A panel data analysis of the effects of voluntary and involuntary separations on unit performance in

We studied the relationship between collective involuntary and voluntary turnover over unit-level performance. Our unique dataset, consisting of panel data with 24-month observations for 232 stores of a large Spanish fashion retail company, enabled us to compare different time lags between the moment of departures and the performance measures using fixed effects regressions. Such analyses revealed that both types of turnover had different consequences. Voluntary turnover was unrelated to store performance. Otherwise, involuntary turnover was significantly and negatively related both to store productivity and efficiency when variables are measured simultaneously, though this effect disappears