Globalization, the knowledge society, and the

Network State: Poulantzas at the millennium

MARTIN CARNOY and MANUEL CASTELLS

Abstract In State, Power, Socialism, Nicos Poulantzas conceptualized a state that

materializes and concentrates power and displaces the class struggle from the econ-
omic to the political arena. In the past twenty years, much has changed. We argue
that economic relations have been transformed by economic globalization, work
reorganization, and the compression of space, time, and knowledge transmission
through an information and communications revolution. Knowledge is far more
central to production, and the locus of the relation between power and knowledge has
moved out of the nation state that was so fundamental to Poulantzas’ analysis.

In his innovative last work, State, Power, Socialism, Nicos Poulantzas conceptualized
a state that materializes and concentrates power and displaces the class struggle from
the economic to the political arena. At the same time, Poulantzas’ state is ‘organically
present in the generation of class powers’ (Poulantzas 1978; English edition, 1980:
45). Thus, the state is neither just political nor juridical in the sense that it reproduces
or enforces the legal bases of capitalist exchange. Rather, it is fundamental to the
conditions under which the bourgeoisie can accumulate and control capital, displacing
struggle and conflict to the political from the economic sphere.
Poulantzas’ brilliant analysis of late capitalism identifies the state as both the
crystallization and locus of class struggle. On the one hand, in the Gramscian tra-
dition, this is a struggle dominated by the unequal power of the bourgeoisie over the
division of knowledge, over the organization of production, over the juridical
apparatuses of the state, and over the definition of space and time where the nation
state presides – a definition that unifies segmented, separated, individualized, and
isolated workers. The state reintegrates separated and individualized workers into the
people-nation under a set of institutions that homogenize and normalize them, differ-
entiating them under a new set of rules, norms, values, history, tradition, language,
and concepts of knowledge that emanate from the dominant class and its fractions. On
the other hand – and this is Poulantzas’ particular contribution to theories of the state
– this integration takes place in the context of class struggle, and all the institutions of
society, including the state, are the product of that struggle. The capitalist state
provides the framework for struggles among fractions of the dominant class,
integrates the working class into a nation and a unifying set of rules and institutions.
But at the same time, the state provides the political space for class struggle, and so –
just as the capitalist state emerged from a struggle – the state becomes shaped by class
struggle. The state is key to the reintegration of workers (and bourgeoisie) into a

Global Networks 1, 1 (2001) 1–18. ISSN 1470–2266 1

Martin Carnoy and Manuel Castells

unified whole that will be reproduced as a class society from generation to generation.
Yet, contradictions arise in the superstructure itself – in the state – even as workers
are reintegrated through state power. Class struggle takes place in the heart of the
state even as it tries to maintain an externally dominant class hegemony.
Crucial to this concept of the state is that it is autonomous vis-à-vis the economy.
As in the case of Gramsci, Offe, and others, Poulantzas viewed the dominant capi-
talist class as fractionalized and internally competitive, hence reliant on the state and
politics to organize the reproduction of capitalist relations of production. The legit-
imacy of the class state depends on its autonomy, and, at least in the recent history of
capitalist development, the reproduction of capitalist relations of production has
depended on the legitimacy of the class state. Social policies were developed by the
state profiting certain dominated classes and cutting into the dominant classes’
economic power without ever threatening their political power. In Poulantzas’ words,
‘… this State by its very structure, gives to the economic interests of certain
dominated classes guarantees which may even be contrary to the short term economic
interests of the dominant classes, but which are compatible with their political
interests and their hegemonic domination’ (Poulantzas 1974: 190).
Thus, the state for Poulantzas was clearly a national class state, but it was neither
an instrumentalist depository (object) of dominant-class power, nor a subject that
possessed an abstract power of its own outside the class structure. It was a place and
centre of the exercise of power, including the exercise of power to reorganize the
national capitalist economy during periods of economic crisis.
In the twenty years since Poulantzas’ untimely death, much has changed. We will
argue that economic relations have been transformed by economic globalization, the
reorganization of work, and the compression of space, time, and knowledge trans-
mission through an information and communications revolution. Knowledge and
information have become far more central to economic production and social
relations, but the locus of the relation between power and knowledge has moved out
of the nation state that was so fundamental to Poulantzas’ analysis. Poulantzas wrote
that ‘It is permanent monopolization of knowledge by this scientist-state, by its
apparatuses and its agents, which also determines the organizational functions and the
direction of the State’ (1978: 61–2, our translation). Knowledge production and
transmission have become even more central to any hegemonic project. Yet, at
the millennium, their organization has begun to move away from nation state
control. And Poulantzas’ temporal and geographic conception of nation has also
changed, as has the role of identity in integrating or separating individuals into the
dominant definition of nation. Because of these changes, we need to reconsider the
meaning and role of the state in the new global economy. We do so in a five-part
analysis:

• the new globalization and the state;

• the reorganization of work and the separation of the worker from jobs and the
reintegrative institutions of the state;
• the decline of the nation state and the intensification of cultural identity;
• the reorganization of knowledge and power; and
• the emergence of the Network State.

