Sifting and sorting: Personal contacts and hiring in a retail bank

Using unique data from a large retail bank, we investigate the theoretical mechanisms by which preexisting social ties affect the hiring process. By focusing on a single, large employer; we are able To identify, the recruitment practices and hiring criteria used during screening for entry-level positions. This method allows us to assemble data for the pool of candidates at multiple phases of the hiring process and to conduct empirical tests of the various roles that personal contacts might play at each stage. Because we are able to treat hiring as a process, rather than as an event, we can also consider the possible selection biases introduced by the multistage screening process. More specifically, we study how employee referral (i.e., being recommended by a current bank employee) affects an applicant's success at multiple stages of the recruitment process, and we examine the cumulative effects of referral status on the chance of being offered a job. Results of probit models indicate that, controlling for other factors, referrals have advantages at both the interview and job-offer stages compared to external nonreferral applicants. Consistent with theoretical arguments that referrals are prescreened by current employees, our results show that referral applicants present more appropriate resume's than do nonreferral applicants. Referral applicants also are more likely than nonreferrals to apply when marker conditions are more favorable. Nevertheless, resume quality and application timing cannot explain referrals' advantage at the interview and hire phases. We discuss the theoretical implications of these findings.