Research Highlights Featured Chart

March 5, 2026

Curbing the influence of criminals on public officials

Better pay for politicians is associated with less corruption in public procurement.

Tyler Smith

Campaign posters of candidates for mayor in Italy hang on message boards.

Source: bestravelvideo

A growing body of research suggests that paying public officials more reduces corruption. However, increasing pay may backfire in areas where organized crime is active.

In a paper in the American Economic Journal: Economic Policy, authors Massimo Pulejo and Pablo Querubín show that higher wages reduced corrupt procurement practices among Italian municipal executives, but increased the likelihood that those officials were subjected to criminal attacks.

The study leverages a feature of Italian law that awards a 28.6 percent monthly wage increase to mayors and other executives in municipalities with populations above 5,000 residents, as measured by the most recent census. Because this threshold creates a sharp discontinuity in compensation, the authors were able to employ a regression discontinuity design to estimate the causal effect of municipal executive wages on criminal attacks. 

Using data collected by the NGO Avviso Pubblico on 2,464 criminal attacks against public officials between 2014 and 2020, the authors found that members of the municipal executive in towns just above the 5,000-resident cutoff were 6.3 percentage points more likely to be targeted by criminal groups than their counterparts in towns just below it—an effect of approximately 0.28 standard deviations.

Figure 2 from the authors’ paper depicts the evidence underlying this result.

Figure 2 from Pulejo and Querubín (2026)

Panel A shows the probability that municipal executives experienced at least one criminal attack between 2014 and 2020. Panel B shows the average number of attacks over the same period. In both panels, observations are grouped into population bins along the x-axis. The solid lines are quadratic polynomials fitted separately on each side of the cutoff, with 95 percent confidence intervals indicated by dashed lines. 

In both panels, the polynomial fitted to municipalities above the threshold sits higher than the one fitted to municipalities below it, with an upward shift at the 5,000-resident mark. The confidence intervals widen around the cutoff, reflecting greater variance in the estimated attack probability and number of attacks. But there is still little overlap between the confidence intervals in the immediate vicinity of the threshold. This discontinuity suggests that higher executive wages led directly to higher levels of attacks by criminal groups. 

The findings suggest that wage increases for public officials in procurement-sensitive roles reduced malfeasance, but in environments with organized crime, criminal groups shifted from bribery to violence. As a result, a durable anti-corruption policy would need to address both the incentives of public officials and their safety.