Abstract
Technological advances are transforming how firms screen and evaluate job applicants. This paper studies how automating the collection and evaluation of interviews reshapes the applicant pool and the downstream consequences for evaluation and hiring. In a natural field experiment with over 3,000 applicants, we randomize candidates into asynchronous audio or video interviews, live online interviews, and a no-screening benchmark. We find that asynchronous interviews decrease application continuation by over 50 percent, particularly among the most qualified applicants and women, implying that interview technologies systematically change who enters the hiring process. A complementary vignette experiment shows that this deterrence is driven by perceptions of increased competition and reduced fairness. Automating interview evaluation via AI, which has made asynchronous interviewing scalable, generates systematically different evaluations from humans – human evaluators assign lower scores to women and underrepresented minorities – and are more predictive of subsequent labor market outcomes. Taken together, our results show that interview automation not only changes how applicants are assessed, but also who enters and remains in the hiring process.