Working Paper: NBER ID: w22093
Authors: Alberto Alesina; Salvatore Piccolo; Paolo Pinotti
Abstract: We investigate how criminal organizations strategically use violence to influence elections in order to get captured politicians elected. The model offers novel testable implications about the use of pre-electoral violence under different types of electoral systems and different degrees of electoral competition. We test these implications by exploiting data on homicide rates in Italy since 1887, comparing the extent of ‘electoral-violence cycles’ between areas with a higher and lower presence of organized crime, under democratic and non-democratic regimes, proportional and majoritarian elections, and between contested and non-contested districts. We provide additional evidence on the influence of organized crime on politics using parliamentary speeches of politicians elected in Sicily during the period 1945-2013.
Keywords: Organized Crime; Violence; Elections; Political Influence; Italy
JEL Codes: D72; K42
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| electoral systems (K16) | strategic deployment of violence (D74) |
| electoral violence (K16) | likelihood of discussing organized crime in parliament (K42) |
| proportional systems (P50) | negative correlation with electoral support for criminal organization-favored parties (K16) |
| majoritarian systems (D72) | violence concentrated in contested districts (D74) |
| organized crime violence (K42) | electoral outcomes (K16) |
| higher presence of organized crime (K42) | spikes in homicides during election years (K16) |