In Wednesday’s FT (“Politicians will always succumb to the need to bend data,” March 13), John Kay confesses that, as a young researcher, he had confidence that “the statistical information that accompanies government pronouncements was reliable.” Mr. Kay “no longer think[s] that.”
Allow me to suggest that Mr. Kay’s fit of youthful enthusiasm for government stats would have been reined in, had his tutors assigned the great Oskar Morgenstern’s classic, On the Accuracy of Economic Observations, which was first published in 1950. Morgenstern’s gem lays out just how the incompetence and trickery of many governments (and businesses) often render data, well, less than reliable.
I require my students to study Morgenstern with the most careful and anxious attention. Now, I am assigning Kay, too.