Towards a Concept of Organized Crime

Ambiguity, confusion, and disputes about the status of concepts are rife in the social sciences. Giovanni Sartori’s taxonomical model, proposed in the 1970s, suggests that concepts should be organized using a hierarchy of categories defined by their extension (the set of entities they apply to) and intension (the attributes defining them). Moving down the hierarchy, categories become more specific; moving up, they become more general.Footnote ¹ Following Sartori’s precepts, at its most general the concept of “crime” is usually defined as “an intentional act in violation of the criminal law, committed without defence or excuse, and penalized by the state” [Tappan Reference Tappan1947: 100]. As we move down the generality ladder, we inevitably construct classifications. Corruption, fraud, and theft come to mind as more specific types of crime. OC is another such example. As will become clear below, I define OC as a set of three activities that actors engage in: namely, illegal production, trade and/or governance.

In developing this concept, I will be guided by three principles. First, OC should not be just an attribute of other concepts but rather stand alongside equally general constructs, such as theft. Second, the concept should be used to test competing theories, rather than being tied to a single one. In other words, and contrary to the view of Carl G. Hempel [Reference Hempel1965: 113], concept formation is distinct from theory formation. However, this exercise involves the creation of classifications that are inevitably an interpretation of the world, so it is not entirely neutral or a-theoretical. Third, concepts should be relatively easy to operationalize effectively, and a broad consensus should be reached on this process. One solution would be to incorporate elements into the concept that are well established and have already been operationalized in other fields of study. Keeping these questions in mind, the first decision we need to take when trying to develop a definition of OC is whether we consider it as an attribute of an ethnic group, as an organization, or as particular activities defined as illegal by the state. All three perspectives exist in the literature [for a discussion, see Paoli Reference Paoli2002]. The “alien conspiracy theory” appeared after World War II and was prevalent in US debates from the 1940s to the 1960s. Originating within the Federal Bureau of Narcotics, it alleged that the majority of narcotics dealers in the USA were “Italian speaking people” [cit. in Smith Reference Smith2016: 63]. Endorsed by the FBI after the so-called Apalachin meeting of Italian mobsters in November 1957, it quickly developed into a widely accepted position. The theory was predicated on the assumption that OC was a foreign import from Italy that threatened American values, and it equated it with La Cosa Nostra [Smith Reference Smith2016: 63; Kleemans Reference Kleemans and Paoli2014: 33–34; Potter Reference Potter1994]. Clearly, this position had xenophobic connotations, which underlines the point that definitions are not neutral. This view ignored the role of local conditions and structural factors in the emergence of OC in the USA.Footnote ² Although deeply flawed, this approach has never entirely disappeared from American public discourse. At the same time, the target group has changed, shifting from Italian to Turkish, Chinese, or Russian immigrant communities [Kleemans Reference Kleemans and Paoli2014: 34]. Latin American cartels seem to be today’s equivalent on the basis of recent US policy decisions. Ultimately, the “alien conspiracy theory” considers OC an attribute of the ethnicity of the offender: thus a foreign-born person committing a given crime would qualify as engaging in organized crime, while a “native” committing the same act would not.

A second position found in the literature is that OC is an attribute of an organization. The view that this particular type of offence is structured as a hierarchical set of relations with an internal division of labour dates back at least to the early twentieth century, yet it was the US Committees investigation of La Cosa Nostra in the 1950s and 1960s that cemented the interpretation of OC as a highly organized conspiracy.Footnote ³ In 1963, Joe Valachi, a “soldier” in the Genovese Mafia “family,” gave a televised testimony to the Permanent Subcommittee on Investigations of the US Senate Committee on Government Operations. The authorities used charts of Mafia “families” to present Joe Valachi’s evidence. In this rendering, the Mafia had a hierarchical structure, with clearly defined roles and a boss at the top. Rather than specific crimes, OC was about a set of top-down interactions. A consultant on the 1967 US President’s Commission on Organized Crime, distinguished criminologist Donald Cressey, built on Valachi’s evidence and methods, namely the organizational charts. He described OC as relating to an organization that is “rationally designed to maximise profits by performing illegal services and providing goods that [are] demanded by society” [cit. in Varese (Reference Varese, Carnevale, Forlati and Giolo2010) 2017: 31]. Again, OC was interpreted not as a type of offence, but as an attribute of the organization committing the offence.

