. 8
( 11)


6.2 percent (Britain), and 11.6 percent (Euro area). Currently, the GDP growth rates
are 3.6 percent (U.S.), 1.7 percent (Britain), and 1.2 percent (Euro area) and the
unemployment rates are 5.0 percent (U.S.), 4.8 percent (Britain), and 8.7 percent
(Euro area). The continuing high unemployment rates in Italy (7.8 percent), Ger-
many (11.6 percent), and France (10.0 percent) suggests the EU countries face a
structural rather than temporary problem. These data are taken from the website
http://www.oecd.org/linklist. See also Chapter 2, Tables 2.2 and 2.3.

Architects of Political Change

the way in which the quandary of the late 1970s was understood, the
arguments of the prophets of chaos, with regard to political instability,
are also accepted. Thus collapse, a few years ago, in Russia, Indonesia,
Malaysia, and foreseen for Mexico, and other Latin American countries,
is generally regarded as a consequence of political “imperfections.” How-
ever, the dif¬culty that Japanese leaders face, in attempting to create the
conditions for economic growth, does pose more serious problems of
explanation.5 There may indeed be something peculiar about the con-
nections between Japanese politicians and corporate actors, effectively
forbidding resolution of what appears to be an economic crisis. Nonethe-
less, it is inconceivable that the cause of these disparate economic crises,
in so many parts of the world, is fundamentally political.6
This chapter argues that the strong form of the economic equilibrium
hypothesis is generally invalid. This strong hypothesis asserts that both
commodities and asset markets will, in fact, typically be in equilibrium,
and will result in outcomes that are Pareto optimal.7 I use the term com-
modities market for what Keynes referred to as a market governed by
prospective yield, or risk. This is the typical market studied in general
equilibrium theory. The individuals in such a market are characterized
by utility or preference; each individual can readily rank bundles of such
commodities by subjective worth.8
As discussed at greater length in Section 7.6, Keynes took pains to
distinguish the behavior of such commodities markets from what I term
asset markets, namely those governed by speculation. Keynes believed that
asset markets were characterized by uncertainty, rather than risk. The term
uncertainty is used for those situations where it is intrinsically impossible

During the crisis in late 1998, GDP “growth” rates were strongly negative: Japan

(’3.6 percent), Indonesia (’17.4 percent), Malaysia (’8.6 percent), South Korea
(’6.8 percent), and Russia (’9.9 percent), and Hong Kong (’5.2 percent). By August
2005, these economies had generally recovered: Indonesia (6.4 percent), Malaysia
(5.7 percent), South Korea (3.3 percent), and Russia (5.2 percent). Japan still has a
growth rate of only 1.3 percent. Again, the unemployment for Japan is 4.2 percent,
suggesting that the problem for Japan is structural. (These data are taken from the
website http://www.oecd.org/linklist.)
The great differences in political regimes among the affected countries clearly suggests

that economics rather than politics is the essential cause. Political inef¬ciencies could,
of course, exacerbate economic instabilities.
As usual, the outcome, x, is Pareto optimal if and only if there is no other feasible

state y that all economic agents weakly prefer to x.
It may be necessary to date these commodities by time; equilibrium then requires

complete future markets. Recent extension of the standard theory then can be used
to demonstrate existence of equilibrium, as long as there are complete future markets.

Keynes and the Atlantic Constitution

for any individual to assign probabilities to various eventualities, in any
coherent form.
To illustrate the notion of uncertainty, consider Russia™s choice to re-
nege on its interest payments due on its public debt in August 1998. The
event surprised even such an experienced international ¬nancier as George
Soros.9 The effects of the Russian decision induced uncertainty into the
international economy because the effects were due not simply to the eco-
nomic rami¬cations, but to the interpretation of this event by other actors
in the global economy. An immediate consequence was a considerable de-
cline in the U.S. stock market as fund managers ¬‚ed to “safer” havens.10
Although the stock market recovered, it started to fall from the high in
May 2000. After the election of Bush, and the uncertainty associated with
the events in the Middle East, the market fell further. At the time of writ-
ing (December, 2005), there is considerable uncertainty over the effects
of global economic and political change, not just in the Middle East, but
also in Latin America.
The scale of international asset markets has increased markedly in
recent years, and technological developments have dramatically reduced
transaction costs (and, of course, time of transaction). The consequence
is that the “virtual,” or speculative component, of world markets has
become increasingly important in comparison with the “real” components
of labor and of ¬‚ows of traded goods and services. I concur with Keynes™
argument that markets in commodities, especially traded goods, are likely
to be governed by what we regard as the law of supply and demand.
Such commodities markets may well exhibit equilibrium. What concerned
Keynes was the degree to which instability or chaos in asset markets

See the interview with Soros in the New York Review of Books (January 14, 1999).

Soros is reputed to have gained over 1 billion pounds sterling by betting 10 billion
pounds sterling that the British pound would be forced to devalue in 1992. This
1992 situation was one of risk only, since it was fairly clear that the overvaluation
of the pound could not be sustained.
Hamish McRae (in an article in The World Press Review in January, 1999) noted

the irony that the 1997 Nobelists in economics, Myron Scholes and Robert Mer-
ton, were involved in long-term capital management (LCM). The theoretic work
of Scholes and Merton was predicated on risk analysis and proved entirely useless
at preventing the loss of at least a billion dollars by LCM. McRae applauded the
Nobel Committee for recognizing, in 1998, Amartya Sen for his work in welfare
economics and social choice theory (Sen, 1970). It is obvious that the LCM debacle
occurred because market-based risk analyses proved utterly incompetent to deal
with situations of uncertainty. In contrast, much of Sen™s work has to do with the
degree to which political institutions may either mitigate, or possibly exacerbate,
economic uncertainty.

Architects of Political Change

could undermine the stability of commodities markets. Given the events
that had occurred in Keynes™ lifetime, his preoccupation was with effects
of this kind not only in the labor market (where the result is persistent
unemployment), but also in the international polity (and the attendant
competitive devaluations).
Keynes accepted this weak version of the equilibrium hypothesis (only
for commodities markets), because he saw a terrible danger to the Atlantic
Constitution. Consider for a moment a world in which it is generally
believed that all markets, both asset and commodity, are chaotic. In such a
world, the returns to capitalists and the wages of labor have no legitimate
basis. To escape this chaos, the citizens of a nation would be rational
in giving up their freedom to the agents of the state. Bound by such a
Hobbesian contract, the citizens could at least hope for some certainty
in their lives. It is true that the political economists of the 1930s, who
studied a socialist state of this kind, generally came to the conclusion
that the “calculation problem” made it impossible for the state to set
prices in such a way as to ensure economic ef¬ciency (see, e.g., von Mises,
1935: 87“130). Keynes was keenly aware that authoritarian state systems
could solve the problem of unemployment by paying the price of ef¬ciency
while necessarily depriving their citizens of their freedom. It seemed all
too probable in the 1930s that citizens would be willing to pay the double
price of inef¬ciency and loss of freedom to avoid the great and apparent
risk of unemployment.
The solution that Keynes sought to the quandary of the 1930s was
based on the logical distinction between the two kinds of markets. He
proposed limited government intervention only in the potentially unstable
asset market in order to create stability in the commodity market and the
return of full employment.
Lest we feel that Keynes™ concerns are of little interest in a world of
relatively low unemployment (at least in the developed nations), consider
again the rational calculus of citizens of Russia, or of the countries of
Southeast Asia, Latin America, or the Middle East. In Russia, in partic-
ular, although the so-called market was introduced with great fanfare it
has only led to ma¬a domination, currency collapse, government enfee-
blement, barter exchange, and varieties of unemployment and underem-
ployment. It would be no surprise at all if the majority of the Russian
people was to choose an authoritarian regime, under the plausible belief
that they were paying for release from chaos with their freedom.
As I discuss in later sections of this chapter, the success of the developed
countries of the OECD in solving the quandaries they faced at the end

Keynes and the Atlantic Constitution

of World War II and in the 1970s, has helped persuade the citizens of
other parts of the world that capitalism can work. More important, many
countries in Latin America and Asia have been willing to experiment with
democratic reforms. The uncertainty generated by recent events could
induce the leaders of such countries to renounce both free market notions
and democratic freedom.
To preserve the vitality of the Atlantic Constitution, it makes sense
therefore to probe, in some detail, the validity and relevance of the equi-
librium hypothesis in the more general context of global political econ-
omy. To do this, the next two sections of the chapter describes in greater
detail what is meant by a constitution. In particular, the notion of a core
belief is outlined, using, where appropriate, recent concepts from game
theory. To pursue the manner in which the core economic equilibrium be-
lief has evolved during this century, in Section 7.5 I consider the collapse
of U.S. hegemony in the 1970s. I then follow the notion back in time by
considering, in Sections 7.6 and 7.7, the quandaries that were apparent in
the 1930s and in 1944, over how to structure the post-war international
economy. Section 7.8 examines the debates in the U.S. and the ¬nal policy
decisions that led to a solution of this quandary. Section 7.9 attempts to
give some conclusions that are relevant for today.

