Accomplice Witnesses and Organized Crime: Theory and Evidence from Italy

• Date: 01 October 2014
• DOI: http://doi.org/10.1111/sjoe.12080
• Published date: 01 October 2014

Scand. J. of Economics 116(4), 1116–1159, 2014

DOI: 10.1111/sjoe.12080

Accomplice Witnesses and Organized

Crime: Theory and Evidence from Italy∗

Antonio Acconcia

University of Naples Federico II, 80126 Naples, Italy

antonio.acconcia@unina.it

Giovanni Immordino

University of Salerno, 84084 Fisciano Salerno, Italy

giimmo@tin.it

Salvatore Piccolo

Catholic University of Milan, 20123 Milan, Italy

salvapiccolo@gmail.com

Patrick Rey

Toulouse School of Economics, FR-31015 Toulouse Cedex 6, France

patrick.rey@tse-fr.eu

Abstract

We develop an agency model of organized crime accounting for the main trade-offs involved

in the introduction of an accomplice-witness program. We characterize the optimal policy and

identify its main determinants in a framework where public officials can be dishonest. Our

predictions are tested by using data for Italy before and after the introduction of the 1991

accomplice-witness program. As predicted by the model and the earlier antitrust literature,

the program appears to have strengthened deterrence and enhanced prosecution. Moreover,

consistent with a novel prediction of our theory, the evidence suggests that the program

efficacy is affected by the judicial system efficiency.

Keywords: Deterrence; leniency; Mafia; whistle-blower

JEL classification:D73; D78; K14; K42

∗We have benefited from comments by Ran Abramitzki, Orazio Attanasio, Ciro Avitabile,

Sandeep Baliga, Richard Blundell, Dimitrios Christelis, Roberta Dessi, Luigi Guiso, Elis-

abbetta Iossa, Tullio Jappelli, Jakub Kastl, Alessandro Lizzeri, Riccardo Martina, Dilip

Mookherjee, Marco Pagano, Erasmo Papagni, Nicola Pavoni, Adam Rosen, Giancarlo Spag-

nolo, an anonymous referee, and seminar participants at BU, UCL, Cattolica University,

Federico II University, Stanford University, the SIE, EALE, and CSEF-IGIER Conferences

as well as the Bonn and Paris Workshop on The Empirics of Crime in 2009. Financial sup-

port from Toulouse School of Economics is gratefully acknowledged. Finally, we are grateful

to Iolanda Barone, for valuable research assistance, and to Marco Antonucci, who serves as

capitan in the Guardia di Finanza, for useful discussions.

CThe editors of The Scandinavian Journal of Economics 2014.

A. Acconcia et al. 1117

I. Introduction

Since Becker (1968), organized crime has attracted considerable attention

from economists, and for good reasons. The diffusion of criminal organiza-

tions, whose internal structure is built on cohesive and militarized echelons,

has forced governments all over the world to reform their legal and judi-

cial systems to enhance the performance of investigation agencies and to

strengthen deterrence. These reforms have promoted the approval of special

laws that have radically changed conviction and imprisonment procedures.

A notable example is the introduction in many countries of accomplice-

witness regulations (also known as leniency programs), whose aim is to

encourage former mobsters to blow the whistle and cooperate with pros-

ecutors in exchange for legal benefits. The prevailing rationale for these

polices rests largely on the argument that the most culpable and dangerous

criminals rarely do the “dirty job”. Even if these individuals are primarily

responsible for the crimes committed by their “soldiers”, they are hardly

ever sentenced because their role is typically confined to behind-the-scenes

control and guidance. This explains why allowing low-tier criminals to

flip and turn informants seems to be a potentially key channel to fight

organized crime.

Yet, so far, very few models have studied the basic trade-offs that shape

the optimal design of cooperation amnesties in criminal proceedings where

whistle-blowers and defendants have belonged to the same organization

and have been in a principal–agent type of relationship. Moreover, despite

the substantial economic costs of these programs and the ethic concerns

triggered by their approval, the evidence on their performance is still quite

limited.1

Why should lawbreakers be rewarded with lighter sanctions when they

decide to blow the whistle and cooperate with the justice? What are the

key determinants of the amnesties they receive as a prize for cooperation?

What is the relationship between the costs and benefits of accomplice-

witness programs, the hierarchical structure of criminal organizations, the

efficiency of the judicial system, and the amount of corruption in the

economy? How do these programs perform in practice?

First, to address these issues, we develop a model where (endogenous)

legal benefits are granted to criminals turned informants who belong to

a hierarchical criminal organization, in order to break down the “code of

silence” (omert`

a) that typically makes it hard to fight organized crime.

Second, building upon the main insights of the model, we draw a number

of empirical predictions for the effects of these programs that are tested by

using the evidence available for Italy before and after the introduction of

the 1991 accomplice-witness program.

1See the related literature in Section V.

CThe editors of The Scandinavian Journal of Economics 2014.

1118 Accomplice witnesses and organized crime

Hence, in the paper, we build on two main blocks. In the first part,

we analyze a simple model, which shows that granting an amnesty to for-

mer criminals willing to cooperate with the justice has a clear economic

logic. We argue that two countervailing effects contribute to determine

these amnesties. On the one hand, rewarding criminals turned informants

with lower punishments exacerbates conflicts within criminal organizations.

More precisely, a more generous amnesty might induce criminals under in-

vestigation to cooperate more often, thereby increasing the prosecution risk

faced by their boss; this is the bright side of leniency programs. On the

other hand, it encourages the entry of new fellows into the illegal business

by lowering their expected sanction and thus the compensation that the boss

has to secure them at the hiring stage. This effect brings out a potential dark

side of leniency programs, whose strength depends upon several features

of the criminal organization such as, for instance, the degree of cohesion

among its members, the boss’s retribution power, the amount of corrup-

tion in the economy, and the corroborating value of the whistle-blowers’

testimony.2Building on the interplay between these two effects, we then

show that an optimal policy (i.e., selected in order to minimize crimes)

reduces the crime rate and increases the conviction probability of the top

echelons of the organization. Moreover, the incentives to blow the whistle

are determined, among other things, by the expected conviction probability

that a low-tier criminal faces when he does not cooperate and obeys the

code of silence imposed by his partners. It is also shown that more gener-

ous amnesties (or even monetary rewards) are necessary to fight organized

crime when the judicial system is not very effective, when criminal organi-

zations feature strong internal cohesion between their members, when the

information provided by an accomplice witness has a valuable investigative

content, and when there exist strong external complicities between public

officials and criminal organizations (corruption).

In the second part of the paper, we use the model predictions to guide

an empirical investigation on the accomplice-witness program introduced

by the Italian parliament in 1991 to fight the Mafia. The objective of

our empirical investigation is twofold. First, we wish to provide evidence

supporting the idea that these programs are likely to have a beneficial

effect on deterrence and prosecution, which can hopefully mitigate the

political and ethical prejudices against them. Second, we wish to use the

model predictions, together with the available evidence, in order to better

understand the forces that shape the performance of such programs in

practice.

2In contrast to our model where the dark side of a leniency program is a feature emerging at

the optimum, Buccirossi and Spagnolo (2006) discuss the effects of badly designed leniency

programs.

CThe editors of The Scandinavian Journal of Economics 2014.