Robust policy schemes for R&D games with asymmetric information

We consider an abstract setting of the differential r&d game, where participating firms are allowed for strategic behavior. We assume the information asymmetry across those firms and the government, which seeks to support newer technologies in a socially optimal manner. We develop a general theory of robust subsidies under such one-sided uncertainty and establish results on relative optimality, duration and size of different policy tools available to the government. It turns out that there might exist multiple sets of second-best robust policies, but there always exist a naturally induced ordering across such sets, implying the optimal choice of a policy exists for the government under different uncertainty levels.