7.3 prophets of chaos
For by Art is created that great Leviathan called a Common-Wealth, or
State . . . which is but an Arti¬ciall Man.
(Hobbes, 1960 [1651]: 81)

Those of us in the developed economies are all fortunate to have lived
in a post-1945 world within which no global wars have occurred and
economic growth has gathered apace. We sometimes forget that the eco-
nomic basis for the international component of the Atlantic Constitution
was laid down in a system of institutions, rules, and understandings de-
vised at the Bretton Woods Conference (of July 1944). The Bretton Woods
System was an “institution” in the sense used by Douglass North. Such an
institution can be understood in terms of “the rules of the game in a soci-
ety, or more formally the humanly devised constraints that shape human
interaction . . . [that] structure incentives in human exchange, whether po-
litical, social or economic” (North, 1990: 3).
The Smithsonian Agreement of December 1971 is usually taken to
signal the collapse of the Bretton Woods System. It would be more correct
to say that the institution of Bretton Woods had been found inadequate to

Architects of Political Change

deal with the complicated Hobbesian problem of maintaining cooperation
and rapid growth in an evolving global economy. During the 1970s, the
voices of the prophets of chaos were loud and generally heeded. Problems
of equilibrating the monetary system, and of the apparent incompatibility
of democracy and economic competitiveness, were increasingly perceived.
To a considerable extent, the solutions were sought in the arguments
of what I call the anarchic philosophers. The strategy of regulating the
domestic and international economy was increasingly seen as impossible,
indeed irrelevant. Instead of attempting to control exchange rates, the ¬‚ow
of capital, and so forth, the developed polities “distanced” themselves, to
a degree, from managing the global economy. The occasional interference
(as in Britain™s later attempt to maintain an over-valued exchange rate for
sterling in 1992“3) made it apparent that the costs of so doing, by a single
polity, could be enormous, and the effort self-defeating.
These changes in the institutional arrangements implied modi¬cations
in the Atlantic Constitution, in the implicit rules that governed economic
behavior at the international level. These were conditioned by changes
in beliefs, particularly about the behavior of markets. The notion of a
natural equilibrium in an international system of exchange economies was
generally accepted, and the possibility of economic chaos was seen as less
credible. This change in belief was entirely rational (indeed “Bayesian”).
From an empirical standpoint, a lessening of the degree of intervention by
government in the macroeconomy, after 1982 or so, certainly appeared to
produce a drop in overall in¬‚ation. The arguments that had been made
during the 1970s by political economists seemed to be borne out. By
changing the rules of political action, by credibly committing politicians
to a less interventionist role, it was possible to reduce in¬‚ation to negligible
levels. In essence, these changes in the “rules” were induced by changes
in beliefs over the validity of the Keynesian framework. It is important
to note, however, that these changes in beliefs were not just based on
empirical analysis. They were substantiated by plausible models, both of
the economy and of the polity. Indeed these models became part of the
accepted language of discourse.
This suggests that what I have called the “Atlantic Constitution” is
not just a set of rules of behavior, nor is it simply a set of “rules of the
game,” as North suggests. It is rooted in beliefs, and these beliefs have
both empirical and theoretical bases. When the “rules of the game” of the
constitution produce effects that are intolerable, then the philosophers,
the prophets of chaos, become engaged to seek both the reasons and cures.
The beliefs that underpin the constitution necessarily evolve, in a way that

Keynes and the Atlantic Constitution

might be termed “Popperian” (Popper, 1957). In fact, Popper (1972) has
asked the question: “How [can] such non-physical things, as purposes,
deliberations, plans, decisions, theories, intentions, and values . . . play a
part in bringing about physical changes in the physical world?”
That such non-physical things are part of the constitution suggests that
the constitution has embedded within it a formal language of discourse.
Some of the anarchic philosophers follow Hume in regarding language
simply as a system of conventions. As Sugden (1980) says, “We all wish to
speak and write in a form that is comprehensible to the people with whom
we wish to communicate, and so there is a self-reinforcing tendency for
communities to evolve common languages.”
From this point of view, a language as convention is a particular equi-
librium solution to a coordination game. But a language is clearly more
than this. It has a grammatical structure that is internally consistent, to
some degree.
In my view, the constitution does provide a language through which
citizens, corporations, and countries can communicate. But there is a fur-
ther element of this constitutional language which should be emphasized.
As I wrote a number of years ago,

[T]he fundamental theoretical problem underlying the question of cooperation is
the manner by which individuals attain knowledge of each others™ preferences and
likely behavior. Moreover, the problem is one of common knowledge, since each
individual, i, is required not only to have information about others™ preferences,
but also to know that the others have knowledge about i™s own preferences and
(Scho¬eld, 1985b)

This constitutional language aims not only at internal consistency, and
empirical relevance, but also provides the framework within which the
acts of agents (whether citizens, corporations, or countries) are mutually
intelligible. That is, it makes it possible, at least potentially, for us to
understand what it is we, and others, are doing.
These remarks lead us to a conception of the constitution that is much
more general than that of an institution. Clearly, the “rules of the game,”
since they must be regarded as humanly devised constraints, are open
to human choice. Because the rules can be socially chosen, the internal
consistency and stability of the constitution depends on the degree to
which it is in equilibrium with respect to the beliefs of its citizens. At
certain times in the evolution of the constitution, there exists one, or a
number of core beliefs that are stable or persist in the population. These

Architects of Political Change

provide the basis for the consistency or coherence of the constitutional
However, the game theorists, who have studied the prisoners™ dilemma
model of cooperation and war, have realized that the evolution of such
Hobbesian societies depends almost entirely on the way beliefs are created
and destroyed.11 In contrast to the earlier work of the “neo-institutional”
political economists, who tended to see political or social equilibrium
resulting directly from institutional constraints,12 the more recent per-
spective has emphasized the way in which equilibria may be destroyed or
recreated by a belief cascade.
Denzau and North (1994) mention two different modes of belief cas-
cades that have been discussed in entirely different contexts. First there
is the situation studied by Bikhchandani, Hirschleifer, and Welch (1992)
where a small group of decision makers “change their minds” on the
basis of their private information, inducing the rest of the population to
“free ride” by following suit. Second, there is the notion of a scienti¬c
revolution destroying an existing paradigm (Kuhn, 1962).
Chapter 8 presents a more detailed examination of versions of these
two modes of belief cascades. To illustrate the ¬rst type, suppose individu-
als hold two hypotheses, say “the market is a bull” as against “the market
is a bear,” but assign a higher probability to the former. Obviously enough,
this is a self-ful¬lling social belief. If, however, in a domain of uncertainty,
market leaders act in a way that is consistent with their increasing belief
in the latter hypothesis, then a belief cascade may ensue. Whether or not it
will depends on the common knowledge foundation of the collective be-
liefs: Is it the case that the acts of the market leaders are clearly intelligible
to the market followers? Recent results in game theory suggest that the
common knowledge foundation, the basis of intelligibility, may, in fact,
collapse near the onset of the cascade. Another way of expressing this is
by the observation that the outcome, as well as the onset, of the belief
cascade may very well be unknowable in principle. It is this argument
which sustains the claim that asset markets are potentially chaotic.13

There is enormous literature on this topic. For a very useful survey, see Denzau and

North (1994).
For a discussion of these perspectives, see Calvert (1995).

