This paper finds that returns from computer use do not vary with the i
ntensity of computer use. This is evidence against the productivity in
terpretation of these returns and supports the view that returns from
computer use can be attributed to unobserved heterogeneity. (C) 1997 E
lsevier Science S.A.