Amartya Sen Economic Methodology: Heterogeneity and Relevance THIS IS AN ISSUE-RAISING, QUESTION-MONGERING PAPER. MANY MORE questions will be asked than will, in fact, be answered. The intention is to propose a broad-based critique of contemporary economic method ology starting off from the recognition of the deep-seated heterogeneity of the subject matter of economics. There is a good deal of discontent about the methods and tradi tions in vogue in contemporary economic theory. I shall argue that while parts of that disquiet are well grounded, they have to be sepa rated from others that seem to take insufficient note of the particular nature of the exercises under attack. In particular, some reproaches are based on an inadequate recognition of the diversity of motivations and concerns underlying different types of economic theory. In reading the literature, I am sometimes reminded of an old story concerning two estranged brothers—one a general and the other a bishop—who meet after many years in a desolate railway station and find their old dislikes revived. The bishop asks the general: “Assistant Station Master, tell me when is the next train to London?” The general replies to the bishop: “Madam, in your condition, do you think it is safe to travel?” Our ability to misrepresent or misunderstand what others are up to is quite striking. As a consequence oddly inappropriate broad sides can come mixed with pertinent and penetrating criticisms. The need for sorting out is, thus, quite central. To some extent, this is what ORIGINALLY PUBLISHED IN SOCIAL RESEARCH VOL. 5 6 , NO. 2 (SUM M ER 1989) social research Vol 71 : No 3 : Fall 200 4 583 will be attempted in this paper, though given the vast varieties of meth odologies used and the criticisms made, the ground I shall be able to cover will, of necessity, be relatively tiny. THE HETEROGENEITY OF ECONOMICS There are many different types of problems with which economics as a subject is concerned. This simple and homely thought is worth stating only because the diversities are so often overlooked, insisting on a narrow approach confined to addressing a particular category of problems, ignoring or denying the legitimacy of others. For exam ple, prediction is undoubtedly a major exercise in economics, but an overconcentration on this one function can lead to undue neglect of other exercises—for example, doing economic evaluation or providing adequate description. The narrowness of that prediction-centered approach can be illus trated by considering Milton Friedman’s otherwise illuminating contri bution to “the methodology of positive economics” in which he proposes that “realism” may have to be judged “by seeing whether the theory works, which means whether it yields sufficiently accurate predictions” (Friedman, 1953: 15). The ability of a theoiy to yield sufficiently accu rate predictions is undoubtedly important, and Friedman does indeed throw much light on this aspect of the problem. But there are many other worthy economic exercises that cannot possibly be judged in this light (Sen, 1982; Hicks, 1983). For example, descriptive statements such as “unemployment has increased” or “inflation is slowing down” call for some way of assessing the reality of what is occurring. They are primar ily concerned with reflecting adequately what has happened rather than with foretelling what will happen or what would have happened had the circumstances been different. Similarly, evaluative statements involv ing valuation and assessment may depart from purely predictive exer cises and the methodology related to that exercise. Whether or not one accepts John Hicks’s powerful argument to the effect that economics is “a discipline, not a science” (Hicks, 1983), we have to acknowledge that there is much more to economics than predictive efficiency. 584 social research At the veiy least, the subject of economics includes three diverse, though interrelated, exercises: (1) predicting the future and causally explaining past events, (2) choosing appropriate descriptions of states and events in the past and the present, and (3) providing normative evaluation of states, institutions, and policies. Predictive economics is directly concerned only with the first of these exercises. The “method ology of economics” has to admit enough diversity to be able to deal with the other classes of problems as well. These...
Stored only when both papers are in the corpus — 98 citations exist upstream.