This paper examines the wage premium to computer use in a developing country:Ecuador. We use different approaches to establish whether the premium is causal.Controlling for a large set of observables we find a wage difference between users andnon-users of around 27%. Using first differences, the...