Almost immediately, this view came under criticism from scholars who argued that Cosa Nostra families had a degree of informality that did not quite fit the strict hierarchy model; they had no fixed roles, and no overarching coordinating body such as the Commission.Footnote ⁴ Rather than a rationally designed organization resembling the Bell Telephone Company, the Mafia was a traditional social system, the product of culture rather than design [Ianni Reference Ianni1973: 133–134]. Over the years, this position was enriched by novel concepts emphasizing that OC is “socially embedded” [Kleemans and van de Bunt Reference Kleemans and van de Bunt1999] and “networked” [Morselli Reference Morselli2005]. According to this view, family ties and geographical position are what help explain patterns in international crime, rather than the presence of a top-down structure. More generally, these scholars do not assume the existence of a predefined structure but derive it from empirical observation. This perspective is a significant improvement on previous models and opens up the possibility that the structure is a function of other variables, such as family ties, geographical position, money and, more generally, the overall context. Still, the focus remains on the structure (however this is defined) as the essential feature of OC.

Should OC be seen through the lens of its structure? Organizations are normally described as instruments used to achieve goals that would be difficult or impossible to obtain by individuals acting on their own [Handel Reference Handel2003: 2]. Classic definitions of OC, such as Cressey’s, are predicated on Weber’s rational-bureaucratic model of organization. Weber defines this model as an “established impersonal order,” founded on the idea of “domination” [Herrschaft] [Weber (Reference Weber1924) 1978: 215 and 212]. This type of domination establishes a hierarchy enforced by formal legality and which is responsible for the efficient delivery of the product to be supplied. Legal-rational authority is a form of domination that is superior to other kinds, namely traditional and charismatic domination. Charismatic authority is unique and non-replicable and therefore cannot be structured. Traditional authority evokes habit and convention and is considered to be less efficient than a more rational mode of organization. For instance, it does not adapt when circumstances change.Footnote ⁵ When adopting the rational organization model, offenders’ crimes are more harmful and/or extensive than when a less structured model is used. The relaxation of the principles of stringent division of labour, hierarchy, and clearly defined roles within OC overstretches the concept. It ends up covering almost any form of co-offending, from one-time partnerships to membership of a secret cabal. Most crucially, such relaxation does away with the Weberian assumption that an intentional decision has been made to create an organization in order to efficiently achieve clearly defined goals. Although this was not openly stated, Donald Cressey’s definition was predicated on the existence of a form of domination that was not only rational and superior to any other, but also considered legitimate by the actors involved (hence the original emphasis on their being both Italian and alienated from mainstream society). If OC can take any form, its structure cannot be its defining feature.

An alternative way to think about OC is to focus on activities. In 1977, US Attorney General Griffin Bell remarked that OC investigations by the Justice Department “would in the future concentrate on activities in a particular industry or enterprise and would move away from the practice of picking syndicate families or individuals as targets” [cit. in Smith Reference Smith1980: 360] [emphasis added]. In his book The Mafia Mystique, D. C. Smith elaborated the view that OC is a business activity operating in conditions of illegality. This “enterprise model” suggests that OC groups are illicit firms producing goods and services, such as gambling, loan sharking, narcotics, and sexual services [Smith Reference Smith1975]. First, this perspective does away with the focus on the organization, which is now a function of the activity. Second, it excludes criminal behaviours that do not involve the creation of a product or provision of a service. Domestic violence, theft, and the purchase of stolen goods do not add production value to the objects (or persons) concerned. Finally, OC is not an attribute of other crimes, of an ethnic group, or of a particular organizational structure.

