EXPLORING THE RELATIONSHIP BETWEEN RESEARCH-AND-DEVELOPMENT AND PRODUCTIVITY IN FRENCH MANUFACTURING FIRMS

Using a newly available dataset on the R&D investment of individual Fr ench manufacturing firms for the 1980s, we replicate and update a seri es of studies on French R&D and productivity at the firm level from th e 1970s, and evaluate the robustness of methods currently used to meas ure the private returns to R&D. Our main findings are: Having a longer history of R&D expenditures helps improve the quality of the R&D elas ticity estimates, but the choice of depreciation rate for R&D capital makes little difference. The correction for double-counting of R&D exp enditures in capital and labor is important and may be interpreted und er certain conditions as converting a measured 'excess' rate of return to a total rate of return to R&D. We show that the direct production function approach to measure returns to R&D capital is preferred on se veral grounds over the rate of return variation used in the past. Fina lly, as in the 1970s, the productivity of R&D capital for French manuf acturing firms in the 1980s is positive; how strong and robust depends on whether we control for potential industry and firm effects.