Socio-Economic Review (2005) 3, 577587

10.1093/SER/mwi026

Wolfgang Streeck
Max Planck Institute for the Study of Societies, Paulstr. 3, 50676 Koln, Germany
Correspondence: E-mail: [email protected]

Authors receiving such impressive comments on their work as we did must be

grateful even if some of the comments are critical. Since criticism is essential to
scholarly progress, they might even be tempted to respond that critical debate
was all they ever wanted to achieve. But given the significance of the themes raised
by Herrigel, OSullivan and Pempel, this would be all too complacent. Respecting
inevitable limitations of space, I want to focus on three issues that to me are the
most important. First, I will try to defend our choice of terminology, pointing out
what its purpose was and what it was not. Second, I will address the relationship
between economical functionalism and historical institutionalism, hoping to be
able to bring out how one might try to reconcile the two. And third, I will briefly
register agreement with Pempel that our books are still far from offering a convincing political account of the complex process of economic liberalization currently under way in the two countries we have studied.

1. The conceptual framework

In this respect we were not nearly as ambitious as some of our readers seem to
believe. In fact, the decisions we made here were of a rather ad hoc kind, driven
more or less pragmatically by the empirical puzzle that we were trying to address.
(Is this, perhaps, how conceptual decisions should generally be made?) That
puzzle is conveniently summarized in three observations that were at the time
shared by many others: (a) that the economic orders of Germany and Japan
were considerably less market-driven than those of other countrieslike the
US and, to some extent at least (e.g. see Dore, 1973, 1987), the United Kingdom;
(b) that, astonishingly and indeed provocatively to many, they had nevertheless
for some time done significantly better than more market-driven national
Oxford University Press and the Society for the Advancement of Socio-Economics 2005. All rights reserved.
For Permissions, please email: [email protected]

Downloaded from ser.oxfordjournals.org at University of North Carolina at Chapel Hill on October 17, 2010

Rejoinder: on terminology, functionalism,

(historical) institutionalism and
liberalization

578

W. Streeck

Downloaded from ser.oxfordjournals.org at University of North Carolina at Chapel Hill on October 17, 2010

economies; and (c) that they had now both moved into stagnation and crisis,
as a result of which they had come under pressure to reorganize themselves in a
more market-driven fashion, ironically in the image of the very countries
that they had outperformed for almost two decades (for a somewhat
different phrasing of the same problem see Streeck, 2001, p. 1, 3; Streeck and
Yamamura, 2003, pp. 12).
To formulate our puzzle we needed proper concepts. Capitalism, which was
what we were dealing with, we conceived, not very originally, to be an economic
system based on private property and the allocation of values through selfregulating markets. Of course there exist in capitalism also states and public
property, and some exchanges take place at administered prices instead of by
free allocationand in some countries, like Germany and Japan, more than in
others. The greater the role of private property and free markets, the more
capitalist a capitalist system could be said to be. But one could also say the
more liberal, given that private property and free exchange are the historical
essence of liberalism. So we decided to distinguish between different versions of
capitalism in terms of their degree of liberalism, with Germany and Japan coming
out historically less liberal than the US or the UK. To simplify further, or so we
hoped, we polarized our property space and moved from the gradual distinction
between more and less liberal to the categorical distinction between liberal and
non-liberal, assuming that everybody would understand what this was intended
to mean and what was not.
With hindsight the flaws of this construction are obvious, and not surprisingly
they become all the more visible if one attempts to make the construction bear
a heavier burden than it was proposed for. It is quite true, as Mary OSullivan
reminds us, that we did not inquire into the details of American and British capitalism and the many differences that no doubt exist between them. It is also true
that we did not discuss the unquestioned importance of non-liberal elements
in American capitalism, such as the NIH research budget or the military
industrial complex. Similarly, we are clearly guilty sometimes of stylizing the facts
especially of US capitalism in terms of what we, somewhat ironically, call the
standard capitalism of standard economics (Streeck, 2001, p. 5). Had our
books been intended to provide an empirical comparison between Germany
and Japan on the one hand and, say, the US and the UK on the other, this would
be unforgivable. But this was not our intention. All we needed, or so we thought,
was to provide some sort of background awareness, based on what we felt was
widespread consensus among political economists, that there were capitalisms
out there that were more capitalist, or liberal, than the Japanese model or
the German model of the 1970s and 1980s; and that these had gained ascendancy while the non-liberal, less capitalist systems of Germany and Japan had
moved from dominance to decline (cf. Streeck, 2001, p. 5).

