chapter  7
14 Pages

Three conventionalist responses: Machlup, Samuelson and Friedman

From the early 1930s to the late 1940s one can readily admit that “Margin-

alism faced a scientific credibility problem” (Hands 2009: 840). This

chapter maintains that the disconcerting anti-marginalist attack was effect-

ively overcome thanks to the combined methodological stratagems of the

most prominent priests of neoclassical orthodoxy of the post-war period,

namely Fritz Machlup, Paul Samuelson and Milton Friedman. Machlup

(1946, 1955) first attempted to eschew the empiricist critique by claiming

that the fundamental assumptions of economics are not directly testable

and cannot be refuted by empirical investigation, reintroducing thus con-

ventionalism to economic methodology (Latsis 1976). Samuelson in 1947

gave a comparable response: a neoclassical theory based upon economic

rationality produces “operationally meaningful theorems”, that is, “hypo-

theses about empirical data which could conceivably be refuted if only

under ideal conditions”. His “Revealed Preference” theory is a thorough application of this methodological point of view. But the most celebrated

conventionalist response to the empiricist critique is undoubtedly Fried-

man’s celebrated essay of 1953. According to Friedman “theories are good

for predictions only”. Therefore, it is useless to criticize the unrealistic nature of economic assumptions like economic rationality, since the aim of

any assumption is only to provide the basis for successful predictions. This

is the meaning of his famous F-twist: “the more significant the theory, the

more unrealistic the assumptions”.1