2
 Poulantzas at the millennium

Globalization and the state

The multidimensional process of globalization has changed the dynamics of the state
in the last quarter of the twentieth century. However, globalization is, at the same
time, a historical process, and an ideology. The neo-liberal assertion of the inevita-
bility of globalization, and of the consequent demise of the state, has triggered a
symmetrically opposite ideology, denying the novelty of globalization, thus reiterat-
ing the old paradigm of the relationship between capitalism and the state. Therefore, it
is analytically indispensable to define globalization, on the basis of the observation of
actual social and economic trends. Only then can we consider the transformation of
the state through the transformation of its framework of reference.
For the sake of clarity, let us start with economic globalization, then extend the
analysis to other dimensions of the process. We define a global economy as the
economy whose core, strategic activities have the technological, organizational, and
institutional capacity to work as a unit in real time, or in chosen time, on a planetary
scale. As has been empirically documented by Held, McGrew, Goldblatt, and
Perraton (1999), a global economy is a new reality, different from processes of inter-
nationalization in previous times, for one simple reason: only at this point in history
was a technological infrastructure available to make it possible. This infrastructure
includes networked computer systems, advanced telecommunications, information-
based technology, fast transportation systems for people, goods and services, with a
planetary reach, and the information processing capacity to manage the complexity of
the whole system. True, most firms and most jobs in the world are not global, in fact
they are local and regional. But most, if not all, economies are dependent upon their
performance of their globalized core. This globalized core includes financial markets,
international trade in goods and services, transnational production and distribution of
goods and services, science and technology, and specialty labour. Let us focus on
global financial markets, since they condition the performance of all economies (for
data sources and analysis, see Giddens and Hutton 2000). Between 1983 and 1995,
calculating in average annual rates of change, while world real GDP grew by 3.4 per
cent, and world exports volume by 6.0 per cent, total issues of bonds and loans grew
by 8.2 per cent, and total outstanding bonds and loans increased by 9.8 per cent. As a
result, in 1998, stocks of outstanding loans and bonds amounted to about US$ 7.6
trillion, that is the equivalent of over 5 times the French GDP. A major expression of
financial globalization is the explosion of currency trading. The daily turnover of
currency markets around the world in 1998 was US$ 1.5 trillion, that is the equivalent
of 110 per cent of UK’s GDP. This volume of trading represented an increase by a
factor of 8 between 1986 and 1998. It was, by and large, unrelated to the expansion of
trade, and mainly induced by speculative movements. Cross border transactions of
bonds and equities, measured as proportion of domestic GDP, between 1970 and
1996, increased by a factor of 54 in the USA, of 55 in Japan, and of almost 60 in
Germany. Starting in the late 1980s, there has been an extraordinary increase of
financial investment between countries, including between industrialized and indus-
trializing countries (so-called emerging markets): acquisitions of overseas stocks by
investors in industrialized economies increased by a factor of 197 between 1970 and 1997.
Most investment in developing economies is concentrated in a few countries. Yet, all
countries’ economies are strongly dependent upon movements of financial investment.

3
Martin Carnoy and Manuel Castells

Global interdependence of financial markets is the result of several developments.

The first one is the deregulation of financial markets, and the liberalization of cross-
border transactions in most countries. The second is the development of a techno-
logical infrastructure that allows transactions of billions of dollars in seconds, as well
as the management of complex patterns of financial interaction throughout the world.
The third is the nature of new financial products, such as derivatives: synthetic
securities which combine the present and future value of stocks, options, commodi-
ties, currencies, and are traded in various markets. Some estimates put the value of
derivatives markets in 1997 at about US$ 360 trillion, or the equivalent of 12 times
the world’s GDP. Variation of the value of one of the components of the derivative in
one market have an impact on the value in another market: this was one important
mechanism of the diffusion of the Asian financial crisis in 1997/98. A fourth source
of integration are speculative movements of financial flows, using computer forecast-
ing models, and moving swiftly in and out of the currency or security to take
advantage of small differences in valuation. Hedge funds, and other speculative
financial organizations, often escape regulation in offshore locations. They link
markets, and amplify market trends through their movements. A fifth source of
integration is represented by market valuation firms, such as Standard & Poor or
Moody’s. By rating securities, and even national economies, according to their
standards, they assert common rules on markets around the world. Sometimes these
valuations are triggers of financial crisis. The fundamental issue is that financial
markets are very imperfect markets that only partly work according to supply and
demand. Most often, they act upon receiving economic signals that we can call
information turbulence: political events, manipulation of information, subjective
opinions of so-called experts, statements from international financial institutions, or
crowd psychology. Thus, ultimately, they are unpredictable. Since capital markets and
currencies are globally interdependent, so are monetary policies and interest rates,
creating major constraints to national economic policies everywhere.
The global economy is also characterized by transnational production of goods
and services (for sources: Held et al. 1999). There are about 53,000 multinational
corporations, with 415,000 foreign subsidiaries. They account for about 25 per cent of
global output, and for two-thirds of global trade. Indeed, one-third of world trade is in
fact intra-firm trade, reflecting cross-border production flows. Moreover, these multi-
national corporations anchor a much wider web of production networks made of
small- and medium-sized businesses. Since corporations are internally decentralized,
and since they often work in strategic partnerships with other corporations, the
organization of production in global capitalism is made of transnational production
networks, with multinational corporations at their core. All economies ultimately
depend upon the performance of these networks.
Economic globalization also includes the selective, asymmetrical integration of
science and technology throughout the world, a global labour market for highly
skilled labour, and the increasing role of international trade and investment in
economic growth (UNDP 1999).
A specific segment of the global economy is the global criminal economy
(Castells 1998/2000: Chapter 3). It includes transnational networks of illicit trade, and
money laundering, with their ancillary activities of corruption, intimidation and
killing, to protect/impose the business interests of criminal associates. The Latin

