The two projects of microeconomics

The Mazzoleni and Nelson paper published in Industrial and Corporate Change 22:6 (Mazzoleni and Nelson, 2013) is a graceful and informative excursion into the ambiguities of intellectual history. As I can lay no claim to special knowledge about the intellectual history of economics, I can only be impressed. The paper is also an examination of the possible future of microeconomics. Mazzoleni and Nelson have prejudices, so do I. This recitation of my prejudices has been stimulated by reading theirs, but it can be interpreted as a comment on their paper only in the sense in which an after dinner speech is commentary on the food.

Modern microeconomics can be seen as a conflation of two projects. The first project is a normative one: inducing humans and artificial actors to make choices that exhibit intelligence of a particular kind. Within the litany of contemporary thought, it draws its framework from theories of choice, particularly those associated with modern statistics. It leads an acolyte to the pleasures of gambling, decision trees, Bayesian estimates, discounting, risk preference, derivatives, game theory, and the other refinements of rational action. It is encapsulated not only in introductory economics but also in statistics, operations research, finance, and management science. Its holy grail is the maximization of expected value as determined by a personal, exogenous, stable, consistent utility function; its principal technologies are the technology of estimation and the technology of gambling. It purports to provide an unassailable template for intelligent action.

The second project is a positive one: describing the actual choices of clusters of intelligent humans when they are organized into markets, organizations, or other forms of interaction. It seeks to describe how decisions happen in individuals, groups, organizations, or societies. It explores the consumption choices of consumers, the production and pricing decisions of firms, the investment decisions of holders of capital, the tactical decisions of military organizations, the social policy decisions of political systems, and myriads of other actions that can be fit into a choice framework.

Insofar as the first project is successful, the second is derivative of the first. If all actors have personal, exogenous, stable, consistent utility functions, if they all maximize expected value, and if they all understand the subtleties involved in the process, or if competition among actors eliminates any who do not, then a theory of intelligent choice will also serve as a theory of actual choice; and generating predictions about decisions is indistinguishable from deriving theorems from the axioms of intelligence. Much of modern microeconomic theory, when imagined to be predictive of observed behavior, reflects that spirit. It consists in theorems derived by assuming the pervasiveness of intelligent, rational action.

The evidence is overpowering, however, that the first project has been only partly successful. The ideology of rational, consequential action is widely endorsed in modern cultures, but the detailed pursuit of its canons is much less commonly observed. Situations exist in which individual actors have learned to be guided powerfully by calculations of expected value, but they are not typical. There is little indication that human adaptation inexorably leads to habitual behavior consistent with either the process or the outcome of rationality as defined by decision theory. As a result, the second project is not reliably derivative of the first. Except in limited domains, you cannot find the reality of human choice behavior, either individual or collective, in the theorems of consistent utility and maximum expected value.

By their penchant for ignoring this fact, the core journals and figures of modern microeconomics have, for most practical purposes, abandoned the second project. The field has largely discarded direct empirical observation of firms. Such empirical studies have been replaced in large part by the pursuit of a grand hypothetical question: if the world were one in which Project 1 were universally successful, what would happen?

Abandoning Project 2 by making it derivative of Project 1 is neither a crime nor a total catastrophe. The elaboration of a rational theory of consequential choice is one of the more impressive products of twentieth century social thought, and persistence in it is hardly to be faulted. The resulting theorems have elements of grace and beauty that are much to be admired, regardless of their limited empirical reality. Moreover, there are domains in which Project 1 has been successful to a substantial enough degree that the theorems of microeconomics provide elements of insight. Both the Mazzoleni and Nelson papers and the Handbook of Organizational Economics edited by Robert Gibbons and John Roberts (Gibbons and Roberts, 2013) pay homage to distinguished economists (e.g. Oliver Hart, Bengt Holmström, Edward Lazear, Paul Milgrom, Jean Tirole, and Oliver Williamson) who have, for the most part, labored in that tradition, and who have made important additions to an understanding of rationality in organizations.

As Mazzoleni and Nelson point out, deviant economists have persistently tried to nudge economic thinking in the direction of greater reality. They identify three clusters of ideas that keep intruding (without lasting impact) on economic consciousness: (i) The idea of organizational decision making as stemming from limits on rationality, conflict, and standard operating procedures. (ii) The idea of institutions as shaping economic behavior through norms and regimes of conventionality, and (iii) the idea of evolution as characterizing the historical development of behavioral precepts. Each of these deviant ideas has a long history and each has made inroads into economic consciousness, among other things having in modern times secured Nobel prizes for Herbert Simon, Daniel Kahneman, and Oliver Williamson, and laid the groundwork for a future Nobel Prize in evolutionary economics.

Each of the deviant ideas can claim some elements of modest success; but each has been resisted by the economics establishment or converted by economic theory into a more rational and less history-driven set of ideas than it inherently is. All three of the deviant Nobel laureates mentioned above, for example, embedded their primary contributions within rational/consequentialist frameworks largely consistent with the economic canon: “bounded rationality,” “prospect theory,” and “transaction costs.”

This inclination to fit deviant ideas into some kind of rational perspective serves to obscure an important feature of the deviant perspectives: each of the deviant views can be seen as shaped importantly by the core idea that individual action is driven less by preferences evaluating a calculation of consequences than by rules built around identities and embedded in roles and institutions. An actor tries to identify a situation and a relevant identity and then tries to apply rules that are recognized as appropriate to that situation and that identity. The intellective processes involved are less calculation and estimation than sorting and matching. The underlying logic of action is logic of appropriateness, rather than a logic of consequences.

Theoretically oriented economists who have noticed the importance of rules have tended to see them as learned or evolved solutions to choice problems, and have been inclined to portray rules as the equilibria of some optimizing process. The solution has a certain charm about it and to some extent mimics well-known classic simplifications of evolutionary biology. However, empirical studies of rules (both in economics and in biology) picture them as arising from processes that are considerably indeterminate in their paths and equilibria, thus not easily predicted simply from their initial conditions. Competition does not assure optimization.

The rules and identities of economic behavior are products of their histories. As they are products of their histories, it is usually possible, looking backward, to see the historical path by which they have evolved; and it is often possible to “rationalize” that path as being the result of the choices of rational actors. It is not, however, generally possible to anticipate the path by which the rules will evolve in the future. The instincts of economists to invent consequentialist/rational reasons for observed history are fairly easily satisfied, but efforts to predict the future by invoking consequentialist/rational calculations encounter significant difficulties.

I think it is unlikely that any enduring theory of human choice can ignore the role of rules and identities and the history-dependent character of their development. It is hard to be confident that extensions of classical micro economics will contribute usefully to understanding such phenomena in the foreseeable future. Neither the history of economic thought, nor the contemporary reality of recruitment and training in economics, nor the culture of academic economics provides a basis for great confidence. However, stranger things have happened in scholarly history, and an optimist for microeconomics can take hope from some aspects of the recent history described by Mazzoleni and Nelson and the parallel histories provided by the Gibbons and Roberts handbook, and from the endless intellectual frustrations generated by the ever more baroque efforts of microeconomics to reconcile the pervasiveness of rule-based behavior with the axioms of rationality.

References