Working Papers
Accounting for Crime: Detecting Mafia-connected Firms Using Machine Learning (R&R at Management Science)
With M. Fabrizi and A. Parbonetti (link to the paper)
Can criminal firms be identified using their own financial statements? We develop a machine learning model to detect firms affiliated with organized crime by leveraging the structural distortions generated by constrained transparency—the tension between mandatory disclosure and illicit objectives. Using judicial evidence from Italy, where companies can be linked to Mafia-related prosecutions, we construct a labeled dataset of 2,082 criminally connected firms. Our gradient-boosted classification algorithm achieves an AUC of 75.8% and a precision of 90.4% on a hold-out sample. The results show that criminal affiliations leave systematic traces in accounting data. By combining accounting theory with predictive analytics, our approach transforms financial disclosures into early warning signals of institutional risk, offering a scalable tool for compliance, enforcement, and policy design.
When the Godfather Falls: Implications for Industry Peers in Non-Traditional Territories (Submitted)
With M. Fabrizi, E. Ipino and A. Parbonetti (link to the paper)
This study examines the economic impact of organized crime using 57 anti-Mafia police operations in Northern and Central Italy from 2013 to 2020. Removing Mafia-connected firms increases non-criminal competitors’ performance by 7.7 percent—equivalent to €1.97 billion in additional EBITDA—driven by revenue growth and lower input costs. The effect rises to 15 percent in low-competition industries with entrenched Mafia influence. The impact varies by the criminal firm’s infiltration strategy and organizational role. Firms engaged in rent extraction harm peers more than those focused on money laundering. Additionally, removing firms tied to high-ranking Mafia members or external enablers, such as politicians or professionals, yields larger gains. These findings show how organized crime distorts competition through coercive and institutional channels, even in advanced economies. Disrupting Mafia influence restores fairness and efficiency, offering new insights into how enforcement actions reshape markets and reduce the economic costs of criminal infiltration.
Tax avoidance implications of employment protection: Evidence from the Italian ‘Jobs Act’ (R&R at JIAAT)
With A. Alexander and L. Menicacci
We investigate the tax avoidance implications of employment protection regulation, focusing on a labor reform enacted in Italy in December 2014 that aimed to rationalize employment protection. Our findings reveal an increase in tax avoidance following the reform. Additionally, we observe a rise in the capital-to-labor ratio, driven by both higher capital investments and a reduction in the labor force. This is consistent with the non-linear relationship between employment protection and capital deepening, where the relationship is negative at high levels of labor protection. We further show that tax avoidance is concentrated in industries with lower depreciation and amortization rates, highlighting the role of limited deductibility of investment costs relative to labor costs in creating tax avoidance incentives.
Selected Work in Progress
Governing Illicit Empires: Hierarchical Control and Organizational Risk in Criminal Business Groups
Solo-Authored
Saving the public good: The Role of Accounting and Procurement Data in Unveiling Mafia Infiltrations in Italian Municipalities
With I. Campagna, F. Longhin, and A. Parbonetti
How Mafia organizations are affected by economic crises: Evidence from a natural experiment in Italy
With Coppola, R., Fabrizi, M., Marchetti, D., Parbonetti, A.