The “enterprise model” is a marked improvement on previous views of OC. And yet, the concept it introduces covers activities that are extremely diverse. It fails to distinguish actors engaged in production activities—such as processing coca leaves, illegal gambling, and running brothels— from those tasked with transporting the product from one location to another, i.e. the trafficker. Human trafficking, human smuggling, drug trafficking, and trafficking in animal parts or stolen data are best understood as forms of trading and can occur in physical or virtual marketplaces. The skills required—and hence the profiles of offenders—are very different from those of the actors involved in production. For instance, Financial Times journalist Mads Nissen travelled to the village of Pueblo Nuevo in Antioquia, Colombia, to meet Ariel Albeiro Muñoz, 19, who picked coca leaves as a day labourer: “Muñoz said he didn’t use cocaine or care much about politics; he just wanted to earn enough money to buy a motorbike.”Footnote ⁶ Workers like Muñoz begin work at 5 a.m., each one collecting about 70 kg of coca leaves, for which they are paid about $10. However, Muñoz has hardly anything in common with those who receive the drugs and ship them across the world.

There is a third activity that needs special attention: the illegal governance of the production and trade of goods and services [Varese (Reference Varese, Carnevale, Forlati and Giolo2010) 2017: 41]. In essence, individuals involved in governance OC perform a function that is normally reserved to the state in legal markets. In order to govern, they need to invest in a special set of resources which are not necessarily available to the illegal producers and traders, such as violence. Here lies the main challenge to the state as the only legitimate user of force [Bates Reference Bates, Kalyvas, Shapiro and Masoud2008].

Obviously, any discussion of OC is predicated on the fact that the state defines what is legal and what is illegal. For instance, paying for organs is legal in Iran and outlawed everywhere else in the world [Steiner Reference Steiner2010]. As pointed out by Beckert and Dewey [Reference Beckert, Dewey, Beckert and Dewey2017], the illegality may comprise every element of the chain—from production to exchange to consumption—or only certain aspects of it. What is legal or illegal, and what counts as OC, is deeply contextual and a function of state decisions [Beckert and Dewey Reference Beckert, Dewey, Beckert and Dewey2017: 6]. In addition, police enforcement differs greatly depending on the product and the country. For instance, under US federal law, the production and sale of marijuana are prosecuted less severely than the production and sale of cocaine. The level of enforcement might also be related to the social acceptance of the illegal act in question [Mayntz Reference Mayntz, Beckert and Dewey2017: 39–40]. And yet the fact that a given product is defined as illegal has an effect regardless of the level of enforcement: it allows state actors to apply the law selectively.Footnote ⁷

Keeping these background points in mind, in the following sections I discuss each of the activities mentioned above, namely production, trade, and governance, with specific references to the papers submitted for this special issue.

Production OC

A viable definition of OC might be based on its three distinct parts: production, trade, and governance. Each requires dedicated attention. By production, I mean transforming raw materials into finished illegal products and/or performing illegal services, and offering these on a market. While, in principle, actors engaged in production could work alone, in most cases of interest they are employed in a criminal firm, which in turn is based on a set of agency relationships of authority: superiors order employees to perform a certain task. Some goods and services will be bought on the market while others internally produced. In any case, “wage employment” and “self-employment” are both covered by this definition. Even a lone producer must, at some point, put the product in an arena where goods and services are exchanged, namely a market. While “trade” and “governance” can also be thought of as products and services, the skills and the resources required for these activities are very different from those needed for the production of, say, drugs, counterfeit medicines, or malware, and this gives rise to different kinds of OC groups.

An example of criminal production at its most basic is erotic fiction writing in China, studied here by Wang and Evelyn. The producers, in this case, are authors of books that are illegal to sell under Chinese law. One crucial element emerges from the work of Wang and Evelyn: the producer may act as the seller of their own product. Thus, in effect, the author has two roles, although the main one is that of writing erotic fiction. Since trading erotic fiction online is relatively simple, it makes sense to internalize this function. And yet even in such a simple case of criminal production, many authors engage agents and promoters to sell online. Division of labour sets in quickly.

Hamill studies a production facility in Tianjin, China, that was used by a small group of producers of counterfeit medicines, which they claimed could treat cancer, psychosis, and heart disease. Some 200,000 boxes of counterfeit medicines were manufactured in five days, claims Kevin Xu, a manager at the firm [Hamill]. Through qualitative social network analysis, Hamill finds that the producers have a clear chain of command and that their network is hierarchical, displaying a high degree of homophily. Given the needs of production in a firm, its organizational structure is, unsurprisingly, a by-product of the activity.

