Sage Journals: Discover world-class research

Abstract

Objectives

This study examines how social embeddedness and multiplex social ties shape criminal collaboration in the Sinaloa Cartel. It investigates how different types of relationships—such as kinship, friendship, meetings, and compadre ties—influence participation in drug crimes and broader forms of collaboration.

Methods

Drawing on trial transcripts from Joaquín “El Chapo” Guzmán Loera's U.S. federal case, we reconstructed social networks among 188 actors. Using social network analysis and Logistic Regression Quadratic Assignment Procedure (LRQAP) models, we assessed the conditional associations of different social ties with drug crimes and criminal collaboration, controlling for demographic and organizational characteristics.

Results

Friendship and meeting ties exhibited the strongest associations with drug crimes and criminal collaboration. Familial ties were initially significant but lost their effect once shared cartel affiliation was considered, suggesting a mediating role of organizational membership. Compadre and prison ties showed no significant associations.

Conclusions

Trust-based relationships such as friendships and meetings play a pivotal role in fostering collaboration within organized crime groups. Familial ties may exert indirect effects through shared cartel affiliation. The findings underscore the importance of distinguishing between types of social ties and demonstrate the value of trial transcripts as a source for advancing empirical research on organized crime.

Keywords

Organized crime criminal networks social network analysis drug trafficking organizations multiplexity

Introduction

Organized crime groups are defined by their illicit operations and perhaps even more by their resilience in hostile environments. A key reason for this resilience lies in the web of social relationships that binds individuals in these organizations. Social ties—family ties, friendships, or long-standing acquaintances—facilitate recruitment, foster trust, and reduce the risk of betrayal (Calderoni, Comunale et al. 2022b). The notion of social embeddedness captures this reality, emphasizing that economic and criminal action is deeply rooted in social relationships (Granovetter 1985; Van de Bunt, Siegel and Zaitch 2014).

While the concept of embeddedness has become a dominant framework, important gaps remain in our understanding of how social ties influence illegal collaboration within large-scale, criminal organizations. Much of the literature tends to treat social ties in broad and undifferentiated terms, without examining whether specific types of relationships play distinct roles in shaping organized crime cooperation. This approach risks overlooking how overlapping or multiplex relationships—where individuals are connected through overlapping ties such as kinship, friendship, and criminal partnerships—affect trust, cooperation, and organizational resilience (Smith and Papachristos 2016; Verbrugge 1979). Moreover, existing research has largely focused on small, localized networks, mostly based in advanced economies. Conversely, major drug trafficking organizations—particularly those based in Latin America—are underexplored in empirical studies. Mexican drug trafficking organizations, in particular, play a central role in regional violence and illicit markets, while also influencing political and institutional dynamics. Examining cooperation and cohesion within criminal networks may help explain why large-scale and complex organizations, such as Mexican drug trafficking syndicates, persist despite sustained and heightened enforcement efforts. Yet, systematic empirical research on their internal social dynamics remains limited.

To address these gaps, this study examines how different social ties relate to criminal collaboration in a key portion of the Sinaloa Cartel, one of the largest and most influential drug-trafficking organizations in the Americas. Leveraging extensive trial transcripts from Joaquín “El Chapo” Guzmán Loera's U.S. federal case, we systematically reconstruct networks of 188 individuals connected through various relationships such as family, friendship, compadre (a tie between a child's parent and godparent—particularly significant in Mexican culture; see Gill-Hopple and Brage-Hudson 2012), joint meetings, and shared criminal activities, and assess how each type of tie influences criminal collaboration. While the dataset reflects the trial's focus on Guzmán and his closest associates, it offers an unusually detailed account of interpersonal dynamics within the core of the organization. Using social network analysis and Logistic Regression Quadratic Assignment Procedure (LRQAP) models, we quantify the impact of these multiplex social ties on both drug crimes and broader criminal cooperation. Our results show that friendship and meeting ties are the strongest predictors of criminal collaboration, while family ties lose significance when controlling for shared group membership, suggesting indirect effects through organizational affiliation. In contrast, other ties such as compadre relationships, prison connections, and joint violent crimes do not show significant associations. These findings underscore the pivotal role of trust-based relationships in fostering cooperation within organized crime and demonstrate the methodological value of trial transcripts for uncovering the social dynamics that sustain criminal enterprises.

By concentrating on the relational dimensions of criminal collaboration, our study offers a complementary perspective to existing scholarship on Mexican drug trafficking organizations. While political science and economics have highlighted the role of state responses, institutional dynamics, and market conditions in shaping organized crime, a criminological approach grounded in social network analysis provides fresh insights into how these organizations are built and sustained from within. This perspective underscores the importance of interpersonal ties and overlapping relationships as mechanisms of trust and cohesion in large-scale, high-risk criminal groups.

Background

The Embeddedness of Organized Crime

The concept of embeddedness, introduced by Polanyi ([1944] 2001) and redefined by Granovetter (1985, 1992), emphasizes that economic action is rooted in social relations. Notably, Granovetter specified that group offending relies on a level of internal trust “that usually follows preexisting lines of relationship” (1985:492)—a point that criminologists largely overlooked for years.¹

At the turn of the twenty-first century, criminologists revived embeddedness to explain how criminal organizations emerge and operate across different contexts (Kleemans and van de Bunt 2008; Kleemans and de Poot 2008; Kleemans and Van De Bunt 1999; Paoli 2002: 84; Van de Bunt et al. 2014; Von Lampe 2016). This revival aligned with classic organized crime studies, who showed how kinship, friendship, and community ties sustain trust, collaboration, and resilience criminal organizations (Albini 1971; Ianni and Reuss-Ianni 1972; Hess [1970] 1973; Blok 1974).

Over time, embeddedness has broadened to encompass the idea that all economic action holds social relevance, overlooking Granovetter's original focus on the direct interplay between social ties and economic behaviour (Joseph and Smith 2021; Krippner et al. 2004). Applied to organized crime, it highlights how criminal organizations are deeply embedded in overlapping social ties—family, kinship, friendship, and professional networks—that shape their structure and operations (Van de Bunt et al. 2014).

Empirical studies using social network analysis have demonstrated how relational embeddedness defines criminal group boundaries, exposes legal–illegal overlaps, and maps cross-border collaborations (Aziani, Berlusconi and Giommoni 2021; Bouchard 2020; Morselli 2009; Papachristos and Smith 2014; Waring 2002). Social ties also affect recruitment (Calderoni, Comunale et al. 2022a; Decker and Van Winkle 1996; Kleemans and Van De Bunt 1999), shape roles (Calderoni 2012; Calderoni and Superchi 2019; DellaPosta 2023; Morselli 2009; Morselli and Petit 2007), and career opportunities, persistence, and desistance (Morselli 2001, 2003, 2005; Pyrooz, Sweeten and Piquero 2013).

Overall, research has shown that diverse social ties serve both as the breeding ground and the binding force of criminal organizations. However, what often remains underexplored is how these ties intersect. A particularly revealing dimension of embeddedness lies in the multiplexity of relationships—when familial, social, and economic ties overlap within the same dyads. These multi-layered connections are especially relevant, as they amplify trust, facilitate coordination, and strengthen organizational resilience.

Multiplexity in Criminal Networks

Multiplexity refers to the phenomenon wherein individuals maintain overlapping types of relationships—familial, social, and economic—within the same network. Originating in broader sociological and anthropological discussions on role plurality and social organization (see Simmel 1950), the concept underscores how people can simultaneously inhabit multiple roles (e.g., relatives, friends, and business partners), thereby intensifying mutual obligations and norms of reciprocity. These layered relationships resonate with Granovetter's (1985) notion of embeddedness and with Coleman's (1988) conception of social capital, where dense, multiplex ties promote collaboration, trust, and information flow. Multiplexity has gained traction as a lens through which to understand how these entwined connections can bolster cooperation, trust, and group cohesion, particularly in contexts where formal legal institutions do not reliably enforce agreements (Smith and Papachristos 2016; Wippell and Haynie 2025).

Criminal collaboration unfolds in contexts where formal legal mechanisms—contracts, courts, and state enforcement—are weak or absent (Gambetta 2009; Paoli 2002; Reuter 1983). In such high-risk environments, trust and informal enforcement become essential to prevent defection. Scholars applying the concept of relational embeddedness to organized crime have shown that layered, multi-dimensional social bonds—such as kinship, friendship, or ritual obligations—are critical for fostering trust, internal cohesion, and resilience. These multiplex relationships unify offenders around shared norms and enable informal conflict resolution where legal institutions are lacking or distrusted. Yet despite highlighting the importance of these overlapping ties, prior research has rarely measured how specific types of social bonds interact and affect distinct dimensions of criminal collaboration—leaving their relative influence largely unexplored.

Few studies have empirically measured the effects of different types of relationships on criminal collaboration. Recent scholarship has advanced this area by describing multiplexity and assessing its impact on organized crime networks (Bright et al. 2015; Malm, Bichler and Van De Walle 2010). Campana and Varese (2013) tested whether kinship ties and the disclosure of past violent acts influenced the likelihood of criminal collaboration. They found that both significantly increased the probability of cooperation, demonstrating how overlapping relational dimensions—such as familial bonds and reciprocal trust—enhance collaboration within organized crime networks. Similarly, Smith and Papachristos (Papachristos and Smith 2014; Smith and Papachristos 2016) found that offenders with overlapping ties, such as familial connections, neighbourhood links, and joint criminal ventures, maintained more durable collaborations in Prohibition Era Chicago's organized crime networks. Diviák et al. (2019) demonstrated that diverse relational channels stabilize a high-risk Czech corruption network by providing redundant communication pathways and reinforcing trust. Together, these studies highlight how multiplex relationships deepen social embeddedness, promoting cooperation and enhancing resilience against both internal conflicts and external threats.

