Judging Foreign Startups

Can accelerators and investors pick the most promising startup ideas no matter their provenance? We investigate this question using unique data from a global accelerator in which judges across international regions are randomly assigned to evaluate startups headquartered across the globe. Our analysis of this natural experiment shows that judges are less likely to recommend startups headquartered outside their home region by 4 percentage points. Additional information about a startup does not appear to attenuate this bias. Back-of-the-envelope calculations suggest that this discount leads judges to pass over 1 in 10 high-potential startups. Our results reveal that judges are able to discern high from low quality startups regardless of their origin, but they systematically discount foreign ideas relative to local ones