Working Paper: NBER ID: w14552
Authors: Jacques Mairesse; Benoit Mulkay
Abstract: This paper is an attempt to assess the existence and magnitude of local research spillovers in France. We rely on the model of an extended production function (Cobb-Douglas and Translog) with both local and neighborhood R&D capital stocks. We estimate this model on 312 employment areas as of 1999, first for the whole economy, then separately for five large manufacturing industries. The estimated elasticities of productivity with respect to R&D capital are significant and plausible, both within own-area and across neighboring areas as well as within own-industry, but they are weaker across different industries.
Keywords: R&D Spillovers; Productivity; Local Knowledge Spillovers; Spatial Econometrics
JEL Codes: C21; O30; O32; O47
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Local R&D capital (O39) | Productivity (O49) |
| Neighborhood R&D capital (D29) | Productivity (O49) |
| Local R&D capital (same industry) (O39) | Productivity (O49) |
| Local R&D capital (different industry) (O39) | Productivity (O49) |
| Local R&D capital (O39) | Local spillovers (F69) |
| Neighborhood R&D capital (D29) | Local spillovers (F69) |