Essays on innovation, R&D investment, and productivity

Chapter 1 describes innovation and research activity of two cross-sections of Italian manufacturing firms over the periods 1992--94 and 1995--97. Italy's productivity growth values over the 1990s had been not worse than any big OECD country, and yet its research expenditures were much lower. Provided that innovation is the engine of growth, Italian firms must be innovative through different channels. However, research firms get a payoff over non-active. Factors positively linked to innovations are geography, exports, and size. The relationship between human capital as educated workers and innovations is puzzling. Only product innovations in 92--94 have a positive correlation with the stock of human capital of the firms.;Chapter 2 estimates the effect of innovations on productivity growth and the productivity elasticity to R&D capital. Firms need to introduce or improve products or processes over time to satisfy market needs and because diffusion phenomena contribute to increase competition. We introduce explicitly innovations into a production function and evaluate their contribute to growth. Innovations push productivity growth up to 18%. We then calculate the TFP growth elasticity to R&D through the innovation effect (0.011). Furthermore, we allow R&D capital or labor to enter the technological coefficient as conventional, and estimate TFP elasticity. R&D's changes increase TFP level more than TFP growth. TFP level elasticity however is still smaller than existing evidence for French and German firms, but comparable to US elasticity in the 1980s. We finally compute the marginal product of R&D capital. On average, the estimated rate of return to R&D is comparable to French and German cases.;In Chapter 3 evidence is presented on the elasticity of private R&D spending on its price. A censored panel-data regression model with random effects is applied to a balanced panel of 726 firms. Implied estimates point out that firms' response to policy measures (including tax credits), aimed at reducing the user cost of R&D capital, is likely to be substantial (1.50--1.77). We also find that the elasticity of R&D spending is higher in recession than in expansion.