4
 Poulantzas at the millennium

American drug cartels, the Sicilian and American Mafias, the Russian Mafiyas, the
Chinese Triads, the Japanese Yakuzas, the Turkish/Albanian heroin cartels, the
Nigerian crime lords, and a myriad of local gangs around the world, have established
economic, political, and crime-enforcement empires that deeply penetrate political
institutions and business firms. IMF estimates propose a wide bracket for money
laundering from criminal proceedings, somewhere between US$ 750 billion and US$
1.5 trillion per year. In either case, it is well over the value of global oil trade. No
analysis of the state at this turn of the century can ignore this fundamental source of
influence and instability on the workings of political institutions.
Globalization also concerns the media system. Giant multimedia groups integrate
television, in its multiple forms, video, films, radio, entertainment, publishing, news-
papers, magazines, and the Internet, in an oligopolistic, trans-border system which
literally frames and shapes information and cultural representation. America-On-Line/
Time-Warner, Viacom/CBS, Disney/ABC, Sony, Seagram, Bertelsmann, Murdoch,
constitute the core of global communication, followed by a number of smaller, but
powerful players in Europe, Japan, and Latin America (Canal Plus, NKK, O Globo,
Televisa). While there is no evidence of any global conspiracy in this fiercely
competitive industry, as a whole, global media business concentrates historically
unparalleled capacity to produce and distribute images, sounds, and text around the
world. While the coming of the Internet as a universal medium of horizontal,
interactive communication could change the oligopolization of media, a similar
process is now taking place in the Internet. Major companies, such as America-On-
Line are increasing their market share of Internet communication, while all
multimedia oligopolies are quickly establishing their own Internet division, to link up
the multimedia system into a global hypertext (Schiller 1999).
Globalization could only proceed because of the new technological paradigm
that developed from the 1970s onwards, on the basis of the revolution in information
and communication technologies. But technology was not the cause, only the
medium. The source of globalization was the process of capitalist restructuring that
sought to overcome the crisis of the mid-1970s. The main actors in this restructuring
were both the corporations and the state. Furthermore, without decisive state
intervention, globalization could not have taken place. Deregulation, liberalization,
and privatization, both domestically, and internationally were the institutional basis
that paved the way for new business strategies with a global reach. Reagan’s and
Thatcher’s policies were decisive in breaking the ideological ground for this to
happen. But the 1990s were the moment when the new rules of the game diffused
around the world. The Clinton administration and the International Monetary Fund,
(strongly influenced by the US Treasury), were decisive in fostering globalization, by
imposing policies to reluctant countries, threatened with exclusion from the new,
dynamic economy.
The Russian transition, and the Asian developmental state, were the most impor-
tant victims of this enforced liberalization: they both collapsed when confronted with
global financial markets, engulfing their countries into major economic and social
crises. European governments also played their own version of globalization: the
Maastricht Treaty, and convergence policies, seen as the only possible way to resist
competition from both the United States and Asia. By so doing, the European Union
increased its strength, but submitted to the new rules of the game, the global game,

5
Martin Carnoy and Manuel Castells

while adding to it a human touch: the partial preservation of the heritage of the
welfare state.
Thus, globalization was, in fact, induced by the state, as a way out of the crisis.
Not under the command of corporations, but certainly with corporate interests as a
fundamental concern: this is the kind of policy that Poulantzas could have character-
ized as an expression of the relative autonomy of the state. Yet, once the process of
globalization was set in motion, it slipped largely out of control of states. First, for
most states in the world, the more countries that were added to this global network of
streamlined capitalism, the fewer the chances of economic survival for countries left
out of the network. Second, the unification of economies in large areas of the world,
such as the European Union, homogenize economic policies by making them increas-
ingly dependent on the free movement of capital, goods, and services, which follow
the rules of profit-making on a global scale. Third, the United States, and the G-7
group, developed an institutional framework to impose the respect for strict market
rules around the world: the International Monetary Fund, World Bank, and the World
Trade Organization were strengthened in their role as watchdogs of the global
economy. Fourth, economies of all countries, including the USA have become largely
dependent upon the performance of their financial markets, which are globally
integrated. Thus, to a large extent, the state has lost economic control over both
monetary policies and interest rates. These are now conditioned by global finance,
and decided by independent central banks – which act as interpreters of signals from
financial markets. Fifth, in a world of global production and distribution, governments
have, by and large, abandoned industrial policy, instead concentrating on providing
the technological infrastructure and human resources to attract and stimulate invest-
ment from producers of goods and services.
Unable to control wealth and information, undermined in many cases by a thriving
criminal economy, de-legitimized by its own actions and ideology, the state seems to
be overwhelmed by the global forces it decisively helped to unleash. Yet, ideologies
notwithstanding, the state does not disappear. It transforms itself. This transformation
is induced not only by globalization, but by structural changes in the work process,
and in the relationship between knowledge and power.

The reorganization of work

The individualization of class members – their separation from their class by the
capitalist state, specifically by its juridical system and its ideological apparatus – is a
fundamental tenet of Poulantzas’ exploration of why the bourgeoisie has chosen the
modern ‘democratic’ state as the expression of class power. The state is not only a
reflection of the division of labour in capitalist society; it is a crucial factor in the
organization of the social division of labour by reproducing the social ‘fractional-
ization – individualization’ inherent in that division. ‘This ideology of individual-
ization not only serves to mask and obscure class relations (the capitalist State never
presents itself as a class State), but also plays an active part in the division and
isolation (individualization) of the popular masses’ (Poulantzas 1980: 66).
The state’s role in individualizing workers and separating them from their class
was crucial in industrial capitalism precisely because of the organization of industrial
work. The factory system separated workers from their tools, but put them into a new

6
 Poulantzas at the millennium

organization of work that gave them new social identity. Despite employers’ efforts to
pit different groups of workers in the factory against each other (Gordon, Edwards,
and Reich 1982), the factory system enabled workers to organize associations around
particular workplaces and long-term employment. They created job-based, firm-
based, and industry-based unions that reproduced class identification. Job and
occupational stability in the industrial system helped produce a new working class
consciousness. The development of this consciousness in the context of industrial
work organization is a crucial fact of late nineteenth century and much of twentieth
century political life. So is the democratic capitalist nation state’s role in combating
and recasting this consciousness through creating individual citizen-nationals,
individual voters, and individual subjects of the juridical process.
Today, a major change is taking place in the workplace. Globalization and
information technology is transforming work. The new organization of work that
characterizes this transformation incorporates individualization of workers directly
into labour markets and the structure of production. This was made possible largely
because of the success of democratic capitalist states in internalizing individualization
of workers through the ideological apparatuses of the state. Globalized firms in those
advanced capitalist countries where the ideology of individualization was more
developed and class consciousness more weakly developed are finding it easier to
restructure their work organizations on the basis of individualized workers. Nonetheless,
the transformation of work is occurring in all capitalist economies. What is the nature
of this transformation and what does it imply for the ideological role of the nation state?
More intense competition on a worldwide scale makes firms acutely aware of
costs and productivity. The ‘solution’ they have settled on is to reorganize work
around decentralized management, work differentiation, and customized products,
thereby individualizing work tasks and differentiating individual workers in their
relationship to supervisors and employers. This has made subcontracting, part-time
work, and hiring temporary labour much easier, since a lot of work can be narrowed
down to specific tasks, even as other, ‘core’ work is conducted in teams and is multi-
tasked. Workers are gradually being socially defined, less by a particular long-term
job they hold, than by the knowledge they have acquired by studying and working.
This knowledge ‘portfolio’ allows them to move across firms and even across types of
work, as jobs get redefined.
This transformation is occurring rapidly and with profound impact on labour
markets. According to our estimates, ‘non-standard’ employment (part-time, tempor-
ary, and self-employed) increased in almost all countries in the OECD between the
mid-1980s and the mid-1990s, rising to 37 per cent in the UK, 30 per cent in France
and Germany, almost 40 per cent in Italy, more than 40 per cent in Holland, almost 50
per cent in Japan and 50 per cent in Australia (Carnoy 2000; see also ILO 1996). In
the USA, already characterized by a high degree of flexibility in industrial-era
capitalist labour markets compared to European countries, the proportion of non-
standard employment is relatively low, at about 27 per cent. But according to a 1999
survey conducted by the University of California at San Francisco, the proportion in
California, a region at the cutting age of the new economy, is much higher, at about
57 per cent (UCSF 1999).
In addition to the growth of non-standard employment, the new workplace is
marked by less attachment to a particular job. Turnover rates are increasing in a