The counterfeit products studied by Hamill entered the same market in which the authentic medicines were being sold, and were falsely advertised as genuine. Between 5 and 7% of the global pharmaceutical market is made up of such medicines, with some countries, such as Ghana, having rates as high as 37% [Beckert and Dewey Reference Beckert, Dewey, Beckert and Dewey2017; Hamill:8]. Illegality poses a problem for producers, who find it hard to build a reputation for quality. Producers must avoid attracting too much attention while at the same time striving to promote their product. Other things being equal, this constraint leads operations to be smaller in size and scope [Reuter Reference Reuter1983]. In order to increase product visibility, some producers try to get around the difficulties of advertising, for example by branding their products, as in the case of cocaine producers who stamp a symbol on the drug package or “brick” [Curcio Reference Curcio2003]. Authors of erotic fiction use online nicknames to maintain their professional reputation [Wang and Evelyn]. Sometimes, however, people engaging in illegal activities online are forced to change their nickname to avoid being identified by the authorities, and therefore risks losing the reputational capital accumulated by their brand name. One cybercriminal was so upset at the prospect of changing his moniker that he wrote in a forum: “Here my name is solv3nt, but in other forums I am known by the name of dentrino” [Lusthaus Reference Lusthaus2012: 81]. He was trying to preserve the credibility he had accumulated in the past. In the case studied by Wang and Evelyn, buyers and sellers try to send credible (hard-to-fake) signals that help the market to function, albeit not perfectly.

By focusing on production as a separate activity, scholars can explore the roles of two key variables: namely, level of enforcement and technology. It is likely that both affect prices, location decisions, distribution chain (long/short), share of total revenues, the labour force (who is hired), the size of capital markets, the financial arrangements, size of inventories, specialization (a single product vs more products) and, as mentioned above, advertisement. Most crucially, ceteris paribus the size of the firm depends on the level of enforcement. In addition, the interplay of production and governance (see below) varies depending on local circumstances. In conclusion, the standard tools used by economic and political sociologists to study the context of legal production can be fruitfully applied in this field too.

Trade OC

The second building block of a viable definition of OC is trade. When I refer to trade OC, I mean moving goods or people from one location to another illegally. As mentioned above, it is most likely that these individuals will be part of a group. Trade OC covers, for instance, drug trafficking, smuggling and trafficking of people and/or endangered species, smuggling of arms or stolen vehicles, money laundering, and a great deal of the illegal buying and selling that takes place online. Groups involved in such activities do not aim to control economic sectors or territories. For instance, different gangs involved in human smuggling share the same routes [Campana Reference Campana2020]. They operate the equivalent of flexible supply chains [Zhang and Chin Reference Zhang and Chin2003].

Since the actors involved in trade OC do not try to exclude others from a particular market, we would expect the potential for competition to be great, although high levels of investment and some structural advantages might be necessary to enter the market. For instance, the trader must be able to forge contacts with both producers and buyers: having the necessary background helps. Moreover, the use of violence will be minimal, as there should be no conflict over territory [for a recent study that supports this view, see Baradel and Breuer Reference Baradel and Breuer2024]. In another recent paper, Baker and Duncan [Reference Baker and Duncan2025: 2] state that transnational drug-trafficking organizations—such as those that have been operating in Medellín, Colombia since the 1970s—“are not inherently political actors”, as their aim is to move drugs, yet in order to carry out their operations they seek either state or criminal protection. To the extent that they seek the former, they influence electoral politics through massive cash payments which in turn can influence electoral outcomes; if they seek the latter, they strengthen criminal governance [Baker and Duncan Reference Baker and Duncan2025].

Thus, the organizational structure of trade OC is expected to differ from that of the groups involved in production or governance (see below). While production OC has a tendency to display a high degree of hierarchy, with a top-down chain of command and homophily ties, trade OC exhibits low levels of demographic homophily, namely a tendency for nodes with different characteristics to form ties with one another and for the network to be quite centralized around key players who connect the others [Breuer and Varese Reference Breuer and Varese2023: 870]. The central actors are nodes connecting parts of the network that would otherwise be separate, allowing for the flow of information and resources. As noted by several criminologists [e.g., Morselli Reference Morselli2005; Kleemans Reference Kleemans, Tonry and Bijleveld2007], these actors operate as a bridge between “structural holes” [Burt Reference Burt1992]. In a study of OC in the city of Liverpool, the authors argue that drug trafficking (the activity) drives the patterns of interactions among groups as well as outside them [Campana and Giovannetti Reference Campana and Giovannetti2025].

