The number of parties in government is usually considered to increase spending. We show that this is not necessarily the case. Using a new method to detect close election outcomes in multi-party systems, we isolate truly exogenous variation in the type of government. With data from municipalities in the German state of Bavaria, we show in regression discontinuity-type estimations that absolute majorities spend more, not less, and increase the property tax rate. We also find weakly significant results for increases in debt. Politically, our results show that the mayor that heads an absolute majority of his own party gains the most, but the party itself does not.
fiscal spending local election absolute majority municipality data regression discontinuity