This study argues that an evaluation of the relationship between CEO turnover and corporate performance can be usefully approached in three steps. (1) An empirical longitudinal analysis of 88 German corporations which had experienced a change in CEO shows that different kinds of CEO turnover are associated with different developments in published performance during the succession period. (2) Using a structural equations model it is possible to demonstrate the aptness of a hypothesis on performance variation based on agency theory and using empirical data on involuntary CEO turnover. (3) By introducing an indirect method of measurement it is shown that CEO "influence" explains a substantial part of the observed variation in performance during the succession period.