7
Martin Carnoy and Manuel Castells

number of OECD countries, particularly among younger and the middle-aged workers
(OECD 1997: Chapter 5; Carnoy 2000: Chapter 3). Younger workers in the United
States change jobs more frequently in order to increase their earnings; older workers
are ‘restructured’ out of jobs, usually taking income losses. The result is flatter age-
income profiles than in the past for all education groups after age forty. The
‘traditional’ relationship between earnings and seniority and the security that went
with it are disappearing (Carnoy 2000).
The conception underlying this relationship of a family-work structure that had
men earning income and women taking care of children is also disappearing (see
Hochschild 1997). In its place, a new work system is developing that includes women
as part-time and full-time workers, a family that is generally much smaller, and jobs
no longer imbedded in a stable working class network dominated by male workers.
The effect of individualization and differentiation is to separate workers from the
‘permanent,’ full-time jobs in stable businesses that characterized post-Second World
War development in Europe, Japan, the United States, and other industrialized
countries. Just as an earlier factory revolution drove a wedge between workers and
products they made, the new transformation is dissolving the identity that workers
developed with industrial organizations such as the corporation and the trade union.
Workers are being individualized, separated from their ‘traditional’ identities built
over more than a century, and from the social networks that enabled them to find
economic security. The ‘job’ and everything organized around the job – the group of
work friends in the company, the after-work hangouts, the trade union – lose much of
their social function. They are as ‘permanently temporary’ as the work itself.
Some, mainly highly educated, professional and technical workers are building
new networks. Instead of just talking to colleagues in the company where they work,
they build electronic mail and informal information relations across companies and
across countries (Saxenian 1994; Wellman 1999). Network technology such as the
Internet helps. Information exchanged in after-work, upmarket hangouts attract
professionals from a broad range of firms and serve the same purpose. The main
question is what happens to the vast majority of workers who do not have easy access
to information about other companies or to workers in other companies, or those
highly skilled workers who fall ‘out of the loop.’ They tend to be left in an
individualized limbo, ‘disaggregated’ from traditional networks but not ‘integrated’
into new ones. New, private networks, such as temp agencies, are emerging to fill this
void (Benner 2000). But, except for some striking exceptions, such as construction
unions that traditionally allocate temp jobs among their members, these new networks
are not organized for or by workers. They miss satisfying the need for social inte-
gration served by stable jobs, unions, and professional associations.
If workers are now being successfully disaggregated from class identity at the
workplace itself, the crucial role of the nation state in individualizing class-conscious
workers becomes less central to capitalist hegemony. Rather, the ideological appar-
atuses of the nation state – the juridical system, education, and political parties – take
on roles that have less to do with preserving capitalist class relations against an active
potential of working-class consciousness, than in trying to keep isolated individuals in
the global economic and political environment. The main project of the capitalist state
changes from one of separating the worker from his class-based identity to one of
bringing the isolated, individualized worker into a global, market identity. The

8
 Poulantzas at the millennium

legitimacy of the nation state itself hinges on its capacity to reintegrate workers into a
globalized notion of community and identity. Poulantzas’ notion of the state as a
locus of struggle remains, but the struggle inside the state shifts from one based solely
on class to one based on group (including class) identity and their various and multi-
faceted conceptions of reintegration. At the same time, the role of political struggle in
the ideological apparatuses of the nation state is not as critical as it used to be.

The reorganization of knowledge and power

For Poulantzas, scientific knowledge, controlled by the state, was fundamental to the
relations of power in capitalist society. In his analysis, state apparatuses control
knowledge and discourse. At the same time, the state is shaped by the way it uses
knowledge to reproduce unequal relations of power. ‘It is the permanent monopol-
ization of knowledge by this scientist State, by its apparatuses and its agents, which
also determines the organizational functions and the direction of the State, functions
which are centralized in their specific separation from the masses’ (Poulantzas 1978:
61–2; our translation). In his analysis, the state uses knowledge to legitimize a
particular ideology – the dominant bourgeois values and norms – by changing that
juridical–political ideology into a set of technocratic ‘facts’ and decisions based on
factors like ‘scientific’ studies, and ‘expertise’. The state not only has an important
hold on the generation of new knowledge in the society, but also on how that
knowledge is used. It is the state and its control over intellectuals that is crucial to the
new formation of the division of knowledge and its uses. One therefore cannot talk
about technology or knowledge without talking about power.
In the developed countries during the industrial era, this definition of knowledge
could be characterized as largely national. The state could define that ‘science’ and
those values that served the reproduction of power within national boundaries.
Poulantzas recognized that scientific knowledge had become central to capitalist
production in late capitalism. Much of this centrality emerged as part of the military-
industrial complex that dominated post-1945 scientific development, not only in
Western capitalist countries but also in the Soviet Union and China, the main statist
economies. Beyond the militarization of scientific innovation, the post-Second World
War welfare state, with its vast public resources, was also able to define the social
incorporation of dominated classes through its knowledge apparatuses, namely the
educational system, and the health system.
In the globalized, information technology environment, however, knowledge
formation and power over knowledge moves out of the control of the nation state.
This is true for three reasons.
First, dominant capitalist values and norms are in the process of being globalized.
They are increasingly organized around knowledge and information that circulates
globally and serves a globalized innovation and profit-making structure (Gordon and
Kimball 1998; Schiller 1999). The highest valued information and knowledge now
has its locus in the global economy, not in any single national site, even though many
kinds of knowledge with global value are produced in national locations or certain
local aggregations of innovation, such as Silicon Valley. This transformation of
knowledge formation from national into global space is still in its incipient stages, but
the trend is clear. It results mainly from the globalization of science-based, innovation

