Elected or appointed? How the nomination scheme of the city manager influences the effects of government fragmentation

Empirical research on the causal effect of government fragmentation derives diverse results. One reason for this diversity might be that studies are typically settled in a variety of institutional environments. To assess the extent to which the local political system might shape the effects of fragmentation, this study measures the causal effect of a change in the nomination scheme of the city manager on the council size effect. I combine a regression discontinuity design with a difference-in-differences approach in a large panel dataset of German municipalities. The study finds that when the manager is appointed by council, there is no significant council size effect, but that there is a negative effect when the manager is elected by voters for those expenditure categories over which the manager has the most discretion. These findings suggest that the nature of the political system does indeed matter.