The argument on unintelligibility is based on a Godel-Turning paradox, that each

agent, i, in interpreting the market must assume that the other agents in the market
are different in kind from i. When this realization hits, as it must in general, then
the basis for rational choice by i is destroyed. The onset of this “irrationality” by

Keynes and the Atlantic Constitution

Chapter 8 also considers an important example of a Kuhnian paradigm
shift, namely the transformations in physics that occurred after 1905,
when Einstein “banished” the core scienti¬c belief in the absolute exis-
tence of time and space. This led to the shattering of the subdisciplines of
physics, and the creation of entirely new disciplines, including quantum
mechanics, among others. Indeed, physics during this century has thrown
up numerous profound quandaries, not the least of which is the problem
of the integration of gravitation with quantum theory. The point about
the belief in absolute space and time was that it created barriers to thought
that hindered the development of our understanding of the world.
I suggest that the economic equilibrium hypothesis has been a core
belief of the Atlantic Constitution in much the same way that the be-
lief in absolute space and time was at the core of nineteenth century
physics. Keynes himself found it extremely dif¬cult to overcome the bar-
rier presented by the “Marshallian orthodoxy” of a belief in the strong
equilibrium hypothesis.
A total rejection of the validity of the equilibrium thesis, for both com-
modity as well as asset markets, calls into question the fundamental po-
litical beliefs that underpin the Atlantic Constitution. For this reason,
policy makers have retained this belief in one form or another. This does
not mean to say that the belief has remained unchanged throughout the
entire period. Section 7.7 suggests that, in attempting to solve the consti-
tutional quandary of 1944, the architects of change of that time adopted
some of Keynes™ insights, particularly with regard to the problem of coop-
eration in the international economic arena. A complex belief about the
compatibility of the Bretton Woods international institution and a Keyne-
sian macroeconomic equilibrium thesis became generally accepted during
the 1950s and 1960s. In turn, the events of the 1970s persuaded ¬rst the
prophets of chaos of that time, and then the international policy makers,
that the only resolution to their particular quandary was a return to the
economic belief based on the strong form of the equilibrium hypothesis.14

market leaders cannot be determined beforehand. When it does occur, it may well
trigger a belief cascade into market instability. It is apparent that Keynes™ intuition
on the collapse of market bubbles was similar, though not, of course, framed in
precisely these game theoretic terms. Chapter 8 discusses this further.
This chapter only sketches the nature of the constitutional quandaries of 1944 and

of the 1970s. A point to note is that the solution of the 1944 quandary was still
rooted in an equilibrium thesis but also involved certain beliefs about the nature of
a bipolar world split between communism and capitalism.

Architects of Political Change

The theoretical arguments just made suggest that the form of the res-
olution of these two quandaries could not, in principle, be determined
beforehand, precisely because belief cascades were involved in both cases.
The same holds for the current quandary. However, examination of such
quandaries does permit a clearer focus on the underlying core beliefs pre-
ceding the onset of the cascade.
It should be remarked that the use of the word core in core belief is
meant to designate a generally accepted principle. When a quandary is
universally perceived, then there is no consensus in belief,15 and the com-
mon knowledge foundation of choice and action dissipates. Even so, what
beliefs there are in the collectivity will still induce barriers to thought and
to behavior. As the core beliefs fragment, it is not implausible that acts
become mutually unintelligible, indeed “irrational.” This phenomenon
can be observed in some of the laboratory experiments based on the
prisoners™ dilemma (Richards, 1990). Indeed the “resource wars” and
aggressive currency devaluations of the 1970s may have had a similar
Social choice theory indicates that the absence of a core belief may in-
duce cyclical or apparently random behavior. The set of all such behavior
is termed the heart. Thus the heart of the Atlantic Constitution, at any
time, is the set of collective actions that may be “rationally” entertained
by the citizens or organizations within the community. The social choice
framework underpinning this book suggests that as core beliefs fragment
at the onset of a quandary, then the heart of the constitution expands to
include behavior that was previously inconceivable.16 The next section
brie¬‚y discusses the extent to which the Atlantic Constitution does truly
face a quandary at present, and offers some observations as to possible
consequences if core beliefs do not adapt to the situation before us.

7.4 political and economic beliefs
in the constitution

This is the biggest ¬nancial challenge facing the world in half a century, and the
United States has an absolutely inescapable obligation to lead and lead in a way
that™s consistent with our values[.]

Throughout this book, the use of the word core in core belief is adapted from social

choice usage. When there is no consensus, then the core is empty, and the core belief
is void.
To speculate, it is possible that the irrationality of World War II had at its source

the collapse of the core beliefs during the 1920s and 1930s.
Keynes and the Atlantic Constitution

We know that our future prosperity depends on whether we can, with others,
restore con¬dence [and] stabilize the ¬nancial system. We will urge the major
industrial economies to stand ready to use the $15 billion I.M.F. emergency
funds to help stop the ¬nancial contagion from spreading to Latin America and
elsewhere. . . .
Today I have asked Secretary Rubin and Federal Reserve Board Chairman
Greenspan to convene a major meeting of their counterparts within the next 30
days to recommend ways to adapt the international ¬nancial architecture to the
21st Century.
(President Bill Clinton, address to the Council on Foreign Relations,
New York, September 15, 1998)
As I have already suggested, the transformations in the Atlantic Con-
stitution after the late 1970s had shifted our understanding of the balance
between politics and economics. Prior to about 1970, a Keynesian under-
standing of the necessity for intervention in the economy was compatible
with a general view of politicians as social welfare maximizers. One could
say that polities were “simple” and economies “complex.” By the 1980s,
theory and reality had suggested that politicians were complicated agents,
potentially predatory and requiring constant vigilance. Models of politics
necessarily become increasingly complex. In contrast, the general accep-
tance of the belief in an equilibrium tendency of economies meant that
the science of economics became simpler.
As noted in the previous section, there is a theoretical problem with the
economic equilibrium belief. The theorem on which the belief is grounded
deals with commodities whose value for each individual is well de¬ned.
The very success of the transformed constitution since 1982 has led to a
vastly increased international ¬‚ow of monetary assets, and to enormously
enhanced activity on the world™s stock exchanges. These ¬‚ows are derived
not from the intrinsic value (utility) of assets, but from beliefs of agents
about the future value of the assets. To fully understand such asset mar-
kets, it is necessary to model the behavior of the agents involved in the
exchange, and this means understanding their beliefs. As Arrow (1986)
has noted, this means solving an in¬nitely complex common knowledge
problem.17 In the fashion described above, belief cascades can destroy
economic equilibrium within such markets. In practical terms, a cascade
from one equilibrium to another could involve a collapse ¬rst of the Mex-
ican peso, then the Southeast Asian currencies and stock markets, then the
Japanese economy, and then the Russian. There is no theoretical reason
at all to suppose that such events are impossible. In fact, formal analyses

Another way of seeing the problem is in terms of the incompleteness of future

Architects of Political Change

of such belief cascades have suggested that it is impossible to know how,
or by what, the cascade is triggered. It is for this reason that such cascades
have been called “chaotic.”18
The current predisposition to view polities as complex, and economies
as simple has guided commentators on the current chaotic events to focus
on “politics.” It is easy to point to nepotism in Indonesia, to senility in
the Liberal Democratic Party in Japan, and to autocrats and the ma¬a in
Russia. It is dif¬cult, however, to see why such political incapacity has
only now made itself felt.
The monetary institutions of the Atlantic Constitution have deployed
their resources of many tens of billions of dollars in the last few years
to assist the ¬‚oundering economies of Russia and Southeast Asia. It is
possible that the Latin American economies will follow the earlier collapse
in Mexico and that the resources of the World Bank and IMF will be
completely overwhelmed.
The point to be made, of course, is that these cascades have occurred
despite the strongly held belief, implicit within the current Atlantic Con-
stitution, that the global economy can generally be left to fend for itself.
I consider this belief to be unfounded for the reasons already mentioned.
At the same time, the belief (held strongly prior to 1965) that the interna-
tional monetary institutions can control the global economy appears just
as unfounded.19
If this quandary is not solved, certain consequences for the structure of
the Atlantic Constitution appear inevitable. The pronounced triumphal-
ism of the last decade will surely disappear, if events play out the way
suggested. Pessimism will again rule. The earlier pessimisms of the inter-
war era and of the decade of the oil crisis had given way, by 1989, to

It is not so much that such belief cascades are sensitive to initial conditions (the

usual de¬nition of chaos). Rather, the resulting behavior can “go anywhere.”
Attempts to regulate any market by controlling price ¬‚uctuations necessarily re-

quires a reserve fund of a scale that is related to (1) the volume of the potential
speculative assets ¬‚ow, (2) the degree of correlation between the individual com-
modity markets, and (3) the inverse of the chosen price band. Although there may
well be welfare bene¬ts from reducing risk through such restriction of volatility,
these should be set against the cost of maintaining the reserve fund. For a discussion
of the theory as applied to commodity markets, see Newberry and Stiglitz (1981).
As section 7.5 mentions, attempts to regulate the international monetary system
and the world commodity markets (under the auspices of the UNCTAD integrated
commodities program) both failed in the decade after the Smithsonian Agreement.
The reasons fairly obviously were to do with the scale of speculative ¬‚ows and the
extent of the correlation. This does not mean to say that it is impossible to regulate,
but only in a much weaker sense than these earlier experiments.

