Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Recent papers have proposed a split-sample prediction method to test for structural stability in GMM estimation when the potential break date is treated as known. In this note, the author derives the limiting distribution of the supremum of this test over all possible break dates, thus allowing for endogenous determination of the potential break date. Copyright 1997 by Blackwell Publishing Ltd