Rejoinder

579

For a somewhat less pragmatic discussion of the concept of non-liberal capitalism, which draws
heavily on the notion of social embeddedness of economic transactions while, again, making no
claim whatsoever to originality, see Streeck (2001, pp. 23, 7) and Streeck and Yamamura (2003,
pp. 23). We also occasionally speak of nationally embedded or nationally organized capitalism,
which indicates how little attached we are to our main concepts.
2

Every countrys Weg, we said in an earlier draft, is a Sonderweg. Unfortunately, this line was
deleted by a copy editor under orders, presumably, to minimize the use of outdated languages.

Downloaded from ser.oxfordjournals.org at University of North Carolina at Chapel Hill on October 17, 2010

In other words, our typology, or more modestly, our simplified categorical

distinction was devised only for the purpose at hand. It was certainly not devised
as a theory to compete, for example, with the theory we believe is behind the
apparently related distinction between liberal and coordinated market economies
(Hall and Soskice, 2001).1 That theory, as we read it, implies refutable hypotheses
to the effect that the world of possible capitalisms divides into two and only
two classes, and that because of internal and external pressures for functional
complementarity between its elements, every capitalist system must sooner or
later locate itself into one of the two camps. Nothing like this is implied in our
two books, or at least no such implication is intended.
In fact, if one were to draw out the assumptions that more or less visibly
underlie our concepts, our strong emphasis on the differences between the two
countries in our non-liberal, or less liberal, category (see Streeck, 2001, pp. 34,
25) would stand out even more than it apparently does now. Obviously we do
regard it as an important similarity between Germany and Japan that both are,
and have long been, less liberal, in our terms, than imagined or prescribed by
standard capitalist economics. However, nowhere do we suggest that this common characteristic should be of such pervasive causal power as to determine all
other characteristics of the two economic and political systems. Certainly, our
account of the enormous differences of the two countries reactions to their
current calamities (Streeck and Yamamura, 2003, pp. 517) implies that simply
being non-liberal is not enough to produce identical or similar responses to
pressures for liberalization. When we repeatedly emphasize the differences, not
just between Germany and Japan, but within any conceivable category of capitalisms, we implicitly but unmistakably subscribe to Max Webers view of social
formations as historical individuals (Weber, 1904). While shaped by causal
mechanisms that can be generalized and ideal-typically modelled, historical
individuals are unique; they must be understood and explained on their own
terms since the forces that form them are both at crossroads with each other
and operate in historically contingent combinations and circumstances.2
An important implication of this approach is, incidentally, that the national
models of (non-liberal) capitalism that we are dealing with resist reification
as they are continually in flux. Historical individuals can only be conceived as
dynamic. Rather than systems in equilibrium, they are the temporary and

580

W. Streeck

2. Functionalism or historical institutionalism?

In her penetrating comment Mary OSullivan notes a strong and recurrent
tension in the two books between two rather different types of institutional
analysis, which she calls functionalism and historical institutionalism.
The functionalist part of the analysis she characterizes as claiming that particular
systems of capitalism provide institutional support for different types of
competitive strategies and economic performance while the historicalinstitutionalist approach emphasizes the constant, unintended, contingent and
3

Recently Arndt Sorge has, in a beautiful book on the societal logic of globalization drawing on the
longue duree history of Germany, introduced the concept of a national meta-tradition to account
for similarities over long periods of time in the responses of societies to the continuous expansion
of their social and economic horizons (Sorge, 2005).

