University of Sussex
Rolla, Pierfrancesco.pdf (2.5 MB)

The social consequences of organized crime groups

Download (2.5 MB)
posted on 2023-06-10, 04:29 authored by PierFrancesco Rolla
Organized crime has adverse effects on security, development and democracy, but little is known about its social consequences. This thesis investigates empirically the effect of individual exposure to the presence of organized crime groups in Italy on a set of key social outcomes: political participation, civic engagement, institutional trust and interpersonal trust. The thesis includes one introductory chapter, three empirical chapters and one concluding chapter. In the introductory chapter, I review the academic literature related to my thesis. My work builds upon three main bodies of literature on: (a) the causes and consequences of social capital, (b) the economic, political and social effects of rebel groups governance, gangs presence and crime and conflict victimisation, and (c) a qualitative sociological literature on how organized crime groups manipulate the social capital of those that surround them to expand and govern areas. This chapter examines these bodies of literature, discusses my contributions to each, and shows how my focus on the social consequences of organised crime in Italy brings them together. The first empirical chapter starts by providing the historical context and existing theories on the origins of the Mafia and other organised crime groups in Italy. I pay particular attention to one theory directly related to my main research question, namely that the emergence of organized crime groups can be viewed as a response to endemic distrust in southern Italy. This ‘cultural explanation’ for the emergence of organized crime has come under scrutiny recently. While it is unquestionable that organized crime groups emerged in areas in southern Italy which in the 1800s experienced a pervasiveness of distrust and very low levels of social capital, these arguments overlook two important points. I present both historical and empirical evidence that (i) organized crime groups did not appear in many areas in southern Italy with very low levels of social capital and (ii) organized crime groups also emerged in areas with very high levels of social capital. This historical review allows me to clarify the direction of causality of my research by attenuating the probability that the direction of causality goes from (low) social capital to organized crime. This then allows me to reverse the question and dedicate the rest of my thesis to study the effect of organized crime on social capital. In the second half of this first empirical chapter, I then present a conceptualisation of social capital, with a focus on the four dimensions I focus on, and organized crime in Italy and describe the data I use to study this relationship. In the second empirical chapter, I attempt to provide a causal explanation for the social consequences of organized crime groups using two identification strategies. First, I leverage a survey of almost 800,000 respondents on social capital and exposure to organized crime conducted in Italy between 2000 and 2018. I use a pooled OLS with province-year fixed effects. I address possible omitted variable biases through the use of a rich set of controls included in the dataset. Second, I rely on a newly digitised dataset, which includes data at the regional level on social capital in Italy since 1861, covering 160 years. I use this dataset to exploit the exogenous arrival of organized crime in Apulia in the 1970s, a region in southern Italy where organised crime was largely absent before the 1970s. I compare the level of social capital in Apulia with a synthetic control of regions with no organized crime presence. The pooled OLS results show that exposure to organized crime reduces social capital by - 0.066SD. In percentage terms, this results in a reduction in social capital by 8.1% on average across Italy between 2000 and 2018. This negative effect of the social capital aggregate index is explained largely by three of its components: lower political participation (-0.01SD), lower institutional trust (-0.20SD) and lower interpersonal trust (-0.11SD). In contrast, being exposed to the presence of organized crime groups has a positive effect on civic engagement (+0.02SD). Results from the synthetic control approach shows that in the decade between 1971 and 1981 the level of social capital in Apulia drastically fell by approximately 30%. The gap between the actual and counterfactual social capital remains constant until today. This drop in social capital happens in the same decade in which organized crime groups settled in Apulia. Results from the synthetic control confirm the pooled OLS results, but are of larger magnitude given the focus on a region with heavy presence of organized crime since the 1970s. In addition to the main results, I also present a quantile analysis, detailed heterogeneity analysis and a battery of robustness checks (coefficient stability test, multiple hypothesis testing, alternative indicators of organized crime and social capital, social desirability bias and the use of alternative estimators - such as matching techniques - to study the same relationship). All results are robust to the use of these robustness tests. In the third empirical chapter, I provide qualitative and quantitative evidence on key mechanisms that may explain the results. To analyse plausible mechanisms, I firstly conducted a series of interviews and conversations with professionals with extensive experience on organized crime in Italy (policemen, army officers and prosecutors) and a number of research specialists on organized crime (criminologists, political scientists and economists) to theorise inductively the possible mechanisms at play. I then tested these mechanisms using a number of detailed key informant interviews with community leaders in Genoa, which were exposed directly to organized crime. The key informant interviews focus on the city of Genoa for two important reasons. First, for budgetary reasons I was not able to extend this work to the whole of the Italian territory. I needed therefore to find a location which would represent the ’median’ experience of Italians with organized crime. According to available socio-economic data, Genoa is the ideal representation of this ’median’ view. Second, the choice of Genoa allows for some degree of external validity as it is a city much more similar in terms of organized crime presence to cities in other developed contexts, in contrast with more common studies focused on rural Sicily where only one particular form of the Mafia (Cosa Nostra) has persisted. Genoa’s historic city centre is home to criminals with linkages to the Mafia families in the South of Italy, Mafia groups with specific business interests (international drug trafficking and money laundering), largely connected to the Ndrangheta, and foreign organized crime groups with interests in local illicit drug trade and pimping and pandering. The interviews to the community leaders were centred on the effect of the exposure to organized crime groups on their social capital and the social capital of their community. In particular, I was interested in answering the ’why’ behind the effect and I coded the transcripts deductively according to theorized mechanisms. This detailed qualitative work was conducted alongside an empirical mediation analysis that quantifies the relative importance of each mechanism. By combining experts’ interviews, oral testimonies of community leaders and the quantitative mediation analysis, I find that two key mechanisms are mostly relevant in explaining the negative effect of organised crime on social capital. The first is a psychological mechanism, in the form of fear and resignation, which explains almost 45% of the overall effect. The second is the state (in)capacity in dealing with the problems created by the presence of an organized crime group with adverse effects on institutional trust and political participation. The positive effect on civic engagement seems to be


File Version

  • Published version



Department affiliated with

  • Institute of Development Studies Theses

Qualification level

  • doctoral

Qualification name

  • phd


  • eng


University of Sussex

Full text available

  • Yes

Legacy Posted Date


Usage metrics

    University of Sussex (Theses)


    No categories selected