9
Martin Carnoy and Manuel Castells

industries, the rapid demilitarization of high tech companies beginning in mid-1980s

(Gordon and Kimball 1998), and, finally, the rapid, global growth of telecommuni-
cations and information technology themselves. Thus, even though the most advanced
forms of knowledge production still take place in the private and public sectors of a
few countries, the public role is proportionately declining and knowledge production
capacity to other countries is spreading, first to the newly industrializing countries,
then further out (Carnoy 1998). And because ‘global values’ are primarily those
rooted in global finance and production, knowledge is increasingly defined in terms of
economic value. Since the nation state’s monopoly of knowledge was at least partly
determined by its ability to interpret and transmit bourgeois behavioral norms,
including language and culture, this shift to greater valuation of profit-producing
skills again means the reduction of the nation state’s control over knowledge.
Second, many of the same forces are decentralizing the control of knowledge
production and transmission within countries away from the centrality of nation
states. Nation state-sponsored and controlled scientific innovation for military
purposes, while still important, is declining rapidly compared to innovation that
originates totally outside state control. For example, the Internet, which was originally
state sponsored as part of a university/military networking program (Arpanet), is
spreading worldwide outside the state knowledge system (Abbate 1999; Naughton
1999; Zook 2000). The near-ubiquity of personal computers and Internet access in the
developed countries is gradually making it possible to access education, particularly
at the university level, through privately financed and managed courses of study,
many of them promising better training in globally-valued knowledge than provided
by the state educational system.
Third, because of increased economic competition and ageing populations, the
industrial nation states have been unable to sustain the level of welfare characterizing
the post-1945 world known to Poulantzas. This has decreased the influence of these
states over those segments of the population already most marginalized in industrial
capitalism. Global competition marginalizes these segments – namely those with the
least formal education, the least socialization into the nation state, and the least
capacity to increase the value of their human capital – even more than before.
Because they have difficulty accessing knowledge deemed valuable by global
markets, their members search for other forms of knowledge that reinforces their
identity both outside global markets and outside the nation state knowledge system
(Castells 1997/2000).
Thus, the modern capitalist state developed into a successful market ‘softener.’
But the decline of that role in the face of powerful global marketization of national
economies pushes the ‘dispossessed’ to seek refuge in new and more exclusive
collectives. These collectives generally do not have the power or the funds to help the
dispossessed financially nor to develop the skills and knowledge valued by global
markets. They can help develop self-knowledge and therefore self-confidence. They can
provide community and therefore a sense of belonging. Cultural identity, or what we
call ‘self-knowledge’ (Carnoy 2000), whether religious, ethnic, racial, or gender, and
whether local, regional or more global, is an antidote to the complexity and harshness
of the global market as the judge of a person’s worth (Castells 1997/2000). Although
the production of such self-knowledge may be highly centralized in formal, some-
times global, religious organizations, it is generally disconnected from nation states.

10
 Poulantzas at the millennium

Thus, globalization has eroded the nation state’s monopoly of scientific knowl-
edge and its ability to use that knowledge to reproduce class power, even as the nature
of class power relations itself moves away from nation state control. Even as
knowledge becomes increasingly important in shaping power relations, the state’s role
in this process is declining for the reasons discussed. Yet, that said, the state still has
an important, albeit changing, role in defining knowledge, distributing it, and using it
to shape power relations. The state-financed and run educational system continues to
dominate the educational process, hence the transmission of knowledge to the young
as well as the production of new knowledge in universities. Those who want to
acquire new knowledge therefore still have to pass through the apparatuses of the
state.
The difference is that in the global environment, global markets, not national
markets and moral values, determine the knowledge transmitted in the nation state’s
educational system. The nation state increasingly uses its remaining control over
knowledge production and transmission to develop a global ideology rooted in the
singular value of productive skills – an ideology that tends systematically to
undermine the importance and legitimacy of the nation state even as the nation state
attempts to use the educational system in this way to maintain political legitimacy.
Further, because knowledge is becoming more important in shaping power, those
groups that were not well served by the state’s education system in the past are now
more willing to abandon the state in their drive to acquire globally-valued skills. In
the industrial era, mere graduation from secondary school or vocational school was
sufficient to get a good, high-paying, permanent job in a manufacturing plant. In
today’s global, knowledge-based service economy, access to university education has
become crucial to getting a high-paying job, and to establishing the networks needed
for social mobility. Poor and many middle-class parents reason that if the state was
not able or willing to deliver the knowledge needed for entry to university education
in the past, why would it do so now. Although they continue to pressure the state to
give their children better access to knowledge, they also push for alternative, privately
run forms of education not under the direct control of the state.
In sum, knowledge formation and power over knowledge in the global economy is
moving out of the control of the nation state, because innovation is globalized,
because the discourse on knowledge is outside the state’s control, and because information
is much more accessible than it was before thanks to technology and communications.
Self-knowledge, in the form of cultural identity, also pulls the discourse out of the
‘scientific.’ The nation state is also losing control over the educational system, as
education decentralizes and privatizes in response to pressures by parents to get the
kind of education for their children that is needed to acquire globally-valued skills.
At the same time, knowledge and information continue to be distributed very
unequally, probably more unequally than when it was a state ‘monopoly.’ The new
relation between power and knowledge is negotiated increasingly in the global
innovation system and global markets. Ironically, national and regional states, which
previously monopolized knowledge and shaped its distribution, are still a major site of
organizing knowledge transmission and production, but now for the global economy.
And it is mainly through this knowledge production and transmission that the state
both maintains its legitimacy and shapes the national economic/political space in
terms of global investment and production. The better the state can ‘reintegrate’ its

11
Martin Carnoy and Manuel Castells

disaggregated workers into a smoothly functioning knowledge-based society, the

higher the potential return to global capital in those national and regional sites, and
the more rapid the economic development at those sites (Carnoy 2000, Chapter 5;
Scott 1998).