In conclusion, although more recent scholarship has begun to empirically probe multiplexity, existing research is still limited, and questions persist regarding how different social dimensions intersect to shape trust, cooperation, and resilience in criminal networks. Our current understanding lacks empirical, quantitative evidence into whether different relationships such as family, friendship, prior interactions, or joint violence affect criminal cooperation in general and drug trafficking collaborations more specifically.

The Current Study

In this study, we advance the examination of multiplexity in organized crime networks by systematically examining how different social relationships affect the commission of drug crimes and criminal collaboration in a large-scale, high-risk, criminal organization. More specifically, we aim to fill three gaps in the current knowledge.

First, there is a clear need for empirical research that directly examines how multiplex social relations affect criminal collaboration. While criminologists have embraced the concept of embeddedness—extending it beyond its original focus on purely economic exchanges (Joseph and Smith 2021)—much of this work remains theoretical or qualitative. In particular, many studies analysed the importance of layered social bonds (e.g., familial, economic, and community ties) in shaping offender decisions, yet relatively few systematically disentangled the influence of each type of network tie in networks of criminal collaboration outcomes such as trust, cooperation, and long-term partnership (Gambetta 2009; Granovetter 1985). This gap is significant because understanding the specific contributions of different social dimensions can illuminate the mechanisms that strengthen or undermine illicit ventures.

Second, although numerous studies investigate the role of social relations in criminal collaboration, few focus on large-scale drug trafficking syndicates. Existing criminal network research often centres on street gangs (Ouellet, Bouchard, and Charette 2019), mafias (Calderoni 2012; DellaPosta 2023), or outlaw motorcycle groups (Morselli 2009; Van Deuren, Diviák, and Blokland 2025), each of which has wide-ranging social and cultural objectives that may extend beyond profit-making. Even within studies of drug trafficking, the emphasis typically falls on small-scale networks (Bright, Hughes, and Chalmers 2012), leaving the organizational strategies and relational dynamics of high-level drug trafficking organizations relatively understudied. Some exceptions note that significant players in large-scale trafficking prefer indirect ties to limit exposure (Desroches 2005; Dorn, Oette, and White 1998; Gundur 2022), and that any direct connections often stem from long-term personal relations linked to kinship or ethnicity (Browne, Mason, and Murphy 2003; Reuter and Haaga 1989). These insights suggest that bigger, more diverse networks may face unique challenges in sustaining criminal cooperation—a phenomenon that warrants closer examination.

Third, to date, much of the literature on organized crime has concentrated on a small number of countries with advanced economies, often overlooking the diverse sociopolitical, economic, and cultural factors that shape criminal collaboration in other regions (Hosford et al. 2021). This narrow focus has limited the field's empirical base and constrained our understanding of how different institutional and social contexts influence the formation, structure, and operation of criminal networks (Pereda 2022).

Latin America exemplifies this gap. Despite the central role organized crime has played in driving violence across the region—making it the most violent globally (UNODC 2023)—research on organized crime in Latin America remains relatively sparse compared to studies conducted in wealthier regions. Yet, Latin America is not merely a geographic “blind spot” in the literature; it is a region where organized crime exhibits distinctive patterns and produces complex political, economic, and social effects that merit closer examination.

Mexico is a particularly significant case. The country's organized crime landscape is deeply intertwined with its historical processes of state formation (Pansters 2018; Smith 2021), the evolution of its political institutions (Durán-Martínez 2018; Snyder and Durán-Martínez 2009; Trejo and Ley 2020), and the development of specific illicit markets (Aguirre and Gomez 2020; Le Cour Grandmaison, Morris, and Smith 2019). These factors have contributed to the development of organized crime dynamics that reflect the particularities of Mexico's social, institutional, and economic environment (Correa-Cabrera 2017; Durán-Martínez 2015; Lessing 2015).

The consequences are far-reaching. Mexico continues to experience high levels of organized crime related violence, with a homicide rate of 24.86 per 100,000 inhabitants as of 2023 (UNODC 2023). Between 2000 and 2024, the Mexican government reported over 121,675 missing persons (San Juan Flores and Guillén 2025). Beyond the direct human toll, organized crime has weakened democratic institutions and the rule of law (Gutiérrez-Romero and Iturbe 2024; Ley 2018), constrained economic growth (Garriga and Phillips 2023), and contributed to the expansion of transnational drugs and firearms trafficking networks (Beittel 2022; Perez Esparza, Johnson, and Gill 2020).

Over the past two decades, political scientists and economists have made important contributions to the study of organized crime in Mexico. Their research has examined the effects of various state responses, including large-scale military deployments (Espinosa and Rubin 2015; Treviño-Rangel et al. 2022), kingpin strategies (Calderón et al. 2015; Estévez-Soto and Lecona Esteban 2024), drug-related arrests (Atuesta 2024), and alternative policing models, including indigenous and community-based approaches (Ley, Mattiace and Trejo 2019; Magaloni, Gosztonyi, and Thompson 2022).

However, criminology has yet to fully engage with these developments. We suggest that criminological approaches—particularly those grounded in social network analysis—can provide a valuable complement to existing research. By focusing on the relational dimensions of criminal collaboration, such approaches can deepen our understanding of how criminal networks are built and sustained over time. Specifically, a network perspective can shed light on which types of social ties—whether based on kinship, trust, or mutual obligation—are most effective in maintaining group cohesion and operational capacity within criminal organizations.

This article addresses these gaps by examining the intersection of social relationships and criminal collaboration in the context of the Sinaloa Cartel, a prominent, large-scale drug trafficking syndicate based in Mexico.

The Sinaloa Cartel emerged in the late 1980s and has gone on to become one of the most prolific drug-trafficking syndicates in the Americas (Valdés Castellanos 2013). The U.S. Drug Enforcement Administration characterizes the cartel as “one of the most violent and prolific polydrug-trafficking cartels in the world [at the heart] of the most dangerous and deadly drug crisis the U.S. has ever faced” (DEA 2024:4).² Estimates of the group's membership vary significantly, ranging from approximately 17,825 individuals (Prieto-Curiel, Campedelli and Hope 2023) to as many as 100,000 (Bunker 2010). Despite its prominence, the Sinaloa Cartel has, until recently, received scarce academic attention, especially in comparison to the Italian and American mafias.

Much of what is currently known about the Sinaloa Cartel derives from journalistic accounts rather than academic research. These accounts often concentrate upon the organization's historical evolution (Grillo 2011; Osorno 2012), explore specific aspects of individual members’ lives (Beith 2020; Hernández 2019), or analyse high-profile events, most notably the trial of former leader Guzmán Loera (Esquivel 2019; Ibarra Chaoul 2019). Whilst many journalistic accounts touch upon elements of the Sinaloa Cartel's social embeddedness, they typically do not do so from a systematic scholarly perspective.

The relatively scant academic literature on the Sinaloa Cartel has yet to examine it explicitly through a social embeddedness framework. Nevertheless, existing research offers expedient insights for exploring the cartel's structural and relational embeddedness. For example, historical analyses of the cartel (Astorga 2016; Enciso Higuera 2015; Smith 2020; Valdés Castellanos 2013) underscored the evolving interplay between drug-trafficking activities and shifts within Mexico's political institutions. Other scholars have investigated specific dimensions of the cartel's organizational functioning, including its reliance upon mechanisms of illegal governance (Felbab-Brown 2022; Pereda and Décary-Hetu 2023), the extent to which intergroup cooperation and internal fragmentation shape its patterns of violence (Atuesta and Pérez-Dávila 2018), and its similarities with and differences to traditional mafia organizations (Paoli, Peters, and Reuter 2023), including its reliance upon kinship for recruitment (Reuter and Paoli 2020). Whilst not all large-scale Mexican crime syndicates appear to rely predominantly on blood ties (e.g., Los Zetas) (Correa-Cabrera 2017), the Sinaloa crime syndicate, and other offshoots of the now-defunct Guadalajara Cartel (e.g., the Arellano Félix and Amado Carrillo organizations, commonly known as the Tijuana and Juárez cartels), exhibit a marked reliance upon familial and kinship networks (Astorga 2016; Paoli et al. 2023; Valdés Castellanos 2013).

However, no study to date has systematically and empirically examined how the Sinaloa Cartel's social relationships—particularly through quantitative network analyses of multiplexity—shape collaboration in drug trafficking activities. Addressing this gap is the principal aim of the present study.

Due to the scarce empirical research on how multiplex social ties shape collaboration in criminal networks—particularly in complex structures like Mexican drug trafficking cartels—no established theoretical or empirical baseline currently exists for evaluating the effects of multiplexity. Accordingly, we adopt a pragmatic and exploratory approach. Our analysis is informed by a limited but growing body of network-based studies specifically addressing multiplexity, as well as by broader theoretical and empirical contributions on the social embeddedness of gangs, mafias, and drug trafficking organizations. On this basis, we propose two simple, common-sense hypotheses to orient our analysis.

H1: Ties associated with trust are positively associated with collaboration, while ties associated with distrust are negatively associated.

Specifically, we expect that personal or operational ties—such as family, friendship, prison, joint violence, and meetings—facilitate collaboration by building trust and reducing uncertainty. In contrast, ties characterized by conflict, such as violence and rivalry, are expected to hinder cooperation.

H2: The strength of association between different social ties and criminal collaboration varies, with deeper, trust-based relationships exerting a stronger influence than more situational or transactional ties.

Not all social relationships equally foster cooperation within criminal networks. Ties rooted in long-term personal bonds—such as kinship, close friendship, and compadre relations—are typically embedded in shared histories, cultural norms, and affective obligations. These ties are more likely to generate enduring trust and informal mechanisms for resolving disputes or ensuring loyalty. In the case of the Sinaloa Cartel, available evidence highlights the strategic use of such relationships to manage internal cohesion, secure cooperation, and prevent defection. By contrast, situational or transactional ties—such as co-participation in violent acts, shared prison experiences, or meetings—may emerge from immediate operational needs rather than from enduring personal trust. While these ties may facilitate contact or coordination, they are less likely to produce the strong normative bonds that underpin sustained collaboration. Therefore, we expect trust-based ties to have a more robust and consistent association with criminal collaboration than ties formed through discrete or context-dependent interactions.

