Milton Friedman’s ‘The Methodology of Positive Economics’, which is the lead article in that economist’s 1953 book Essays in Positive Economics, is probably the best known piece of methodological writing in economics. It is also a marketing masterpiece. Never before has one short article on methodology been able to generate so much controversy. It has been reviewed often, usually negatively. Yet ironically, the methodological prescriptions advanced in his essay have become widely accepted among many working economists. And this has happened without Friedman ever having directly responded to his critics! What follows is still another attempt at an analysis of Friedman’s classic. A brief summary of the article reveals it to be a statement of prescriptivist methodological principles that reflects, in many respects, what we have called the mature positivist view of economic science. In at least one respect, however, Friedman’s position is unique: his belief that the ‘realism’ of assumptions of theories does not matter. This position, when coupled with his emphasis on prediction in science, allows us to interpret Friedman not as a positivist, but as a ‘methodological instrumentalist’.
Friedman begins with the assertion that many disagreements in economics, and especially those which concern policy decisions, arise from confusion over the distinction between positive and normative economics. He feels that more agreement on the positive effects of a certain policy action would lead to more clarity (and perhaps even more agreement) on normative issues. The stated goal of his article, then, is to provide our discipline with a more solid foundation and understanding of what properly constitutes positive