Downloaded from ser.oxfordjournals.org at University of North Carolina at Chapel Hill on October 17, 2010

idiosyncratic products of fleeting configurations of competing causal forces

entering an unending permutation of local compromises with one another.
German and Japanese capitalism were never doneand this, not the limitations
of our empirical research, is the reason for our difficulties, as pointed out by
OSullivan, in dating them. As especially our second volume shows, what was
called the German model 20 years ago was a post-war synthesis that was already
coming apart at the time it was discovered, and much the same can be said of the
Japanese model (see Streeck, 2001, p. 36). Moreover, both systems differed
greatly from their respective predecessors in the 1930s, which were again different
from, say, what existed in the 1890s. Equally importantly, there still seem to have
been at each point in time differences between the two countries on the one hand
and the Anglo-American world on the other that by and large seemed to fit the
crude distinction between liberal and non-liberal that we adopted to organize
our material. Moreover, although it appears to have been generally impossible
to foresee in the two countries histories how their national models of capitalism
would develop in the future (Streeck, 2001, p. 30), with hindsight one finds
non-trivial elements of historical identity in spite of, and inside, a steady stream
of continuous and often disruptive change. These elements we have tentatively
characterized as the corporatist and the ie traditions, respectively (Streeck,
2001, p. 25), which over a long time seem to have provided the two societies
of our study with different and specific blueprints for the reorganization of
political-economic institutions in crisis.3 The presence of such continuities within
continuous change is one of the reasons for our expectation, with all due caution,
that Germany and Japan will come out of the present process of liberalization
remaining distinct from other countries, although this may be in ways that are
impossible to know beforehand.

Rejoinder

581

path-dependent characteristics of the historical process of institutional change.

OSullivan continues:

Gary Herrigel, in his comment, makes a similar point, although he tends

to identify institutionalism in general with what to OSullivan is only the functionalist subtype of institutionalism. The two books, Herrigel says, view the institutional systems in Japan and Germany as highly coherent, unitary systems of
interconnected and complementary institutional realms of governance. The
modelsof non-liberal capitalismthat such a perspective enables one to
construct may be ideologically useful. Nevertheless, their coherence
is ultimately an abstraction that blends out a great deal of anomalous
relationships, habits, dispositions and institutional practices in both
societies . . . It also creates the impression that such institutional systems exist on the ground as clear bright line rules that guide
behaviour. . . . The bedrock institutionalist commitments of the
authors and editors stand here unadorned. Their coherent models
of complementary institutions rely on the traditional institutionalist
idea that institutions impose constraints and enforceable obligations
on actors.
But this, according to Herrigel, is fundamentally misleading:
The actual systems in Germany and Japan . . . are much more incoherent, non-unitary and provisional than they are portrayed in either of
the volumes. Rather, they are composed of a patchwork of different
institutional solutions to a wide array of political economic problems.
The range of solutions work alongside on another not only (or
not even) in complementary ways, but also in relations of nonparalysing juxtaposition. Indeed, it is difficult, on the ground, to
identify a coherent, stable system of constraining rules in Japan and
Germany (or anywhere else for that matter).
I have quoted Herrigel so extensively for two reasons: to show how close
his self-perceived anti-institutionalism is to OSullivans historical institutionalism, and because I feel so sympathetic with the way he describes the world.

Downloaded from ser.oxfordjournals.org at University of North Carolina at Chapel Hill on October 17, 2010

From this (latter) perspective, institutions come together in some

kind of logical way through fortune rather than by design. As such,
the historical institutionalist strand in the books describes the process
of institutional change as much less orderly, more chaotic, than most
functionalists would assume it to be, at least if they were to seek to
explain the origins of the institutions whose functional characteristics
concern them.

582

W. Streeck

4
In addition to Hall and Soskice (2001), see Crouch and Streeck (1997), Aoki (2001) and Amable
(2003), among others.
5
For more on this see Streeck (2004a) and my contribution to the recent debate in this journal on
the concept of complementarity (Streeck, 2005).

