Changing Aims in Economics.

This book is an expansion of the 1990 Hennipman Lectures that Professor Hutchison delivered in Amsterdam. The text is only 103 pages and they are small pages with large type at that. But there are 65 pages of notes, using smaller type, which are far more interesting than the text. The book is not a scholarly treatise but rather the wise ruminations of a leading scholar in economic methodology. I found the book delightful reading even when I disagreed with his judgments and despite the convoluted sentence structure.

His thesis is simple and straight-forward: the aim of economics since at least Sir William Petty has been ". . . to contribute usefully to less unsuccessful policy-making . . .". To do this ". . . economists need to be able to produce . . . predictions which, on the average, are slightly but significantly less inaccurate and unreliable than would be forthcoming without their input of systematic, more or less disciplined economic knowledge".

Hutchison's first target is "hyper-abstract" theory building which he attributes to the growing dominance of American academic economists since 1945. He claims they have turned away from the traditional aim of economics for the fools gold of a formal theory based on propositions that are not even in principle empirically testable. One reason for this is their reliance on the fundamental theorem of perfect rationality (with perfect information and no uncertainty) on which to erect an edifice of pure theory that has no relevance for economic policy.

The charge is too sweeping. Certainly pure theorists have become the stars amongst academic economists but policy oriented work has continued apace both in and out of the academy. Also he ignores the work done by theorists of the imperfect information school such as George Akerloff, Edmund Phelps, and Joseph Stiglitz.

Hutchison claims that the dominance of the hyper-abstraction theorists paved the way for the rise of his second target--the post-modernist movement composed of permissive pluralists, Marxists, and conversationalists. Their rejection of any methodological roles or standards leads to "anything goes." To quote Hutchison:

If the main aim of academic economics is to be the carrying on of an amusing, but essentially aimless conversation; or the propagation of some kind of mathematical aestheticism, or nebulous "understanding"; or, alternatively, uncontrained individual self-promotion, either in financial terms or in terms of "high status", then indeed, "anything goes", or something like it, may serve as an adequate methodological basis or maxim.

But if the aim of economics is to provide policy guidance, that goal is seriously jeopardized by the disregard or rejection of methodological standards.

Hutchison claims that these anti-positivist, post-modernists have called upon faddish philosophers--Feyerabend, Quine, Rorty--to overawe economists into believing that their time-honored methodological practices are fatally flawed. Thus, Hutchison defends three methodological principles that he sees as under attack.

First, he claims the positive-normative distinction is essential to carry-on the work of economics. This is not the strong claim that economics is value-free but rather a weaker claim that it is useful and indeed essential to be clear about what is being claimed for a proposition in economics when the positive and moral aspects are closely intertwined.

Second, he defends, despite Quine, the distinction between analytic (definitional or tautological) statements and synthetic (empirical-historical) propositions. He argues this is necessary for clarity and takes issue with those who claim that assumptions do not have to be "realistic" in the sense of having independent empirical support.

Third, any theory must result in empirical predictions if it is to have use in pursuing the aim of economics--policy guidance. Again this is only a low-level claim that these predictions are better than can be obtained without the theory. This is not a defense of the much criticized hypothetical-deductive model of explanation which claims that explanation and prediction are symmetrical.

Hutchison's methodological position will appeal to applied economists because it emphasizes the importance of being approximately accurate on important policy issues. However, it will not satisfy methodologists or philosophers of science because it does not attempt to answer the conceptual and philosophical charges leveled against the positivist methodological position.

If the choice is between generating ever more abstract theoretical models or having nice conversations that are policy irrelevant on the one hand and generating low-level empirical generalizations that have some relevance for policy guidance on the other, as an economist of the institutionalist persuasion, I have no doubt which to choose.

Charles K. Wilber University of Notre Dame