Count-data regression models of the time to adopt new technologies

This paper presents a framework for interpreting and using the count-data model for estimating the time of technology adoption. The Bernoulli trials of the negative binomial model are interpreted as the stages involved in a potential adopter learning and updating information relevant to a new technology. Empirically, the paper estimates the Poisson, the generalized negative binomial, and the geometric models in order to identify the determinants of computer adoption on farms in California.