But one of the first things taught to economics gradu

 students, even back in my day, is to be wary of the perils of data mining. And the middle-income tra

p is a classic example of the pitfalls of endless number crunching. Give me a database and a po

werful computer, and I can “validate” almost any economic relationship masquerading as an analytical co

njecture. There are five key reasons to dismiss the widespread diagnosis that China is ensnared in the middle-income trap.

First, a middle-income trap may not even exist. That is the conclusion of a rigorous empirical stud

y by Lant Pritchett and Lawrence Summers that covers a broad cross section of 125 economies from 1950 to 20

10. The best they could come up with is a strong tendency for growth discontinuities and mean reversion. At the recent China Development Forum in Be

ijing, Summers went further in assessing likely outcomes in rapidly growing developing economies, dubbing any

mean-reverting slowdown as merely a tendency to close a “post-miracle gap”. Needless to say, the statistical regularity of s

uch periodic growth gaps is very different from the permanent quagmire of a growth trap.