Indeed, Hamill finds that in Operation Singapore, “medicines are passed along in a sequence of exchanges” [20] and the network is constituted by “temporary alliances crafted to accomplish certain objectives” [21]. The trade network is “adaptable, temporary, and transactional” [21]. This dense, clustered network structure allows for more effective, short-term coordination; the paths between nodes are short in order to make information transmission easier. Rapid and effective communication between the main players is essential: thus we expect a high volume of electronic communication across significant distances, where in-person meetings are not possible.

Arguably, there is a lot of money in trade—more than there is in production. During his trial in France, the main organizer of the distribution network of counterfeit medicines studied by Hamill, a man called Arnaud Bellavoine, declared in court: “I was hungry for money. I had big financial problems… I had the idea of Plavix and Zyprexa. They were very fashionable things and large volumes were prescribed in Europe” [Hamill: 18]. The medicines smuggled into the UK from November 2006 to May 2007 netted the criminals some £7 million. Such high profits (and margins) for trade are to be found in other illegal markets too. For instance, the wholesale price for cocaine in the Netherlands is ten times that in Colombia [Kleemans Reference Kleemans, Tonry and Bijleveld2007: 177].Footnote ⁸ Similarly, high profits can be made by those who smuggle either arms, animals, and people, or synthetic drugs. As noted by Peter Reuter, the entire cocaine supply chain resembles an hourglass: a few intermediaries make the most money, while the many who produce and retail the product make relatively little [Reuter Reference Reuter2003; see also Feltran: 4].

Promises are normally not enforced by third parties. Thus, references to the binding power of one’s word and invocations of mutual respect are a regular feature of the conversations wiretapped by the police, as in Xu’s chats with Bellavoine, a manager of offshore companies and a pivotal actor in his trade network [Hamill: 22]. Since mafia-style third-party enforcement is hard to implement when transactions are complex and take place across different continents, referring to moral obligations is a crucial mechanism [Hamill et al. Reference Hamill, Hampshire, Mariwah, Amoako-Sakyi, Kyei and Castelli2019]. Words are not always cheap.

The mechanisms that eventually cause the functions of production and trade to separate are growth in an organization’s size, complexity, and profits. The more complex the production process, the more specialized the skill set needed; the more complex the distribution chain, the harder it is for producers to oversee the shipment of goods from one part of the world to another. Hamill shows that the distribution network for counterfeit medicines in Operation Singapore is clearly separate from its production network. Trading across continents, from China to the UK, requires skills that individuals working in a plant in Northern China do not have. As expected, those involved in distribution are middle-aged British and French businesspeople who have vast competence in international trade and travel, with connections to Singapore, Mauritius, Tunisia, France, and Luxembourg, and who are proficient in more than one language. Conversely, the producers are all from the same location and speak the same language [Hamill: 18]. Some countries, such as the Netherlands, appear to have become hubs for trade OC; in the case of the Netherlands, this may possibly be due to its geographical location, population diversity, and good infrastructures. As Kleemans points out: “Criminal groups [in the Netherlands] are primarily involved in international illegal trade, using the same opportunity structure that facilitates legal economic activities” [Kleemans Reference Kleemans, Tonry and Bijleveld2007: 176]; the Netherlands’ legitimate economy is also specialized in transportation.

A focus on trade OC opens up new areas for research. Scholars might study how long-distance as opposed to short-distance trade affects the opportunity to govern. While in the former case it is almost impossible to envisage a third-party enforcer operating between, e.g., Colombia and Calabria, within a region where there is a strong governance OC, trade might be managed by a traditional mafia. In this case, the activity affects not only the internal organization but the interrelations among groups. In addition, trade OC appears to be the area in which the interplay between the legal and illegal spheres is greatest.