These hypotheses are intentionally simple. They reflect the exploratory nature of this study and the need to establish foundational empirical insights into how social relationships shape criminal collaboration in complex, high-risk networks like the Sinaloa Cartel.

Methodology

Data and Measures

Our principal data source was the official transcript from the U.S. federal criminal trial against Joaquín “El Chapo” Guzmán Loera, held before a jury in Brooklyn, New York, between November 2018 and February 2019. We accessed the hearing transcripts via the U.S. Public Access to Court Electronic Records (PACER) database. The complete set of transcripts includes 55 files from January 20, 2017, the day of the arraignment, to February 12, 2019, the day of the verdict. Our analysis focused upon the portion of the trial where evidence was presented to the jury, which covered hearings from November 13, 2018, to January 30, 2019, corresponding to 34 documents spanning over 6,800 pages.

Guzmán was put on trial for ten counts, including engaging in a Continuing Criminal Enterprise, international drug trafficking, use of firearms, and money laundering conspiracy. These charges covered activities spanning from the late 1980s through 2014, reflecting his leadership of the Sinaloa Cartel over more than two decades.

As is customary in U.S. federal trials, the evidence was primarily presented through live witness testimonies, including cooperating witnesses, law enforcement agents, and other individuals with direct or indirect knowledge of drug trafficking operations. These testimonies, delivered under oath in open court and transcribed verbatim, served as the primary source for extracting relational data. To strengthen reliability and contextualize the information, we complemented the transcripts academic studies and journalistic accounts of the trial, the Sinaloa Cartel, and biographies of the cartel's leaders (Beith 2020; Esquivel 2019; Hernández 2019; Ibarra Chaoul 2019; Smith 2021).

We manually extracted data on actors and their relationships from the transcripts into spreadsheets. One researcher performed the initial coding, another validated it, and a third conducted additional consistency checks, including additional online and archival searches to verify, correct, and supplement information such as exact names, dates of birth, kinship ties, and cartel affiliations, when available. For each actor, we recorded attributes such as name, surname, pseudonym, sex, year of birth, nationality, cartel affiliations and dates of affiliation, and role or function within drug trafficking organizations (Table 1). We coded eleven types of relationships. Drug crime captures joint involvement in drug-related activities such as production, transport, or distribution. Illegal transaction refers to the exchange of illicit goods or services for money, including acts of bribery. Worked together denotes collaboration on operational tasks not otherwise classified. Meeting ties are based on shared participation in events or gatherings. Family ties include close blood or marital relations, excluding indirect kinship. Friendship is recorded when one actor refers to another as a friend, distinct from family ties. Compadre ties reflect godparent relationships, typically established through events such as baptisms or weddings, where one actor serves as godparent to the other's child. We distinguish between Violence, where one actor targets another, and Joint violence, where both actors participate together in acts such as assault or murder. Finally, we identify ties based on Prison interactions and sustained Rivalry.

Table 1.

Summary of the Actor Attributes (n = 188).

Variable	Level	Count	%
Sex	Male	168	89
	Female	20	11
Decade of birth	1960s	24	13
	1970s	21	11
	1980s	21	11
	1950s	13	7
	1940s	6	3
	Other	3	2
	Missing	100	53
Origin	Mexico	95	51
	Colombia	34	18
	U.S.A	11	6
	Other	11	6
	Missing	37	20
Cartel^a	Sinaloa Cartel	74	39
	Beltran Leyva	12	6
	North Valle Cartel	12	6
	Cifuentes	9	5
	Juarez Cartel	9	5
	Tijuana Cartel	8	4
	Los Zetas	3	2
	Gulf Cartel	2	1
	La Linea	2	1
	Medellin Cartel	2	1
	Cali Cartel	1	1
	No Cartel	73	39
Role/function	Trafficker	87	46
	Group leader	22	12
	Middle-level management	22	12
	Family member	18	10
	LEA	15	8
	Sicario	6	3
	Missing	18	10

Individuals could be affiliated to one or two drug trafficking organizations. The statistic reports the frequency of affiliations by organization.

All relationships were documented as a tie list, which means they were recorded as connections between two actors. If a relationship involved more than two people, then we documented all possible actor combinations. The frequency of the recorded relationships varied significantly across types, and, except for engaging in violence against someone, all relationships were considered undirected. Overall, we identified at least one relationship for 188 actors.³

We constructed valued adjacency matrices that recorded the frequency with which two actors shared each type of relationship. Given the nature of the data sources as well as the potential bias towards a few specific individuals (such as the defendant and his closest associates), we binarized and reciprocated these matrices. This approach enabled us to capture whether any two actors shared at least one relationship, without indicating a specific direction. Additionally, we created a combined relationship category called “criminal collaboration” by merging instances of drug-related crime, illicit transactions, and collaborative activities. The matrices facilitated the analysis of binary undirected networks, each with distinct topological features and structural characteristics (Table 2 and Figure 1).

Figure 1.

Graphs for the different networks.

Table 2.

Descriptive Statistics of the Networks.

Relation	Actors	Recorded relations	Edges	Density	Avg. degree	Clustering coeff	Diameter
Criminal collaboration	109	1,586	247	0.04	4.53	0.20	7
Drug crime	81	1,078	170	0.05	4.20	0.22	4
Meeting	92	923	337	0.08	7.33	0.38	5
Illegal transaction	49	388	64	0.05	2.61	0.05	5
Family	56	166	164	0.11	5.86	0.91	2
Worked together	62	120	118	0.06	3.81	0.22	6
Violence	68	104	88	0.04	2.59	0.03	8
Joint violence	31	94	50	0.11	3.23	0.46	4
Friendship	56	72	72	0.05	2.57	0.16	6
Rivalry	38	72	72	0.1	3.79	0.03	5
Prison	17	28	17	0.12	2.00	0.06	5
Compadre	17	16	16	0.12	1.88	0.00	4

The set of actors was mostly composed of men (89%), many of whom were in their forties, fifties and sixties (35%). Age was the least reliable and most difficult variable to capture, in as far as we could not ascertain the decade of birth for around half of the actors. Most actors originated from Mexico (51%), although actors from Columbia (origin of drugs) and the U.S. (destination of drugs) were also common (24%). As the trial centred around El Chapo and his connections, many of the actors (39%) were part of the Sinaloa Cartel. However, many other drug trafficking organizations were mentioned during the trial, thus highlighting the many collaborations and conflicting ties that bring these groups together. The data also stress the importance of non-affiliated actors (39%) who either facilitate the activities or try to impede them. This is consistent with other network research showing that non-affiliated actors can constitute a substantial share of criminal networks. For example, Morselli's study on Hells Angels’ involvement in drug trafficking found that up to 72% of actors were not members of biker gangs (Morselli 2009:123). Almost half of the actors were presented as traffickers of various levels. The leadership (12%) and middle-level management (12%) of drug trafficking organizations represented around a quarter of all the actors, suggesting comparable patterns of hierarchical concentration with other criminal networks. For example, Calderoni and Superchi (2019) studied four mafia networks and found that leaders made up 5–14% of members, while middle-level affiliates accounted for 6–11%.

The number of ties coded for each type varied greatly. During the trial, witnesses were primarily asked about their illicit activities in connection with El Chapo, and not so much about their rivalries, their time in prison or their friendship ties. This explains why a lot of information was collected about the drug crimes and illicit transactions, and not so much about who their compadres are. Violent ties were also surprisingly low, perhaps indicating that the prosecution focused more upon the drug crimes, and that El Chapo was too removed from the acts of violence to be connected to them directly beyond reasonable doubt.

Figure 1 and Table 2 present the topology and descriptive statistics of the different networks. They show marked differences in density, cohesion, and centralization across relationship types. Our main outcome variables, criminal collaboration and drug crime, generate large and connected networks, with over 1,000 recorded relations and average degrees around 4.5 and 4.2, respectively. Meeting ties exhibit the highest average degree (7.33) and a high clustering coefficient (0.38), indicating dense local connectivity, while family ties, though smaller in size, are highly cohesive with the highest clustering (0.91). In contrast, networks like compadre, prison, and rivalry are much smaller, sparser, and more fragmented, with lower edge counts, clustering, and average degrees. The joint violence network, while limited in scope, also shows moderate cohesion. These patterns—also visually evident in the network diagrams—underscore the substantial heterogeneity across types of ties, both in how they shape the overall connectivity and structure of the Sinaloa Cartel's relational landscape and in how they are captured and reported within the source material.

Analytical Strategy

To assess the impact of different relationships upon drug crimes and broader criminal collaboration, we conducted both descriptive and inferential analyses. The descriptive analyses encompassed not only the examination of the topological features of each of the networks reported above but also an assessment of the networks’ multiplexity. We measured the multiplexity of different relationships by calculating the Jaccard index for each pair of networks. This index quantifies the overlap between two networks, providing insights into how frequently actors share multiple types of relationships. The Jaccard index, commonly used to assess the degree of overlap between two sets, is calculated as the ratio of the intersection to the union of the two sets:

J (A, B) = \frac{| A \cap B |}{| A \cup B |}

where A and B represent two distinct networks, each composed of a set of ties between actors. This approach enabled us to systematically evaluate the extent to which different types of relationships co-occur across the networks (Figure 2).

Figure 2.

Heatmap of the Jaccard Indexes for each pair of networks.