Keynes and the Atlantic Constitution

the optimistic view that liberal democracy may constitute the “endpoint
of mankind™s ideological evolution and the ¬nal form of human govern-
ment” (Fukuyama, 1992).
This “triumphalism” was based on the fading away both of commu-
nism in Russia and Eastern Europe and of autocratic regimes in Latin
America and Asia. The reason for their decline and disappearance clearly
seems to have been the effect that the success of the Atlantic Constitution
had on the beliefs of populations in these countries. On the one hand, if
the Atlantic Coalition is manifestly unable to solve its economic dilemmas,
then a citizen of a predatory regime will be willing to accept the depreda-
tions of that regime. On the other hand, if the risk of economic collapse
under the Atlantic Constitution falls relative to the obvious bene¬ts of
citizen and property rights, then the depredations become unbearable.
Cascades of citizen unrest can then overthrow the autocracies. But liberal
democracies can also fall. If beliefs in the likelihood of chaos increases,
then politicians, in their pursuit of power, will rationally rise up to insti-
tute predatory autocracies, and citizens may well fear the relative cost of
The events in Russia in late 1998 give us all cause to fear that this fall
into autocracy could be unbelievably swift. An even greater fear is that the
expected amelioration of the regime in China may very well be reversed.
In the developed countries of the Atlantic Coalition, the belief in the
legitimacy of the market could be eroded equally rapidly. It is not sur-
prising that those countries, particularly Britain and the United States,
where the economic equilibrium belief has been most readily accepted,
are also those at present experiencing low unemployment, and, perhaps,
an increasing degree of wealth inequality. When this equilibrium belief
was less potent, both Britain, most obviously, but also the United States,
to some degree, were beset by labor unrest.
In many of the countries of the European Union, the consequences
of the economic equilibrium hypothesis have met with substantial polit-
ical resistance. It seems reasonable to suppose that the effect has been to
weaken the desire or will to impose restrictions on government interven-
tion in the economy.21

It is clear how to model this formalism. I need only assume that citizens and politi-

cians are rational, but with different motivations. The essence of the model is given
in Weingast (1997a, b). I have simply added the idea of risk to the cost calculation
of the citizen.
I have argued elsewhere that the interrelationship between beliefs and constraints on

government is mediated by the electoral system. Thus, in electoral systems based on

Architects of Political Change

Obviously, beliefs about the appropriate relationship between politics
and economics do indeed differ, as I have suggested, among the United
States, Britain, and the European polities. This is entirely consistent with
the notion of an Atlantic Constitution, since I assume only that the broad
pattern of the beliefs will still be compatible. However, my understanding
of the constitution as based on a coherent language suggests that if the
fundamental beliefs start to diverge, then policy coordination between the
countries will become increasingly dif¬cult.
Suppose the degree of perceived risk in the global economy increases,
while attempts to solve the quandary fail. It is obviously impossible to
read the future, but it would seem all too plausible that the European
Union, in implementing monetary union, the Euro, between twelve of its
members, will also become increasingly protectionist. I have no proof of
this assertion, but my colleague, Andrew Sobel, has kindly brought to my
attention the comments of the French Prime Minister, Edouard Ballader,
in 1993:

Can we [West Europeans] take it for granted that we will remain suf¬cient leaders
in suf¬cient numbers of sectors to survive in the face of countries with populations
in¬nitely larger than ours and with levels of social protection in¬nitely smaller? I
say we should leave this to the market, but only up to a certain point. What is the
market? It is the law of the jungle, the law of nature. And what is civilization? It
is the struggle against nature.22

If the coherence of the core political and economic beliefs is called into
question to such a degree that countries fall into autocracy or protection-
ism, then the consequences would be profoundly unpleasant. It seems to
me that this is not simply an empirical policy matter. It is not just a question
of juggling the ¬‚ows from the monetary institutions to the countries at risk
to enable them to overcome a temporary economic dif¬culty. It is a ques-
tion of probing the core beliefs of the constitution more vigorously than
we have chosen to do since the onset of the last quandary in the late 1970s.
As I have noted, the earlier crisis originated in the fall of the Bretton
Woods system. The later sections of this chapter use the notion of the
Atlantic Constitution in an attempt to gain a better understanding of
the proper balance between politics and economics. To do this I detail my

proportional representation, parties involved in coalition government are unwilling,
as it were, to increase the risk their voters face if labor markets are made more
competitive. Majoritarian systems, such as Britain, appear to have been more risk
preferring since 1980. See Chapter 2.
Quoted in Sobel (1998).

Keynes and the Atlantic Constitution

view of the nature of the problem, as perceived by the architects of change
circa 1944 and circa 1980. Their beliefs were conditioned by the empirical
reality that they had already experienced, and by the theoretical apparatus
that they had at their disposal. I see these architects as concerned actors of
political and economic reason, fully aware of the dangers to the Atlantic
Constitution, and determined to protect it and the pattern of citizen rights
and responsibilities that had been created over many centuries.
Whereas Popper recommends piece-meal engineering over utopian vi-
sions, it is obvious that, at times, the stakes can be so high that vigorous
social “experimentation” is rational. By focusing on the quandaries as
perceived in 1944 and in 1980 and the solutions that were devised, we
may better understand the recent evolution of the Atlantic Constitution.23
To do so, we must ¬rst recreate the fears, as presented by the prophets of
chaos, in terms of the theoretical and empirical framework within which
they were expressed. Second, I discuss, in each case, how the framework
was reinterpreted by architects of change, so that a solution could be
constructed, through transformations in the fundamental core beliefs un-
derlying the Atlantic Constitution.24

7.5 the collapse of hegemony in the 1970s
Hegemon (from ·γ μιν, Greek) leader; hegemony, preponderance, especially of
one state of a confederacy or union, over others. (O.E.D.)

As far as I am aware, Kindleberger (1973) gave the ¬rst interpretation
of the international economic system of states as a “Hobbesian” prisoners™
dilemma, which could be solved by a leader, or “hegemon.”

A symmetric system with rules for counterbalancing, such as the gold standard
is supposed to provide, may give way to a system with each participant seek-
ing to maximize its short-term gain. . . . But a world of a few actors (countries)
is not like [the competitive system envisaged by Adam Smith]. . . . In advancing
its own economic good by a tariff, currency depreciation, or foreign exchange
control, a country may worsen the welfare of its partners by more than its gain.

Implicit in my understanding is that the Atlantic Constitution evolves through

“punctuated equilibria.” See Eldredge and Gould (1972). For a general argument
to this effect, see Chapter 8.
This kind of analytical narrative, focusing on deadly quandaries within the Atlantic

Constitution, has been discussed in the earlier chapters. See also Rakove, Rutten, and
Weingast (1999) for the period of 1760“1776, prior to the American Revolution, and
the essays in Analytical Narratives (1998) by Robert Bates, Avner Greif, Margaret
Levi, Jean-Laurent Rosenthal, and Barry Weingast.

