Why Do Job Referrals Matter?

Evidence from three field experiments

Most workers in the U.S. find their jobs through friends and relatives, and companies are more likely to hire applicants who have been referred. In a series of three experiments in a real labor market, Harvard economist Amanda Pallais and I show that this is not just nepotism: referred workers are actually more productive and have lower turnover than non-referred workers.

By hiring referred and non-referred applicants and comparing their performance on different tasks, we show that applicants who are referred are significantly better workers in ways that are not apparent from resumes alone. In fact, the referral itself signals that the worker is on average 10% more productive and 20% less likely to churn. This performance edge holds even at companies to which they were not referred.

Being referred also makes workers more productive. For example, when referred workers get to work directly with their referrers, they are more productive still — and work 13% longer hours.

Not all referrals are created equal, however. The most valuable referrals come from workers who are themselves productive, and who have close relationships with their referrals.

Full results are published in the Journal of Political Economy; an ungated copy is available here.