City Manager Turnover Revisited: Effects of the Institutional Structure and Length of Tenure on City Manager Turnover

This study attempts to test a well-known model of city manager turnover. We use a Binary TSCS (Time-Series Cross-Section) analysis to address the problem of duration dependence, which typically occurs when the assumption is that a turnover decision at a certain point of time is entirely independent of previous decisions in the turnover data. In essence, manager turnover should be understood not as a single event at a given point in time but rather as being affected by the duration of tenure. Our empirical analysis of turnover data from 156 large cities in the United States for the period 1984–1999 indicate that, particularly when the length of manager tenure was controlled for, political push factors had significant effects on city manager turnover that were not detected by means of the ordinary logit analysis. The result on time duration suggests that longer tenure can reduce the likelihood of city manager turnover.