Abstract:

We estimate how local femicides affect voting and turnout in subsequent Italian parliamentary elections. We merge a geocoded registry of femicides (2011–2022) with municipality-level results for the national elections (2013, 2018, 2022) and exploit staggered exposure to a municipality’s first femicide. Using difference-in-differences estimators designed for staggered adoption, we find an electoral backlash against the Center-Right coalition: its vote share falls by about 2.2 percentage points, with offsetting gains for the Five Star Movement (+1.0 pp) and the Center-Left (+0.5 pp). Femicides also reduce participation: turnout declines by about 0.8 percentage points, with larger declines among women than men. Heterogeneity by nationality suggests that backlash is concentrated in cases involving Italian victims or Italian perpetrators; when the perpetrator is foreign, the Center-Right effect is small and statistically indistinguishable from zero. The results are consistent with electoral sanctioning and demobilization following salient failures of protection, with attribution shaped by identity cues.

Presentations:

• 38th Annual Conference of the Italian Society for Public Economics (SIEP), University of Modena and Reggio Emilia (Italy), 2026

Abstract:

We study how local governments adjust fiscal policy following extreme acts of gender-based violence, and whether these adjustments are shaped by migration-related concerns. Combining administrative data on femicides with municipal budgets for Italian municipalities, we exploit femicides as shocks to the local salience of violence in staggered difference-in-differences models. Femicides trigger a sizable and persistent increase in security expenditures, concentrated on local police, with little corresponding expansion of prevention-oriented or social services for women at risk. The magnitude and persistence of the security response depend strongly on the perpetrator's nationality: when the perpetrator is foreign-born, the increase in police spending is much larger and more persistent. These patterns are consistent with high-profile crimes involving foreigners being more likely to be politicized and framed as public-order threats, reallocating scarce local resources toward visible security measures and away from policies more directly targeted at reducing gender-based violence.

Presentations:

• 37th Annual Conference of the Italian Society for Public Economics (SIEP), University of Naples Federico II (Italy), 2025

• 40th Annual Meeting of the European Economic Association (EEA), Bordeaux School of Economics (France), 2025

Abstract:

We merge original daily data on admissions to an Anti-Violence Center (AVC) in Southern Italy with the local soccer team’s match results from 2013–2018 to test whether sporting events influence violence against women. We find that home defeats lead to a significant increase in AVC admissions, with effects amplified when the team was favored to win according to betting odds. The impact is concentrated in cases involving partners or ex-partners and among Italian perpetrators. Overall, the results indicate that salient emotional events can exacerbate dynamics within abusive relationships. Importantly, our evidence suggests that women’s decisions to seek formal support and begin a pathway out of violence are often made immediately after a tipping point in ongoing abuse.

Presentations:

• 37th Annual Conference of the Italian Society for Public Economics (SIEP), University of Naples Federico II (Italy), 2025

Abstract:

This paper examines whether female political representation affects the use of discretion in local public procurement. Combining Italian municipal political records with contract-level procurement data for 2010–2019, we exploit Law 56/2014, which mandated gender balance in executive committees above 3,000 inhabitants. We instrument the female share of aldermen with reform exposure, jointly determined by population eligibility and first post-reform election timing. A one-standard-deviation increase in women’s representation lowers direct awards and threshold bunching by 0.19 and 0.33 standard deviations, respectively. It also increases the number of bidders, without affecting repeat winners or worsening contract governance. Direct-award effects extend across several sectors, while disability-related procurement shows both a larger contract share and less threshold bunching. The findings are not explained by educational selection or mayoral gender, are larger in the South, and remain robust across alternative samples and specifications, placebo tests, and staggered difference-in-differences estimates.

Presentations:

• 38th Annual Conference of the Italian Society for Public Economics (SIEP), University of Modena and Reggio Emilia (Italy), 2026