Political fragmentation and fiscal policy : evidence from German municipalities

The government fragmentation hypothesis (GFH) states that coalition governments spend more than single-party governments due to an underlying common pool problem. Using a large panel data set on 604 local governments in the German state of Baden-Württemberg for the 1994-2014 period, I test the GFH for tax rates, the growth in government debt as well as expenditures and its sub-categories. Studies using standard regression methods fail to identify causal effect as the type of government is generally not random. I apply a RDD, that exploits quasi-random variation generated by close elections. I add external validity to the recent quasi-experimental literature by investigating the GFH for a mayor-council system. I find that contrary to the theoretical prediction, coalition governments do not increase taxes and government debt. There is a non-robust, negative effect on total expenditures, which is mainly driven by administrative expenditures and material expenditures.