Crisis and reconstruction of the state: the Network State

Towards the end of the twentieth century the state was confronted with a series of
challenges: the process of globalization; the transformation of the work process, and
its consequences for the welfare state, one of the pillars of state legitimacy; the
centrality of knowledge in economy, and society, thus affecting the education system,
a key source of control for the modern state; and a deep crisis of legitimacy, resulting
from all the above, plus the damage inflicted upon the state by the rise of identity
politics (Castells 1997/2000; Guehenno 1993). Let us briefly review these different
dimensions of the contemporary crisis of the state without reiterating the arguments
already presented in this paper.
Globalization limits the sovereignty of the state. But it also does something else: it
redefines the social boundaries of the state (Held 1991). If nations are intertwined, the
community to which the state is accountable becomes blurred. How can a strictly
national state respond to transnational communities? The classic nation state requires
a bounded national community as its frame of reference. When globalization forces
the state to redefine this frame of reference, then national communities lose their
channel of political representation. What follows is the development of nationalism
against the state. The separation between nation and the state is a fundamental process
characteristic of our time (Calhoun 1998).
The individualization of work, the development of networking and flexibility as
forms of economic activity, the instability of employment, all undermine the
institutions of the welfare state that were built on stable partnerships between
representatives of capital, workers, and the state. The state reduces its role as a
guarantor of social protection, and individualizes its relationship with most citizens.
The ensuing crisis of the traditional welfare state undermines the legitimacy of the
state because it loses its appeal as the provider of last resort.
The centrality of knowledge in the new economy enhances the role of schools as
productive forces. Consequently, the role of the school as a national ideological and
domination apparatus recedes, undermining one of the key elements of social control
and ideological reproduction on the part of the state.
The rise of identity politics challenges citizenship as the sole source of political
legitimacy. The state has to respond to a series of cultural demands, often
contradictory, from a plurality of sources that supersede the individual citizen as the
basis of political representation.
In addition to structurally induced crisis of legitimacy, the state also suffers from
the crisis of legitimacy of the political system. This crisis is induced by the practice of
media politics, closely associated to the politics of scandal. This is, the widespread
use of character assassination and diffusion of damaging information as the main
weapons in a political competition simplified into snap shots, and sound bites in the
media, as media become the decisive space of politics (Castells 1997/2000; Rose-
Ackerman 1999).

12
 Poulantzas at the millennium

States, however, react to this multidimensional crisis. They react by reconfiguring

themselves to try to accommodate to new pressures and new demands. This recon-
figuration develops, primarily along two axes. First, nation states build international,
supra-national, and co-national institutions, in order to manage together the process of
globalization that threatens to overwhelm individual states. Let us take, as an
example, the European states as the clearest expression of this new historical trend.
They have constituted the European Union, with the historical project of incorpor-
ating the whole of Europe, minus Russia. The 15 member European Union is already
a new co-national state. It is co-national because the decision-making power is in the
hands of the Council of Governments, in which all member states are represented,
with veto power over key decisions. But this is not the extension of the member
nation states. It is a new instance of shared sovereignty. As European states they have
also created supranational institutions – institutions that, while ultimately dependent
upon the member states, have their autonomy (European Commission, European
Parliament), and even their independence (the European Central Bank). They have
also trusted their defence to a supra-national organization with a high degree of
autonomy in the conduct of the war: NATO.
A similar process is taking place at the world level, with institutions overlapping
across areas of the world in key areas of policy making. The G-7 group (a co-national,
informal institution) takes major decisions in managing the global economy. The
execution of these decisions is entrusted to the International Monetary Fund, the
World Bank, and the World Trade Organization. But these institutions also have a
large degree of autonomy vis-à-vis their member states at large, although they work in
close cooperation with the US government. International institutions also play an
increasing role in handling social, environmental, and political problems around the
world. These are institutions created by the nation states, with no real power, but with
considerable influence in public affairs, as they become the space of negotiation and
co-intervention for governments. They are the United Nations agencies, the regional
associations (such as the Organization of American States, Organization of African
Unity, Asian Pacific Economic Cooperation, ASEAN, the Arab League etc.). These
organizations are half a century old, but their role is increasing, as shown by the
growing number of military interventions and the observation of political processes
under their flags. Besides, critical questions for humankind, such as environmental
problems, public health, human rights, and emergency relief, are being treated
internationally by a complex web of ad hoc agreements and organizations.
The other axis of the nation state’s reconfiguration is its attempt to regain
legitimacy and to represent the social diversity of its constituency through the process
of decentralization and devolution of power and resources. This translates primarily
into revitalizing sub-state national governments (such as Scotland or Catalonia),
regional governments, local governments, and non-governmental organizations. Indeed,
the dramatic expansion of non-governmental organizations around the world, most of
them subsidized and supported by the state, can be interpreted as the extension of the
state into civil society, in an effort to diffuse conflict and increase legitimacy by shifting
resources and responsibility to the grassroots. In the 1990s, Europe has seen an extra-
ordinary development of regional and local politics, together with the expansion of citizen
participation, and the growing recognition of national cultures and specific identities.
These movements can be observed throughout the world (Borja and Castells 1997).

