The Effects of Double-Counting and Expensing on the Measured Returns to R&D

This article focuses on the effects of double-counting and expensing on the measured returns to R&D. The contribution of research and development (R&D) to economic growth has been measured in two general ways. The first is to compute total factor productivity in a growth accounting framework and to attribute this growth to R&D. The prevailing view is that, in the presence of double-counting, the measured contribution of R&D represents the return above and beyond the normal remuneration to traditional capital. Author demonstrate that this excess returns interpretation (ERI) is essentially correct in the growth accounting framework and that the resulting bias in the measured contribution of R&D to growth is large. In postwar U.S. manufacturing the measured residual is biased downward by as much as 30%. The expensing bias is downward and reinforces the excess returns bias in this case, and the total bias is large. In both the growth accounting and econometric contexts, the magnitude of the biases may vary across samples and over time. There is simply no substitute for properly measured variables.