then, however, opinion has taken
a decided turn for the worse, though the decline was not in any sense demanded by
experimental results. Subjects did not suddenly become any less adept at experimental
tasks nor did experimentalists begin to grade their performance against a tougher
standard. Instead, researchers began selectively to emphasize some results at the
expense of others . . . The viewthat people are irrational is real in the sense that people
hold it to be true. But the reality is mostly in the rhetoric (Lopes, 1991, pp. 66, 80).
Well before the 1970s, work in decision theory under uncertainty had been challenged
by some economists as a theory that was unable to account for certain observations, such
as individual purchases of lotteries and insurance, requiring explanation. They sought to
redene the domain of the utility function to get a better representation (Markowitz, 1952;
Friedman and Savage, 1948). Also, externalities in choice had long been recognized as
potentially important qualications of standard theory. But it is the neoclassical hypothesis
of a particular denition of self-interested agents that led to some of the most productive
results and therefore was a prominent and easy target of criticism.
V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150 145
Kahneman clearly does not see people as irrational except in the narrow con-
structivist context used in economic modeling based on dominant choice and
arising from the abstract mental concept of the context-free one shot isolated
decision or game. In fact he describes his empirical ndings contradicting ratio-
nality, in this particular representation of the SSSM, as having been easy, thanks
to the implausibility to any psychologist of this rational model.
(See the interview of Kahneman and Smith on the Nobel web site.) Unfortunately,
the popular pressby focusing on the newsworthy rhetoric of behavioral economic
studieshas often interpreted the contributions of Kahneman as proving that people
are irrational, in the popular sense of stupid, so elementary and transparent are
their errors, in the popular sense of mistakes. The pejorative meaning attached
to words that dene theoretical concepts and hypotheses has been all too prominent
in professional interactions, and this allows easy public access to the muck raking
fray in public discussions. Thus, errors in the sense of deviations from the SSSM
predictions, are referred to in the psychology and behavioral science literature as
cognitive errors, meaning mistakes as deviations from what should be observed.
This description implicitly accepts the undoubted and un-doubtable truth value of the
SSSMas a representation of optimality, and therefore those subjects are indeed making
mistakes transparently contrary to their own rational best interest. In science, however,
observed deviations from theoretical predictions normally mean that either the theory
is wrongi.e. needs modication as a predictive theory, which may mean altering
the conception of optimalityor, as in tests of decision theory, human decision making
is awed. Since, as Kahneman has noted, if human decision making is as awed as
indicated by these representations it is hard to understand our species success over the
last 2 million years of survival and occupation of the planet. Kahneman here raises
a natural question: to what kind of optimal decision making process have human
beings become adapted?
But looking beneath the rhetoric, many psychologists to their credit, have maintained
an intensive program examining the contradictions between observed behavior and the pre-
dictions of classical models of choice, bargaining and competition, and have moved the
argument from an account of theoretical possibilities and anomalies to deeper empirical
investigations. For example Siegel (1959) and Fouraker and Siegel (1963) reported both
conrmations and contradictions, and used the pattern to propose improved models. These
were among the important contributions dening work prior to the 1970s and described by
Lopes above, but which were ignored in the rush to produce anomalies and to revise earlier
work in the discredited behaviorist tradition. Similarly, in prospect theory Kahneman and
Tversky (1979) have proposed modications in both the utility and probability weighting
functions of standard expected utility theory, and thus revised the specication of opti-
mality in expected utility theory. Research strategies that focus on the study of errorsas
mistakes, rather than deviations from predictionscan distort professional beliefs, to
say nothing of popular representations, if the rhetorical emphasis is on the failures, to the
146 V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150
exclusion of the predictive successes, of the theory. New theory needs to be able to embrace
the old theory where it is accurate, improve its performance where it is inaccurate, and have
new predictive consequences.
As I see it experimental market economics and behavioral economics are in princi-
ple complementary. Experimental economists study market performance given individual
valuations, while cognitive psychologists study the performance consistency of individual
decision making. If the objects traded are prospects the appropriate valuations are their
cash values, whether based on expected utility, prospect theory (Kahneman and Tver-
sky, 1979), or some other representation. Thus, Plott and Agha (1983) study experimental
markets in which speculators have the capacity to buy in a market with certainty of de-
mand and supply and resell in a second market with demand uncertaintyin effect the
resale values were simple two-outcome gambles in the second market. They report conver-
gence in the second market to a CE dened by demand and supply based on the expected
values of the gambles (also see Plott and Turocy, 1996). But the connective interface be-
tween rationality at the individual and at the market level and how institutions modulate
the interface is yet to be fully explored. It appears that markets may yield equilibrium out-
comes given induced cash values, whatever are the cash valuesrational, irrational, or non-
rationalthat are provided by individuals. If those cash values are not individually stable
or consistent this may just mean that market equilibrium predicted outcomes are subject to
uncertainty.
