Labour turnover and its effects on performance: an empirical test using firm data

In this article we test the hypothesis that the relationship between labour turnover and the economic performance of the firm is bell-shaped: a turnover level too low has a negative effect and likewise does a level too high. Our analysis is based on economic performance data of 110 offices of a temp agency. Since these offices vary highly in labour turnover but are similar in product and operational management, the data enabled us to control for a number of important intervening variables. From a regression analysis it could be shown that labour turnover indeed is related to office performance in a curvilinear way, indicating that it is especially excessive turnover that matters. This result proved robust for both performance level and change of performance as the dependent variables.