A demonstration of the importance of temporal considerations in the prediction of newcomer turnover

Abstract

We theorize about and examine empirically the increase in predictive strength of antecedents of voluntary turnover over the first three years of employment using survival analysis with time-varying covariates and period effects. On the basis of employee survey data gathered from 240 newcomers working in a retail bank and organizational turnover records collected over multiple years, we find that job embeddedness and job satisfaction increase in their ability to predict voluntary turnover as tenure increases. The results of this study emphasize the importance of integrating time into theories to better understand the dynamics of the turnover process. Specifically, additional theorizing about when different effects may be manifest and how long they will persist is critical. Equally important is the development and application of research methods that are suited to dynamic modeling, such as those introduced here.