Japanese R&D Profitability: Industry Production Function Estimation Using Panel Data

This article examines the contribution to profit from research and development (R&D) using industry-level accounting panel data for eight industries in Japan from 1986 to 2012. Problematically, simple least-squares regression estimation of production functions, where the authors specify sales or value-added as the explained variable and investment in R&D as the explanatory variable, involve endogeneity. Two possible ways of addressing this are the instrumental variables method and another that utilizes the orthogonality between error terms and appropriately time-lagged explanatory variables. The authors compare how both these methods eliminate endogeneity in the estimated production function and thus improve the accuracy of estimates of the rate of return on R&D. These findings thus contribute to the managerial decision-making process on R&D strategy by providing insights into the precise contribution of firm R&D to profit.