Abstract

There is an abundant empirical literature demonstrating party favoritism whereby the central government is prone to disburse financial transfers to favor aligned local governments. This contrasts with much scarcer evidence on reverse party favoritism, i.e. aligned local governments offering non-pecuniary support to the central government in times of elections. In this paper I show that such reverse party favoritism exists. To demonstrate it, I exploit the fact that during the SARS-CoV-2 pandemic crisis, the Polish government was keen to launch postal voting in the presidential elections scheduled for May 2020. The organization of these elections hinged on the Polish Post getting access to the lists of voters, which were in the possession of heads of the municipal executive (mayors). Since the relevant legislation on postal voting had not been enacted on time, the vast majority of local executives refused to share their lists of voters. Nonetheless, numerous mayors did transfer the lists to the Post. By employing a set of standard (linear probability and logistic) regression models and regression discontinuity design, I show that the political alignment of mayors with the central government leads to approximately 20–25 percentage points greater likelihood of transferring the lists of voters to the Polish Post. This difference in probabilities tends to be smaller in cases of divided governments but not in cases of higher political contestability.