Architects of Political Change

Beggar-thy-neighbor tactics may lead to retaliation so that each century ends up
in a worse position from having pursued its own gain. . . .
This is a typical non-zero sum game, in which any player undertaking to adopt
a long range solution by itself will ¬nd other countries taking advantage of it. . . .
In these circumstances, the international economic and monetary system needs
leadership, a country that is prepared, consciously or unconsciously, under some
system of rules that it has internalized, to set standards of conduct for other coun-
tries and to seek to get others to follow them. . . . Britain performed this role in the
century to 1913; the United States in the period after the Second World War until,
say . . . 1963. . . . [P]art of the reason for the length of . . . the world depression
was the inability of the British to continue their role of underwriter . . . and the
reluctance of the U.S. to take it on until 1936.25

In the 1970s, Robert Keohane and Joseph Nye (1977) rejected “realist”
theory in international politics, and made use of the idea of a hegemonic
power in a context of “complex interdependence” of the kind envisaged
by Kindleberger. Although they did not refer to the formalism of the
prisoners™ dilemma, it would appear that this notion does capture elements
of complex interdependence. To some extent, their concept of a hegemon
is taken from realist theory rather than deriving from the game-theoretic
However, it is very easy to adapt the notion of a symmetric prisoners™
dilemma, so as to clarify the concept of a hegemon. A non-symmetric
n-agent prisoners™ dilemma (nPD) can be constructed as follows. Let di
∈[0, 1] be the strategy of the i th country (di = 0 means defect, di = 1
means cooperate). Each country was a weight (proportional to its GDP),
ai say. The total collective good of the system, N, of states is
B(N) = ai di . (7.1)

The payoff ui to state i, when it adopts strategy di is
ui (di ) = B(N) ’ di . (7.2)
n i=1

For a prisoners™ dilemma, 1 < r < n. The term in di is
(ai di ) ’ di . (7.3)

Kindleberger, 1973: 11. In the next section, I discuss earlier remarks in this vein by

Keynes in 1936.

Keynes and the Atlantic Constitution

Clearly if
(ai ) < 1, (7.4)
then ui is maximized at di = 0. If
(ai ) > 1, (7.5)
then ui is maximized at di = 1.
In the symmetric game, ai = 1 for all i, so the “rational” strategy for
each country is to defect, by choosing di = 0. If
ai > , (7.6)
then this country, i, rationally must cooperate, irrespective of the strategies
of other countries. To keep things simple, suppose a j = 1 for all j other
than this hegemon, i. In this very trivial formulation, some things are
obvious. If more states join the game (so n increases), it becomes more
“dif¬cult” for ai to be large enough for cooperation. The coef¬cient r is
the “rate of return on cooperation.” As r falls it becomes more dif¬cult
again for i to remain the cooperative hegemon. In this formulation the
term hegemon is something of a misnomer, since i is simply a rational
cooperator. However, if coalitions are permitted and a hegemonic power
leads a coalition M of states, dictating policy to these states, then the
optimality condition for the joint cooperation of the states in the coalition
M is
ai > . (7.7)

The collective bene¬ts of the coalition M can then be redistributed by
the hegemon in some way, to keep the coalition intact.
A number of years ago, I analyzed a prisoners™ dilemma model of
this kind and observed that it could be used to understand international
economic cooperation. The conclusion was pessimistic.

In the postwar years we have seen the development of a dominant cooperative
coalition: the Atlantic Community. At the core of this cooperative coalition was
the United States; through its size it was able to generate collective goods for
this community, ¬rst of all through the Marshall Plan and then in the context
of the Defense Alliance. . . . Since the sterling devaluation of 1967 we have seen
intermittent stop-go or in¬‚ation-de¬‚ation strategies by many of the developed

Architects of Political Change

To simplify, we may regard such strategies as bargaining strategies of the type
considered here. These strategies necessitated coalitional readjustments or agree-
ments over distribution, such as the Smithsonian Agreement of 1971. . . . In a
sense the United States has found it costly to be the dominant core of the coali-
tion . . . the Atlantic coalition may be in the process of fragmentation, with its
individual members oriented to individually rational, or “beggar-my-neighbor”

Obviously these comments were made at a time when the size of the
U.S. economy had declined relative to the overall GDP of the OECD.
Following the logic of the prisoners™ dilemma, it was apparent that co-
operation within the Atlantic Community would become more dif¬cult.
Throughout the late 1960s, economic behavior had indeed become in-
creasingly chaotic. The problem had, of course, been noted early on by
Trif¬n (1960), who asked how stability could be maintained by a constant
¬‚ow of dollar assets out from the United States. The round of devalua-
tions, including the sterling devaluation of November 1976 and of the
French franc in August 1969, made it clear that the Bretton Woods sys-
tem could not be retained. The monetary instability of that period may
also have created the context for the “political business cycle.” Coincident
elections in many of the OECD countries induced incumbent governments
to simultaneous re¬‚ation. Average in¬‚ation moved up from about 4 per-
cent to nearly 8 percent in two years in the early 1970s (McCracken,
1977); the rest of the decade of the 1970s was taken up with the two
oil crises, with the arguments in the context of UNCTAD over changing
the relationship between developed and less developed countries, and (by
the early 1980s) with the third world debt crisis.27 Although the earlier
analysis was based on the one-shot n-person prisoners™ dilemma (nPD), it
is possible to pursue the Hobbesian metaphor in the context of an iterated
nPD. Analyses of such games suggest that they are chaotic, that anything
can, in principle, occur.28 In other words, while the one-shot nPD will tend
to fall into Hobbesian war when there is no hegemon, in the iterated ver-
sion cooperative coalitions may rise and fall in an indeterminate fashion.
The formal problem presented by the iterated nPD (with n agents)
is that the strategy space is extremely complex. Although the results of

Scho¬eld (1975). Dynamic aspects of such a game were considered in Scho¬eld

(1977a, b).
These crises have been discussed by many authors. An overall view of their interac-

tions is given in Scho¬eld (1984b).
This is my interpretation of the so-called folk theorem for in¬nitely repeated games.

See Fudenberg and Tirole (1992).

Keynes and the Atlantic Constitution

Axelrod (1984) are generally seen as providing an avenue of escape from
defection into cooperation, via tit-for-tat strategies, his work was only
concerned with two agent models. With many countries participating,
strategies may well involve complicated history-dependant punishments
against other countries, or groups of countries. Keohane (1984), in his
more recent work, has attempted to provide an explanation for interna-
tional cooperation in the post-hegemonic world of the 1980s. In essence,
his argument is that a new regime came into being, based on new expecta-
tions and mutual assumptions. A formal interpretation of his argument is
that a convention”a self-reinforcing “Nash” equilibrium in the iterated
nPD”was created. But what exactly was this new regime, this equilib-
rium convention?
As I have intimated previously, an international economic regime is
more than a convention. It is a system of rules and modes of meaning,
a language, embedded in a framework”a constitution”that is intelligi-
ble to the participants. Most importantly, it must provide the common-
knowledge background so the agents know the game and can interpret
the actions of others. I have argued in the previous sections that the shift
in the constitution circa 1980 involved an acceptance of the economic
equilibrium thesis.
By the end of the 1970s, it had become obvious to all that global capital
was capable of swamping the resources of any individual country, even
the United States. Already by 1973, the Long Report29 had estimated
that private Eurodollar assets were of the order of 270 billion dollars (in
current terms over 1 trillion dollars). By 1983, the capital requirements
of the developing nations were being met principally by private capital
sources. It is true that this had precipitated the so-called debt crisis.30
However, by the end of the 1980s this crisis had been accommodated to
some degree.
Given that no individual country could immunize itself against inter-
national capital, it made sense to accept the fact and consider strategies
that were consistent with the logic of the market. As I observed earlier,
this is not to say that all countries adopted identical strategies. Differences
in electoral and political patterns in the developed countries resulted in
varied microeconomic strategies in labor markets, and capital markets

The U.S. Government Printing Of¬ce: The Long Report to the U.S. Senate (1973).

For example, by 1983 it was estimated that the Latin American countries owed

approximately 300 billion dollars. Debt service at that time exceeded earnings in
many of these countries.

