Animal Spirits or Complex Adaptive Dynamics?
A Review of
George Akerlof and Robert J. Schiller
Animal Spirits (Princeton, 2009)
March 31, 2009
In his epoch-making General Theory (1936), John Maynard Keynes noted that
concerning investment decisions, “most, probably, of our decisions to do some-
thing positive, the full consequences of which will be drawn out over many days
to come, can only be taken as the result of animal spirits–a spontaneous urge to
action rather than inaction, and not as the outcome of a weighted average of quan-
titative bene?ts multiplied by quantitative probabilities.” Because of the propen-
sity of investors to base decisions on variables other than “market fundamentals,”
the aggregate investment function of an economy will tend to be erratic. Indeed,
even today, it is virtually impossible to predict aggregate investment successfully,
although the other sources of aggregate demand and supply are relatively well un-
derstood. Keynesian economic policy suggested that government use anti-cyclical
spending and taxation to counter the vicissitudes of aggregate investment. While
countercyclical monetary policy might also have the same effect via the interest
rate, Keynes’ theory of the “liquidity trap” suggested that investment is quite in-
sensitive to the interest rate. Experience bears out Keynes on this point, at least for
large shortfalls in aggregate demand. In an economic downturn, it is critical that
the monetary authority ensure a high level of liquidity to avoid arti?cial curbs on
the willingness of businesses to invest, but liquidity itself is not suf?cient to restart
It is not hard to see that Keynesian analysis, if correct, applies to all sorts of
shocks to the economy, not just ?uctuations in investment. Indeed, such automatic
Santa Fe Institute and Central European University.
stabilizers as progressive income taxation, unemployment compensation, and ac-
celerated depreciation schedules can act as shock absorbers, smoothing out eco-
nomic ?uctuations. In the past, at least for normal-size ?uctuations, active govern-
ment intervention has been generally ineffective because it takes the government
too long to get into action, and by the time it passes the appropriate legislation
and the effects have reaced producers and consumers, the economy has already
come out the other end of the business cycle. Moreover, Keynesianism was com-
promised by the long stag?ationary period of the 1970’s: according to Keynesian
theory, there is a trade-off between in?ation and unemployment, whereas during
this period there was both high unemployment and rampant in?ation in the United
A whole new school of “rational expectations” emerged in the 1970’s that held
that markets always are in equilibrium, unemployment was always voluntary (be-
cause all workers had to do is accept lower wages), and active government inter-
vention is part of the problem, not the solution Lucas (1981). While this brand
of macroeconomic theorizing became dominant in the profession, it is important
to understand that there is nothing in standard economic theory that proves that
markets are always in equilibrium, or that they are necessarily ef?cient, or that
regulation is unnecessary. It is true that a cabal of Chicago economists, thrust into
the limelight by the electoral success of Ronald Regan, offered a far-fetched free-
market ideology that had no solid theoretical support, and was rejected by most
The subprime mortgage meltdown that began in 2007 and dominates the macroe-
conomy today shows to the general public that the Chicago crowd was wrong, but
most economists knew that all along. The really stunning fact about the current
macroeconomy is that disequilibrium in the home mortgage market could so se-
riously compromise the American ?nancial system. Even those who foresaw the
housing crisis did not predict so massive a credit collapse, one leading to levels
of government intervention that would have been inconceivable even in the recent
past. The basic contribution of Akerlof and Shiller’s book is to show the impor-
tance not only of Keynesian animal spirits, but also other ways in which human
decision-making affects the macroeconomy that violate the canons of neoclassical
I think Animal Spirits is true to the spirit of John Maynard Keynes, if not the
letter, in stressing that we cannot understand the macroeconomy without having a
theory of how humans make decisions. The reader who is only interested in the
current crisis and has read the New York Times or Wall Street Journal regularly,
however, will not learning much this book, since the current crisis is discussed in
more than passing only in an Appendix to Chapter 7 and in Chapter 12. Rather
than writing a complete contemporary analysis, the Akerlof and Shiller put to-
gether papers mostly from the 1990’s, with up-to-date commentary. This strategy
works well, at least for non-experts. Indeed, all readers can gain from Akerlof and
Shiller’s defense of behavioral macroeconomics.
