Working Paper: NBER ID: w0916
Authors: Kim B. Clark; Zvi Griliches
Abstract: This paper presents the results of a study of productivity growth and R&D in the 1970s using data on narrowly defined 'business units within a firm. Estimates are developed under different assumptions about technology ,industry effects, and changes in the return to R&D over time. The R&D data are broken down into process and product expenditures, and some information is available on past success in developing proprietary technology, andontheincidenceofma3or changes in technology in the recent past. The results suggest a significant relationship between R&D and the growth of productivity; in versions using total factor productivity as the dependent variable, the estimated rate of return to R&D investment is about 20 percent. We find some evidence that R&D has its biggest effect on productivity in those markets where major changes in technology have occurred in the recent past. Previous success in developing proprietary process technology affects total factor productivity directly, but appears to have little effect on estimated returns to R&D. The notion that the productivity of R&D declined in the l970sfinds Little support in this data. Irrespective of model specification, trends in the R&D coefficient are substantively arid statistically insignificant. Our calculations suggest that reduced investment in R&D may have accounted for at least 10 percent of the decline in total factor productivity growth in the l970s.
Keywords: productivity growth; research and development; business level; PIMS database
JEL Codes: O32; D24
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
Cause | Effect |
---|---|
R&D expenditures (O32) | total factor productivity (D24) |
previous success in developing proprietary process technology (O39) | total factor productivity (D24) |
R&D expenditures (O32) | productivity of R&D (O32) |
shift towards more product R&D (O39) | measured growth rate of productivity (O49) |