To explore the factors influencing involvement in drug crimes and broader criminal collaboration, we conducted a series of Logistic Regression Quadratic Assignment Procedure (LRQAP) analyses. The LRQAP is a statistical method designed for analyzing relational data within networks, where observations are interdependent and non-independent by nature. Traditional regression techniques assume that observations are independent of one another. However, in network data, this assumption is violated because the presence or absence of a relationship between one pair of actors can be influenced by the relationships between other pairs. LRQAP addresses this issue by utilizing a permutation-based approach that accounts for the non-independence of network data.⁴ LRQAP adapts the principles of linear regression to network data by using a matrix regression approach, where the dependent and independent variables are represented as adjacency matrices. These characteristics make LRQAP particularly well-suited for analyzing criminal networks, as evidenced by previous studies (Bichler and Norris 2024; Campana and Varese 2013; Carrington and Graham 2022; Grund and Densley 2012; Krajewski, DellaPosta and Felmlee 2022; Wippell and Haynie 2025).⁵

We developed two LRQAP models to predict the most common tie type, drug crime, as well as the broader criminal collaboration network, which aggregates drug crimes, illegal transactions, and cooperative activities. We focus primarily on drug crimes as the dependent variable following both theoretical and empirical considerations. In line with the criminological literature on embeddedness, drug trafficking constitutes the most analytically meaningful and stable outcome for assessing the role of social ties in illicit economic activity. It was also the most frequently reported type of collaboration in the trial (1,078 ties), producing a denser and more balanced network for analysis. Other forms of cooperation, such as illegal transactions (388 ties) and “worked together” relations (120 ties), were mentioned less systematically and result in highly sparse networks that are less suitable for separate modelling. To account for potential measurement error and to ensure robustness, we also estimated models using a broader construct of criminal collaboration that aggregates drug crimes, illegal transactions, and cooperative activities. This allows us to test whether the observed patterns hold under a more inclusive definition of collaboration while maintaining analytical coherence.

The primary independent variables comprised the other relationship types that were not part of the criminal collaboration variable. To further account for potential confounding factors, we included variables that captured shared characteristics between actors: same sex, same nationality, same function or role within the cartels, and similar age (within ±3 years). These control variables were incorporated into the model to account for demographic and role-based homophily factors that might influence the likelihood of criminal connections independent of relational ties.

We conducted separate LRQAP analyses for each dependent variable, running 1,000 random permutations for each model to obtain robust significance levels. This permutation-based approach tested the null hypothesis that there is no association between the independent variables (relationship types and control variables) and the dependent variables (drug crime and criminal collaboration). The LRQAP analyses allowed us to determine whether actors who shared specific relationships or demographic characteristics were statistically more (or less) likely to engage in drug crimes or criminal collaborations together. In line with our focus on multiplexity, we estimated models that included multiple tie layers simultaneously, so that each coefficient reflected the contribution of a given relationship conditional on the others.⁶ We first ran bivariate models, reported in the Appendix. Subsequently, we performed a step-by-step analysis in which we incorporated the independent variables in successive runs. In order, we incorporated: (1) distrust, the conflictual relationships (violence and rivalry); (2) trust relationships (compadre, family, friendship, prison, meeting, joint violence); (3) both conflictual and trust relationships; (4) same as the last, whilst controlling for same age, same country, same function and same sex and; (5) same as the last, whilst also controlling for same cartel. We interpreted the findings through each of these steps to identify when the significance and direction of the relationship changed. All analyses were conducted in R (R Core Team 2025) with packages igraph (Csárdi et al. 2024) and sna (Butts 2024).

Results

The analysis of the Jaccard Indexes revealed substantial overlap among specific networks, particularly those involving transactional and collaborative ties. For example, drug crime, illegal transaction and worked together show notable intersections, reflected in higher Jaccard values with one another. This indicates that the actors involved in these types of relationships often share multiple ties across these criminal activities. The overlap between these relations also justifies the creation of the criminal collaboration network that was obtained by merging these relations. Conversely, relationship types like family, compadre, and joint violence display lower Jaccard Index values across most of the pairings, thus suggesting that these networks are relatively distinct and less likely to co-occur with other relationship types. For example, the non-overlap between family and compadre relations was expected as godfathering relations are normally established with individuals outside of blood family. Still, the overall scores in most cases can be considered as modest or low given that they are under 0.30. This suggests that there is little overlap between the different networks that we collected, albeit this is possibly due to the context in which the data was collected.

The results of the LRQAP models with drug crimes as the dependent variable assessed the likelihood of the joint commission of drug crimes based on other social relationships (Table 3). With regards to model fit, the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) indicated that the fourth and fifth models provided the best fit to the data.

Table 3.

Results of LRQAP Models. Dependent Variable = Drug Crime.

Measure	Distrust	Trust	Trust & distrust, no controls	Trust & distrust, controls no cartel	Trust & distrust, all controls
Violence	1.357* (1.108)		0.720 (1.126)	0.581 (1.103)	0.875 (1.100)
Rivalry	1.981** (1.156)		−0.647 (1.062)	−0.842 (1.118)	−0.744 (1.085)
Compadre		1.795* (1.114)	1.648 (1.082)	1.598 (1.034)	1.523 (1.006)
Family		0.968* (1.114)	0.944* (1.109)	0.818 (1.114)	0.591 (1.126)
Friendship		2.181*** (1.060)	2.305*** (1.027)	2.212*** (1.053)	2.097 (1.088)***
Prison		0.045 (1.081)	−0.062 (1.098)	0.017 (1.041)	−0.195 (0.975)
Meeting		3.285*** (1.199)	3.304 *** (1.156)	3.160*** (1.173)	2.650 (1.124)***
Joint violence		−0.122 (1.063)	−0.089 (1.085)	−0.273 (1.076)	−0.264 (1.090)
Same age				0.570 (1.125)	0.628 (1.068)
Same country				0.339 (1.762)	−0.041 (1.663)
Same function				0.123 (1.671)	0.026 (1.606)
Same sex				1.735* (2.143)	1.575 (2.054)
Same cartel					1.663 (1.785)***
Intercept	−4.679 (0.140)	−5.181 (0.418)	−5.186 (0.393)	−6.911 (2.320)	−7.150 (2.343)
AIC	1895	1523	1525	1497	1422
BIC	1918	1578	1595	1599	1531
Degrees of freedom	17578	17578	17578	17578	17578
Null deviance	24368	24368	24368	24368	24368
Observations	17578	17578	17578	17578	17578
Residual deviance	1889	1509	1507	1471	1421
Replications	1000	1000	1000	1000	1000

Note: Reported values are unstandardized log-odds coefficients from LRQAP logistic regressions, with standard errors from the netlogit model shown in parentheses. Standard errors are model-derived and included to describe variability. However, due to the non-independence of network data, they should not be interpreted as in standard regression models based on independent observations.

Significance is based on permutation-derived p-values: *p < .05; **p < .01; ***p < .001.

The first model showed that conflictual relationships (violence and rivalry) increased the likelihood of a drug crime tie. However, these associations disappeared in more complex models. The second model reported a positive and statistically significant association between drug crimes and compadre, family, friendship, and meeting ties, with the last two reporting the greatest effect. For instance, individuals connected through friendship ties were approximately nine times more likely (coefficient = 2.181, odds ratios, or OR = 8.85) to engage in drug crimes, whilst those with meeting ties were around 27 times more likely (coefficient = 3.285, OR = 26.76). Family ties and compadre ties also increased the likelihood of drug crimes, making individuals about 2.6 times (coefficient = 0.968, OR = 2.63) and six times (coefficient = 1.795, OR = 6.02) more likely, respectively, compared to those without such ties. However, in models 3 to 5, compadre and family relationship progressively lost their statistical significance whilst controlling for same age, country, function, sex, and cartel.

In the last and most complete model, friendship, meeting, same age, and same cartel ties best predicted drug crime ties, with all other variables not having a significant relationship with the drug crime network.

Prison ties and joint violence were not statistically significant across all the models.

The models with criminal collaboration as the dependent variable (Table 4) confirmed the results presented for drug crimes. Overall, the goodness-of-fit of these models were high, albeit slightly lower than those reported in Table 3.

Table 4.

Results of LRQAP Models. Dependent Variable = Criminal Collaboration.

Measure	Distrust	Trust	Trust and distrust, no controls	Trust and distrust, controls no cartel	Trust and distrust, all controls
Violence	0.988*(1.078)		0.446(1.064)	0.336(1.033)	0.615(1.082)
Rivalry	2.874***(1.143)		0.519(1.064)	0.255(1.104)	0.539(1.042)
Compadre		1.035(1.140)	1.070(1.049)	1.001(1.062)	0.820(1.041)
Family		1.075**(1.081)	1.146**(1.137)	1.064*(1.082)	0.818(1.097)
Friendship		3.154***(1.062)	3.113***(1.056)	3.002***(1.015)	2.884***(1.085)
Prison		0.361(1.123)	0.443(1.064)	0.472(1.016)	0.321 (1.006)
Meeting		3.391***(1.217)	3.299***(1.139)	3.180***(1.125)	2.709*** (1.133)
Joint violence		−0.393(1.093)	−0.402(1.126)	−0.618(1.107)	−0.573(1.035)
Same age				0.440(1.139)	0.487(1.092)
Same country				0.360(1.748)	−0.018(1.725)
Same function				0.038(1.741)	−0.077(1.639)
Same sex				1.676*(2.117)	1.518*(2.026)
Same cartel					1.628***(1.764)
Intercept	−4.326(0.113)	−4.806(0.331)	−4.811(0.310)	−6.457* (2.373)	−6.673 (2.454)
AIC	2530	2019	2019	1982	1878
BIC	2553	2073	2089	2083	1987
Degrees of freedom	17578	17578	17578	17578	17578
null deviance	24368	24368	24368	24368	24368
Observations	17578	17578	17578	17578	17578
Residual deviance	2524	2005	2001	1956	1850
Replications	1000	1000	1000	1000	1000

Significance is based on permutation-derived p-values: *p < .05; **p < .01; ***p < .001.

As in the previous models, distrust relationships became insignificant as the model complexity increased. Moreover, family, friendship, and meeting ties were consistently positively associated with the likelihood of a criminal collaboration tie. Friendship ties reported greater effect than in the models on drug crime alone. However, family ties lost their statistical significance in the most complex model, which included the same cartel control variable. Furthermore, prison and joint violence ties lacked significance in all the models. Contrary to the previous models, compadre relationships were never statistically significant across all the models. The final model indicated that friendship ties greatly increased the likelihood of criminal collaboration, making individuals approximately 18 times more likely to engage in such activities (coefficient = 2.913, OR = 18.4). Similarly, meeting ties were associated with a 19-fold increase in the likelihood of criminal collaboration (coefficient = 2.954, OR = 19.2). It also reported positive and statistically significant effects for same sex and same cartel.

