Economic Theory: Structural Abstraction or Behavioral Reduction?
Yale University - School of Management; Yale University - Cowles Foundation
AGREEMENT ON DEMAND: CONSUMER THEORY IN THE TWENTIETH CENTURY, Philip Mirowski and D. Wade Hands, eds., History of Political Economy (Annual Supplement), Vol. 38, pp. 322-342, Durham, NC: Duke University Press, 2006
In physics, optimization is an organizing principle for natural phenomena. Entropy tends toward its maximum and marbles roll toward minimum potential energy, all without intent or purpose. Injection of this principle into economics initially followed the physicists' organizing perspective and helped develop the powerful insights of the abstract equilibrium theory. However, humans and their institutions being the unit of analysis, economists could not long resist the temptation to give optimization a behavioral spin. Photons may travel along paths that minimize their travel time without intention or purpose; but economists were all too human to think in a similar vein of the people buying ice cream or cars. Once optimization was posited as a behavioral principle of individual human beings, it was easy for cognitive sciences to show that it lacked descriptive validity. Individual behavior is more complex and less predictable. The aggregate characterizations of Walrasian abstraction could not be derived starting from such complex micro-level behavior. If psychology and equilibrium theory were to be reduced into a single science, something had to give. Given the cognitive limitations humans share with all organisms, validity and relevance of the conclusions of equilibrium theory became suspect.
The marriage of economics and computers led to a serendipitous discovery: there is no internal contradiction in suboptimal behavior of individuals yielding aggregate-level outcomes derivable from assuming individual optimization. Individual behavior and aggregate outcomes are related but distinct phenomena. Science does not require integration of adjacent disciplines into a single logical structure. As the early-twentieth-century unity of science movement discovered, if we insist on reducing all sciences to a single integrated structure, we may have no science at all. In Herbert Simon's (1996, 16) words: This skyhook-skyscraper construction of science from the roof down to the yet unconstructed foundations was possible because the behavior of the system at each level depended on only a very approximate, simplified, abstracted characterization of the system at the level next beneath. This is lucky; else the safety of bridges and airplanes might depend on the correctness of the 'Eightfold Way' of looking at elementary particles.
This is the story of how we found that economists can have their cake while psychologists eat it too. Willingness to abandon the reductionist agenda in social sciences - something natural sciences did in the early twentieth century - reveals that the predictive validity of Marshallian supply and demand theory need not depend on the theory's assumptions being literally descriptive of cognitively bounded human agents. The next section discusses the role computers have played in the design and operation of markets, and in modeling and facilitating the decisions of market participants. These developments, combined with the growth of experimental tradition in economics research and the use of such experiments for classroom instruction, set the stage for the discovery that structural properties of markets can be isolated from the behavioral patterns of their participants. The final section explores the scientific antecedents and consequences of this finding.
Number of Pages in PDF File: 21
Keywords: Optimization principle, individual behavior, aggregate outcomes, reductionism
JEL Classification: B29, B49, C91
Date posted: August 31, 2007