Beyond the Market: The Social Foundations of Economic Efficiency: Jens Beckert
Princeton University Press | ISBN: 0691049076 | 2002-09-03 | PDF (OCR) | 368 pages | 1.67 Mb
Beyond the Market launches a sociological investigation into economic efficiency. Prevailing economic theory, which explains efficiency using formalized rational choice models, often simplifies human behavior to the point of distortion. Jens Beckert finds such theory to be particularly weak in explaining such crucial forms of economic behavior as cooperation, innovation, and action under conditions of uncertainty--phenomena he identifies as the proper starting point for a sociology of economic action. Beckert levels an enlightened critique at neoclassical economics, arguing that understanding efficiency requires looking well beyond the market to the social, cultural, political, and cognitive factors that influence the coordination of economic action. Beckert searches social theory for the components of an alternative theory of action, one that accounts for the social embedding of economic behavior. In Durkheim and Parsons he finds especially useful approaches to cooperation; in Luhmann, a way to understand how people act under highly contingent conditions; and in Giddens, an understanding of creative action and innovation. Together, these provide building blocks for a research program that will yield a theoretically sophisticated understanding of how economic processes are coordinated and the ways that markets are embedded in social, cultural, and cognitive structures. Containing one of the most fully informed critiques of the neoclassical analysis of economic efficiency--as well as one of the most thoughtful blueprints for economic sociology--this book reclaims for sociology the study of one of the most important arenas of human action.
Summary: Preaching to the Sociological Choir
Rating: 2
Jens Beckert writes intelligently and lucidly, never hiding weakness of conception behind a flurry of big, vague words. Beckert is also respectful of the reader, never leaving in doubt exactly where is stands in the course of his argument. Beckert's task, reformulating and respecifying economic sociology, is high theory, however, and contemporary sociology is devoid of high theory. This leaves Beckert with little to work with, filling the bulk of this book with rehashes of the work of Durkheim, Parsons, and Giddens, and relegating his alternative framework to the final ten pages of the book.
Beckert's alternative is a version of the embedded markets position so ably developed by Granovetter (1985), as well as Beckert's mentors in the Princeton sociology department. This approach has led to excellent analyses of real-live markets that should be read by all economic theorists and policy-makers. However, this approach does not pretend to be general sociological theory, and it is an important supplement to, not alterative to, traditional economic analysis.
The embedded markets approach is has interesting lessons for the interrelations among the behavioral disciplines. Sociological theorists have been perennially preoccupied with the question of the appropriate boundary-line between economics and sociology. Virtually every sociological theorist (except James Coleman and friends) include the rejection of the "utilitarian" model of economic theory. Granovetter calls the economist's rational actor "undersocialized," and sees sociology as providing a better-socialized actor. On the other hand, Granovetter joins the herd in considering the traditional sociological model, that of Talcott Parsons' theory of action, as "oversocialized." (Wrong 1961) Granovetter defines the proper actor for economic sociology as striking the balance between over- and -undersocialization. In fact, Granovetter's actor engages in repeated games with other, monitoring their activities, and using the threat of non-renewal to ensure contract compliance in situations where the complete contracts of Walrasian economic cannot be written or enforced.
The interesting thing about the actor of embedded markets theory is that he is simply a rational actor involved in repeated games. Thus, Granovetter (1985) is proposing a shift from Walrasian economics to game theory, which is exactly what economists were carrying out in this period (e.g., the famous Fudenberg, Levine, Maskin paper of 1994). Embedded markets theory is, in fact, perfectly compatible with modern economic theory, and certainly not an alternative to it. The ability of sociologists in this tradition to deepen our understanding of such markets through historiographic and highly descriptive studies is an important complement to economic theory, but certainly not an alternative. This causes deep problems for Beckert, who consistently criticizes the rational actor model, while in fact never deviating from it in any significant way, except for avoided any mathematical formulation of it.
It follows that the notion that embedded markets theory is an alternative to the general equilibrium model is untenable, because it is completely non-quantitative, and deals with a single market at a time. By contrast, general equilibrium theory can be used to model real economic activity, and supplies insights as to why market economies tend to aggregate individual activities into a relatively efficient overall system. This is not to say that there are not severe problems with Walrasian theory---there are. Most important (not mentioned by Beckert), the theory has no serious dynamics, and provides not a clue how prices move to their equilibrium levels. Beckert trots out several of the usual critiques of general equilibrium theory, but he seems never to have heard of the vast body of economic theory developed in the post-WW II era, to deal with market externalities and volatility, nor has heard of the theory of the second best. The notion that embedded markets theory can provide a policy alternative to this body of research is patently absurd.
