Rank effect in bureaucrat recruitment

Bureaucrats are a crucial part of states, but only recently has a pioneer literature focused on the determinants of their recruitment. In this article, we study the role of rank in bureaucratic selection in the context of the Chilean civil service, where a decision maker (the President or agency director) chooses a candidate from a shortlist ranked according to a final interview score. Using a regression discontinuity design, we document that first-place candidates have a large advantage in the probability of being selected for top positions in the public sector relative to those in second place. In contrast, all other ranks seem to be irrelevant. Our results are robust to the use of several methods of estimation that take into account the discrete nature of the running variable. Moreover, the effect is associated with a selection process with low levels of competition and low-quality final candidates. •We study the role of rank in bureaucratic selection in the Chilean civil service.•We use a regression discontinuity design based on interview score differences.•We find that barely ranking first gives a large advantage to a candidate.•Results are robust to methods that are consistent to a discrete running variable.•The effect is related with low levels of competition and low-quality final candidates.