But there is a more serious problem with the narrow focus on the alleged rationality
or irrationality of individual choice. Its not a failure to see the forest for the trees; its a
failure to see either the forest or the trees for the leaves. Individual choice either in the
sense of commodities and services based on preference ordering, or under uncertainty is
not where the action is in understanding economic performance. The fundamental theorem
in economics is that wealth is created by task specialization across individuals, groups, pop-
ulations, regions and climates; and specialization is determined by the depth and breadth of
the marketpersonal or impersonal exchange systems. What are important about individual
choices are the decisions that cause people across time and generations to change tasks, loca-
tions, and directions in an effort to better themselves in response to market prices. This does
not require individuals to use their gains from specialization and trade for self-interested
economic ends in the sense of always choosing dominant outcomes.
To focus on the rationality of choice in the narrow frames dened to give it precision
does not provide guidance toward a better understanding of this specialization process.
People can make a lot of cognitive errors on the way to creating a new marketsay in
off-the-shelf software programs that allowing new rms to specialize and create much new
wealth. All the work of economics is in specialization and the exchange systems that make
possible the wealth they create. That work is beyond the eld of vision of the individual,
but it should not be beyond the vision of the scientist.
4. Fairness
What is fairness? The descriptor fairness has so many meanings in different contexts
that I believe it is best to avoid the term entirely in experimental science except where
V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150 147
it is explicitly modeled and the model tested in environments where subjects make deci-
sions on the basis of the dening parameters of the proffered fairness model; then the
descriptor fair and its ambiguity can be avoided altogether. This is the way it was used
in the utilitarian denitions by Franciosi et al. (1995), then by Fehr and Schmidt (1999),
and Bolton and Ockenfels (2000). Of course it is appropriate to use the descriptor if the
purpose is to see how its instructional use might have an emotive affect on behavior, but
this is not what is done in Kahneman, Knetch and Thaler (KKT) (1986). The emotive con-
tent of fairness is clear in the important work of Zajac (1995), who has also examined
the rhetoric of fairness arguments as self-interest serving in the Florida, 2000, election
controversy where each side charged the other with gross violation of fairness (Zajac,
2002).
KKTstate . . . the phrase it is fair, is simply an abbreviation for a substantial majority of
the population studied thinks it is fair (KKT, 1986, p. 201). KKT thereby decentralizes the
denition to whoever is responding to a question that uses the phrase it is fair. Response
volatility will therefore compound variation in self-denition with variation in individual
responses. This is not good instrument design. Here are some alternative concepts of fairness
each anchored in SSSM theories or in cultural norms and which therefore may provide a
more precise connection with principleseither constructivist or emergent:
(1) A utility for the reward of other(s) as well as self. This is the concept underlying the
market tests of the effect of fairness in experiments.
(2) Equality of outcomes. This is the imputation rule found by social psychologists to be
preferred by people in situations where they have no means by which they can identify
differences in individual merit or in their contributions to the total to be apportioned to
individuals.
(3) Equality of opportunity to achieve outcomes: To each in proportion to their merit or
contribution to the total to be allocated. This is the equity principle, the preferred impu-
tation reported by social psychologists when individual contributions can be identied.
It is also the ancient rst harpoon principle in Inuit, Arctic and other cultures (see
Freuchen, 1960, pp. 5353).
(4) To each in proportion to his or her contribution to the net surplus of society. This derives
from the equilibrium market allocation principle. The achievement of such allocations
and the fact that they are unintended, are both conrmed by over four decades of
experimental market tests. Indeed, the fact that they are unintended can explain why
the perceived fairness of the imputations is denied by people from the same population
that conrms fairness in the sense of the equity principle.
(5) Thou shalt not steal, or covet the possessions of thy neighbor. These historical Judaic
laws are the property right norms that support markets, and surely underlie the equity
principle that emerges in social psychological ndings.
(6) Reciprocity: it is fair behavior to return favors and fair to expect others to return
yours. This also has historical Judaic-Christian roots in the Golden Rule: Do unto others
as you would have others do unto you, which helps to explain its emergence in two
person extensive form games in the laboratory, even when subjects are anonymously
paired.
