Targeted Spending: The Effect of the Electoral Timing on the Distribution of Discretionary Jobs within the Italian Bureaucracy

Abstract

Under certain circumstances, Italian public offices can appoint temporary employees by passing the standard procedure, which requires public competition. The 2020 pandemic shock led first to the deferral of seven scheduled regional elections, and then to the announcement of a new election date. I estimate a DID model with time-varying ATTs and wild bootstrapping of the standard errors. My results are robust to a different estimation strategy (RDD) and model specification, and show a sharp negative effect of the deferral on the percentage of employees hired by the regions where they were born. However, this effect is only specific to the time period immediately following the deferral, and vanishes as soon as the new election date is announced. Therefore, it seems that an adaptation strategy to the new course of the political events has been in place in the treated group. As only residents can vote in regional ballots, my finding suggests that- in the proximity of an election- incumbents may tailor discretionary public expenditures towards people that are ex-ante more likely to be pivotal.

Keywords: Discretionary public spending; Pandemic-related election deferral; Tentative vote-buying strategy

JEL Classification: H72, P00, R50