Animal Spirits vs. Rational Expectations
The behavioral economists’ idea that there is something quirky about human decision-
making that explains macroeconomic dynamics lies in diametrical opposition to
the hitherto dominant Chicago “rational expectations” school of macroeconomic
theory, whose major critique of traditional Keynesianism was the latter’s incom-
patibility with the rational actor model. The Keynesian notion that there is a stable
marginal proensity to consume out of income, that investment is based on animal
spirits, that prices are downwardly rigid because of money illusion, and that social
convention prevents unemployed workers from driving down wages by compet-
ing with employed workers, were, according to the Chicago School, ill-founded
assumptions incompatible with rational choice.
The school of behavioral macroeconomics to which Akerlof and Shiller be-
long was waiting in the wings for a chance to tell their (compelling) story, and the
current crisis is just what they were waiting for. There is much data in support
of Keynesian-like economic behavior, and the supposed superiority of “rational
expectations” theory indeed rings hollow.
For instance Stein and Song (1998), based on the sample period 1955-1973,
show that the consumption function exhibits a marginal propensity to consume of
0.86, while for the sample period 1974 to 1991, the marginal propensity to consume
rises to close to unity. Of course, what matters more for policy purposes is how
families react to tax rebates and other episodic changes in income. Here, the aggre-
gate marginal propensity to consume appears to be an overstatement. Less than half
the 2001 tax rebates, for instance, were spent, the remainder going into saving and
debt reduction. The rebates of 2008 indicate a somewhat higher marginal propen-
sity to consume. Indeed, between May and July of 2008, $90 billion of stimulus
payments were distributed to U.S. families. Broda and Parker (2008) found that
families that received these payments increased consumption spending by 3.5%,
and the overall effect on nondurable consumption in the second quarter of 2008
alone was 2.4%, and they estimate that this would rise to 4.1% in the third quar-
ter of 2008. These facts do not support the “rational expectations” consumption
function, according to which spending should not be affected at all, given rational
life-cycle consumption planning. Rather, it indicates that counter-cyclical tax pol-
icy can be a potent stabilizing force, provided government can respond quickly to
new economic conditions.
Behavioral game theory has also given us critical insights into human behav-
ior, many of mesh well with the Keynesian perspective. Daniel Ellsberg (1961)
has shown that human decision-makers have a strong distaste for uncertainty, as
opposed to riskiness, a propensity that might well explain the reluctance to in-
vest when the future becomes uncertain (Segal 1987). Similarly, loss aversion
(Kahneman et al. 1991, Tversky and Kahneman 1981, Genesove and Mayer 2001)
may explain why some prices, such as residential houses, remain sticky downward
and transactions decline instead of prices. Finally, we know a lot more about the
psychological aspect of wages than in Keynes’ time. Not only are wages affected
by gift-exchange (Akerlof 1982, Fehr and Tougareva 1995, Fehr et al. 1998a, Fehr
et al. 1998b, Charness and Haruvy 2002, Fehr and Schmidt 2006) and the desire
to maintain labor discipline (Gintis 1976, Shapiro and Stiglitz 1984, Bowles 1985,
Gintis and Ishikawa 1987, Bowles and Gintis 1993), but also are an employer’s sig-
nal of satisfaction with work performance. For instance Loewenstein and Sicher-
man (1991) showed that in a experimental setting, workers prefer a rising wage
schedule over time to either a ?at or a declining wage schedule, despite the fact
that for a given total amount of wages paid, the declining schedule has a higher
present value. Moreover, in an especially thorough survey of employers, Bewley
(2000) corroborated a deep reluctance of ?rms to cut wages in a recession, on the
grounds of hurting worker morale.
Complexity and Regulation
Despite this corroborating evidence, however, the major thesis of the book is only
partially correct in attributing macroeconomic instability to human foibles. The
authors say in conclusion that “if we thought that people were totally rational, and
that they acted almost entirely out of economic motives, we too would believe
that government should play little role in the regulation of ?nancial markets, and
perhaps even in determining the level of aggregate demand.” (p. 173). In fact,
there is nothing in economic theory that says that rational individuals interacting
on markets will produce either stable or socially ef?cient outcomes. The Wal-
rasian general equilibrium model, which is the canonical framework for investigat-
ing macroeconomic behavior on a theoretical level, shows that in the absence of
market externalities, there are market-clearing equilibria that are Pareto-ef?cient.