Discussion and Conclusions

Our findings suggest that social relationships significantly increase the likelihood of both drug crime ties and broader criminal collaboration within the Sinaloa Cartel network, as reconstructed from El Chapo's trial transcripts. By modelling multiple relationships simultaneously, our analyses capture the multiplexity of actors’ connections and highlight which ties retain independent associations with collaboration. Yet not all ties exhibit the same impact t (Figure 3). Specifically, friendship and meeting relationships exhibited the strongest associations with drug-related ties, while compadre or family bonds diminished in significance once other relational and organizational factors were considered, suggesting that their effects are often absorbed by overlapping ties such as friendship or shared cartel membership. Conflictual relationships, including direct violence, similarly showed no impact on collaboration. These outcomes underscore that trust is vital in organized crime, yet different social ties may produce distinct pathways to cooperation.

Figure 3.

Coefficient estimates by model and dependent variable.

These results align with longstanding views on the pivotal role of trust in criminal organizations (Calderoni et al. 2020; Calderoni, Campedelli et al. 2022a; Decker and Van Winkle 1996; Ianni and Reuss-Ianni 1972; Kleemans and Van De Bunt 1999; Krebs 2002; Morselli 2003). Yet, they also highlight how certain social ties may be more effective than others in fostering collaboration within large-scale drug trafficking networks.

Friendship and meetings significantly impact the likelihood of drug crimes within the Sinaloa Cartel, aligning with studies showing economic activity rooted in social practices and relations, including organized crime (Joseph and Smith 2021). Social relations are instrumental in maintaining trust and criminal collaborations within hostile environments like those organized crime groups operate within (Campana and Varese 2013; Gambetta 1993). Meetings are often necessary and unavoidable for coordinating activities under heavy law enforcement surveillance. In large criminal organizations, social gatherings and recurring face-to-face interactions help preserve trust (Calderoni and Superchi 2019).

Family ties, although theoretically expected to bolster cooperation, showed weaker effects. This may be due to the mediating role of cartel affiliation: many males from the same blood family were affiliated with the same cartel (see Figure A1 in the Appendix), suggesting that the influence of family ties might be indirectly captured through shared cartel membership.⁷ Qualitative evidence from the trial is abundant. El Chapo's brothers and sons occupied key leadership and operational roles; El Mayo Zambada's sons followed similar trajectories; and the Beltrán Leyva brothers, the Cifuentes family, and the Flores twins all exemplify how collaboration often follows familial or quasi-familial lines. As Paoli and colleagues (2023:168) note, the Sinaloa Cartel “should be understood as a loose, not formalized alliance of largely autonomous drug-selling enterprises, which are each run by the males of a blood family.” These dynamics help explain why family and compadre ties lose statistical significance once shared cartel membership is controlled for: they often signal the same organizational embeddedness. Lastly, we emphasize that the family network included a higher proportion of women (23%) than other networks (ranging between 3% and 5%). This disparity may reflect gendered patterns in the Sinaloa Cartel with women operating in critical yet less visible roles, such as financial and logistical tasks, resulting in fewer overt acts liable to prosecution and thus less ties in the drug crime network (Allum 2024; Bonello 2024; Selmini 2020).

We found weaker evidence for the roles of compadre relationships, prison, and joint violences. We suggest that the effects of these variables may have been absorbed by more general and frequently reported relations, such as meetings, friendship, and family. In particular, in the Sinaloa Cartel godparents were often chosen from friends or affiliates in the same organization. Evidence from Joaquín Guzmán Loera's trial shows compadre ties are formed between individuals who already work together (Esquivel, 2019).⁸ Our findings may also reflect the sequencing of relationships: Sinaloa Cartel members often become compadres after engaging in joint criminal activities. Consequently, the positive effect of compadre relations on drug crimes in bivariate models was absorbed in multivariate models. Furthermore, we suggest that the limited impact of prison and joint violence ties may reflect the nature and scope of our data—these ties are likely underreported or underdocumented in our sources—so we urge caution in interpreting their marginal role in the models as evidence of substantive irrelevance.

With regards to the effect of co-participation in violence and kinship relations on criminal collaboration, we note that Campana and Varese found that sharing information about violence had a greater impact on the likelihood of criminal collaboration than kinship ties (Campana and Varese 2013:281–82). However, there were differences between their measure of sharing information and our measure of participating in violent actions jointly with someone. These differences testify to the complexity of interpreting the role of different social relations in fostering criminal collaborations and underscore the need for both additional studies and more granular measurements.

Our study highlights the untapped potential of trial transcripts from common law systems as rich data sources for analyzing criminal networks. Unlike civil law systems, where detailed judicial records are readily available, common-law systems rely on oral proceedings, limiting written documentation. However, extended jury trial transcripts provide detailed insights into social ties, surpassing the brevity of indictments, sentencing comments, or judge decisions used in prior studies (Bichler et al. 2019; Bichler, Norris and Ibarra 2020; Bright et al. 2012; Hughes, Bright and Chalmers 2017; Jones et al. 2018).

Our study has limitations. First, the manual data extraction process carries risks of oversight and interpretive bias, often requiring knowledge of Mexican drug trafficking and the Sinaloa Cartel. To enhance accuracy, we supplemented this process with academic and journalistic sources in Spanish and English, triangulating information on key details like birth dates, nationalities, cartel affiliations, and family ties. Multiple coders and researchers systematically checked and validated the data to ensure consistency and reliability.

Second, the adversarial nature of U.S. trials may raise concerns about the validity and reliability of the information, as most witnesses were called by the prosecution, some under cooperation agreements. However, key relationships, such as family and friendship ties, are difficult to fabricate or conceal, as they are often verifiable through multiple sources. Moreover, central issues like drug crimes and patterns of criminal collaboration were subject to rigorous scrutiny during the trial, with testimony delivered under oath, subject to cross-examination, and exposed to legal consequences for false statements. The trial's high-profile status also ensured extensive media coverage by experienced journalists, providing an additional layer of public oversight. While the possibility of errors or intentional misrepresentations cannot be entirely excluded, the consistency of the testimonies and external verification through media and legal records support the data's reliability, making it comparable to that of qualitative interviews.

Third, our data do not provide a comprehensive representation of the Sinaloa Cartel, which is impossible given its vast size and complexity. With tens of thousands of members operating across multiple countries and diverse roles—from low-level operatives to top leaders—capturing the full extent of the cartel's network is unfeasible (Paoli et al. 2023). Even well-informed criminals or law enforcement officials cannot fully account for all relationships within such an expansive structure. Moreover, mapping the Sinaloa Cartel's network would require detailed insights into its “upperworld” connections, involving individuals in legal spheres—such as bankers, lawyers, politicians, and law enforcement—who facilitate illicit activities without formal ties to the organization. Systemic deficiencies in Mexico's justice system further obscure these connections, complicating efforts to fully analyse the cartel's relational structure (Pereda 2023).

While the trial provided detailed insights, the evidence and testimonies primarily focused on “El Chapo,” highlighting his role, activities, and interactions within the Sinaloa Cartel. This defendant-centric perspective may underrepresent other key figures, regional dynamics, and the broader organizational structure. Additionally, the emphasis on Guzmán's leadership, struggles, escapes, and capture may skew the portrayal of the cartel's activities and networks toward those directly connected to him.

However, we argue that our data provides a reliable basis for examining the social structure, embeddedness, and multiplexity of a significant part of one of the largest drug trafficking organizations. The main challenge lies in the potential for false negatives—unreported relationships—rather than false positives, as the relationships documented are unlikely to have been fabricated. We believe that missing ties do not significantly affect our findings, as the trial offered extensive, often corroborated, testimony focused on Guzmán's central role, ensuring the documentation of many critical connections. Additionally, the presence of multiplex ties (e.g., family, friendship, criminal collaboration) strengthens the robustness of our analysis. Our findings align with existing research on criminal networks, highlighting well-documented dynamics such as the role of social ties in fostering trust and collaboration. Thus, even if some ties were overlooked, their absence is unlikely to alter the broader patterns observed.

In conclusion, our analyses provide four contributions. First, criminologists have expanded the concept of embeddedness to include the overlap of various social relationships within organized crime. Our findings demonstrate that multiplexity matters: by measuring both the overlap among different ties and their conditional contributions when included simultaneously in the models, we show that multiple social ties within El Chapo's network significantly influenced the likelihood of engaging in illegal economic activities. This aligns closely with the original notion of embeddedness (Joseph and Smith 2021). Second, our network-based approach to analyzing the Sinaloa Cartel offers a valuable complement to existing research focused on the effects of anti-crime strategies on Mexico's illegal markets. Various scholars have examined the outcomes of targeted enforcement measures, particularly the “kingpin strategy,” which aims to dismantle criminal organizations by capturing or killing their top leaders. A growing body of evidence suggests that this approach has been counterproductive in Mexico, often correlating with increased levels of violence (Calderón et al. 2015; Estévez-Soto and Lecona Esteban 2024; Jones 2013).

Our network perspective contributes to a broader understanding of why such strategies have failed. First, it shows that formal hierarchies and leadership titles within Mexican organized crime groups are not necessarily the glue that holds them together. Consequently, removing high-ranking individuals does not automatically sever the underlying social connections that sustain group cohesion. Second, in groups where members are linked through multiple, overlapping relationships—such as kinship, friendship, business ties, or shared rituals—targeting a single type of tie (e.g., business links) may have limited impact. Effective disruption, therefore, requires a nuanced understanding of which relational structures most strongly support cooperation within the group.