148 V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150
5. Expected wealth maximization or survival?
KahnemanandTverskys (1979) most widelyrecognizedcontributions inprospect theory
were in choice exercises demonstrating the relevance of two propositions: the idea that the
theory applies to changes in wealth (that is, income) relative to the individuals current asset
state, and that people choose as if they are risk preferring in losses and risk averse in gains.
No one has stated these propositions more eloquently than Adam Smith:
We suffer more . . . when we fall from a better to a worse situation, than we ever
enjoy when we rise from a worse to a better. Security, therefore, is the rst and the
principal object of prudence. It is averse to expose our health, our fortune, our rank,
or reputation, to any sort of hazard. It is rather cautious than enterprising, and more
anxious to preserve the advantages which we already possess, than forward to prompt
us to the acquisition of still greater advantages. (Adam Smith, The Theory of Moral
Sentiments, p. 213).
This much is not inconsistent with the axioms of standard expected utility theory, which
requires only that the prizes of choice can be ordered, and therefore applies either to wealth
or changes in wealth, since either can be ordered. Also, the theory does not preclude risk
preferring preferences so long as utility, kinked or smooth, is bounded.
But I intend no defense of the standard theory; quite the opposite: the weakness of the
theory is that it places so little restriction on choice. Hence, to which prizes the theory is best
applied has always seemed to me to be inherently a subject for empirical determination, with
the objective of tightening up the wide breadth of the theory. If applied to wealth, the theory
starts to infringe on preference theory over time, recognized from the start as especially
difcult modeling terrain where tractability strains plausibility, and rejected empirically by
Binswanger (1980) in his large-stakes tests using Indian farmers.
These deeper considerations have been obscured by the narrowfocus on errors relative to
the arbitrary interpretation that utility is necessarily only about wealth, and on the rhetoric
of errors as human mistakes, rather than deviations from the predictions of a theory that
itself may be the mistake.
This brings me to the economics of survival on which I will close. Is it rational to
maximize the expected value of wealth? Work on the economics of survival by Roy Radner
and his associates offers an alternative:
Suppose the investor or rm is assumed in continuous time to choose from the set
of available investments to maximize the probability that she will survive forever.
A basic theorem shows that there exists a critical level of wealth below which the
investor appears in the short run as if he is risk preferring and above which risk averse.
But this prediction has noting to do with risk aversion. Below the critical wealth level
the best prospect for avoiding bankruptcy is to take very high variance risks because
they offer a small chance of getting back in the gameincreasing wealth enough to
enable survival.
Moreover it is shown that the investor who chooses to maximize expected prot
(discounted total withdrawals) fails in nite time. Moreover, there exist a variety of
V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150 149
non-prot-maximizing behaviors that have a positive probability of never failing. In
fact it is shown that rms that maximize prots are the least likely to be the market
survivors.
My point is simple: when experimental results are contrary to standard concepts of
rationality, assume not just that people are irrational, but that you may not have the right
model of rational behavior. Listen to what your subjects may be trying to tell you.
Think of it this way. If you could choose your ancestors, would you want them to be
survivalists or to be expected wealth maximizers?
References
Arrow, K., 1987. Rationality of self and others in an economic system. In: Hogarth, R., Reder, M. (Eds.), Rational
Choice. University of Chicago Press, Chicago, pp. 201215.
Binswanger, H.P., 1980. Attitudes toward risk: experimental measurement in rural India. American Journal of
Agricultural Economics 63, 395407.
Bolton, G.E., 1991. A comparative model of bargaining: theory and evidence. American Economic Review 81,
10961136.
Bolton, G., Ockenfels, A., 2000. ERC: a theory of equity, reciprocity, and competition. American Economic
Review 90, 166193.
Chu, Y.P., Chu, R.L., 1990. The subsidence of preference reversals in simplied and marketlike experimental
settings. American Economic Review 80, 902911.
Cox, J., Grether, D., 1996. The preference reversal phenomenon: response mode, markets and incentives. Economic
Theory 7, 381405.
Fehr, E., Schmidt, K.M., 1999. Atheory of fairness, competition, and cooperation. Quarterly Journal of Economics
114, 817868.
Fouraker, L., Sidney, S., 1963. Bargaining Behavior. McGraw Hill, New York.
Franciosi, R., Kujal, P., Michelitsch, R.V.L., Smith, V.L., Deng, G., 1995. Fairness: effect on temporary and
equilibrium prices in posted offer markets. Economic Journal 105, 938950.