However, as has been long understood, this model has absolutely no attractive dy-
In Walras’ original description of general equilibrium (Walras 1954 ),
market clearing was effected by a central authority. This authority, which has
come to be known as the “auctioneer,” remains today because no one has suc-
ceeded in producing a plausible decentralized dynamic model of producers and
consumers engaged in market interaction in which prices and quantities move to-
wards market-clearing levels. Only under implausible assumptions can the contin-
uous ‘auctioneer’ dynamic be shown to be stable (Fisher 1983), and in a discrete
model, even these assumptions (gross substitutability, for instance) do not preclude
instability and chaos in price movements (Saari 1985, Bala and Majumdar 1992).
Moreover, contemporary analysis of excess demand functions suggests that restric-
tions on preferences are unlikely to entail the stability of tˆatonement (Sonnenschein
1972,1973; Debreu 1974; Kirman and Koch 1986).
It has been a half century since Debreu (1952) and Arrow and Debreu (1954)
provided a satisfactory analysis of the equilibrium properties market economies,
yet we know virtually nothing systematic about Walrasian dynamics. This sug-
gests that we lack understanding of one or more fundamental properties of market
exchange. There are thus slim grounds for Akerlof and Shiller to attribute macroe-
conomic ?uctuations wholly to “animal spirits” that would not exist were economic
An alternative perspective that deserves consideration is that the market econ-
omy is a complex nonlinear system (Blume and Durlauf 2005, Beinhocker 2006,
Miller and Page 2007), which by its very nature is subject to volatility because the
probability distributions underlying stochastic behavior have the “fat tails” char-
acteristic of complex systems (Crutch?eld et al. 1986, Saari 1995, Farmer and
Lillo 2004). In effect, as Axel Leijonhufvud once remarked, neoclassical eco-
nomic theory models “smart people in unbelievably simple situations,” while the
real world is populated by “simple people [coping] with incredibly complex situa-
tions.” The complex adaptive economy is never in equilibrium, but is continually
subjected to shocks, both exogenous and endogenous, that affect its short-term
movements. There are frequent local nonlinear resonances that lead to signi?cant
deviations of economic variables (prices, quantities, wages, asset prices) from their
equilibrium values even in the absence of strong aggregate or systematic perturba-
tions to the system.
There have been notable contributions to the notion of economies as com-
plex systems in recent years. These include Brian Arthur’s work on increasing
returns (Arthur 1994), Peyton Young and Mary Burke’s analysis of crop shar-
ing (Young and Burke 2001), evolutionary models inspired by Nelson and Win-
ter (1982) and Hodgson (1998), William Brock and Stephen Durlauf’s study of
social interaction (Brock and Durlauf 2001), Edward Glaeser, Bruce Sacerdote
and Jose Scheinkman’s treatment of crime (Glaeser et al. 1996), Samuel Bowles’
treatment of institutional evolution (Bowles 2004), Robert Axtell’s study of ?rm
size (Axtell 2001), Alan Kerman and his colleagues models of ?nancial markets
(Kirman et al. 2005), and models of the evolution of other-regarding preferences
(Gintis 2000, Bowles et al. 2003) and the agent-based simulation of general equi-
librium and barter exchange
Inspired by this literature, and with the knowledge that a key means of un-
derstanding complex systems with emergent properties that cannot (yet) be ana-
lytically modeled, I turned to computer modeling, which can provide important
insights. When I subjected the Walrasian general equilibrium model to an agent
based simulation (Gintis 2007), I found that there is a robust tendency towards mar-
ket clearing equilibrium, but this is always offset by highly volatile movements in
prices, wages, capital demand, and other macroeconomic variables. This volatility
is due to the inherent stochasticity of complex systems, not to the “animal spirits”
of economic actors. For instance, Figure 1 illustrates the ?uctuations in demand
and supply in my agent-based simulation with no exogenous shocks.
Figure 1: Demand and Supply in Sector 1 of a Multi-Sector Economy. Note that there
is considerable period-to-period volatility. Excess supply averages about 8% of average
This analysis suggests that when the simulated economy experiences a system-
wide shock of moderate proportions, it should return to its long-run state after a
certain number of periods, which we may call the relaxation time of the dynamical
system. This is in fact the case. For instance, I simulated a four-sector economy
with 2000 agents, and ran the economy for 3,000 periods. After each 500 periods,
the ?rms in the economy were all subjected to a technological shock, taking the
form of the optimal ?rm sized k falling from 35 to 14. This shock persisted for 10
periods, after which the original value of k was reestablished. Figure 2 suggests
that the economy recovers its high ef?ciency price structure after a few hundred
rounds. Figure 2 shows that ef?ciency is severely compromised by the shocks, but
is restored within two hundred rounds.
