We show that tax authorities can stimulate tax compliance by strategically releasing audit-relevant information. We focus on audit policies that disclose to taxpayers that audit risk discretely drops above a threshold. In our theory, we derive conditions for the existence of compliance gains over undisclosed flat audit rules and build a test for such improvements that relies on a change in the probability jump at the threshold. Our empirical analysis relies on the Sector Studies, an Italian auditing system targeted at small firms and based on a disclosed threshold design. We leverage more than 26 million Sector Study files submitted between 2007 and 2016. First, we show that taxpayers bunch at the threshold to a great extent, and that bunching intensity is related to evasion proxies, availability of evasion technologies, and tax incentives. Then, we examine a staggered Sector Studies reform that widens the initial audit risk discontinuity. In line with our theory, taxpayers who benefit from audit exemptions above the threshold reduce their relative compliance, while those below the threshold improve it. However, mean reported profits increase by 16.2% in treated sectors over six years, suggesting – in light of our test – that a disclosed rule performs better than a flat undisclosed one.

Awarded the 2022 Innovative Policy Research Award by the Asian Development Bank and the International Economic Association. Featured at the APPAM 42nd Annual Fall 2020 Research Conference, the NTA 114th Annual Conference on Taxation, the 2022 CESifo Area Conference on Public Economics, the 15th RIDGE Forum - Workshop on Public Economics, the 2022 European Economic Association Congress, the 2023 EU Tax Observatory Workshop, and in the 2023 IMF Fiscal Affairs Department Seminar series.

How disruptive is tax auditing for small firms and the self-employed? How willing are taxpayers to avoid the costs associated with fiscal inspections? Even when audits detect no evasion, they can place undue burden on the day-to-day activity of small taxpayers. With micro-level administrative data over the 2007-2016 period, I study the revenue response of Italian firms and the self-employed across two types of auditing policies. On one hand, I explore the effect of a reduction in the length of on-site audits mandated since 2011. On the other, I assess the extent of bunching below the exemption threshold of Sector Studies as it evolves across sectors and years. Observed responses can inform us on the relative costs that firms associate to tax audits, and guide policy choices for the optimal design of modern tax enforcement systems. 

We study the effects of a unique Italian reform incentivizing voluntary tax compliance among the self-employed and small businesses. Starting in 2011, taxpayers in a growing number of sectors were promised an increase in audit exemptions upon achieving a set of desirable conditions defined by the Revenue Agency. While policy rewards might induce a tax base rise among previously non-compliant filers, curbing audit risks for broad categories of the taxpaying population might prove revenue reducing. Over the first six years of implementation, our event-study analysis of more than 9 million anonymized records reveals a substantial expansion of average declared revenues, total costs, and gross profits, with little heterogeneity across macro-regions. Although aggregate compliance does not seem to improve by policy metrics, our distributional analysis shows that large gains obtain among taxpayers appearing non-compliant in the year before their sector's reform. We also provide a dynamic perspective on bunching at salient, audit-relevant revenue thresholds generated by the system. Relative revenue reshuffling from above and below these thresholds provide evidence that bunching in our context may emerge from both desirable and adversarial updating in compliance behavior. 

Presented at the Italian tax administration meeting "Incontro ISA - Analisi dei dati ed iniziative per il futuro", June 17, 2020.

My upcoming projects in collaboration with the Italian Tax Police (Guardia di Finanza) explore the potential of machine learning for audit targeting, the determinants of evasion heterogeneity across firm types, sectors, and places, and the role of optimal audit strategies and information provision to taxpayers for business tax compliance.

Recipient of the 2021/2022 Molly and Domenic Ferrante Economics Research Fund at Harvard.

In this paper, we develop and test a conceptual framework to link diversity in small-scale communities and local resource management, intended as the ability of a community to protect common resources. Diversity could reduce this ability by its adverse impact on the potential for collective action and the extent of individual attachment to the community. Using geocoded data on 1,600 Grade-3 gas leaks in 2016 across Boston and Cambridge, we show that block groups displaying higher degrees of ethno-racial and linguistic fractionalization in the most recent Census and American Community Survey enjoy a lower share of reparations of the local pool of leaks. In particular, the more robust and conservative estimates imply that moving from the 10th to the 90th percentile of the linguistic fractionalization distribution is associated with a decrease in the reparation share by 6.3 percentage points, compared to the baseline average of 3.7%. We address a number of empirical challenges inherent to the spatial nature of our data by estimating models at the leak level and accounting for spatial autocorrelation. Although we are able to confirm, through quantitative and qualitative evidence, that more fractionalized communities appear to be less endowed with social capital and individual attachment, measures of these concepts do not explain away the baseline results. We conclude by discussing possible reasons for this puzzle in the form of alternative mechanisms and other challenges to our approach. 

Is there a link between ethnic favoritism and political business cycles? I test this hypothesis with micro-level data from 20 Sub-Saharan African countries over the 1999-2013 period. Preliminary results show that a set of individual economic well-being indicators from the Afrobarometer vary over the electoral cycle depending on the ethnic distance between survey respondents and their national leader, as proxied by a novel measure of linguistic proximity. The relationship is nonlinear for some consumption goods during election years and more generally in their immediate aftermath, suggesting the need for the ethnic favoritism literature to move beyond the use of binary indicators of individual-leader coethnicity. The presence of an “ethnic business cycle” in the data calls for further exploration of the political roots of Africa’s underdevelopment. 

Do political budget cycles (PBCs) exist across starkly different regimes? The bulk of the empirical literature on electoral manipulation focuses on democratic settings and motives, but competitive elections are set up in a wide variety of political contexts, albeit with different meanings. I exploit a large panel of 94 developing countries over the 1989-2004 period to assess the intensity of budget deficit manipulation over the full range of an aggregate regime measure (the POLITY index). PBCs behave nonlinearly and peak among hybrid systems, that is, those regimes that mix autocratic and democratic features in the structure of central authority. Given that competition in these settings is often real but unfair, I explore the possibility that such cycles are induced by the risk of political instability and the threat of unrest faced by the incumbents. Probit and Poisson analyses reveal that electoral time is an especially volatile moment in these already fragile systems. Budget manipulation thus emerges as part of a set of instability management tools ensuring incumbent survival.