Governance OC

The third building block of a viable definition of OC is governance. Here, governance OC means the unlawful regulation and control of the production and trade of goods and services. This type of OC seeks to control who is allowed to produce or trade. The entities can take at least two forms. Some might limit themselves to governing the production and sale of a particular commodity, like drugs or sexual services. Others aspire to protect several markets. Thus, governance can be limited or extensive.Footnote ⁹

Governance OC covers activities such as dispute settlement, protection against competition, protection against thieves, labour racketeering, intimidation of lawful right holders, debt recovery, enforcement of cartel agreements, and governance of prisoners. With the development of cybercrime, some of these functions have been taken up by the administrators of electronic platforms on which buyers and sellers interact, the punishment normally being exclusion from future transactions [Odabaş et al. Reference Odabaş, Holt, Breiger, Beckert and Dewey2017; Lin and Wang Reference Lin and Wang2024]. A group may engage in a combination of such activities. When it engages in most of them in a certain territory, it aspires to enforce a mafia order, and thus acquires a political dimension too. When the ability of the state to govern vanishes or is absent, criminal governance becomes a form of rebel rule and eventually takes the place of the state [for a relevant example, see Winn Reference Winn2024; for academic studies, see Arjona et al. Reference Arjona, Kasfir and Mampilly2015]. Thus, the two phenomena—of OC and of rebel activities leading to civil war—are related, but should not be conflated [Kalyvas Reference Kalyvas2015].

Governance-based OC structures are likely to exhibit decentralized networks with low clustering, long path lengths, and high homophily based on demographic attributes. We expect this type of OC to take time to emerge and stabilize; thus it operates with a longer time horizon, as its participants are likely to wait patiently for the right moment to act rather than to require immediate rewards [Breuer and Varese Reference Breuer and Varese2023].

There is a tendency to confuse governance OC with corruption. It should be made clear that governance OC is autonomous from the state. When state agents ensure market stability illegally, they engage in corruption, breaching their contract with their official employer [Varese Reference Varese, Ledeneva and Kurkchiyan2000]. Similarly, the phenomenon observed in China of thugs hired by businesspeople colluding with state officials to remove peasants from profitable land is not a form of governance OC [Varese and Wong Reference Varese and Rebecca2018]. Unlike cases of corrupt officials or paid thugs, the youths known as “les petits” who operate in the Félix-Pyat neighbourhood of Marseille are an example of a rudimentary form of governance OC. They first attempted to regulate only drug distribution, and then evolved into a group imposing social order on the neighbourhood at large; a compelling mosaic of their relations emerges in Rodgers and Jensen’s account.

The drug market in Félix-Pyat has been dominated by Comorians since the mid-2010s. At first it was controlled by a group known as “les grands” (the big ones). When key members of this group were sent to prison, the youngsters took over [Rodgers and Jensen: 11]. In order to build up their reputation, they initially engaged in behaviour which caused them to come across as “unpredictable and dangerous”, as an informant told Rodgers and Jensen [Ibid.: 11]. This strategy is used by many in governance OC, in stark contrast to the practice in trade OC [for the deployment of “sadism as an instrument of cartel warfare” by Los Zetas in Mexico, see Grayson Reference Grayson2014]. In 2019, when a runaway youth from Northern France tried to sell drugs as a freelancer in Félix-Pyat, he was kidnapped and tortured by “les petits”: they burned his genitals with a blowtorch, almost killing him [Ibid.: 12]. Their actions were intended to send a clear message to anyone who might think of challenging the group’s prerogative over who could sell drugs in Félix-Pyat. The episode shocked the neighbourhood and led to a high-profile trial. Arguably, this was the intended outcome.

Criminal governance was not always present in Félix-Pyat. It appeared only at the very beginning of the twenty-first century. In the mid-l 990s, independent dealers were still able to trade: for example, the man known as Disco, who had arrived in Félix-Pyat as a young man in the early 1980s. Disco sold drugs from his first-floor apartment, having bought his supplies from corrupt policemen. Somebody informed on him to honest police officers, and he was jailed in 1999. This example also suggests a lack of order and predictability in the market: anybody can rat on their competitors. The appearance of a rudimentary form of governance makes the market more predictable, ensuring that no one informs the authorities about the activities of protected dealers.

