Friedman’s t-ratios were overstated by nearly 100%

From Prof Sir David F. Hendry.

Sir, Wolfgang Münchau asserts, in “Perils of placing faith in a thin theory” (April 22): “Unlike Profs Reinhart and Rogoff, Friedman did have a theory when he pushed monetarism in the late 1960s and the 1970s. There were no fat-finger errors in the 1960s equivalent of an Excel spreadsheet. He had solid empirical evidence.”

I regret to inform him that his statement is false on two counts. First, there were serious errors resulting from Milton Friedman’s failure to understand the computer output produced by his “phase averaging” process, such that the t-ratios he reported for UK money demand were overstated by nearly 100 per cent.

Second, as documented by Neil R. Ericsson and myself in the American Economic Review, 1991, “An Econometric Analysis of UK Money Demand in Monetary Trends in the United States and the United Kingdom by Milton Friedman and Anna J. Schwartz”, almost every empirical claim they made about UK money demand was refuted.

David F. Hendry, Oxford University, UK