The Effect of Quits on Worker Recruitment: Theory and Evidence

Recruitment effort by a firm can signify one of two things: a desire to expand or a need to replace workers who have quit profitable positions. Standard matching models with on-the-job search treat these two recruitment activities as the same. Yet, we provide empirical evidence that suggests these two activities differ in the sense that, all else equal, an establishment is much more likely to post a vacancy and hire a worker if a worker has quit a position at the firm. Our evidence is robust to a variety of controls, including establishment fixed effects. One natural explanation for this is that workers who quit leave behind firm-specific physical and organizational capital, thereby making replacement hiring less costly than the creation of a new position. To this end, we develop a matching model with on-the-job search and multi-worker firms that differentiates between the cost of creating a new position and the cost of adverting for an existing opening. The model naturally creates a distinction between worker and job flows and, through endogenously-determined thresholds for separations, worker replacement and position creation, produces rich firm-level employment dynamics that are broadly consistent with our empirical evidence.