Rodgers and Jensen returned to Félix-Pyat in 2023, some eleven months after their first trip, and noticed that a transformation had taken place. “Les petits” were more friendly, they displayed sartorial uniformity, and most significantly, they had written a letter to the people of Félix-Pyat. In the text, they apologized for past misunderstandings, pleaded with the residents not to go to the police, threatened those who did, and promised to take care of the neighbourhood and help the community [Rodgers and Jensen: 12]. Many saw the letter in a positive light. Shortly afterwards, youths dressed in the gang’s outfits were seen cleaning the streets. The group even intervened to settle a brawl between two families, calming everybody down. This parallels the case Zeng and I documented recently of a neighbourhood gang in the city of Nottingham [Varese and Zeng Reference Varese and Zeng2024]; as in Félix-Pyat, here too the young gang members had initially only been involved in one market (drugs), but over time they had taken on the role of community governance (for similar dynamics in the case of London, see Campana et al. Reference Campana, Varese and Meneghini2025]. “Les petits” were connected to larger groups and became the target of citywide conflicts. While criminal governance can bring a degree of order, this may also be accompanied by more violence.

Governance should be thought of as a continuum of complexity. It can be ad hoc, informal, precarious, and small-scale, nesting in deprived neighbourhoods within relatively well-functioning states where policing is generally effective. At the other end of the spectrum are the complex and far-reaching forms seen in parts of Latin America [Lessing Reference Lessing2021; Feltran Reference Feltran2020], the Caribbean, and in traditional mafia territories [Gambetta Reference Gambetta1993; Varese Reference Varese2001]. These groups might be able to provide infrastructural services, jobs, loans, and even healthcare [Rodgers and Jensen: 14]. Yet they all belong to the same category; the differences are in degree and not in kind. This is why the classification of OC into three types of functions suggested here is valuable.

Feltran’s study focuses on an extraordinary governance system that ensures more efficient delivery of drugs across the world. The Brazilian Primeiro Comando da Capital (PCC) has developed a system—known as the Agência—that connects independent producers in Colombia, Bolivia, and Peru, cocaine traffickers in Latin America, and buyers in various parts of the world [Feltran: 5]. Entry into the system is voluntary, and the driving motives for all those involved are higher profits and greater certainty. The PCC’s rise to prominence generated a conflict with a competing criminal governance structure—the Comando Vermelho (Red Command) [Feltran: 20]. While many criminologists have endorsed Reuter’s [Reference Reuter1983] finding that criminal firms remain small and eventually fragment, they failed to take into account that this result was predicated on a high degree of state repression. When the state proves unable or unwilling to control significant sectors of society (including its prisons), governance OC grows in complexity and sophistication, as was the case with the PCC or, further afield, the Japanese mafia [Baradel Reference Baradel2025]. The effects of the PCC’s system are that the market for cocaine runs more smoothly, prices are low, and supply is high, as noted by several recent official reports cited in Feltran [Ibid.: 1]. In a nutshell, the PCC ensures stability and governance in the cocaine market, from production to distribution (trade), ultimately backed by the ability to use violence.

As a result of his extraordinary level of access to participants, Feltran can confirm that the “myriad of individual or collective entrepreneurs” (known as companheiros) remain autonomous. The PCC would not be able to take over them all. Instead, its system provides them with infrastructure, access to ports, secure routes, and reliable payment methods. The PCC is able to make this system work because it retains a degree of territorial power, for instance on the Bolivian border (Caceres), at the Port of Santos, and at “many other Brazilian, Argentinian, and Uruguayan ports” [Feltran: 6]. Its logistical hubs extend to French Guiana and the port of Le Havre. The key innovation of the system is that participants keep their profits and do not pay a fee to join. Instead, they agree to produce or deliver drugs that are “owned” by the PCC. Thus, institutional revenues are clearly distinct from individual ventures. In return, independent traders get a better deal on their own drugs. Everybody wins, if they stick to the plan. If disputes arise, adjudication is carried out by the PCC’s internal justice system [Feltran: 11].

