Structural Discrete Probability Models Derived From Theories Of Choice

Daniel McFadden

An object can have no value unless it has utility. No one will give anything for an article unless it yield him satisfaction. Doubtless people are sometimes foolish, and buy things, as children do, to please a moment's fancy; but at least they think at the moment that there is a wish to be gratified.

—F, M. Taussig, Principles of Economics, 1912 5.1 Economic Man

The classical economists made the assumption of homus economicus virtually tautological: if an object were chosen, then it must have maximized the utility of the decision maker. By contrast, contemporary economic analysis of consumer behavior has focused on the objective market environment of economic decisions and has excluded whim and perception from any formal role in the utility maximization process.1

From the standpoint of the observer unmeasured psychological factors introduce a random element in economic decisions. The result is a probabilistic theory of choice which has many features in common with psychophysical models of judgment (Coombs 1964, Luce and Suppes 1965, Bock and Jones 1968, Krantz, Luce, Suppes, and Tversky 1971, Krantz 1974).

Probabilistic choice models lend themselves readily to econometric implementation, particularly for choices among discrete alternatives. This chapter develops and compares a number of these models in forms suitable for econometric applications.

This research was supported in part by the National Science Foundation, through grant SOC75-22657 to the University of California, Berkeley. Portions of this chapter were written while the author was an Irving Fisher Visiting Professor of Economics at the Cowles Foundation for Research in Economics, Yale University. An early version was presented at the Third World Congress of the Econometric Society, Toronto, Canada, 1975. The author has benefited greatly from discussions with Amos Tversky at the formative stage of this chapter and has borrowed freely from his ideas. Charles Manski and Steven Cosslett have also provided useful comments. The author retains sole responsibility for errors.

1. Also excluded in the conventional consumer analysis is consideration of procedural rationality, the question of how an organism with perceptual and computational limits makes a decision; see Simon (1978). This chapter will not take up the question of probabilistic choice theory in the presence of bounded rationality. However, we note that the distributions of demand attributed in this chapter to taste variation or errors in judgment could often be reinterpreted as a consequence of bounded rationality, and vice versa.

Was this article helpful?

0 0

Post a comment