Selection Or Self-Rejection? Applications Into A Treatment

We develop a new method to estimate the returns to R&D, their distribution, and their determinants. We model a continuous optimal treatment with outcome heterogeneity, where the treatment outcome depends on the applicant’s investment. The model takes into account application costs, and isolates the effects of the treatment on the public agency running the treatment program. Under the assumption of a welfare-maximizing agency, we identify general equilibrium treatment effects and social returns to R&D. The model yields a restriction on the application equation that helps identify the parameters of the cost-of- application function. The proposed estimation strategy is applied to project level data from the granting process of R&D subsidies. We find that larger firms have higher marginal profitability of R&D. Rates of return on R&D are high and their distribution skew. Agency specific returns are non-monotonic in private returns. Project level spillovers are linear in R&D. The median increase from subsidies in the agency’s utility not appropriated by the applicant is 16 000€. Application costs increase with the profitability shock and ignoring application costs severely biases the estimated rates of return upwards.