Research

Working Papers

Context and the Perception of Harm: Evidence from Online and In-Person Sexual Harassment (with V. Tonei & M. Vecchi), Draft soon!

We provide causal evidence that the context in which harmful conduct occurs shapes how it is evaluated. We study this using a vignette-based survey experiment that holds behaviour constant while randomising whether a sexual harassment incident occurs online or in person. Online settings generate a systematic discount in perceived seriousness (7% of the mean), with larger effects on willingness to report (13% of the mean) and preferred sanctions. The discount is concentrated in image-based harassment and larger among male respondents. In addition, it is not moderated by either direct or vicarious experience of harassment, suggesting that it may reflect normative perceptions of online harm as less serious rather than lack of exposure alone. These context-dependent distortions have implications for the enforcement of emerging legal protections, victim support, and the design of public communication around digital abuse.

Crisis Narratives and Judicial Enforcement: Evidence from the Greek Fiscal Crisis (with R. Megalokonomou & M. Vlassopoulos), IZA DP No. 18489

This paper investigates whether crisis narratives affect the judiciary's handling of tax evasion. We study this question in the context of the Greek debt crisis, in which tax evasion was publicly blamed for the fiscal collapse, and judges themselves experienced substantial salary cuts as part of the resulting austerity programme. Using a novel dataset compiled from Greek Supreme Court rulings between 2006 and 2014, we compare tax evasion appeals with appeals in other serious crimes not directly related to the fiscal crisis, such as homicide and rape, in a difference-in-differences framework. We find that the probability that the Supreme Court rejects tax-evasion appeals increases by about 25 percentage points relative to these control offences after January 2010-about a 43% increase relative to the pre-crisis baseline. Effects are larger in months with greater public attention to tax evasion, as measured by Google Trends, suggesting a role for salience. Our findings suggest that crisis narratives, particularly when coupled with personal economic shocks to judges, can influence the judicial treatment of tax offences.
Previous version: SSRN 6428438

Air Supremacy Is Not Enough: The Effect of Drone and Air Strikes on Terrorist Attacks in Somalia and Yemen (with A. Baronchelli, R. Ricciuti, F. Santi and T. Tefera), CESifo Working Paper No. 12242

Drones are among the most recent evolutions in warfare, used to respond to or deter insurgent activity and to serve as substitutes for direct troop engagement. We empirically evaluate the effect of drone and air strikes as a deterrent for terrorist attacks, using Somali and Yemeni data from 2016 to 2020. We construct a novel dataset from different sources, such as the Armed Conflict Location and Event Data Project, the Global Terrorist Database, and the New America Foundation. Given the specific features of the data (spatial heterogeneity, interaction between phenomena, and low intensity), we implement a Cox model with different specifications. Results show very limited effects of both drone and air strikes, indicating that air supremacy is not sufficient in fighting against non-state actors.

Female Mayors and Corruption Scandals: An RDD Approach with Italian Data, SSRN 5372480

This paper investigates the influence of mayors’ gender on corruption scandals and investigations in Italy. Utilising a novel dataset, I apply a regression discontinuity design to mixed-gender close elections. The results indicate that mayors’ gender does not exhibit a statistically significant impact on the occurrence of corruption scandals or investigations at the local level.

Impact of depenalization on drug deaths in England and Wales. An instrumental variable approach (with A. Pickering), NICEP Working Paper: 2023-03

This article investigates the role of drug depenalization on drug-related deaths in England and Wales. We use an instrumental variable approach, based on Police and Crime Commissioners elections and voters' left-wing preferences, which corresponds to a decrease in drug-related arrests. The IV results indicate that a decrease in our instrumented variables generates an increase in deaths related to drug poisoning/drug misuse. Specifically, a 1% decrease in our instrumented variables corresponds to an increase of 0.04% to 0.07% in the drug poisoning/misuse deaths ratio. We replicate our analysis using different definitions of political preferences, lag specifications, dependent and independent variables and the findings are similar.