In the case of the Sinaloa Cartel, our findings indicate that interpersonal relationships—particularly friendships—and the physical spaces where these ties are cultivated and sustained play a key role in maintaining the group's internal cohesion. Identifying and monitoring these friendships and in-situ meeting points may offer a more effective strategy for disrupting the criminal networks that have emerged from the Sinaloa Cartel's fragmentation.

Third, much of the existing research on embeddedness in organized crime has relied heavily on qualitative data, which limits the ability to disentangle the effects of different relationship types and quantify their relative influence at the individual level. Our results show that specific social ties are differentially associated with the probability of criminal collaboration, which underscores that illegal economic activity is not embedded uniformly across all social relations but rather is strongly linked to certain key connections. Fourth, we have demonstrated that the transcripts from long-lasting and complex jury trials in common law jurisdictions can also serve as a valuable and underutilized source for network data collection, offering rich insights into the dynamics of organized crime.

Footnotes

Acknowledgements

All authors contributed to the overall design and writing of the article. FC and DDH conducted the analyses, while FC and VP wrote the introduction and background. FC and DDH prepared the methods section, and FC wrote the results and discussion. All authors reviewed and approved the final version of the manuscript. The authors gratefully acknowledges the contributions of Valeria Gerosa and Claire Bernandon to the data collection phase.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Francesco Calderoni

Valentin Pereda

David Décary-Hétu

Notes

Author Biographies

Francesco Calderoni is a Professor of Criminology. His areas of research are mafias, organized crime networks and illegal markets.

Valentin Pereda is an Assistant Professor of Criminology. His area of research is organized crime.

David Décary-Hétu is a Full Professor of Criminology. His area of research is the impact of technology on crime and the regulation of crime.

Appendix

Table A2.

Results of Bivariate LRQAP Models. Dependent Variable Criminal Collaboration.

Measure	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
Compadre (1)	4.803*** (0.791)
Violence (2)		2.111** (0.880)
Family (3)			2.156*** (1.019)
Friendship (4)				4.649*** (0.870)
Prison (5)					4.165*** (0.837)
Meeting (6)						3.893*** (1.504)
Rivalry (7)							3.145*** (0.902)
Joint violence (8)								3.363*** (0.831)
Same Cartel (9)									2.253*** (2.418)
Intercept	−4.292**(0.037)	−4.283*(0.099)	−4.314(0.166)	−4.426(0.088)	−4.283 **(0.038)	−4.756(0.322)	−4.319(0.087)	−4.307(0.073)	−5.109(1.002)
AIC	2536	2580	2560	2351	2556	2107	2532	2539	2430
BIC	2552	2595	2576	2367	2571	2122	2548	2555	2446
Degrees of freedom	17578	17578	17578	17578	17578	17578	17578	17578	17766
Null deviance	24368	24368	24368	24368	24368	24368	24368	24368	24628
Observations	17578	17578	17578	17578	17578	17578	17578	17578	17766
Residual deviance	2532	2576	2556	2347	2552	2103	2528	2535	2426
Replications	1000	1000	1000	1000	1000	1000	1000	1000	1000

Significance is based on permutation-derived p-values: *p < .05; **p < .01; ***p < .001.

References

Aguirre

Jerjes

Gomez

Mario

. 2020. “Competitive Strategies in Contexts of Organized Crime: The Case of the Avocado Industry in Mexico.” International Journal of Criminal Justice Sciences 15(1):114–26. doi:10.5281/zenodo.3822110. https://ijcjs.com/menu-script/index.php/ijcjs/article/view/290 .

Albini

Joseph L.

1971. The American Mafia Genesis of a Legend. Sociology Series. New York, NY: Appleton-Century-Crofts.

Allum

Felia.

2024. Women of the Mafia: Power and Influence in the Neapolitan Camorra. 1st ed. Ithaca, NY: Cornell University Press.

Astorga

Luis.

2016. El siglo de las drogas: Del Porfiriato al nuevo milenio. Mexico City: Penguin Random House Grupo Editorial.

Atuesta

Laura H.

2024. “The Connection Between Drug-Related Detentions and Drug-Related Violence: Another Failure of the War on Drugs in Mexico.” Deviant Behavior 46(7):821–37. doi:10.1080/01639625.2024.2372421

Atuesta

Laura H.

Samantha Pérez-Dávila

Yocelyn

. 2018. “Fragmentation and Cooperation: The Evolution of Organized Crime in Mexico.” Trends in Organized Crime 21(3):235–61. doi:10.1007/s12117-017-9301-z

Aziani

Alberto

Berlusconi

Giulia

Giommoni

Luca

. 2021. “A Quantitative Application of Enterprise and Social Embeddedness Theories to the Transnational Trafficking of Cocaine in Europe.” Deviant behavior 42(2):245–67. doi:10.1080/01639625.2019.1666606

Beith

Malcolm.

2020. The Last Narco: Updated and Revised. New York, NY: Grove.

Beittel

June S

. 2022. Mexico: Organized Crime and Drug Trafficking Organizations. Washington, DC: Congressional Research Service. https://www.congress.gov/crs-product/R41576 .

10.

Bichler

Gisela

Norris

Alexis

. 2024. “Crossing Lines: Structural Advantages of Inter-Racial Criminal Street Gang Violence.” Trends in Organized Crime 27(3):235–62. doi:10.1007/s12117-022-09468-9

11.

Bichler

Gisela

Norris

Alexis

Dmello

Jared R.

Randle

Jasmin

. 2019. “The Impact of Civil Gang Injunctions on Networked Violence Between the Bloods and the Crips.” Crime & Delinquency 65(7):875–915. doi:10.1177/0011128717739607

12.

Bichler

Gisela

Norris

Alexis

Ibarra

Citlalik

. 2020. “Evolving Patterns of Aggression: Investigating the Structure of Gang Violence During the Era of Civil Gang Injunctions.” Social Sciences 9(11):203. doi:10.3390/socsci9110203

13.

Blok

Anton.

1974. The Mafia of a Sicilian Village 1860-1960: A Study of Violent Peasant Entrepreneurs. New York, NY: Harper & Row.

14.

Bonello

Deborah.

2024. Narcas: The Secret Rise of Women in Latin America’s Cartels. Boston, MA: Beacon Press.

15.

Bouchard

Martin.

2020. “Collaboration and Boundaries in Organized Crime: A Network Perspective.” Crime and Justice 49:425–69. doi:10.1086/708435

16.

Bright

David

Greenhill

Catherine

Ritter

Alison

Morselli

Carlo

. 2015. “Networks Within Networks: Using Multiple Link Types to Examine Network Structure and Identify Key Actors in a Drug Trafficking Operation.” Global Crime 16(3):219–37. doi:10.1080/17440572.2015.1039164

17.

Bright

David A.

Hughes

Caitlin E.

Jenny

Chalmers.

. 2012. “Illuminating Dark Networks: A Social Network Analysis of an Australian Drug Trafficking Syndicate.” Crime, Law and Social Change 57(2):151–76. doi:10.1007/s10611-011-9336-z

18.

Browne

Deborah

Mason

Mark

Murphy

Rachel

. 2003. “Drug Supply and Trafficking: An Overview.” The Howard Journal of Criminal Justice 42(4):324–34. doi:10.1111/1468-2311.00289

19.

Bunker

Robert J.

2010. “Strategic Threat: Narcos and Narcotics Overview.” Small Wars & Insurgencies 21:8–29. doi:10.1080/09592311003589229

20.

Butts

Carter T.

2024. “Sna: Tools for Social Network Analysis.” Retrieved from https://cran.r-project.org/web/packages/sna/

21.

Calderoni

Francesco.

2012. “The Structure of Drug Trafficking Mafias: The ‘Ndrangheta and Cocaine.” Crime, Law and Social Change 58(3):321–49. doi:https://doi.org/10.1007/s10611-012-9387-9

22.

Calderoni

Francesco

Campedelli

Gian Maria

Comunale

Tommaso

Marchesi

Martina E.

Savona

Ernesto U.

. 2020. “Recruitment into Organised Criminal Groups: A Systematic Review.” Trends and Issues in Crime and Criminal Justice (583):1–28.

23.

Calderoni

Francesco

Campedelli

Gian Maria

Szekely

Aron

Paolucci

Mario

Andrighetto

Giulia

. 2022a. “Recruitment into Organized Crime: An Agent-Based Approach Testing the Impact of Different Policies.” Journal of Quantitative Criminology 38(1):197–237. doi:10.1007/s10940-020-09489-z

24.

Calderoni

Francesco

Comunale

Tommaso

Campedelli

Gian Maria

Marchesi

Martina

Manzi

Deborah

Frualdo

Niccolò

. 2022b. “Organized Crime Groups: A Systematic Review of Individual-Level Risk Factors Related to Recruitment.” Campbell Systematic Reviews 18(1):e1218. doi:10.1002/cl2.1218

25.

Calderoni

Francesco

Superchi

Elisa

. 2019. “The Nature of Organized Crime Leadership: Criminal Leaders in Meeting and Wiretap Networks.” Crime, Law and Social Change 72(4):419–44. doi:10/gt47vs

26.

Calderón

Gabriela

Robles

Gustavo

Díaz-Cayeros

Alberto

Magaloni

Beatriz

. 2015. “The Beheading of Criminal Organizations and the Dynamics of Violence in Mexico.” Journal of Conflict Resolution 59(8):1455–85. doi:10.1177/0022002715587053

27.

Campana

Paolo

Varese

Federico

. 2013. “Cooperation in Criminal Organizations: Kinship and Violence as Credible Commitments.” Rationality and Society 25(3):263–89. doi:10.1177/1043463113481202

28.

Carrington

Peter J.

Graham

Alexander V.

. 2022. “The Interurban Network of Criminal Collaboration in Canada.” Canadian Journal of Criminology and Criminal Justice 64(2):101–23. doi:10.3138/cjccj.2022-0004

29.

Coleman

James S.

1988. “Social Capital in the Creation of Human Capital.” American Journal of Sociology 94:S95–S120.