Architects of Political Change

responded in obvious ways. For third world countries in particular, polit-
ical chaos or internal strategies of excessive manipulation led to increases
in perceived risk, and their capital ¬‚ow declined (Sobel, 1999). Accep-
tance of the logic of the market thus imposed a degree of discipline on
both developed and less developed nations.
The strategic domain available to countries contracted. While the game
still probably retained the fundamental characteristics of the nPD, the
“externalities” (the effect of the strategy of one country or another) di-
minished. Although the U.S. trade de¬cit remained a problem, it was no
longer seen as a fundamental disequilibrating feature of the global econ-
omy, but as a topic to be tackled by bilateral negotiation, usually with
Japan, and later with China.
This modi¬cation to the Atlantic Constitution was put in place, princi-
pally by Reagan and Thatcher, in the early 1980s. These changed beliefs
were accepted by decision makers in the international economy, since
they made both logical and empirical sense.31 The arguments of those
prophets of chaos, who had asserted the fundamental incompatibility
of democracy and economic rationality, appeared unfounded. As I have
indicated, there were bene¬cial unanticipated consequences, including
the democratization of regimes in Latin America and in the old Soviet
Even more surprisingly, the Cassandras who had foreseen the increas-
ingly rapid decline of hegemonic U.S. power, and the recurrence of eco-
nomic “Warre”32 were confounded by the obvious economic vitality both
of the United States and Britain. However, the coherence of the core belief
that was created after 1982 or so depended crucially for its validity on the
economic equilibrium thesis. Although I have argued that there are the-
oretical reasons to reject the strong equilibrium hypothesis, there could,
nonetheless, be reasons to accept its empirical validity. To see whether the
thesis has universal validity, it makes sense to look again at the events
of the interwar period. In the following two sections of the chapter, I in-
terpret these events in terms of Keynes™ understanding of the quandary
implicit within the Atlantic Constitution in the period before and during
World War II.

I have not found a complete account of this belief cascade, but Yergin and Stanis-

law (1998) do provide a readable account. Although the book was written before
the “crisis” of 1998, they do mention in their conclusion the consequences of any
disruption of the international ¬nancial system.
See Kennedy (1987) as well as Thurow (1992, 1996) and Chapter 2.

Keynes and the Atlantic Constitution

7.6 keynes and the quandary of the 1930s
Chaos, (from χ ±oσ, Greek), void, confusion.

The Treaty of Versailles was signed on June 28, 1919. Keynes had
been involved in the negotiations and was appalled by the reparations
required of Germany by the Allies. He calculated that at best Germany
could pay 1500 million pounds sterling over thirty years. As he wrote,
in his Economic Consequences of the Peace (1919), “The fact that we
have no adequate knowledge of Germany™s capacity to pay over a long
period of years is no justi¬cation . . . for the statement that she can pay
ten thousand million pounds.”33
Keynes also outlined his plan for a peace treaty involving “limited repa-
rations as well as cancellation of inter-ally debts; creation of a European
free trade area; . . . an international loan to stabilize the exchanges; and
encouragement of Germany™s natural organizing role in Eastern Europe”
(Skidelsky, 1983: 391).
Without these remedies,

Nothing can then delay for long that ¬nal civil war between the forces of reaction
and the despairing convulsions of revolutions, before which the horrors of the late
German war will fade into nothing, and which will destroy, whoever is victor, the
civilization and progress of our generation.34

There was a distinct Malthusian tone to much of Keynes™ book.

The feeding of the peoples of Central Europe is the fundamental problem in
front of us. . . . Some of the catastrophes of past history, which have thrown back
human progress for centuries, have been due to the reactions following on the
sudden termination of temporarily favourable conditions which have permitted
the growth of populations beyond what could be provided for.35

Kindleberger (1973) suggests that reparations “may not have been di-
rectly responsible for the depression, but together with war debts they
complicated and corrupted the international economy at every stage of
the 1920s and during the depression through to 1933” (23).
Just to indicate the amounts involved, France and Britain owed the
United States 8.7 billion dollars (which should be scaled up by a factor of
more than 30 to convert into current dollars). Interallied debt (including

Quoted in Skidelsky (1983: 391) from Keynes (1919: 128).

Ibid., 391, quoting from Keynes (1919: 170).

Ibid., 387, quoting from Keynes (1919: 146).

Architects of Political Change

Russia™s) was approximately 900 billion dollars in current terms, with
German “reparations” of the order of 300 billion current dollars.
The story of the run-up to the Great Depression is well known. An
important decision, in terms both of the global economy and Keynes™ in-
terpretation of it, was the British decision to return to the gold standard,
at a parity for sterling of 4.86, on May 14, 1925. There is reason to be-
lieve that the Chancellor of the Exchequer, Winston Churchill, was against
this decision. However, the return to the gold standard was seen by nearly
everyone as a necessary move to stabilize the international monetary sys-
tem. Perhaps the Japanese were less sanguine: It was not until 1929 that
attempts were made to return the yen to par. A boom did follow, but
the ¬‚ow of capital in to the U.S. stock market reduced overseas lending
by the United States to the extent of 2 billion dollars (60 billion dollars
current; Kindleberger, 1973: 59). Economists disagree about this effect,
of course, but Kindleberger™s analysis seems justi¬ed. Kindleberger (1973:
76“7) notes that from 1925 to 1931 there was a worldwide de¬‚ation in
commodities. For example, rubber prices fell from 340 cents per lb. in
1925 to 75 cents per lb. in 1931. From early 1929, speculative pressure
on sterling and other currencies rapidly increased (Kindleberger, 1973:
As I argued earlier, speculative asset markets are potentially chaotic:
The particular event that triggers the cascade, the collapse of the bubble, is
impossible to determine. On September 3, 1929, the Dow-Jones Industrial
Average reached 381 (over twice the average 1926 value); it started to slip
on October 3, and to fall seriously on October 24. On October 29, 16.4
million shares were traded, and the Dow fell to 198. It continued to fall
until mid-1932.
It would seem incontestable that the Wall Street crash was the result
of a belief cascade that in turn led to a further belief cascade that had
profound, real, rather than virtual, economic effects. It is still debated
whether the Depression could have been avoided. Whatever the case in
theory, the core beliefs that were still in place at the time made it almost
impossible for the United States to act in any vigorous fashion to stabilize
the international monetary system (certainly while staying on the gold
standard). The logic of the prisoners™ dilemma aspect of international

International pressures, and the fears over monetary instability because of the inad-

equacy of the Young Plan of April“May 1929 led to a further ¬‚ow of gold into the
United States. Sterling™s gold par came under speculative pressure, and the Bank of
England™s gold stock fell 640 million dollars in four months.

Keynes and the Atlantic Constitution

trade became apparent with the signing of the Smoot-Hawley Tariff Act
on June 17, 1930. Kindleberger describes the retaliatory responses ¬rst
of Canada, then France, Australia, and so on. World trade continued to
contract (from about 3 billion dollars in January 1929 to much less than
one-third this value in March 1933). The obvious economic consequences
made it not at all surprising that the incumbent, Hoover lost the U.S.
presidential election to Roosevelt in 1932. Almost the ¬rst acts by Roo-
sevelt were to accept the Thomas amendment on April 18, 1933, to issue
3 billion greenbacks (not backed by gold or silver), and more generally
to announce an abandonment of the gold dollar, during the World Eco-
nomic Conference in June, 1933. Kindleberger (1973) asserted that “[t]he
Democratic administration had little interest in or knowledge of the world
economy. . . . It would be three years before the administration [in 1936]
felt a responsibility for the operation of the international monetary sys-
tem” (229).
An immediate consequence of Roosevelt™s unilateral move was that the
member countries of the British Empire met formally to create the sterling
area. With regard to domestic U.S. policy to restore prosperity, Brinkley
(1998) says, “in truth, the New Dealers [in the U.S.] had no idea how to
end the Depression because they had only the vaguest idea of what caused
it” (18).
Two solutions to the quandary of the Depression were implicit
in Keynes™ The General Theory (Keynes, 1936). They derive not just
from empirical induction, but from a profound theoretical assertion. I
have alluded to this previously. For convenience let us now call it the
Hypothesis of Chaotic Asset Markets. Obviously, this is an anachronistic
phrase, since the term chaos has a formal mathematical meaning that dates
from 1975.37 However, chaos has a meaning in common language that
is close to the technical interpretation. It is of some interest to speculate
on the origins in Keynes™ thought on this hypothesis. To do so, we can be
guided by Skidelsky™s biography of Keynes (Skidelsky, 1983, 1992, 2000).
Keynes went up to King™s College, Cambridge, in 1902, and took the
Tripos in 1905, coming twelfth wrangler. As Skidelsky observes, his ex-
ams involved algebra, optics, elliptic functions, integral calculus, and
so forth. After the exam, Keynes started work on Marshall™s Principles
of Economics (1890). But more than economics, it was G. E. Moore™s

