Innovation Subsidies: Misallocation and Technology Upgrade

This paper examines TFP effects of size-dependent subsidies that generate misallocation but also incentivize technology adoption at the establishment-level in the context of a model with heterogeneous establishments and endogenous technology adoption. I focus on the subsidy program initiated in 2005 in Indian Iron and Steel industry which was directed to larger plants. Using plant-level data, I found that there was an acceleration of TFP growth but only among plants that used more productive technologies. Instead, I observed a larger reallocation across plants with less productive technologies after the policy change. I used a variant of the Lucas (1978) span-of-control framework with heterogeneous plants and endogenous technology choice at the establishment-level to explore if the subsidy policy contributed to the observed productivity growth. My findings indicate that size-dependent subsidies can increase aggregate TFP if the policy encourages technology adoption at the establishment-level. My calibrated model predicts that the subsidy program in Indiaâs Iron and Steel sector accounted for 1/5 of the observed aggregate TFP growth through three direct channels: First, the subsidy introduces idiosyncracy in the prices faced by individual plants causing resource misallocation (misallocation effect: -2%), Second, the subsidized plants with standard technology switch to the more efficient technology (technology upgrade effect: 49%), and finally, plants make new decisions on occupation and technology use (selection effect: 53%).