Work in Progress

“Government Fragmentation and Political Corruption”
with Albert Solé-Ollé and Pilar Sorribas-Navarro.

Abstract: This paper studies whether government type –non-majority v. majority– has an effect on political corruption. We use data from Spain’s local councils, where a proportional electoral rule generates a substantial share of non-majority governments. We focus on the housing boom period of 1999-2007, which witnessed a large surge in political corruption related to local zoning decisions. To identify the treatment effect we compare close-elections non-majorities and majorities that are identical in terms of relevant political traits (i.e., council size, mayor’s ideology, and number of opposition and local parties). Our findings allow us to reject the hypo-thesis that non-majority governments are on average more corrupt than the majority ones. Moreover, we also find that some non-majority governments –those with a pivotal party, which are the ones able to enter agreements with both ideological blocs– are less corrupt. This result is consistent with a story in which coalition partners are more willing to denounce political corruption when they have other agreement options.

“Legislature vs. Government fragmentation: The Political Economy of Fiscal Consolidation under Fiscal Rules”


This paper analyses the causal effect of political fragmentation (both in the legislature and government) on fiscal consolidation under fiscal rules: when deficits and new debt are not feasible. Fragmentation is defined as the number of parties represented. The entrance of an extra party in the legislature affects the way governments pursue consolidation. When fragmentation is low (i.e. majority governments), the priority is on reducing current expenditures. Increasing fragmentation, however, shifts the focus towards increasing current revenues. Political fragmentation thus has a measurable effect on the size of the government budget: the larger the number of parties in the legislature, the bigger the budget. Heterogeneous analysis shows that severe financial situations neutralise this behaviour and governments base consolidation on expenditure cuts indistinctively of their fragmentation. Interestingly, the effect of legislature fragmentation on consolidation remains even when it does not affect government fragmentation. That is, majority governments modify consolidation behaviour when the legislature becomes more fragmented. The reason could be that the entrance of an extra-party increases contestation (discussion, media attention, voters’ information…) or that a government anticipates future electoral results and modifies its behaviour even if the government is strong. The identification strategy is based on a Regression in Discontinuity Design RDD on the extra-party entrance in the city councils for Spanish local governments during 2011-2015 term-of-office.

“Multi-level Tax Blaming: The Political Accountability of Fiscal Adjustments in Multi-level Governments”

Abstract: This study provides empirical evidence of political accountability in multi-level governments when fiscal adjustments are implemented. Under partial decentralisation voters may be uncertain about governments responsibilities reducing political accountability. I find that fiscal adjustments remarkably affect incumbents’ electoral results. Voters reward the application of policies leading to tax reductions while punish those supposing an increase of tax liability. This paper also provides empirical evidence that governments can be held accountable for their actual actions in multi-level governments (local and central government). Clarity of responsibility remains in situations of partial decentralisation and voters are able to identify governments’ performance and attribute credit or blame for their policies. Even though voters’ reactions are limited to the local elections, they consider mayoral alignment with the central government. Only aligned mayors are punished at local elections when the central government is the only responsible for fiscal adjustments. Therefore, voters indirectly punish central government incumbent party. The causal effects are estimated by a DID technique considering Spanish municipalities. The analysis used the application of centrally-planned fiscal adjustments on property taxation, a partially-decentralised tax, affecting local governments in Spain during the 2011-2015 term-of-office.