Beckert is in fact preaching to the choir---the choir of sociologists who do not do theory, who do not like theory, who are deeply scared of mathematical models, and therefore who welcome a critique of economic theory, however ineffective. Also pleasing to these folks is the fact that Beckert is incessantly critical of economic theory, but is never critical of sociological theory (he does treat individual sociologists critically).
There is not an equation in the book, there is barely a reference to a formal analytical model, and Beckert's alternative, which rejects the rational actor model, explicitly rejects formal modeling: "deserting the action model of homo economicus also leads to the far-reaching abandonment of the possibilities of formalization that result from the equilibrium view of economic theory.'' (p. 295) Talk about bizarre! Does Beckert really imagine that we should give up quantitative modeling of economic activity because markets are embedded?
Beckert has a short but fairly informative critique of the repeated game theory of cooperation. He concludes that it fails to explain cooperation because it assumes perfect signaling. He does not mention that imperfect public signaling was handled by Fudenberg, Maskin, and Levine in 1994. The case of private signals is much more difficult, and the proper critique of the Folk Theorem is in terms of private signals, I believe. But Beckert never heard of private signaling, apparently.
Like most sociologists, Beckert does not like the rational actor model and, as illustrated in the previous paragraph, counsels its abandonment. This is also bizarre, because the rational actor models is quite compatible with embedded markets theory. Beckert's arguments, like that of most critics of the rational actor model, are shoddy and ignorant. He does not really understand Bayesian choice theory, which leads him to reject rational choice when "probabilities are unknown." Of course, Bayesian priors do not depend on objective priors, so there is no problem. Beckert argues that individual attitudes toward risk and assessment of probability are not individual, but rather are social. This is very true, and it is an important insight that could possibly be turned into a powerful model of social risk-taking, but surely not by rejecting the rational actor model!
Beckert's intermediate chapters are expositions of the sociology of Durkheim, Parsons, Luhmann, and Giddens. He does a very good job, and provides excellent bibliographies. His critiques are however, standard and not very insightful. For instance, after a fine exposition of Parsons' later work (Economy and Society and later), his criticism is a repeat of the "oversocialized conception of the actor" (Wrong, 1961; Garfinkel, 1967; etc.) This criticism is accurate but does not invalidate Parsons' insights; rather, it suggests that there are principles at work other than socialization that determine individual beliefs and preferences. This is not a sufficient ground for rejecting Parsons' theory of action.
This book was published in 2002, and written between 1994 and 1995, which perhaps explains the complete absence of reference to experimental economics and behavioral game theory. This fact, however, renders his critique of the rational actor model completely obsolete. What is now known as "other regarding behavior" or "virtuous behavior", Becker calls "irrational behavior." Of course, economists have traditionally be guilty of confusing these notions, but even by the year 2000, every economist had learned the difference between irrationality and non-self-regarding preferences. Curiously, Beckert rejects other-regarding behavior as any more than a curiosum, and fully embraces a self-regarding action framework (perhaps because this is compatible with the embedded markets position). Similarly, he denies that the internalization of norms is important for economic sociology, and his critique of this concept is just philosophical drivel. The fact is, internalization is not needed for embedded markets theory, so why bother including it in the sociologist's repertoire?
Sociologists have got to get real. Rejecting game theory and the rational actor model is precisely why sociological theory does not exist. The only real reason sociologists do not accept these tools is that they are hard to learn and they involve a lot of math. The economist's rational actor model is surely too simple in many ways to deal with complex interpersonal interactions. Most important, people are not always selfish, beliefs are not purely subjective, but rather are shared among individuals, and even preferences are not individually given, but rather are a function of social situation (framing). All of these three facts are critical to understanding human strategic interaction in cooperation and conflict. These facts complicate the model of individual human behavior, and will require even deeper and more powerful mathematical tools, not their abandonment.
Similarly, there are deep problems with traditional game theory, and sociological theory, especially the theory of norms is likely to be a critical theoretical tool in repairing the weakness of game theory, but the knee-jerk instinct of sociologists to reject game theory is a major reason for the impotence of sociological theory today. Game theory is not everything, but it surely part of a total theory.
Wednesday 13 April 2011
Beyond the Market: The Social Foundations of Economic Efficiency: Jens Beckert
0 comments:
Post a Comment