Full Program »
Revisiting Friedman’s F53: Hedging, Karl Popper, Frank Knight, And Max WeberThe paper claims that Friedman’s famous 1953 essay “On the methodology of positive economics” (in Essays in Positive Economics, Chicago: University of Chicago Press, 1953, pp. 3-43; henceforth F53) has been seriously misread in the past. In order to argue for this claim, I first show why it is easy to misread F53: F53 excessively uses a particular hedging technique, namely, scare quotes, which makes is very difficult to understand what is really meant in F53. In particular, many of the fundamental concepts of F53 are only used with scare quotes, especially the term “assumptions”. The term “assumptions” occurs 73 times in F53, 37 times with scare quotes. Second, I show that there are several statements in F53 that are rather strange and not compatible with any of the usual readings of F53. Of course, many other authors have noticed these strange statements but usually, they were found to be just untenable and were thus dismissed. The most prominent and often discussed statement of this kind is: “Truly important and significant hypotheses will be found to have “assumptions” that are wildly inaccurate descriptive representations of reality, and, in general, the more significant the theory, the more unrealistic the assumptions (in this sense).” (p. 14) My claim is that as long as one does not understand this statement, one misses a substantial part of F53’s message. Third, I show that an alternative reading of F53 can be achieved if one takes seriously Friedman’s reference to ideal types in Sect. V of F53 (“ideal types” is used six times in this section). Friedman was familiar with Max Weber who introduced the concept of ideal type because in 1935-36, he had attended a seminar on Max Weber given by his teacher Frank Knight at Chicago. Frank Knight was an ardent supporter of Weber’s ideal type methodology for economic theory. Given that in F53’s view ideal types are a fundamental building block of economic theory, it becomes clear why both an instrumentalist and a directly realist reading of F53 is inadequate. It further becomes clear what F53 means by its famous as–if-methodology: it means analyzing a phenomenon in terms of only one ideal type, in spite of contributions of other ideal types to the phenomenon in question. Fourth, the reading of F53 in terms of ideal types makes it intelligible why F53 so emphatically introduces elements from Popper’s falsificationist methodology. Friedman had not read Popper by 1953, but was familiar with Popper’s central methodological ideas due to a personal meeting in 1947 (neither Popper, nor Knight, nor Weber are quoted in F53). Hypotheses that are articulated on the basis of ideal types contain a fundamentally speculative element. In empirical science, they must therefore be severely restrained by empirical control, and Popper’s falsificationism informs F53 how to exert this control. Fifth, I show that most of the strange passages of F53 make very good sense under the new interpretation.
Department of Economics