13
Martin Carnoy and Manuel Castells

The two trends of supranationality and devolution go hand in hand. Nation states
are surviving, and indeed strengthening, their position by going global and local at the
same time and by trying to found their legitimacy both in citizenship and cultural
plurality. In the process, they assure their historical continuity, but they contribute to
the demise of the nation state as it was constituted during the Modern Age, and
exported to the postcolonial world (Guehenno 1993; Hoogvelt 1997). This is because
the modern nation state was based on the twin principles of national representation of
citizens, and territorially-based national sovereignty. With shared sovereignty, and
shared sources of legitimacy, the social foundations of the classic nation state are
irreversibly undermined by supranationality from above, and transnationalism from
below (Smith and Guarnizo 1998).
What emerges is a new form of the state. It is a state made of shared institutions,
and enacted by bargaining and interactive iteration all along the chain of decision
making: national governments, co-national governments, supra-national bodies,
international institutions, governments of nationalities, regional governments, local
governments, and NGOs (in our conception: neo-governmental organizations).
Decision-making and representation take place all along the chain, not necessarily in
the hierarchical, pre-scripted order. This new state functions as a network, in which
all nodes interact, and are equally necessary for the performance of the state’s
functions. The state of the Information Age is a Network State.
Can this Network State accommodate the pressing demands of adapting the
welfare state to the new work process, and ensuring knowledge-production through
the school system? Here is where Poulantzas’ teachings become invaluable. Let us
conceptualize the state as on the one hand performing the functions of facilitating
accumulation of capital and reproduction of labour power and on the other hand
ensuring domination (of social interests), and legitimation (of state institutions).
Based on this conception, we can hypothesize a set of new historical modalities of
fulfilling this complex set of tasks via the Network State. Accumulation and
domination are facilitated globally by co-national and supranational institutions.
Legitimation and reproduction are ensured primarily by regional and local
governments and NGOs. New, global dimensions of legitimation (human rights), and
reproduction (environment, health) are fulfilled by international institutions, under the
hegemony of accumulation-oriented supra-national institutions. The welfare state is
downsized by supra-national, accumulation-oriented institutions (e.g. the International
Monetary Fund), thus lifting the burden of responsibility from the nation state. The
school system is gradually pushed away from national ideological domination toward
generating knowledge based on global, not national values. The ideological functions
of schooling are increasingly localized and customized to subsets of the national
collective. Thus, the state diversifies the mechanisms and levels of its key functions
(accumulation, reproduction, domination and legitimation), and distributes its
performance along the network. The nation state becomes an important, coordinating
node in this interaction, but it does not concentrate either the power or the responsi-
bility to respond to conflicting pressures.
Now, who is ‘the state’? Does the state react by itself and reconfigure itself in full
consciousness of this process, independently from social classes and other social
actors? Here is where Poulantzas’ concept of relative autonomy becomes essential.
Governments acted and reacted under the pressure of economic forces and social

14
 Poulantzas at the millennium

actors in the 1970s–1990s period. They made key decisions that induced global-
ization, and allowed the emergence of a knowledge economy, and they reconfigured
state institutions. Thus, states acted on their own. However, they acted under pressure
from dominant capitalist groups within the framework of preserving/expanding
capitalism and accepting liberalism as the hegemonic ideology. The ideological battle
had been won in society by cultural libertarianism and by the demise of statist
ideologies, associated with the collapse of Communism in the minds of people around
the world. States that did not join this global network were increasingly marginalized.
Governments which tried sharply different, nation-oriented policies were compelled
to change course (e.g. France under Mitterrand in 1981), or were pushed aside by
crisis (e.g. Peru under Alan Garcia in 1986). The Network State was integrated into
global networks of accumulation and domination, while responding to pressures and
demands from national/local societies. State policies were selected by dominant
interests and legitimized by citizens in various degrees. The process of trial and error
determined the course of political transformation. The Network State resulted from
the outcomes of social struggles and geopolitical strategies fought in the transition
period from the industrial era to the information age in the last lap of the millennium.

Conclusion: the theory of the state at the turn of the millennium

The theory of the state must tackle the issues posed by the history of the state. At the
turn of the millennium, the three major issues concerning the state are the following:
how social domination is enforced; how legitimacy is established; and which kind of
autonomy the state has vis-à-vis dominant classes and social actors at large.
Because of the individualization of the work process, the state’s role in disorgan-
izing class struggle has changed. The state is no longer required to individualize
workers as citizens to undermine class consciousness, as Poulantzas brilliantly
perceived to be the case in the era of industrial mass production. The production
process does the job for the state. Class domination is now enforced at two
fundamental levels. Vis-à-vis dominant classes (that is the collective capitalist), class
domination is ensured by managing and spreading globalization. The capitalist state is
the globalizing state. Thus, in order to represent the interests of capitalism, the state
has to overlook the interests of the nation and of its national citizens. Still, it will have
to deal with legitimacy problems arising under this new pattern of domination. But
domination is exercised through globalization, and through the networking of nation
states that become syndicated in defense of globalization. Vis-à-vis domestic social
classes, the state exercises domination through the education system. In a society
where education, information, and knowledge are the critical sources of wealth and
influence, class formation takes place in the classroom. Who gets what in the
education system determines who gets what in capital, communication, and political
influence. But as we have argued, class domination through education is increasingly
passed to more localized political units.
Domination cannot survive for long without legitimacy. People must feel some
degree of allegiance to the state, or domination will become synonymous with
dictatorship. The betrayal of national interests, the rise of media politics, and its close
associate the politics of scandal, all contribute to undermine the legitimacy of the
nation state. Its strategy to escape de-legitimation is two-pronged: it strives to

15
Martin Carnoy and Manuel Castells

stimulate economic growth, national employment, and domestic consumption; and,

simultaneously, it decentralizes political responsibility by increasing local/regional
autonomy.
For most people in the world, fully aware of what is going on, and ready to stay
home, the critical matter is personal security. Security ultimately translates into
economic growth and improving living standards. In this sense, even social inequality
is not a major issue. If people see their lives improving, they will not be ready to lose
what they have only to correct the injustice of the rich getting richer. So, steady
improvement of living standards for the large majority of the population, via
informational productivity, and globalization-induced economic growth, is the main
axis for building state legitimacy. We are not saying this will work. We are simply
observing that this is a widespread state practice, and is, in fact, the only option once
the choice has been made to adapt to the rules of global financial markets. In addition,
redistribution through welfare, pensions, taxation, and other efforts to equalize
income, whenever and wherever possible, would increase legitimacy. But this is an
afterthought for the Network State. Redistribution is only implemented under
substantial pressure from social movements.
The second way to establish legitimacy in the new historical context is decentral-
ization of state power to sub-state levels: to sub-national groupings, to regions, and to
local governments. This increases the probability that citizens will identify with their
institutions and participate in the political process. While nation states cede power,
they also shift responsibility, in the hope of creating buffers between citizens’
disaffection and national governments. Legitimacy through decentralization and
citizen participation in non-governmental organizations seems to be the new frontier
of the state in the twenty-first century.
Still, the state will have to respond to social movements’ demands to avoid
a legitimacy crisis. Some of these demands may not be easy to accommodate within
the existing state institutions. This is particularly the case with demands emerging
from the women’s movement, as the crisis of patriarchalism as a hegemonic
institution will lead to the calling into question of the patriarchal dimension of the
state.
The greatest historical change for the state is the dramatic decline in its autonomy.
The state becomes dependent on the collective capitalist represented by global
financial markets. It becomes dependent on the process of globalization of production,
trade, technology, and communication. It becomes dependent on other states, as it
links up with state institutions to constitute the new, network state. It becomes highly
dependent on the ideological apparatuses constituted around global media. It is
dependent on lower levels of the state, as these levels increasingly perform key
legitimation functions previously in the hands of the nation state. And it continues to
depend on the institutions of patriarchalism, and particularly on the patriarchal family,
the corner stone of ideological hegemony (‘family values’). When and if captured by
specific interest groups, states become predatory states (like in Russia, Mexico and
Nigeria, among so many others), losing all autonomy. Last, but not least, when
captured by fundamentalist identity movements, be it religious or nationalist (as in
Iran or Serbia) they lose all autonomy vis-à-vis religions or ideologies. Overall, the
relative autonomy of the state is fading away, to a large extent because relatively
autonomous states chose their own historical demise.

