We analyze the design of a unique Spanish public program aimed at recruiting high-quality researchers in public research centers: the Ramón y Cajal Program. We claim that, after a number of calls for applicants, the program design changed in response to agents’ needs. Exploiting data on applications and candidates, we find that the new program design led to significant changes in the probability of being awarded a contract. In particular, opportunities for candidates with no attachment to the system were equalized.

Key words: brain gain, government research programs, human capital, policy evaluation.
JEL Classification: O38, D78, C21, C78.