Pubblications

The words that keep people apart: official language and accountability (with A. Baronchelli & R. Ricciuti). Journal of Institutional Economics, 2025;21:e35. [Data]

This paper examines how the distance between a country’s official language and the languages spoken by its citizens influences accountability. Two arguments support this relationship: first, the role of language as a tool for communication between elites and citizens; and second, its role in shaping cultural patterns that underpin social interactions. Using a dataset of 147 countries, we reveal a consistent negative correlation between linguistic distance and levels of accountability across all measures. Higher educational attainment can mitigate the negative impact of a foreign official language on accountability.

Lady Justice: The impact of female judges on trials’ verdicts in US. European Journal of Political Economy, 2025, 102678. [Online Appendix] [Data Available upon Request]

This study evaluates the impact of judges’ gender on jury trial outcomes in the U.S. state of North Carolina. The identification strategy is based on conditional random assignment of judges to cases. Specifically, I take advantage of the compulsory judges’ rotation imposed by the North Carolina Constitution. The results indicate that the presence of a female judge increases of 9.64–13.50 percentage points the probability of having at least one guilty verdict from the jury and of 9.5%–13.45% increase in the proportion of guilty verdicts expressed by the jury. Additionally, I perform a series of robustness and heterogeneity checks. I also investigate the potential mechanisms driving the results, exploring the influence of the jury selection process and women’s attitudes toward the courts and sentencing.

Beyond Reasonable Doubt: The impact of political independent jurors on jury trials in the US. Journal of Law &Economics, 2025, 68.2: 361-386. [Online Appendix] [Data] [Supplementary materials]

This article evaluates the impact of jurors’ political affiliations on trial verdicts in North Carolina. The research design relies on the day-to-day random variation in the composition of jury pools. The results indicate that if there is one additional independent juror in the pool, the percentage of guilty verdicts decreases by 2.93 percent and the conviction rate decreases by 2.85 percentage points. The impact of Democratic jurors is negative but not statistically significant. I also evaluate possible political discrimination in patterns of removing jurors. Democratic jurors are 3.7 percentage points more likely to be removed from a seated jury. The results for independent jurors are positive but not statistically significant. I implement heterogeneity checks and robustness checks. I also use potential jurors’ political affiliations as an instrument for the political affiliations of seated jurors to replicate the analysis and obtain similar results.

Impatience and crime. Evidence from the NLSY97 (with S. Basiglio & G. Turati). Journal of Economic Psychology, 101 (2024): 102711. [Online Appendix]

We empirically test the relationship between crime and impatience at the individual level, exploiting data from the National Longitudinal Survey of Youth 1997 (NLSY97). Besides providing information on violent, property, and drug crimes, NLSY97 allows us to observe different behaviors, sharing impatience as a common latent factor. We use factor analysis to extract this common factor as a measure of impatience. Estimates from a Logit model suggest a positive association between impatience and crime. This relationship differs across violent, property, and drug crimes, but we do not find significant heterogeneities when comparing individuals according to gender, parental education, and ethnic groups. The main result is robust to different factor analysis specifications and controls for risk preferences. Our findings support policies aimed at influencing individual time preferences as an indirect way to combat crime.

The rise of populist parties in the aftermath of a massive corruption scandal. Public Choice, 184, 289–306 (2020). [Media Coverage: Topics in Public Economics (SIEP)]

Populism is on the rise in Europe and America, and understanding its origins and evolution is becoming increasingly important. In this study, I investigate the effects of the exposure of corruption scandals on populist parties in Italy, exploiting a novel dataset based on electoral data at the municipal level. I find that local corruption exposure helps the populist parties in national elections and hurts the incumbents as the scandals are revealed.

Work in Progress

Local Courts and the Supreme Court in Greece (with R. Megalokonomou & M. Vlassopoulos)

The Effect of Mass Shooting (with A. Ghazaryan and M. Zapryanova)

The Effect of PCC Reform on Crime (with A. Pickering)

Google Sites

Report abuse