Downloaded from ser.oxfordjournals.org at University of North Carolina at Chapel Hill on October 17, 2010

OSullivans advice at the end of her comment to play down the functionalist discussion and play up the historical institutionalism . . . relying heavily on concepts
such as path dependence, hybridization and bounded innovation I take to be
pointing in the same direction as Gary Herrigel. As far as I am concerned, this
is a direction in which all that the two commentators can do is crash into open
doorswhat we call in German offene Turen einrennen. My only reservation
with respect to Herrigel is that, in order to get rid of functionalist institutionalism,
he seems to favour discarding institutionalism as such. I also have a problem with
OSullivans endorsement of historical as opposed to functionalist institutionalism, which may however be less important than it appears at first glance.
I begin with OSullivan. My question to her is, in brief, how and to what extent
should functional or efficiency constraints be treated as causal factors shaping
institutions and their (national) configurations. On the basis of what I read,
I would expect her answer to be miles away from the economistic efficiency
theories of social institutions that are so popular in todays rational choice
theorizing. Needless to say I would endorse this very strongly. But in doing so
I want to avoid encouraging her altogether to abandon concepts like competitive advantage, functional coherence and complementarity, national economic
competitiveness, etc. that have inspired such rich and stimulating literature.4
To me this would be like throwing away the theoretical baby with the economistic
bath water. Instead we better combined our efforts to solve the tricky problem
of how to integrate economic concerns and pressures for economizing into a
non-functionalistic, non-economistic, i.e. historical and sociological theory
of institution-building and institutional changebasically what I understand
OSullivan calls historical institutionalism. To me this is the problem
par excellence of any institutionalist political economy, which is another way
of saying of any political economy with the ambition of dealing with the
real world.
How to go about this? At the least, I believe, theories of economic institutions
must dissociate themselves from the series of assumptions that more or less
implicitly underlie most variants of what may be called economistic functionalism. I limit myself to two of these.5 The first is that economic performance and
functional complementarity of institutions result from purposive design, either
from the top down by an interventionist state or from the bottom up by business
firms concerned about their competitiveness. Institutionalist theory, precisely

Rejoinder

583

Downloaded from ser.oxfordjournals.org at University of North Carolina at Chapel Hill on October 17, 2010

when it is dealing with economic institutions, must fully take on board the
realities of limited foresight and limited control, multiple and conflicting objectives, ambiguous external constraints and opportunities, and an unpredictably
changing task environment. Economically functioning institutions are constructed, if at all, under and out of such conditions. Both process and result of
institution-building reflect these conditions and cannot be accounted for without
extensive reference to them.
The other assumption that I believe one might want to avoid is that politicaleconomic institutions, or institutional configurations, are selected by a perfect
market which eliminates the inefficient and leaves the efficient. There are many
reasons why this would seem highly unrealistic. Institutions that are less efficient
than desirable do not normally disappear to be replaced wholesale by customdesigned new institutions. As a rule institutional settings change, not through
institutional breakdown, but through institutional change. Change, however,
proceeds only slowly as institutions are by their nature inert. Moreover, institutional environments are far more volatile and ambiguous than institutions can
possibly be flexible, which makes institutional change in pursuit of economic efficiency a highly risky pursuit of a rapidly moving target. Among other things,
this makes it advisable to preserve as many alternative options as possible.
Also, where real-world institutions compete with each other for comparative
advantage, they measure themselves, not against absolute standards of optimal
performance, but against the actual performance of similar historical, compromised and multi-functional entities like themselves. This leaves those who control
them a broad band of strategic choice between equally satisfactory (or unsatisfactory) functional equivalents. All this goes to suggest that the difference
between the mechanisms that drive institutional change and the market is more
than one of degree, rendering a market model of institutional change not just
unrealistic but fundamentally useless.
In what sense, then, may institutions be regarded as economically functional
in historical institutionalism, and what is it that makes them so? In our two
books we have repeatedly pointed to the importance of chanceof good luck
as opposed to merit or virtueas well as to the historically contingent character
of the challenges faced by institutional systems. Generally it would seem that in
reality, economizing comes in mostly ex post as actors seek to reconcile in a creative, Schumpeterian way the institutionalized givens of their social world with
economic interests and objectives, pragmatically trying to make the best out of
necessity or historical inevitability. I have outlined this perspective in my paper
on Beneficial Constraints (Streeck, 1997) and in a recent debate on that paper
with Wright and Tsakalatos in this journal (Streeck, 2004b). Ex-post economizing
by making-dofitting together from below what nobody was able to design
coherently from abovemay be precisely what happens in the interstices of

