The Effects of Digital-Technology Adoption on Productivity and Factor Demand : Firm-Level Evidence from Developing Countries

This paper presents firm-level estimates of revenue-based total factor productivity premiums of manufacturing firms adopting digital technology in 82 developing economies over 2002-19. The paper estimates productivity using the control function approach and assuming an endogenous revenue-based total factor productivity process, which is a function of multiple firm-choice variables. It estimates the effects of digital technology adoption, learning by exporting, and managerial experience on revenue-based total factor productivity and factor demand. The results reject the 0 hypothesis of an exogenous revenue-based total factor productivity process, in favor of one in which digital technology adoption, along with the other choice variables, affects revenue-based total factor productivity and factor demand. The estimated premiums are positive for 67.3 (email adoption), 54.6 (website adoption), 59.4 (learning by exporting), and 60.6 (managerial experience) percent of the sample. The probability-adjusted median (log) revenue-based total factor productivity premium associated with email adoption is 1.6 percent and that of website adoption is 2.2 percent, with the latter being higher than the premiums corresponding to exporting and managerial experience. On average, changes in digital technology adoption, email, and website are labor and capital augmenting. The paper also explores the role of complementarities among the firm choice variables