Abstract

An important aspect of the study of technological innovations is the explanation of the extent and pace of diffusion. We show that pooling data across vintages of a technology may result in misleading conclusions about the impact of key factors on the duration of time to adoption of the innovation. This is especially important for a technology that affects both product/service quality and a firm's costs of operation to different degrees as the technology evolves over time. Using data on the diffusion of point-of-sale optical scanners between 1974 and 1985, we find that factors such as the stock of prior adopters, household income, family size, the four-firm concentration ratio and item-pricing laws had predictably different effects on the diffusion rate depending on the vintage of the technology. These results are robust to controlling for unobserved heterogeneity among firms, inclusion of additional regressors and a change in functional form.