584

W. Streeck

Actors allow themselves to be constrained by rules when they believe

those rules solve problems. When they do not, creative actors coping
with uncertainty and guided by dispositions that are not reducible to
specific institutional arrangements either modify the rules or agree
simply to ignore them in order to construct new arrangements that
address more directly jointly identified problems.
To me this is going a step too far. Abandoning institutionalist functionalism is
one thing, abandoning institutionalism another. I strongly support Herrigels
effort to soften up institutional analysis and avoid the determinism that often
undermines its credibility, for example by making it impossible to conceive of
continuous institutional change in the absence of exogenous disruption. But
this cannot mean that social action must be conceptualized as completely voluntaristic. Together with Kathleen Thelen I have recently suggested a model of institutional change that emphasizes continuous interaction between rule-makers and
rule-takers puzzling and arguing over the meaning of institutionalized rules in
the diverse and changing conditions in which they are supposed to be applied
(Streeck and Thelen, 2005). I believe that this model does allow, as Herrigel
puts it in the above quote, for actors to modify the rules or agree simply to ignore
them. At the same time, it does not allow to forget that institutionalized rules are
protected by social sanctions that may be effectively applied in their defence.
While sanctions may fail, they may still be strong enough to prevent local actors
from, again in Herrigels words, construct[ing] new arrangements that address
more directly jointly identified problems. Could it be that foreclosure of
pragmatically attractive alternatives by institutionalized sanctions is more characteristic of the non-liberal capitalisms of Europe and Japan than of the US? On
the whole I doubt it.

3. The politics of liberalization in Germany and Japan

At the end of his comment T. J. Pempel expresses the view that our two books pay
too little attention to the political factors, and actors, who have in the past produced the similarities and differences between Germany and Japan and are today
governing the responses of the two countries to the new challenges of globalization. In particular, Pempel mentions the role of organized labour, the political

Downloaded from ser.oxfordjournals.org at University of North Carolina at Chapel Hill on October 17, 2010

functionally interdependent institutions built with distributed power and in

pursuit of particularistic sectoral interests.
Concerning Herrigel, my question to him is whether institutions, in the
way he uses the term, are at all constrainingwhich in Durkheimian language
would amount to asking whether his institutions actually are ones. Herrigel
writes:

Rejoinder

585

Downloaded from ser.oxfordjournals.org at University of North Carolina at Chapel Hill on October 17, 2010

forces behind the liberalizing reforms of today, and the changed role and
interests of the US in the global system. While I do not fully subscribe to his reading, I admit that we did and still do lack an adequate understanding of the present
politics of liberalization in non-liberal capitalism. Pempel himself points to the
contradiction that in social-democratic Germany it was the labour market that
had to bear the brunt of the crisis of the 1990s, whereas in Japans corporatism
without labour (Pempel and Tsunikawa, 1979) the crisis first hit the capital
markets. Presumably this was related to the fact that Japan still is a sovereign
nation with its own monetary policy, unlike Germany which was prevented by
European Monetary Union from using (or, depending on ones perspective,
abusing) fiscal and monetary policies to protect the employment of its industrial
working class. It also had to do with the fact that in Germany there was a public
welfare state ready and ablealthough, as it turned out, within strict limitsto
absorb the labour surplus generated by privatization and lean production, while
in Japan the stability of the national social compact depended on stable employment in that countrys large private firms.
Moreover, Pempel correctly observes that already in the 1970s the United
States was no longer willing to underwrite the post-war illusion of national economic sovereignty and social-political autonomy inside the Free World, as
instituted by the global regime of embedded liberalism. In the late 1990s the
US had begun aggressively to push for a new wave of international liberalization,
and especially for the opening up of the economies of its two main rivals,
Germany and Japan, to international markets and American firms. National governments also contributed, in Germany more than in Japan, by promoting internationalization in order to weaken domestic organized interests that had grown
too powerful for national economies to keep manageable. Globalization, which
is often simplistically described as the inevitable effect of improved means of
cross-border communication and the subsequent increase in cross-border ties,
was to an important extent a defensive reaction of governments who found
increasingly impossible the task of running the mixed economies they had inherited from the post-war era. Capital, tired of the costly national politics of entitlement that had emerged out of post-war democratization, supported this reaction
and often did its best to make it inevitable. Globalization, it seems, was and is
two things at the same time: a political contraption to restore sound money,
free markets and managerial prerogative, and also a structural force more or
less successfully constraining politics to move upwards to a new level above the
nation-state, with complex effects on the way in which the economy is embedded
in stabilizing social institutions or, for that matter, disembedded. Have nationstates lost control through global liberalization, or have they promoted global
liberalization to regain control? Is global governance tantamount to the political
and legal civilization of the nation-state from without, or to the de-civilization of

