The piece C. Hoyt Bleakley wrote about Robert W. Fogel is loving and accurate ("What if railroads didn't change the world?" Fall 2013). Fogel hired me as an assistant professor in 1968 (I stayed until 1980), and we were close colleagues.
Hoyt emphasized the Fogel method as "slogging through the data," and hails him as a pioneer of "data crunching in microeconomics." But "data crunching" these days—and too often Bob went along—means grossly misusing tests of statistical significance.
And "data" were never what Bob collected. The word means in Latin "things given." In none of his work did Bob rely on what other people had given him. The word data embodies an attitude to facts fatal to serious science. We need a new word, capta, which means "things seized."
Finally, Bob's theory of science did not describe what he actually did. He discovered the railroads did transform the economy, but did not much increase its size. Did he discover it by constructing new canals, as he claimed to Hoyt? By "diving deep into data?" No. Without Bob's firm command of Chicago-style price theory, all his work would have been pointless. I remember trying to persuade him that his great book on railroads could be reduced to a three-line proof: Transport was 10% of GDP; railroads were 50% of that; they reduced transport costs by 50%. Voila: 10%x50%x50%=2.5% of GDP, which is what he had discovered.
Bob didn't like it, but it's the real economics in the book. In it he described a linear programming exercise, but then he never did it. Such Baconian "slogging" is not science. Asking brilliant questions is.
Deirdre N. McCloskey
Distinguished Professor of Economics, History, English, and Communication
University of Illinois at Chicago