30.

Correa-Cabrera

Guadalupe

. 2017. Los Zetas Inc: Criminal Corporations, Energy, and Civil War in Mexico. Austin, TX: University of Texas Press.

31.

Csárdi

Gábor

Nepusz

Tamás

Traag

Vincent

Horvát

Szabolcs

Zanini

Fabio

Noom

Daniel

Müller

Kirill

Salmon

Maëlle

Antonov

Michael

Chan Zuckerberg and Initiative igraph author details . 2024. “Igraph: Network Analysis and Visualization.” Retrieved from https://cran.r-project.org/web/packages/igraph/index.html

32.

DEA. 2024. National Drug Threat Assessment. Washington: Drug Enforcement Administration. https://www.dea.gov/documents/2024/2024-05/2024-05-24/national-drug-threat-assessment-2024 .

33.

Decker

Scott H.

Pyrooz

David C.

Sweeten

Gary

Moule

Richard K.

. 2014. “Validating Self-Nomination in Gang Research: Assessing Differences in Gang Embeddedness Across Non-, Current, and Former Gang Members.” Journal of Quantitative Criminology 30(4):577–98. doi:10.1007/s10940-014-9215-8

34.

Decker

Scott H.

Van Winkle

Barrik

. 1996. Life in the Gang: Family, Friends, and Violence. Cambridge Studies in Criminology. Cambridge: Cambridge University Press.

35.

Dekker

David

Krackhardt

David

Snijders

Tom A. B.

. 2007. “Sensitivity of MRQAP Tests to Collinearity and Autocorrelation Conditions.” Psychometrika 72(4):563–81. doi:10.1007/s11336-007-9016-1

36.

DellaPosta

Daniel.

2023. “Bridging the Parochial Divide: Outsider Brokerage in Mafia Families.” Social Science Research 114:102913. doi:10.1016/j.ssresearch.2023.102913

37.

Desroches

Frederick J.

2005. The Crime That Pays: Drug Trafficking and Organized Crime in Canada. Toronto: Canadian Scholars’ Press.

38.

Diviák

Tomáš

Dijkstra

Jan Kornelis

Snijders

Tom A. B.

. 2019. “Structure, Multiplexity, and Centrality in a Corruption Network: The Czech Rath Affair.” Trends in Organized Crime 22(3):274–97. doi:10.1007/s12117-018-9334-y

39.

Dorn

Nicholas

Oette

Lutz

White

Simone

. 1998. “Drugs Importation and the Bifurcation of Risk.” British Journal of Criminology 38(4):537–60.

40.

Durán-Martínez

Angélica

. 2015. “To Kill and Tell? State Power, Criminal Competition, and Drug Violence.” Journal of Conflict Resolution 59(8):1377–402. doi:10.1177/0022002715587047

41.

Durán-Martínez

Angélica

. 2018. The Politics of Drug Violence: Criminals, Cops and Politicians in Colombia and Mexico. New York, NY: Oxford University Press.

42.

Enciso Higuera

Froylan Vladimir

. 2015. “The Origin of Contemporary Drug Contraband.” Ph.D. thesis, Stony Brook University.

43.

Espinosa

Valeria

Rubin

Donald B.

. 2015. “Did the Military Interventions in the Mexican Drug War Increase Violence?” The American Statistician 69(1):17–27. doi:10.1080/00031305.2014.965796

44.

Esquivel

J. Jesús

. 2019. El juicio: crónica de la caída del Chapo. Mexico City: Grijalbo.

45.

Estévez-Soto

Patricio

Lecona Esteban

Reynaldo

. 2024. “The Impact of the Kingpin Strategy on Extortion and Kidnapping.” Criminology & Criminal Justice 25(1):200–27. doi:10.1177/17488958241276178

46.

Felbab-Brown

Vanda.

2022. “How the Sinaloa Cartel Rules.” https://www.brookings.edu/articles/how-the-sinaloa-cartel-rules/.

47.

Gambetta

Diego.

1993. The Sicilian Mafia: The Business of Private Protection. Cambridge, MA: Harvard University Press.

48.

Gambetta

Diego.

2009. Codes of the Underworld: How Criminals Communicate. Princeton, NJ: Princeton University Press.

49.

Garriga

Ana Carolina

Phillips

Brian J.

. 2023. “Organized Crime and Foreign Direct Investment: Evidence From Criminal Groups in Mexico.” Journal of Conflict Resolution 67(9):1675–703. doi:10.1177/00220027221145870

50.

Gill-Hopple

Kathy

Brage-Hudson

Diane

. 2012. “Compadrazgo: A Literature Review.” Journal of Transcultural Nursing 23(2):117–23. doi:10.1177/1043659611433870

51.

Granovetter

Mark S.

1985. “Economic Action and Social Structure: The Problem of Embeddedness.” American Journal of Sociology 91(3):481–510. doi:10.1086/228311

52.

Granovetter

M. S

. 1992. “Problems of Explanation in Economic Sociology.” Pp. 25–56 in Networks and Organizations: Structure, Form and Action edited by N. Nohria and R. G. Eccles. Boston, MA: Harvard Business School Pr.

53.

Grillo

Ioan.

2011. El Narco - the Bloody Rise of Mexican Drug Cartels. London: Bloomsbury.

54.

Grund

Thomas U.

Densley

James A.

. 2012. “Ethnic Heterogeneity in the Activity and Structure of a Black Street Gang.” European Journal of Criminology 9(4):388–406. doi:10.1177/1477370812447738

55.

Gundur

Rajeev Vishwanath

. 2022. Trying to Make It: The Enterprises, Gangs, and People of the American Drug Trade. Ithaca, NY: Cornell University Press.

56.

Gutiérrez-Romero

Roxana

Iturbe

Nayely

. 2024. “Causes and Electoral Consequences of Political Assassinations: The Role of Organized Crime in Mexico.” Political Geography 115:103206. doi:10.1016/j.polgeo.2024.103206

57.

Hagan

John.

1993. “The Social Embeddedness of Crime and Unemployment.” Criminology; An Interdisciplinary Journal 31(4):465–91. doi:10.1111/j.1745-9125.1993.tb01138.x

58.

Hernández

Anabel.

2019. El Traidor. Un Diario Secreto Del Hijo Del Mayo. Mexico City: Grijalbo.

59.

Hess

Henner.

[1970] 1973. Mafia and Mafiosi: The Structure of Power. Lexington, KY: Lexington Books.

60.

Hosford

Kevin

Aqil

Nauman

Windle

James

Gundur

R. V.

Allum

Felia

. 2021. “Who Researches Organised Crime? A Review of Organised Crime Authorship Trends (2004–2019).” Trends in Organized Crime. 25(3):249–71. doi:10.1007/s12117-021-09437-8

61.

Hughes

Caitlin E.

David

Bright

Jenny Chalmers.

. 2017. “Social Network Analysis of Australian Poly-Drug Trafficking Networks: How Do Drug Traffickers Manage Multiple Illicit Drugs?” Social Networks 51:135–47. doi:10.1016/j.socnet.2016.11.004

62.

Ianni

Francis A. J.

Reuss-Ianni

Elizabeth

. 1972. A Family Business: Kinship and Social Control in Organized Crime. New York, NY: Russell Sage Foundation.

63.

Ibarra Chaoul

Alejandra.

2019. El Chapo Guzmán: El juicio del siglo. Miami, FL: Aguilar.

64.

Jones

Nathan.

2013. “The Unintended Consequences of Kingpin Strategies: Kidnap Rates and the Arellano-Félix Organization.” Trends in Organized Crime 16(2):156–76. doi:10.1007/s12117-012-9185-x

65.

Jones

Nathan P.

Layne Dittmann

Jun

Reese

Tyler

. 2018. “A Mixed Methods Social Network Analysis of a Cross-Border Drug Network: The Fernando Sanchez Organization (FSO).” Trends in Organized Crime 23(2):154–82. doi:10.1007/s12117-018-9352-9

66.

Joseph

Jared

Smith

Chris M.

. 2021. “The Ties That Bribe: Corruption’s Embeddedness in Chicago Organized Crime.” Criminology; An Interdisciplinary Journal 59(4):671–703. doi:10.1111/1745-9125.12287

67.

Kleemans

Edward R.

de Poot

Christianne J.

. 2008. “Criminal Careers in Organized Crime and Social Opportunity Structure.” European Journal of Criminology 5(1):69–98.

68.

Kleemans

Edward R.

Van De Bunt

Henk

. 1999. “The Social Embeddedness of Organized Crime.” Transnational Organized Crime 5(1):19–36.

69.

Kleemans

Edward R.

van de Bunt

Henk

. 2008. “Organised Crime, Occupations and Opportunity.” Global Crime 9(3):185–97. doi:10.1080/17440570802254254

70.

Krajewski

Andrew T.

DellaPosta

Daniel

Felmlee

Diane

. 2022. “Vertical Organizations, Flat Networks: Centrality and Criminal Collaboration in the Italian-American Mafia.” Social Networks 68:127–38. doi:10.1016/j.socnet.2021.06.001

71.

Krebs

Valdis E.

2002. “Mapping Networks of Terrorist Cells.” Connections 24(3):43–52.

72.

Krippner

Greta

Granovetter

Mark

Block

Fred

Biggart

Nicole

Beamish

Tom

Hsing

Youtien

Hart

Gillian

Arrighi

Giovanni

Mendell

Margie

Hall

John

Burawoy

Michael

Vogel

Steve

O’Riain

Sean

. 2004. “Polanyi Symposium: A Conversation on Embeddedness.” Socio-Economic Review 2(1):109–35. doi:10.1093/soceco/2.1.109

73.

Le Cour Grandmaison

Romain

Morris

Nathaniel

Smith

Benjamin

. 2019. “The Last Harvest? From the US Fentanyl Boom to the Mexican Opium Crisis.” Journal of Illicit Economies and Development 1(3):312–29. doi:10.31389/jied.45

74.

Lessing

Benjamin.