586

W. Streeck

the economy from within? As so often, the only thing we seem to know for certain
is more research is needed.

References

Aoki, M. (2001) Toward a Comparative Institutional Analysis, Cambridge and London,

The MIT Press.
Crouch, C. and Streeck, W. (eds) (1997) The Political Economy of Modern Capitalism:
Mapping Convergence and Diversity, London, Sage Publishers.
Dore, R. (1973) British Factory, Japanese Factory: The Origins of national Diversity in
Industrial Relations, Berkeley, CA, University of California Press.
Dore, R. (1987) Taking Japan Seriously: A Confucian Perspective on Leading Economic
Issues, Stanford, Stanford University Press.
Hall, P. and Soskice, D. (eds) (2001) Varieties of Capitalism: The Institutional Foundations
of Comparative Advantage, Oxford and New York, Oxford University Press.
Pempel, T. J. and Tsunekawa, K. (1979) Corporatism without Labor? The Japanese
Anomaly. In Schmitter, P. C. and Lehmbruch, G. (eds) Trends Toward Corporatist
Intermediation, London, Sage Publishers.
Sorge, A. (2005) The Global and the Local: Understanding the Dialectics of Business Systems,
Oxford, Oxford University press.
Streeck, W. (1997) Beneficial Constraints: On the Economic Limits of Rational
Voluntarism. In Hollingsworth, J. R. and Boyer, R. (eds) Contemporary Capitalism: The Embeddedness of Institutions, Cambridge, Cambridge University Press,
pp. 197219.
Streeck, W. (2001) Introduction: Explorations into the Origins of Nonliberal Capitalism
in Germany and Japan. In Streeck, W. and Yamamura, K. (eds) The Origins of
Nonliberal Capitalism: Germany and Japan in Comparison, Ithaca and London, Cornell
University Press, pp. 138.
Streeck, W. (2004a) Taking Uncertainty Seriously: Complementarity as a Moving Target,
sterreichische Nationalbank, 1(1), 10115.
Workshop Proceedings of the O
Streeck, W. (2004b) Educating Capitalists: A Rejoinder to Wright and Tsakalatos, SocioEconomic Review, 2(3), 42583.
Streeck, W. (2005) Requirements for a Useful Concept of Complementarity, SocioEconomic Review, 3(3), 36668.
Streeck, W. and Thelen, K. (2005) Introduction: Institutional Change in Advanced
Political Economies. In Streeck, W. and Thelen, K. (eds) Beyond Continuity:
Institutional Change in Advanced Political Economies, Oxford, Oxford University Press,
pp. 239.

Downloaded from ser.oxfordjournals.org at University of North Carolina at Chapel Hill on October 17, 2010

Amable, B. (2003) The Diversity of Modern Capitalism, Oxford, Oxford University

Press.

Rejoinder

587

Streeck, W. and Yamamura, K. (2003) Introduction: Convergence or Diversity?

Stability and Change in German and Japanese Capitalism. In Yamamura, K. and
Streeck, W. (eds) The End of Diversity? Prospects for German and Japanese Capitalism,
Ithaca and New York, Cornell University Press, pp. 150.
Weber, M. (1904) Die Objektivitat sozialwissenschaftlicher und sozialpolitischer
Erkenntnis, Archiv fur Sozialwissenschaft und Sozialpolitik, 19(1), 50374.
Downloaded from ser.oxfordjournals.org at University of North Carolina at Chapel Hill on October 17, 2010