Recent empirical studies point to higher wages enjoyed by workers in e
nvironments where new technologies are intensively used. An examinatio
n of the 1984 and 1990 establishment-based WIRS reveals similar patter
ns. This paper argues that endogeneity bias is endemic in these result
s. Controlling for this endogeneity bias suggests that the estimated i
mpact of new technologies on wages is seriously upward-biased. It is m
ore likely that the earnings-technology correlation is driven by the i
mpact of higher earnings on technical change rather than vice versa.