2015. “Logics of Violence in Criminal War.” Journal of Conflict Resolution 59(8):1486–516. doi:10.1177/0022002715587100

75.

Ley

Sandra.

2018. “To Vote or Not to Vote: How Criminal Violence Shapes Electoral Participation.” Journal of Conflict Resolution 62(9):1963–90. doi:10.1177/0022002717708600

76.

Ley

Sandra

Mattiace

Shannan

Trejo

Guillermo

. 2019. “Indigenous Resistance to Criminal Governance: Why Regional Ethnic Autonomy Institutions Protect Communities from Narco Rule in Mexico.” Latin American Research Review 54(1):181–200. doi:10.25222/larr.377

77.

Magaloni

Gosztonyi

Thompson

. 2022. “State-Evading Solutions to Violence: Organized Crime and Governance in Indigenous Mexico.” (No. WPS-191). Berkeley, CA: Center for Effective Global Action. University of California. https://doi.org/10.5072/FK2KW5H86X

78.

Malm

Aili E.

Bichler

Gisela

Van De Walle

Stephanie

. 2010. “Comparing the Ties That Bind Criminal Networks: Is Blood Thicker Than Water?” Security Journal 23(1):52–74.

79.

Morselli

Carlo.

2001. “Structuring Mr. Nice: Entrepreneurial Opportunities and Brokerage Positioning in the Cannabis Trade.” Crime, Law and Social Change 35(3):203–44.

80.

Morselli

Carlo.

2003. “Career Opportunities and Network-Based Privileges in the Cosa Nostra.” Crime, Law and Social Change 39(4):383–418.

81.

Morselli

Carlo.

2005. Contacts, Opportunities, and Criminal Enterprise. Toronto: University of Toronto Press.

82.

Morselli

Carlo.

2009. Inside Criminal Networks. New York, NY: Springer.

83.

Morselli

Carlo

Petit

Katia

. 2007. “Law-Enforcement Disruption of a Drug Importation Network.” Global Crime 8(2):109–30.

84.

Osorno

Diego Enrique.

2012. El cártel de Sinaloa: una historia del uso político del narco. Mexico City: Grijalbo.

85.

Ouellet

Marie

Bouchard

Martin

Charette

Yanick

. 2019. “One Gang Dies, Another Gains? The Network Dynamics of Criminal Group Persistence.” Criminology; An Interdisciplinary Journal 57(1):5–33. doi:10.1111/1745-9125.12194

86.

Pansters

Wil G.

2018. “Drug Trafficking, the Informal Order, and Caciques . Reflections on the Crime-Governance Nexus in Mexico.” Global Crime 19(3-4):315–38. doi:10.1080/17440572.2018.1471993

87.

Paoli

Letizia.

2002. “The Paradoxes of Organized Crime.” Crime, Law and Social Change 37:51–97.

88.

Paoli

Letizia

Peters

Bryan

Reuter

Peter

. 2023. “Is the Sinaloa Cartel a Mafia?” The British Journal of Criminology 64(1):157–74. doi:10.1093/bjc/azad004

89.

Papachristos

Andrew V,

Smith

Chris M.

. 2014 “The Embedded and Multiplex Nature of Al Capone.” Pp. 97–115 in Crime and Networks, edited by Morselli

. New York, NY: Routledge.

90.

Pereda

Valentin.

2022. “Why Global North Criminology Fails to Explain Organized Crime in Mexico.” Theoretical Criminology 26(4):136248062211045. doi:10.1177/13624806221104562

91.

Pereda

Valentin.

2023. “Working for Both Sides: How Organised Crime Networks Influence Mexican Police Culture.” Policing and Society 33(7):749–66. doi:10.1080/10439463.2023.2195649

92.

Pereda

Valentin

Décary-Hetu

David

. 2023. “Illegal Market Governance and Organized Crime Groups’ Resilience: A Study of The Sinaloa Cartel.” The British Journal of Criminology 64(2):326–42. doi:10.1093/bjc/azad027

93.

Perez Esparza

David

Johnson

Shane D.

Gill

Paul

. 2020. “Why Did Mexico Become a Violent Country?” Security Journal 33(2):179–209. doi:10.1057/s41284-019-00178-6

94.

Polanyi

Karl.

[1944] 2001. The Great Transformation: The Political and Economic Origins of Our Time. 2nd ed. Boston, MA: Beacon Press.

95.

Prieto-Curiel

Rafael

Campedelli

Gian Maria

Hope

Alejandro

. 2023. “Reducing Cartel Recruitment Is the Only Way to Lower Violence in Mexico.” Science 381(6664):1312–16. doi:10.1126/science.adh2888

96.

Pyrooz

David C.

Sweeten

Gary

Piquero

Alex R.

. 2013. “Continuity and Change in Gang Membership and Gang Embeddedness.” Journal of Research in Crime and Delinquency 50(2):239–71. doi:10.1177/0022427811434830

97.

R Core Team. 2025. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from https://www.R-project.org/

98.

Reid

Shannon E.

2015. “Friendship Group Composition and Juvenile Institutional Misconduct.” International Journal of Offender Therapy and Comparative Criminology 61(2):191–209. doi:10.1177/0306624X15589737

99.

Reuter

Peter.

1983. Disorganized Crime: The Economics of the Visible Hand. Cambridge, MA: MIT Press.

100.

Reuter

Peter

Haaga

John

. 1989. The Organization of High-Level Drug Markets: An Exploratory Study. Santa Monica, CA: Rand.

101.

Reuter

Peter

Paoli

Letizia

. 2020. “How Similar Are Modern Criminal Syndicates to Traditional Mafias?” Crime and Justice 49:223–87. doi:10.1086/708869

102.

San Juan Flores

Patricia

Guillén

Beatriz

. 2025. “México, el país que desaparece: sin rastro de 125.000 personas.” https://elpais.com/mexico/2025-03-23/mexico-el-pais-que-desaparece-sin-rastro-de-125000-personas.html.

103.

Selmini

Rossella.

2020. “Women in Organized Crime.” Crime and Justice 49:339–83. doi:10.1086/708622

104.

Simmel

Georg.

1950. The Sociology of Georg Simmel. New York: The Free Press.

105.

Smith

Chris M.

2020. “Mafia Organizations: The Visible Hand of Criminal Enterprise.” Contemporary Sociology 49(5):432–34. doi:10.1177/0094306120946390g

106.

Smith

Benjamin T.

2021. The Dope. New York, NY: W. W. Norton & Company.

107.

Smith

Chris M.

Papachristos

Andrew V.

. 2016. “Trust Thy Crooked Neighbor Multiplexity in Chicago Organized Crime Networks.” American Sociological Review 81(4):644–67. doi:10.1177/0003122416650149

108.

Snyder

Richard

Durán-Martínez

Angelica

. 2009. “Does Illegality Breed Violence? Drug Trafficking and State-Sponsored Protection Rackets.” Crime, Law and Social Change 52(3):253–73. doi:10.1007/s10611-009-9195-z

109.

Thoumi

Francisco E

. 1995. Political Economy and Illegal Drugs in Colombia. Studies on the Impact of the Illegal Drug Trade. Boulder, CO: Lynne Rienner Publishers.

110.

Trejo

Guillermo

Ley

Sandra

. 2020. Votes, Drugs, and Violence: The Political Logic of Criminal Wars in Mexico. First edition. Cambridge Studies in Comparative Politics. New York: Cambridge University Press.

111.

Treviño-Rangel

Javier

Bejarano-Romero

Raúl

Atuesta

Laura H.

Velázquez-Moreno

Sara

. 2022. “Deadly Force and Denial: The Military’s Legacy in Mexico’s ‘War on Drugs’.” The International Journal of Human Rights 26(4):567–90. doi:10.1080/13642987.2021.1947806

112.

UNODC. 2023. Homicide and Organized Crime in Latin America and the Caribbean. Vienna: United Nations Office on Drugs and Crime.

113.

Valdés Castellanos

Guillermo.

2013. Historia del narcotráfico en México. Mexico City: Aguilar.

114.

Van de Bunt

Henk

Siegel

Dina

Zaitch

Damián

. 2014 “The Social Embeddedness of Organized Crime.” Pp. 321–40 in The Oxford Handbook of Organized Crime, edited by Paoli

. Oxford: Oxford University Press.

115.

Van Deuren

Sjoukje

Diviák

Tomáš

Blokland

Arjan

. 2025. “Co-Offending among Outlaw Motorcycle Gang Members: The Role of Social and Geographical Proximity.” Journal of Criminal Justice 96:102340. doi:10.1016/j.jcrimjus.2024.102340

116.

Verbrugge

Lois M.

1979. “Multiplexity in Adult Friendships.” Social Forces 57(4):1286–309. doi:10.1093/sf/57.4.1286

117.

Von Lampe

Klaus.

2016. Organized Crime: Analyzing Illegal Activities, Criminal Structures, and Extra-Legal Governance. Thousand Oaks, CA: Sage.

118.

Waring

Elin.

2002 “Co-Offending as a Network Form of Social Organization.” Pp. 31–47 in Crime and Social Organization, edited by Weisburd

Waring

E. J.

. New Brunswick, NJ: Transaction.

119.

Wippell

Jack G. R.

Haynie

Dana L.

. 2025. “A Multiplex Network Approach to Understanding Extremist Organizations: A Case Study of the Proud Boys.” Journal of Criminal Justice 96:102353. doi:10.1016/j.jcrimjus.2025.102353

Social Embeddedness,Multiplexity,and Criminal Collaboration Within the Sinaloa Cartel

Abstract

Objectives

Methods

Results

Conclusions

Keywords

Introduction

Background

The Embeddedness of Organized Crime

Multiplexity in Criminal Networks

The Current Study

Methodology

Data and Measures

Analytical Strategy

Results

Discussion and Conclusions

Footnotes

Acknowledgements

Declaration of Conflicting Interests

Funding

ORCID iDs

Notes

Author Biographies

Appendix

References