Essays in Public Finance and Political Economy

Abstract

This dissertation consists of three essays in Public Finance and Political Economy dealing with business tax compliance in Italy and the management of natural gas infrastructures in the Greater Boston area.

The first essay, co-authored with Matteo Paradisi, studies the potential role for audit rule disclosure in stimulating business tax compliance. To this end, we use administrative data from an Italian audit system known as Sector Studies. Sector Studies present small firms and the self-employed with relatively higher audit risks when reporting revenues below an idiosyncratic threshold that is revealed to the taxpayers. As a consequence, businesses respond by adjusting their reported revenues to bunch at these thresholds. Accordingly, we show in the data that bunching consistently correlates with measures of tax evasion. We thus estimate a structural model of business revenue underreporting to quantify the effect of disclosure. Relative to scenarios where audit risks are perceived to be constant, our calculations suggest that disclosure induces higher mean reported revenues without substantial welfare costs over the 2007-2010 period. We conclude that audit rule disclosure provides a relatively low-cost policy to reduce revenue underreporting even among the "hard-to-tax".

The second essay investigates the effects of a 2011 reform to the Sector Studies. The policy, known as "reward regime", expanded the audit exemptions conferred to taxpayers complying with their Sector Study. While these rewards might induce a tax base rise among previously non-compliant filers, the provision of audit exemptions might still result in revenue losses on the net. Over the first six years of implementation, I show that the reform is associated with both higher mean reported revenues and gross profits in the affected sectors. These tax base gains are concentrated among taxpayers appearing non-compliant in the year before their sector's reform. The gradual roll-out of the reform also helps to illustrate the dynamics of bunching in a threshold-based, disclosed audit rule setting. Both taxpayers starting below and above their Sector Study revenue threshold before the reform reduce their distance to the threshold following the policy change. This evidence suggests that tax enforcement strategies relying on audit rule disclosure can engender both compliance-improving and -reducing behaviors, calling for close scrutiny of both individual and aggregate effects.

The third essay, co-authored with Felipe Jordán Colzani, investigates the relationship between ethnic diversity and the reparation rate of natural gas leaks in Cambridge and Boston. Local diversity can reduce the ability of a community to protect its common resources by undermining collective action and an individual's attachment to the community. Using recent geocoded data on public-street gas leaks, we show that neighborhoods with higher ethno-racial and linguistic fractionalization enjoy a substantially lower share of reparations of the local leak pool. While more fractionalized communities in the data display lower levels of social capital and individual attachment, measures of the latter concepts fail to explain away our baseline correlations. Possible reasons for this puzzle may lie in the difficulty of isolating the complex social and political dynamics that connect a community to the overall state of its local infrastructure.