Murray Rothbard was a foremost defender of apriorism in economics and a critic of positivism, empiricism, and scientism (1, 2, 3). But he was not opposed to quantification, particularly in the study of economic history. In fact, he strongly supported the use of statistics in doing applied economics. I was reminded of Rothbard’s view of data when rereading Strictly Confidential for this week’s Rothbard Graduate Seminar. Consider this passage from his lengthy and detailed review (1961) of George B. DeHuszar and Thomas Hulbert Stevenson’s History of the American Republic:
[A] critical defect is the almost complete absence of any quantitative or numerical data. It is often difficult to find the dates at which happenings occur, so vague and imprecise is the narrative. Apart from a few references to population figures, there are virtually no statistics of any kind in the work.
Now, I am an open and long-time condemner of the overuse of statistics, and I deplore as much as anyone the new trend in “scientific” economic history to hurl vast quantities of processed statistics at the reader, and conclude that one has captured the “feel” and essence of the past. But some statistics, surely, are necessary; and it becomes annoying to read constant references to “increases” in steel production, or living standards, or whatnot, when not the foggiest quantitative notion is presented to the reader of how large these increases and movements are. There is also an almost desperate need to present government budget statistics, so that the reader will know how large government in relation to the private economy has been in any given era; but neither in this nor in any other area does DeHuszar give a shred of quantitative data.
Rothbard is equally critical of the authors’ conceptual framework (more precisely, their lack of a consistent conceptual framework) and their failure to bring to life the grand sweep of American history. But this emphasis on the need for quantification may surprise some of Rothbard’s critics, and some of his defenders too.