16
 Poulantzas at the millennium

At the same time, the legitimacy of the nation state and its capacity to enforce the
underlying rules and regulation of national market economies through democratic
means and smoothly running political apparatuses, as well as to support a well-
developed market information system, are important to global finance capital. They
lower risk for capital and raise profit/risk ratios (World Bank 1999). Global capital’s
‘co-dependence’ on smoothly functioning civil-political societies offsets some of the
decline in national autonomy we have described; it provides even those nation states
in supranational arrangements some leverage vis-à-vis both global capital and
pressures for more local autonomy.
We are thus entering a world that is very different from the one in which
Poulantzas lived – together with us. Thus, many of his/our former analyses do not
apply to new historical realities. But this is not a major flaw for social theories. Only
metaphysics pretends eternal validity. Social theories are not supposed to provide
answers forever. Instead, their value is tested on the relevance of the questions they
allow us to ask. That we are tentatively able to explore the state at the turn of the
millennium by asking questions inspired and conceptualized by Poulantzas’ theory is
a tribute to the perennial value of his thought. Nicos Poulantzas lives in our minds,
and he will continue to live in the minds of political theorists during the twenty-first
century.

Martin Carnoy is Professor of Education and Economics at Stanford University.

Manuel Castells is Professor of Sociology at the University of California, Berkeley.

Note
This article is a revised and updated version of a paper originally presented at the University of
Athens International Conference on Nicos Poulantzas, Athens, Greece, 30 September – 2
October 1999.

References
Abbate, Jane (1999) Inventing the Internet, Cambridge, MA: MIT Press.
Benner, Chris (2000) ‘Work flexibility and labor institutions: Labor markets in Silicon Valley,’
Berkeley: University of California, Department of City & Regional Planning, Ph.D.
Dissertation (unpublished).
Borja, Jordi and Manuel Castells (1997) Local and global, London: Earthscan.
Calhoun, Craig (1998) Nationalism, Minneapolis: University of Minnesota Press.
Carnoy, Martin (1998) ‘The globalization of innovation, nationalist competition, and the
internationalization of scientific training,’ Competition and Change, 3: 237–62.
Carnoy, Martin (2000) Sustaining the new economy: work, family, and community in the
Information Age, Cambridge, MA: Harvard University Press.
Castells, Manuel (1997/2000) The power of identity. Oxford: Blackwell.
Castells, Manuel (1998/2000) End of millennium. Oxford: Blackwell.
Giddens, Anthony and Will Hutton (eds) (2000) On the edge: Essays on a runaway world,
London: Jonathan Cape.
Gordon, David, Richard Edwards, and Michael Reich (1982) Segmented work: divided
workers, New York: Cambridge University Press.

17
Martin Carnoy and Manuel Castells

Gordon, Richard and Linda Kimball (1998) ‘Globalization, Innovation, and Regional
Development,’ Competition and Change, 3 (1–2): 9–39.
Guehenno, Jean-Marie (1993) La fin de la democratie, Paris: Flammarion.
Held, David (1991) ‘Democracy, the nation-state and the global system,’ Economy and society,
20 (2): 138–72.
Held, David, Antony McGrew, David Goldblatt and Jonathan Perraton (1999) Global
transformations, Stanford: Stanford University Press.
Hochschild, Arlie (1997) Time bind, New York: Metropolitan Books.
Hoogvelt, Ankie (1997) Globalization and the postcolonial state, Basingstoke: Macmillan.
International Labour Organization (1996) World Employment 1996/97, Geneva: ILO.
Naughton, John (1999) A brief history of the future: The origins of Internet, London:
Weidenfeld and Nicolson.
OECD (1997) Employment outlook, Paris: Organization for Economic Cooperation and
Development, July.
Poulantzas, Nicos (1974) Political power and social classes, London: New Left Books.
Poulantzas, Nicos (1978) L’état, le pouvoir, le socialisme, Paris: Presses Universitaires de
France.
Poulantzas, Nicos (1980) State, power, socialism, London: New Left Books, Verso edition.
Rose-Ackerman, Susan (1999) Corruption and government, New York: Cambridge University
Press.
Saxenian, AnnaLee (1994) Regional advantage, Cambridge, MA: Harvard University Press.
Schiller, Dan (1999) Digital capitalism, Cambridge, MA: MIT Press.
Scott, Allen (1998) Regions in the world economy, New York: Oxford University Press.
Smith, Michael and Luis Guarnizo (eds) (1998) Transnationalism from below, London: Sage.
UNDP (1999) Human development report 1999, New York: United Nations Development
Programme.
UCSF (University of California at San Francisco), Institute for Health Policy Studies (1999).
‘The State of Health in California—1999.’ UC San Francisco (Power Point Presentation).
Wellman, Barry (ed.) (1999) Networks in the global village, Boulder, CO: Westview Press.
World Bank (1999) World development report, Washington, DC: World Bank.
Zook, Matthew (2000) ‘Venture capital and regional networks in the development of the
internet industry,’ Berkeley: University of California, Department of City and Regional
Planning, Ph.D. dissertation (unpublished).

18