I remember being in a conference with Ludwig von Mises in the sixties at FEE [the Foundation for Economic Education]. And I asked him about Friedman and economics. And he waved his hand in the typical Austrian way and he said: “Friedman is not an economist. He’s a statistician.”
Now in describing Friedman in these terms, Mises was not name calling but had a very specific meaning in mind. For Mises (pp. 247-48) a “statistician” was someone “who aim[s] at discovering economic laws from the study of economic experience.” But Mises maintained that statistics is not a method useful for research in economic theory because it deals with historical facts. According to Mises:
Statistics is a method for the presentation of historical facts concerning prices and other relevant data of human action. It is not economics and cannot produce economic theorems and theories. The statistics of prices is economic history. The insight that, ceteris paribus, an increase in demand must result in an increase in prices is not derived from experience. Nobody ever was or ever will be in a position to observe a change in one of the market data ceteris paribus. There is no such thing as quantitative economics. All economic quantities we know about are data of economic history.
Indeed in his magnum opus, A Monetary History of the United States, co-authored with Anna Schwartz, Friedman confirmed the accuracy of Mises’s characterization. In their Preface (p. xxii), Friedman and Schwartz stated that their aim in writing the book was “to provide a prologue and a background for a statistical analysis of the secular and cyclical behavior of money in the United States and to exclude any material not relevant to that purpose.” In the final chapter, entitled “A Summing Up,” the authors (Friedman and Schwartz, p. 676) listed three propositions regarding money that they discovered to be “common” to U.S. monetary history and concluded, “These common elements of monetary experience can be expected to characterize our future as they have our past.” It would be difficult to find a better expression of the statistician’s view of the social world.