The modern study of governance OC has a long tradition in the social sciences, starting with the insightful essays by Frederic C. Lane [Reference Lane1958], Thomas C. Schelling [Reference Schelling1967; Reference Schelling1971], Robert Nozick [Reference Nozick1974], Mansur Olson [Reference Olson1993], and the empirical research agenda on protection inaugurated by Diego Gambetta [Reference Gambetta1993], which built on studies by Landesco [(1929) Reference Landesco1968] and Reuter [Reference Reuter1983]. Researchers in empirical economics have refined the theory and conducted extensive empirical tests, creating a rich body of work [see Porreca and Thompson Reference Porreca, Thompson and Zimmermannforthcoming]. For instance, the claim that governance OC tends to emerge during periods of sudden market expansion in the absence effective state regulation is supported by a wide range of cases, including nineteenth-century Sicily, Russia in the 1990s, and the northern Italy in the 1970s [Gambetta Reference Gambetta1993; Bandiera Reference Bandiera2003; Buonanno et al. Reference Buonanno, Durante, Prarolo and Vanin2015; Dimico et al. Reference Dimico, Isopi and Olsson2017; Varese Reference Varese2011]. And yet there are still two crucial areas in which scholars could make additional and original contributions.

First, most studies tend to focus on Latin America and traditional mafia territories. The assumption, either explicit or implicit, is that such groups operate in a context where the state is “infrastructurally weak” [Centeno Reference Centeno2002: 10; see also Feldmann and Luna Reference Feldmann and Luna2022]. However, a few recent investigations have shown that governance OC can nest within strong states and affects immigrant communities that cannot turn to legitimate institutions [e.g. Densley Reference Densley2013]. Is it possible that factors other than recent immigration—such as barriers to accessing local services, affordable housing, public facilities, or social clubs and venues for recreation, in a setting where drug profits and unemployment are high—might give rise to this form of OC [Campana et al. Reference Campana, Varese and Meneghini2025; Varese and Zeng Reference Varese and Zeng2024]? Additional research on governance OC in non-immigrant communities in strong states in the Global North is likely to be extremely fruitful.

Opposition to existing oppressive power structures and adherence to principles of equality are part of the ideology of the PCC, which points to the fact that governance OC can be a political actor. Indeed, states are also in the business of governing markets and territories. Once a criminal group has dominated a market (e.g. for drug trafficking), it often succeeds in expanding its control to neighbouring markets. If this logic is extended, OC groups can evolve into fully developed mafias and, if unchecked, into states. In other words, governance OC—including mafias—and states belong to the same category, the key difference being that the latter are not illegal.

The two studies on governance OC by Rodgers and Jensen and by Feltran capture the phenomenon at almost opposite ends of the spectrum. While in Félix-Pyat the governance of the neighbourhood by “les petits” is rudimentary and fragile, the PCC has developed what might be the most complex and efficient system for the regulation of drug production and trade in existence today. Yet, some key commonalities confirm that governance OC differs from trade-based OC in terms of its structure, skill requirements, and activities. Unlike trade-based groups, which seek quick financial returns, governance-oriented groups operate with long-term goals in mind, investing heavily in violence. Although that violence is rarely used, the credible threat it poses helps to establish control and minimize challenges. These groups also gather extensive intelligence on their territories (not just on a specific product or market), enhancing their ability to govern effectively. The presence of governance OC improves the functioning of the market, offering a degree of protection to dealers. However, we cannot subscribe to a simplistic demand-and-supply model. Governance might not emerge even if there is a demand for it, or, when it does, it might be imposed on traders (as state governance is often imposed on communities), or it might benefit only a minority. In the case narrated by Feltran, the Agência’s platform structure enables “the extreme exploitation of low-level independent operators, who are often from impoverished and marginalised communities” [23].

The framework proposed here also calls for further research on a second area of interest: the intersection of the three factors we have examined, production, trade, and governance. For instance, the `Ndrangheta appears to be involved in both trade of drugs and local governance of territories and markets in the region of Calabria. And yet its ability to dominate the trade of cocaine into Southern Europe is predicated on its control (governance) of the port of Gioia Tauro. The framework I propose allows us to ask questions such as the following: what are the conditions that enable a mafia to govern trade? Can a trade OC group evolve into governance OC, as appears to have been the case with some Mexican “cartels” and the Medellín organization led by Pablo Escobar [for the case of Colombia, see Baker and Duncan Reference Baker and Duncan2025]? Does the nature of the product traded—including humans—affect the chain’s governability? How does cybercrime production intersect with online trade and governance? Answers to such questions could shed more light on the ongoing transformation of OC.