Mayor games in Bavaria: Self selection of local politicians is not influenced by constitutionally defined remuneration increases! Quasi-experimental evidence from Germany

Remuneration for politicians has been discussed frequently and thus, has been the subject of many reforms in recent decades. Remuneration has an influence on whether or not citizens choose to run for office as well as the amount of effort they might invest once elected. In this paper, I investigate the impact of variations in remuneration on the quality of politicians. It is well known that the amount of remuneration of a Bavarian mayor is regulated by constitutionally defined population thresholds. Utilizing this exogenous variation in the remuneration of politicians, I apply the quasiexperimental regression discontinuity design to six different population thresholds. My analysis demonstrates that positive or negative correlations between remuneration and quality of politicians are found for all six thresholds when simple ordinary least squares regression analysis is used. However, further analysis using regression discontinuity design shows that, in fact, no causal relationship between remuneration and quality exists. This result holds when using the Difference-in- Discontinuities approach. When comparing my results to other empirical findings, one possible explanation for the absence of effects might be rooted in the fact that Bavarian mayors do not face term limits.