Figure 2: Relaxation Time in a Four Sector Economy Subject to Macro Shocks. Every 500
periods, the economy sustains a shock whereby each ?rm’s optimal size is reduced from 35
to 14. The shock lasts for 10 periods, after which the original optimal ?rm size is restored.
Note that goods prices stabilize after a few hundred periods, and are approximately equal
across all runs.
Akerlof and Shiller do not have enough evidence to assert con?dently that people
are driven by irrational animal spirits to produce market volatility. People imitate
the successful, both in my agent-based model and in real life. This is generally
quite rational behavior, but it can produce “behavioral cascades” that are destabi-
lizing (Bikhchandani et al. 1992, Edgerton 1992). If someone is doing well and
someone lese is not, the latter copying the former is the rational thing to do. The
fact that behavior this leads to economic volatility suggests the need for market reg-
ulation. As Keynes observed, the investment process is a sort of beauty contest in
which the winner is not the one who picks the most beautiful, but rather is the one
who pick what others consider the most beautiful. This is not because people are
irrational, but rather because the success of one’s investment plans depend on oth-
ers’ investment decisions, and there is no objective measure of the interpenetrating
beliefs of investors.
Consider, for instance, a recent experimental result of Nobel prize-winning
economist Vernon Smith and colleagues (Hussam et al. 2008). The authors re-
port that when they impose a large increase in liquidity and dividend uncertainty to
shock the environment of experienced subjects who have converged to equilibrium,
a ?nancial bubble is rekindled. They suggest that in order for price bubbles to be
extinguished, the environment in which the participants engage in exchange must
be stationary and bounded. Experience alone is not robust to major new environ-
ment changes in determining the characteristics of a price bubble. Of course, there
is no reason for the investment environment in real economies to be ergodic, since
changing technology, social organization, and the con?guration of international
economic forces render economic dynamics a continual structural novelty.
Akerlof and Shiller have given us a useful book, especially for the non-expert,
but it would be sad if the debate over the renovation of the American economy
revolved around the quirkiness of human decision-making, and ignored the inher-
ent incapacity of an improperly regulated economy to produce socially desirable
outcomes, however “rational” the economic agents.
Akerlof, George A., “Labor Contracts as Partial Gift Exchange,” Quarterly Journal
of Economics 97,4 (November 1982):543–569.
Arrow, Kenneth J. and Gerard Debreu, “Existence of an Equilibrium for a Com-
petitive Economy,” Econometrica 22,3 (1954):265–290.
Arthur, Brian, Increasing Returns and Path Dependence in the Economy (Ann Ar-
bor: University of Michigan Press, 1994).
Axtell, Robert, “Zipf Distribution of U.S. Firm Sizes,” Science 293 (2001):1818–
Bala, V. and M. Majumdar, “Chaotic Tatonnement,” Economic Theory 2
Beinhocker, Eric, The Origins of Wealth: Evolution, Complexity, and the Radi-
cal Remaking of Economics (Cambridge, MA: Harvard Business School Press,
Bewley, Truman F., Why Wages Don’t Fall During a Recession (Cambridge: Cam-
bridge University Press, 2000).
Bikhchandani, Sushil, David Hirshleifer, and Ivo Welsh, “A Theory of Fads, Fash-
ion, Custom, and Cultural Change as Informational Cascades,” Journal of Polit-
ical Economy 100 (October 1992):992–1026.
Blume, Lawrence E. and Steven N. Durlauf, The Economy As an Evolving Com-
plex System III (Santa Fe, NM: Santa Fe Institute, 2005).
Bowles, Samuel, “The Production Process in a Competitive Economy: Wal-
rasian, Neo- Hobbesian, and Marxian Models,” American Economic Review
75,1 (March 1985):16–36.
, Microeconomics: Behavior, Institutions, and Evolution (Princeton: Princeton
University Press, 2004).
and Herbert Gintis, “The Revenge of Homo economicus: Contested Exchange
and the Revival of Political Economy,” Journal of Economic Perspectives 7,1
, Jung-kyoo Choi, and Astrid Hopfensitz, “ The Co-evolution of Individual Be-
haviors and Social Institutions,” Journal of Theoretical Biology 223 (2003):135–
Brock, William and Steven N. Durlauf, “Discrete Choice With Social Interactions,”
Review of Economic Studies 68,235 (April 2001):235–260.
