Far from being the glamorous place portrayed in some movies, the underworld is actually a difficult place to live – and conduct business. Thieves often steal from each other, mobsters may 'whack' their partners or turn state witnesses, and business partners may renege on their promises.
In addition, criminals can't rely on a number of mechanisms that have been devised and put at the disposal of individuals who operate in the overworld, including legally binding contracts and an effective court system. Information on reputation travels with difficulty and is often unreliable.
A wealth of anecdotal evidence on criminal organizations tells us that individuals populating the underworld are often untrustworthy and selfish. In his memoirs, Joseph Pistone suggests that trust is in scarce supply even among mafia members who have taken an oath of allegiance to a code of honour.
As an FBI special agent, Pistone managed to infiltrate the notorious Bonanno Family for six years under the fictitious name of Donnie Brasco. Also, the Sicilian Mafioso-turned-state-witness Tommaso Buscetta recalls in his memoirs that mafia members are constantly suspicious of their associates; they tend not to believe the others' word and, if something goes wrong, are quick to accuse their fellow conspirators of not keeping to the agreements.
Mafiosi, and criminals more generally, tend to operate in an environment characterized by generalized mistrust. Yet, organized crime groups do exist and illegal activities – some of them as complex as arranging illegal trading operations between Latin America and North America, Europe or Australia – take place daily. How do criminals ensure co-operation and overcome the severe challenges posed by their environment?
In a study recently published in Rationality & Society, Federico Varese and I explored the mechanisms that enable criminals to co-operate in settings characterized by the absence of law and (to some extent) trust.
Being threatened with violence goes some way to explain why people don't defect on agreements. However, deploying violence is often costly, particularly in settings with effective law enforcement. Violence can attract attention, hinder a group's activities and create a climate of uncertainty among the associates. Above a certain level, it can make an organization utterly dysfunctional.
Taking hostages to guarantee promises
Criminals can rely on strategies to convince their counterparts that they are indeed committed to a given course of action and are trustworthy. For instance, they may want to reassure a counterpart that they won't cheat, walk away with the loot or report on them after they've committed a crime together.
Here's the problem: How can the counterpart be sure that this isn't just cheap talk? After all, these are commitments made by untrustworthy individuals. To commit oneself to some course of action is a rather easy task. The difficult part is to do it in a way that's credible enough for the counterpart to be willing to enter an agreement.
There are multiple ways of making a given commitment a credible one. In the literature, they go under the name of credible commitment strategies. In our paper we explore one of them: hostage-taking.
Taking hostages to guarantee promises is a practice that has a long history: Ancient Romans were already resorting to it as early as the end of the Second Punic War (202 BC).
Also the British monarchy relies on the same strategy to guarantee the safety of the monarch when they travel to parliament to deliver the annual speech. On that day, a prominent member of parliament is held hostage in Buckingham Palace and "freed" only upon the safe return of the monarch. Albeit now merely ceremonial, this procedure is still followed today.
In a context of low trust and in the absence of third-party enforcers, taking hostages is a way to have people commit to a course of action regarded as desirable. In criminal organizations, two specific devices may serve as hostage-taking strategies: violence and kinship.
While violence is an essential element of criminal groups to punish misbehaviour, it's also used as a way to foster co-operation by holding compromising information on the people who've committed it.
If one has committed violence, he will be less likely to defect if he knows he can be reported to the authorities for a previous act of violence. As a by-product, violence thus serves as a credible commitment among mobsters.
An additional strategy to decrease defection – and thus increase co-operation – is to entrust some key tasks to next-of-kin. Kin may have a stake in the future performance of the organization, and therefore may be more likely to align their interests to those of the group.
But there's also an additional feature that makes next-of-kin rather attractive as potential criminal associates: information about the identity and whereabouts of a relative is usually more cheaply available than information about a stranger. As a consequence, it's easier to inflict a punishment on the relatives of a deviant member if the latter shares blood ties with the boss or other associates.
Indeed, during the so-called Second Mafia War (Sicily, 1982-84), eight of Buscetta's relatives – including two sons and a brother – were murdered, although they had no role in the Mafia. After he defected, more were killed. Next-of-kin are in fact hostages of the criminal organization.
However, contrary to the Ancient Romans and the British monarchy, Mafiosi tend not to hold physical hostages, at least in territories where law enforcement is effective and fairly high. What they hold hostage instead is information, specifically information on the whereabouts of next-of-kin. (In a similar fashion, recruiting locally can also be interpreted as a weaker form of hostage-taking.)
Dissecting organized crime
Federico Varese and I were independently given permission to analyze evidence prepared for two cases that came to court in Italy.
The first involved a Neapolitan Camorra group based near Naples with international branches in Scotland and the Netherlands. The second involved a Russian Mafia group that moved from Moscow to Rome in the 1990s.
Both had been under surveillance for some time, unaware that the telephone conversations of key players were being tapped by the police.
For our study, we coded the information contained in the transcripts, tracking which individuals within the criminal organization were in contact with each other and monitoring how often they were in touch with one another.
We also coded the content of the conversations between two Mafia members, identifying four different tasks: group management, research acquisition, protection activities and investments in the economy (both legal and illegal businesses).
The Neapolitan Camorra dataset includes some 1,400 contacts among 51 members; the Russian Mafia dataset includes the 295 contacts exchanged by 22 members.
We also coded information on the background of each member, including whether they shared a blood tie with the boss (in the case of Neapolitan Camorra) or any other member within the group (for the Russian Mafia).
In addition, by having also coded tasks that don't require the use of violence, we were able to test whether actors who shared information on violence were more likely to co-operate on such non-violent tasks.
After coding all the network matrices of interest, we were finally in the position to test whether having shared compromising information about violent acts did indeed increase co-operation and whether, along the same lines, relatives were indeed more likely to co-operate.
The results confirm both intuitions.
Sharing information about violent acts does have an impact on the frequency with which two actors get in contact and co-operate. Most importantly, this extends to activities that don't require the use of violence to be successfully carried out (e.g. acquisition of resources such as clean bank accounts) or for which violence is even detrimental (e.g. money laundering operations in new territories).
The same applies to blood ties. But there's more. Our study also reveals that sharing information about violence creates an even stronger bond between criminals than kinship – including those criminal groups that are kin-based, as is the case for the Neapolitan Camorra.
While kinship still remains a common device to foster co-operation among criminals, criminal groups are more likely to resort to other non-kin based mechanisms when they're available. The saying "blood is thicker than water" doesn't necessarily apply here.
Criminals do face acute trust problems. Law enforcement agencies could exploit this weakness by introducing additional doses of 'unpredictability' into the system: for instance, by randomly seizing some drugs without publicizing it. This will increase distrust among criminals and may lead to the collapse of already fragile co-operation agreements.
Moreover, hostage-taking strategies can be neutralized by decreasing the value of hostages. In the case of compromising information, this can be done by decreasing the value of this information through plea-bargaining provisions. After all, it's the State that makes a piece of information about a given act a compromising one by attaching a penalty to the act committed.
Finally, witness protection programs across the world have been devised with the intent to decrease, among others, the value of next-of-kin as hostages, by relocating them in a secret location under a new identity.
Paolo Campana is Research Fellow at the Extra-legal Governance Institute, Department of Sociology, University of Oxford. The paper "Cooperation in Criminal Organizations: Kinship and Violence as Credible Commitments," co-authored with Federico Varese, recently appeared in the journal Rationality & Society.