Freuchen, P., 1960. Book of the Eskimos. Cleveland, World Publishing.
Friedman, M., Savage, L.J., 1948. The utility analysis of choices involving risk. Journal of Political Economy 56,
279304.
Gazzaniga, M.S., 1998. The Minds Past. Berkeley, University of California Press.
Hayek, F.A., 1973. Law, Legislation, and Liberty, vol. I: Rules and Order. University of Chicago, Chicago.
Hayek, F., 1988. The Fatal Conceit. Chicago, University of Chicago Press.
Hogarth, R., Melvin, R., 1988. Rational Choice. University of Chicago Press, Chicago.
Holt, C., Laury, S., 2002. Risk aversion and incentive effects in lottery choices. American Economic Review, Dec.
(92), 16441655.
Hume, D., 1739/1985. A Treatise of Human Nature. Penguin Books, London.
Kachelmeier, S.J., Shehata, M., 1992. Examining risk preferences under high monetary incentives: evidence from
the Peoples republic of China. American Economic Review 82, 11201141.
Kahneman, D., Tversky, A., 1979. Prospect theory: an analysis of decision under risk. Econometrica 47, 263291.
Kahneman, D., Knetsch, J., Thaler, R., 1986. Fairness as a constraint on prot seeking: entitlements in the market.
American Economic Review 76, 728741.
Knowlton, B., Mangels, J., Squire, L., 1996. A neostriatal habit learning system in humans. Science 273,
13991402.
Libet, B., 2004. Mind Time. Cambridge, MA, Harvard University Press.
Lopes, L., 1991. The rhetoric of irrationality. Psychology and Theory 1, 6582.
Maquet, P., 2001. Science 294, 10481052.
Markowitz, H., 1952. The utility of wealth. Journal of Political Economy 60, 151158.
Mazzarello, P., 2000. What dreams may come? Nature 408, 523.
150 V.L. Smith / The Journal of Socio-Economics 34 (2005) 135150
Mullainathan, S., Thaler, R., 2001. Behavioral economics. In: International Encyclopedia of the Social and Be-
havioral Sciences. Pergamon, Oxford.
Plott, C., Agha, G., 1983. Intertemporal speculation with a random demand in a speculative market. In: Tietz, R.
(Ed.), Aspiration Levels in Bargaining and Economic Decision Making. Springer-Verlag, Berlin.
Plott, C., Turocy III, T., 1996. Intertemporal speculation under uncertain future demand. In: Albers, W., Guth,
W., Hammerstein, P., Moldovanu, B., van Damme, E. (Eds.), Understanding Strategic InteractionEssays in
Honor of Reinhard Selten. Springer-Verlag, Berlin, pp. 475493.
Poincar e, H., 1913. Foundations of Science. Science Press, New York.
Rabin, M., 1993. Incorporating fairness into game theory and economics. American Economics Review 83 (5),
12811302.
Schwartz, T., Ang, J.S., 1989. Speculative Bubbles in the Asset Market: An Experimental Study. Paper Presented
at the American Finance Association Meeting, Atlanta, December.
Siegel, S., 1959. Theoretical models of choice and srategy behavior. Psychometrik 24, 303316.
Soper, B., Gigiolotti, G., 1993. Intransitive cycles: rational choice or random error. Theory and Decision 35,
311336.
Tversky, A., Kahneman, D., 1987. Rational choice and the framing of decisions. In: Hogarth, R.M., Reder, M.W.
(Eds.), Rational Choice. University of Chicago Press, Chicago.
Wagner, U., Gais, S., Haider, H., Verleger, R., Born, J., 2004. Sleep inspires insight. Nature 427, 352355.
Zajac, E.E., 2002. What fairness-and-denial research could have told the Florida supreme court (and can tell the
rest of us). The Independent Review VI, 377397.
Further Reading
Radner, R., 1998. Economic survival. In: Jacobs, D., Kalai, E., Kamien, M. (Eds.), Frontiers of Research in
Economic Theory: The Nancy L. Schwartz Memorial Lectures, 19831997.
Smith, A., 1759/1982. The Theory of Moral Sentiments. Indianapolis, Liberty Classics.
Sunder, S., 2004. Market as Artifact: Aggregate Efciency from Zero Intelligence Traders, in Models of a Man:
Essays in Memory of Herbert A. Simon. MIT Press, Cambridge, Mass.