Faraway or nearby ? domestic and international spillovers in patenting and product innovation

The diffusion of knowledge plays a central role in endogenous growth theories. Simply put, in these models new knowledge can be generated from preexisting knowledge. In other words, existing knowledge is a pure public good, which can benefit any economic agent anywhere. More generally, endogenous growth theories rely on a broad set of assumptions that have not been tested sufficiently, especially for developing economies. The scope and nature of knowledge spillovers is, however, important for policy, because the presumed positive spillovers can justify government intervention (if the spillovers are localized) or laissez faire (if the spillovers are international). This paper empirically assesses the scope and direction of knowledge spillovers in national patenting and, separately, product innovation by firms. The first set of exercises tests whether the cumulative knowledge specifications of the knowledge production function can explain international patterns of patenting or whether own research and development is necessary to produce patents. The second set of exercises analyzes whether firm product-quality upgrading and the introduction of new products depend on product innovation within industries, within or across countries. The evidence supports the view that existing stocks of knowledge, domestic and foreign, enhance national innovation and entrepreneurship in the form of product innovation. More specifically, the evidence suggests that within-country and international knowledge spillovers are positive, but international spillovers can be negative for firms that are far from innovative firms in terms of productivity. The results depend on the concept of ?distance? between countries and firms.