As noted in Chapter 1, the term chaos was ¬rst used by Li and Yorke (1975), and

then adopted in order to describe indeterminate social choice processes in Scho¬eld

Architects of Political Change

Principia Ethica (1903) that in¬‚uenced Keynes. Moore had based his ar-
gument for the necessity of general rules of conduct (“conventions”) on
the lack of rational grounds for asserting that one of two propositions is
even probably right. In a paper Keynes delivered to the Apostles in 1903,
he argued that saying A is more probable than B implicitly judges the rel-
evance of available evidence on the relation. This interpretation became
the key notion in a dissertation that Keynes submitted in December 1907
for a prize fellowship at King™s. The ¬nal version, after much intense ef-
fort, was not published until 1921.38 In the Michaelmas term of 1905,
Keynes had returned to Cambridge to study economics under Marshall,
reading Jevons, Cournot, Edgeworth, and the like. By December 1905, he
had decided to study instead for the civil service exam in August 1906,
reading psychology, the Greeks, history, mathematics, and logic. He came
in second. Skidelsky notes that Keynes™ worst mark was in economics.
Almost 30 years later, the most important book of the century in eco-
nomics, and possibly in social philosophy, was published. The General
Theory appeared on February 4, 1936. In my judgment, the profound in-
ference from Keynes™ arguments was that the strong economic equilibrium
thesis had to be modi¬ed in order to understand a sophisticated capitalist
economy. In coming to this conclusion, Keynes depended both on the the-
oretical “wrestling,” ¬fteen years previously, that he had devoted to the
task of ¬nishing the Treatise on Probability, and on his effort at induction,
his attempt to make sense of the post-1929 world of the Depression. To
reconceptualize economics, he had to deny the fundamental framework
of Marshallian equilibrium theory.39
Skidelsky intimates, and I concur, that Moore™s Principia Ethica forced
Keynes to think in a new way about rationality. Interestingly, Keynes™
denial of the equilibrium thesis was not put in the form of a theorem, but
rather in common sense, but metaphorical, language:

If I may be allowed to appropriate the term speculation for the activity of fore-
casting the psychology of the market, and the term enterprise for the activity of

Keynes (1921: 341). Keynes takes issue with a consequentialist argument by Moore

that utter ignorance of the far future forbids us from making choices for the greater
good. Keynes argues that it may be possible to make “probabilistic” judgments,
that is, in the presence of risk. Clearly, however, Moore™s argument has validity if
interpreted in terms of uncertainty, and it was this insight that Keynes pursued in
The General Theory.
From my viewpoint (and seemingly from Keynes™ as well), equilibrium theory, as

represented by the formalism of Marshallian economics, was a barrier to under-
standing the economic world. On conceptual barriers, see Margolis (1993).

Keynes and the Atlantic Constitution

forecasting the prospective yield of assets over their whole life, it is by no means
always the case that speculation predominates over enterprise. As the organiza-
tion of investment markets improves, the risk of the predominance of speculation
does, however, increase. . . .
Speculators may do no harm as bubbles on a steady stream of enterprise. But
the position is serious when enterprise becomes the bubble on a whirlpool of
(Keynes, 1936: 158“9)
As Skidelsky (1983: 611) notes, “the mathematization of The General
Theory started immediately” in work by Viner and Hicks.40 In his analysis
of Keynes™ contribution, Minsky (1975) argues that Keynes™ key insight
was that uncertainty is not identical to risk. Moreover, uncertainty vi-
tiated the formal economic models of Keynes™ “teachers and colleagues
Marshall, Edgeworth, and Pigou” (Minsky, 1975: 66). These classic au-
thors had held that
At any given time facts and expectations were assumed to be given in a de¬nite
and calculable form; and risks . . . were supposed to be capable of an exact ac-
tuarial computation. The calculus of probability was supposed to be capable of
reducing uncertainty to the same calculable status as that of certainty itself. . . . By
“uncertain” knowledge, let me explain, I do not mean merely to distinguish what
is known for certain from what is only probable. . . . Even the weather is only
moderately uncertain. The sense in which I am using the term is that in which the
prospect of a European war is uncertain, or the price of copper and the rate of
interest twenty years hence. . . .
(Keynes, 1937)
From uncertainty comes the possibility of speculative booms and
crashes, of “real” effects in the economy, including persistent unemploy-
ment (the obvious specter of the time). Minsky faults Keynes, in a sense,
for not pursuing the full rami¬cations of his insight, and for retaining
much of the equilibrium Marshallian apparatus in determining the likely
consequences of the conjecture for the operation of the macroeconomy.
Keynes himself was well aware of the dif¬culty of working outside the
“Marshallian orthodoxy.” In the preface he notes that “The composition
of this book has been for the author a long struggle of escape . . . from
habitual modes of thought and expression” (Keynes, 1936: xviii).
Hicks™ geometrization of Keynes™ theory and the neoclassical synthesis
that followed, put equilibrium theory back in center stage. As Skidelsky
(1983) observes, “The determinate system in which the ˜authority™ could
act on the multiplier either by monetary or by ¬scal policy also gave
economists a potentially key position at the centre of government” (349).
See Viner (1936) and Hicks (1937).

Architects of Political Change

One of Keynes™ colleagues, Richard Kahn (1984), later wrote:

Keynes™ insistence on the overwhelming importance of expectations, highly subject
to risk and uncertainty, was one of his biggest contributions. This completely
undermines the prevalent idea”for which Keynes™ attempt at simpli¬cation is
responsible”that his schedules can be regarded as stable relationships handed
down from heaven.

In fact, Keynes seems to have lost interest in the ensuing debate. Skidel-
sky observes that it took Keynes six months to reply to Hicks, and then
dif¬dently, after receiving the ¬rst draft of Hicks™ paper. Indeed, both
Minsky and Skidelsky suggest that Keynes wrote The General Theory
to expose the pathology of capitalism, not to suggest a cure through a
modi¬cation of classical equilibrium theory. With hindsight, the events
of the 1970s obviously suggest that Keynes was correct in dismissing the
“equilibrating” capability of the eventual “Keynesian” synthesis.
The conclusions that Keynes drew from his own theory have generally
been ignored (although not by Minsky). Of particular interest is his insight
into the prisoners™ dilemma aspect of the economic game, even when there
is a supposed equilibrating mechanism”the gold standard”in place.
After a surprising discussion of the theoretical and practical consistency
of mercantilism, Keynes (1936) commented

We may criticize them for the apparent indifference with which they accepted
this inevitable consequence [the tendency to promote war] of an international
monetary system. But intellectually their realism is much preferable to the confused
thinking of contemporary advocates of an international ¬xed gold standard and
laissez-faire in international lending who believe that it is precisely these policies
which will best promote peace.
[In this gold standard, laissez-faire system] there is no orthodox means open
to the authorities for countering unemployment at home except by struggling for
an export surplus and an import of the monetary metal at the expense of their
neighbours. Never in history was there a method devised of such ef¬cacy for setting
each country™s advantage at variance with its neighbours™ as the international
gold . . . standard. (349)

In his concluding chapter, Keynes (1936) observed that “authoritarian
state systems . . . seem to solve the problem of unemployment at the ex-
pense of ef¬ciency and of freedom”(381). Against socialism he argued that
“it is not the ownership of the instruments of production which it is im-
portant for the State to assume” (378). Instead he proposed “a somewhat
comprehensive socialization of investment will prove the only means of
securing an approximation to full employment” (378).