Broda, Christian and Jonathan Parker, “The Impact of the 2008 Tax Rebates on
Consumer Spending: Preliminary Evidence,” July 2008. University of Chicago
Graduate School of Business.
Charness, Gary and Ernan Haruvy, “Altruism, Equity, and Reciprocity in a Gift-
Exchange Experiment: An Encompassing Approach,” Games and Economic
Behavior 40 (2002):203–231.
Crutch?eld, James P., J. Doyne Farmer, Norman H. Packard, and Robert S. Shaw,
“Chaos,” Scienti?c American (1986):46–57.
Debreu, G´erard, “A Social Equilibrium Existence Theorem,” Proceedings of the
National Academy of Sciences 38 (1952):886–893.
Debreu, Gerard, “Excess Demand Function,” Journal of Mathematical Economics
Edgerton, Robert B., Sick Societies: Challenging the Myth of Primitive Harmony
(New York: The Free Press, 1992).
Ellsberg, Daniel, “Risk, Ambiguity, and the Savage Axioms,” Quarterly Journal of
Economics 75 (1961):643–649.
Farmer, J. Doyne and F. Lillo, “On the Origin of Power Law Tails in Price Fuctua-
tions,” Quantitative Finance 4,1 (2004):7–11.
Fehr, Ernst and E. Tougareva, “Do Competitive Markets with High Stakes Remove
Reciprocal Fairness? Experimental Evidence from Russia,” 1995. Working
Paper, Institute for Empirical Economic Research, University of Zurich.
and Klaus Schmidt, “The Economics of Fairness, Reciprocity and Altruism–
Experimental Evidence and New Theories,” in Jean Mercier Ythier, Serge Kolm,
and Louis-Andr´e G´erard-Varet (eds.) Handbook of Reciprocity, Gift-Giving,
and Altruism, Vol. 1 (Amsterdam: Elsevier, 2006) pp. 615–692.
, Georg Kirchsteiger, and Arno Riedl, “Gift Exchange and Reciprocity in Com-
petitive Experimental Markets,” European Economic Review 42,1 (1998):1–34.
, Simon G¨achter, Erich Kirchler, and Andreas Weichbold, “When Social Norms
Overpower Competition—Gift Exchange in Labor Markets,” Journal of Labor
Economics 16,2 (April 1998):324–351.
Fisher, Franklin M., Disequilibrium Foundations of Equilibrium Economics (Cam-
bridge, UK: Cambridge University Press, 1983).
Genesove, David and Christopher Mayer, “Loss Aversion and Seller Behavior:
Evidence from the Housing Market,” Quarterly Journal of Economics 116,4
Gintis, Herbert, “The Nature of the Labor Exchange and the Theory of Capitalist
Production,” Review of Radical Political Economics 8,2 (Summer 1976):36–54.
, “Strong Reciprocity and Human Sociality,” Journal of Theoretical Biology
, “The Dynamics of General Equilibrium,” Economic Journal 117 (October
and Tsuneo Ishikawa, “Wages, Work Discipline, and Unemployment,” Journal
of Japanese and International Economies 1 (1987):195–228.
Glaeser, Edward L., B. Sacerdote, and J. Scheinkman, “Crime and Social Interac-
tions,” Quarterly Journal of Economics 111 (May 1996).
Hodgson, Geoffrey M., “The Approach of Institutional Economics,” Journal of
Economic Literature 36,1 (1998):166–192.
Hussam, Reshmaan N., David Porter, and Vernon L. Smith, “Thar She Blows:
Can Bubbles Be Rekindled with Experienced Subjects?,” American Economic
Review 98,3 (June 2008):924–937.
Kahneman, Daniel, Jack L. Knetsch, and Richard H. Thaler, “The Endowment
Effect, Loss Aversion, and Status Quo Bias,” Journal of Economic Perspectives
5,1 (Winter 1991):193–206.
Keynes, John Maynard, The General Theory of Employment, Interest and Money
(New York: Harcourt Brace, 1936).
Kirman, Alan, H. Foellmer, and U. Horst, “Equilibrium in Financial Markets with
Heterogeneous Agents: A New Perspective,” Journal of Mathematical Eco-
nomics 41,1–2 (February 2005):123–155.