The paper analyzes the design of industrial policies, in the form of subsidies to innovation activity or to local production, when domestic firms are inefficient and there is a risk of supply-chain disruption. We first establish a case for research subsidies, since private investment (to improve the inferior technology) is lower than the socially optimal one. We next show the equivalence with subsidies to (inefficient) local production in case of intertemporal economies of scale. Then, within a general frame- work, we analyze profit and welfare maximizing investments and optimal subsidies in case of segmented markets and an integrated market organized as a duopoly, a monopoly or a research joint-venture. We show that research joint ventures or a public research center socially outperform the other environments since they benefit from a larger integrated market and a wider circulation of the innovation while preserving a competitive market. Finally, in large markets with significant technology gaps, it may be convenient to concentrate all the research in a single lab while maintaining a competitive market.
