Abstract:
There is substantial empirical evidence that innovation is geographically concentrated. Unlike what is generally assumed, however, it is not clear that localised knowledge spillovers provide a theoretically valid explanation for this. Studying spillovers of cost-reducing technology between Cournot oligopolists we show that 1) localised knowledge spillovers of any level do encourage agglomeration, but 2) whether this leads to higher levels of effective R&D depends on the type and level of knowledge spillovers, the number of firms, and the industry's R&D efficiency.