Keynes and the Atlantic Constitution

There were two entirely different solutions to the capitalist quandary
as perceived by Keynes in 1936. One was in the direction of “national
Self-Suf¬ciency” as discussed in his paper (Keynes, 1933). The other was
toward an “international socialization of investment.” These two possi-
bilities must have been on Keynes™ mind prior to his participation in the
recon¬guration of the post-war Atlantic Constitution at Bretton Woods
in 1944.

7.7 the constitutional quandary of 1944
Quandary, a state of extreme perplexity or uncertainty.

Before discussing the attempt to resolve the Depression quandary by
the architects of change of 1944, it will be useful to give a more precise
view of Keynes™ signi¬cance for the Atlantic Constitution.
As I have argued in the earlier sections of this chapter, a fully-developed
constitution must necessarily relate political and economic rights in the
context of a credible theoretical and empirical framework. We might in-
deed use the term paradigm for this framework (Kuhn, 1962). The equi-
librium thesis had been of fundamental importance for the constitution
since at least the time of Adam Smith. Once the thesis is accepted, then it
is credible that political liberty is compatible both with economic liberty
and ef¬ciency. If there are strong empirical reasons to deny the equilibrium
thesis, then the step to autocracy is fairly easy. Keynes clearly believed that
ef¬ciency in exchange could only be maintained in a decentralized econ-
omy.41 From this perspective, autocracy trades ef¬ciency against the risk
of the monetary collapse that capitalism seemed to engender. Since a con-
stitution is maintained by credible beliefs, the apparent irrelevance of the
equilibrium thesis during the Depression meant that the entire Constitu-
tion could fail.42 However, the equilibrium thesis was deeply embedded
in what Keynes termed “habitual modes of thought” and the thesis had
proved impossible for policy makers to deny in a fundamental way. As a
consequence they could not proceed to attack the problem of unemploy-

There was an extensive literature in the 1930s on this proposition, with Oskar Lange

(1938) and Michael Kalecki pitted against Ludwig von Mises (1935) and Friedrich
von Hayek.
Note that a constitution involves an equilibrium in beliefs, what I call the core

belief. Just as with asset markets, such a belief equilibrium may crash. I discuss this
possibility further in Chapter 8.

Architects of Political Change

It seems fairly clear, on reading the record, that Keynes welcomed Roo-
sevelt™s pragmatic attempts to relieve unemployment by various means
from 1933 to 1936. It was not obvious in 1936, though it became clear
the next year from the U.S. downturn of 1937, that such pragmatic ef-
forts were not guaranteed success. To attack the unemployment quandary,
Keynes had to overcome the “barrier” of the conceptual power of the
equilibrium thesis. This he did in a sense, by distinguishing commodities
markets from “speculative” asset markets. While accepting the relevance
of the equilibrium theorem for the former, he denied that it was appro-
priate for the latter.43 If the potentially chaotic, speculative characteristic
of asset markets could be overcome, say, by “socialization of investment”
then both the ef¬ciency of the decentralized commodity market and po-
litical freedom could be retained. Thus a modi¬ed version of the Atlantic
Constitution could be preserved.
My discussion, following Minsky, of the Keynesian interpretations
made by Viner and Hicks, suggests how powerful was the Marshallian
equilibrium thesis. It appears obvious that the later “Keynesian ortho-
doxy” was dependent on a version or interpretation of the equilibrium
proposition that governments could reimpose full employment equilib-
rium by appropriate monetary and ¬scal strategies.44
The vision held out by Keynes differed signi¬cantly from the one which
derived from the later Keynesian orthodoxy. However, Keynes™ conclu-
sions seem not so much to expose the capitalist “pathology” as to suggest
that, if the investment quandary could be solved, then capitalism would
fade away, and a more equitable economic system evolve. More signif-
icantly “there would be no important economic forces calculated to set
the interest of one country against that of its neighbors. . . . International
trade would cease to be what it is . . . but [instead become] a willing and
unimpeded exchange of goods and services in conditions of mutual ad-
vantage” (Keynes, 1936: 382).

I have argued in the earlier sections of this chapter that Keynes was correct in this

separation. Although a generally applicable equilibrium theorem for commodity
markets would not be available until 1956, a weaker version (but one more general
than that of Marshallian economic theory) was known in the 1930s. See von Neu-
mann (1945) for the translation of his original paper of 1932 (written in German).
It seems to me that the step by Keynes away from Marshallian equilibrium, followed

by the half step back by Viner (1936) and Hicks (1937) resembles the move by Coper-
nicus away from the Ptolemaic planetary orthodoxy toward the heliocentric system.
This move was followed by a half a step back by Tycho Brahe (to an incoherent
system, where planets orbit the sun, which orbits the earth). See Margolis (1993).

Keynes and the Atlantic Constitution

At the cost of “the enlargement of the functions of government [to
involve] the task of adjusting to one another, the propensity to consume,
and the inducement to invest” (Keynes, 1936: 380), the liberty and ef¬-
ciency derived from the Atlantic Constitution could not only be preserved
but enhanced. The prisoners™ dilemma-like characteristics of international
economics could be made to disappear. As I have observed previously, our
understanding of the international economic system, during the period
from 1960 to the end of the 1970s, is that it was indeed an nPD. Be-
cause most of the OECD countries during that period utilized principles
of neo-Keynesian demand management, either Keynes™ vision was wrong,
or the theoretical principles underlying the neo-Keynesian synthesis were
invalid. The combined evidence from the period suggests the latter hy-
pothesis. This does not, of course, require us to suppose that Keynes was
right. It does, however, make it plausible that the Keynesian orthodoxy
of the 1950s and 1960s had very little to do with Keynes™ vision.
Given that the Atlantic Alliance had been beset by autocratic regimes,
which rejected the foundation of the constitution, it was entirely credible
in 1944 that general belief in the constitution could indeed “collapse.”
My reading of the ¬nal chapter of The General Theory is that Keynes saw
the quandary in terms not just of solving the problem of unemployment,
but of saving what I call the Atlantic Constitution. For this reason, I use
the term the Constitutional Quandary of 1944 to refer to the problem
faced by the architects of change in their attempt to reconstruct the global
political economy after World War II.
As we have seen, in the last two chapters of The General Theory, Keynes
contrasted two differing beliefs about the nature of the international sys-
tem. The “mercantilist” belief was of a zero-sum world. In pursuing na-
tional advantage and relative strength, mercantilists acknowledged the
tendency to promote war. In contrast, Marshallian equilibrium theory re-
jected the relevance of mercantilist belief. Keynes (1936) comments iron-
ically, that “we were brought up to believe that [mercantilism] was little
better than nonsense, so absolutely overwhelming and complete has been
the domination of the classical school” (335).
Marshallian principles asserted that the rule of the gold standard in-
duced a cooperative characteristic to trade, under which laissez-faire
would bene¬t all nations. Keynes™ analysis, and what I have called his
chaos hypothesis, led him to the quite different conclusion that trade was
not, in fact, a cooperative game, but involved mixed motives. As we have
seen above, this game can be considered to be an nPD. The possibility

Architects of Political Change

that a hegemon could induce cooperation seems not to have been con-
sidered a possibility by Keynes in 1936. By 1944, however, Keynes had
developed the idea of an International Clearing Union.45 If we accept the
notion of an nPD, then such a union would appear able “to protect na-
tional economies from de¬‚ationary pressures by providing free access to
an international pool of credit” (Block, 1977). We may call this solution
“International Keynesianism.”
There are two other possibilities that derive from Keynes™ thinking,
both associated with the notion of “national self-suf¬ciency.” In 1933,
Keynes made this famous observation:

Ideas, knowledge, science, hospitality, travel”these are the things which should of
their nature be international. But let goods be homespun whenever it is reasonably
conveniently possible, and, above all, let ¬nance be primarily national. (758)

Such a recommendation assumes, to some degree, a mercantilist structure
to the world. To speculate somewhat, the Sterling Area, and its system
of Imperial preference set up in the British Empire after 1933, is compat-
ible with this view. Such a mercantilist world would, in all probability,
evolve into blocs, potentially antagonistic. Keynes had also argued in The
General Theory that if the investment quandary could be solved within
each nation, then the mutual advantages of trade could be realized. Such
a world need not be mercantilist. These two possibilities for the future I,


. 8
( 11)