Firm performance and ceo turnover in private family firms: evidence from norway

Abstract

In this paper, we examine the relationship between firm performance and CEO turnover within the dynamics of private firms. More specifically, we will compare and analyse the differences of CEO turnover in private family firms and private non-family firms. Our hypotheses revolve around our research question “Is there a difference between private non-family firms and private family firms in the sensitivity of CEO turnover to prior firm performance, and, if so, is it a result of a difference in monitoring?” Each hypothesis builds on existing theories, such as the classical relationship between firm performance and CEO turnover, agency theory and the stewardship versus stagnation perspective. At the heart of our thesis is the analysis of differences in private non-family and private family firms regarding their CEO turnover to performance sensitivity. Surprisingly, we find that private family firms are significantly more likely to replace their CEO if performance is bad than private non-family firms, as measured by lagged return on assets (ROA). The difference becomes even starker when applying Propensity Score Matching, further supporting our results. The results are robust to different empirical models and alternative performance measures. Our findings are surprising given the well-established longer-term perspective in family firms, which includes less frequent CEO turnovers on average. Thus, we believe our results can spur additional discussion on a still limited literature on CEO turnover in private family firms. Moreover, we analyse whether the CEO turnover decision is a result of better monitoring. We find that private family firms are less likely to fire its CEO based on exogenous shocks as measured by industry-wide shocks, and that firm performance increases significantly more in private family firms than in private non-family firms following a turnover. Additionally, we find a significantly negative relationship between prior firm performance and family firms hiring an outside CEO. In our analysis, we use a comprehensive sample of 182 973 private Norwegian non-family firms and 163 758 private Norwegian family firms retrieved from the CCGR database. The logistic model is employed to analyse the relationship between CEO turnover and firm performance, while the GLS linear regression is used to examine post-CEO turnover performance. Lastly, we employ the two-stage regression model to assess relative performance.