A Two-Phase Longitudinal Model of a Turnover Event: Disruption, Recovery Rates, and Moderators of Collective Performance
研究个体离职事件如何分两阶段影响集体绩效：先造成即时破坏，后逐渐恢复；发现离职者的职位（经理或员工）和团队相互依赖程度会调节影响，经理离职的恢复更慢。
Blending theory on collective turnover and group adaptability, this paper develops a two-phase longitudinal model that explains how and why an individual-level turnover event has effects on collective performance. Phase 1 (disruption) is marked by a sudden and negative change in unit-level performance, while Phase 2 (recovery) entails a gradual increase in unit performance over time. We further propose that the positional distribution (manager or employee) of the turnover event and the amount of interdependence within the collective will influence the consequences of a turnover event. Using a sample of 524 branches of a U.S. bank, the results largely support the hypotheses as a turnover event leads to an immediate and negative change in branch-level performance. Notably, there is an interaction between manager and employee turnover events that significantly decreases performance. In the recovery period, branches recover more quickly after losing an employee than they do after losing a manager. However, the moderating effect of branch-level interdependence depends on the position, and there is no statistically significant recovery from a manager turnover event within our performance window. These results highlight the insights of our two-phase model and the need to study collective turnover dynamics over time and across different positions.