Crime in Pieces: The Effects of Mexico’s “War on Drugs”, Explained

In December 2021, the United States federal government offered up to $20 million for information leading to the capture of four of Joaquín “El Chapo” Guzmán’s sons. Following the arrest and extradition of their father in 2016, the young men – collectively known as “los Chapitos” – have engaged in a bloody feud for control of the Sinaloa Cartel with a rival faction.

The Biden administration’s decision to place a bounty on the heads of the Guzmán brothers at the helm of the Sinaloa Cartel follows a long history of making the removal of kingpins a central goal of the so-called war on drugs. But after more than fifteen years of this “kingpin strategy”, Mexico remains ensnared in deadly mass violence involving criminal groups, with more than 30,000 homicides registered each year since 2017. A large share of these killings is estimated to be the result of organised criminal violence.

In total, at least 543 armed groups operated in Mexico between mid-2009 and late 2020.

How has the kingpin strategy contributed to this state of affairs? Crisis Group data demonstrates that as leaders have been removed, the number of armed groups in Mexico has grown. At least 543 armed groups operated in Mexico between mid-2009 and late 2020 – the vast majority being criminal in nature, with some vigilante groups and a small number of organisations with more political aims in the mix. The geography of violence has also changed: criminal activity has reached new areas, reflecting the growing diversification of illicit revenue streams. As the number of groups they are hosting grows, municipalities tend to become more violent, suggesting that the homicide rate and criminal fragmentation are intertwined.

The challenges that Mexico faces as it battles criminal violence have fundamentally changed over the last decade and government policy should change too. Rather than continuing to rely so heavily on military solutions, President Andrés Manuel López Obrador should implement new strategies that focus on protecting the most vulnerable residents of violence-affected areas, tackling corruption and collusion, and redressing the specific causes of conflict in the country’s most violent regions, including by developing licit economic alternatives to armed group recruitment. 

Gathering accurate data on Mexican armed groups poses huge practical challenges, in large part because crime reporting – normally a key source of information – is so difficult. Mexico is the most dangerous country in the Western Hemisphere for journalists. Self-censorship in high-conflict zones is rampant. For example, a newspaper in Ciudad Juárez released an open letter to criminal groups in the aftermath of an attack that killed members of its staff: “We ask you to explain what you want from us, what we should publish or not publish, so we know what to expect”.

A 2011 agreement among certain media outlets to avoid spreading criminal propaganda – one of the major means by which small illegal groups announce their presence –  further complicates learning about their number and relative strength from the mainstream press. Moreover, because criminal outfits increasingly operate very locally, they may not receive coverage from major state or national news, even though they play a crucial role in driving regional violence.

To overcome these issues, Crisis Group developed a dataset of criminal groups in Mexico by drawing on “narcoblogs” – citizen journalism websites dedicated to tracking the activities of criminal groups involved in the drug trade. These have a number of advantages relative to traditional media in tracking criminal groups. Narcoblog contributors are typically anonymous, and the sites rely on a mixture of reporting, press aggregation and information taken directly from citizens, making them less prone to self-censorship. Where necessary, Crisis Group supplemented these data with information from the mainstream press related to the groups’ origins and alliances. The methodology used to develop this dataset is described here.

The above-referenced conclusion that 543 armed groups (only a portion of which were active at any particular time) operated in Mexico between mid-2009 and 2020 was based on information from the Borderland Beat blog. Breaking the data down further, Crisis Group found that the number of criminal groups in Mexico has more than doubled, from 76 to 205 between 2010 – the first full year of data – and late 2020. These crime rings vary greatly in size and scope. The U.S. Drug Enforcement Agency (DEA) has recognized just nine of these groups as major drug-trafficking organisations. At least five of these no longer operate as unified entities.

Of other groups identified in the dataset (ie, groups other than major drug-trafficking organisations), 212 align with a larger organisation, although these connections are often unstable. Most often, these cells are either local bands to which larger groups provide resources, or they are specialised subgroups that operate semi-independently from the larger hierarchy of a bigger group. These groups are particularly important to track because they frequently fight among themselves, switch sides or become independent splinter groups. With 70 past and present cells, the Sinaloa Cartel has by far the most such subgroups in Crisis Group’s data; indeed, it is often referred to as the “Sinaloa Federation”. 

At least 107 of the identified groups began as splinters of other organisations. The Zetas, once among the most important criminal operations in Mexico, now mainly exists as a series of splinter groups, the most powerful being the Cartel of the North East and the Old School Zetas.

Crisis Group’s data also track alliances forged through either permanent or temporary coordination agreements between groups, as described in narcoblogs and other press sources. At least 87 groups in the dataset allied with another at various points, though this is likely an underestimate given that such alliances may be region-specific and short-lived. Ten criminal groups identified are actually umbrella organisations that aggregate multiple outfits operating in a territory, under a single name. The United Cartels in Michoacán, for example, coalesced a number of local groups to repel incursions from the Jalisco New Generation Cartel (CJNG), currently one of two of the most powerful criminal groups in the country alongside the Sinaloa Cartel. In other cases, alliances are more ephemeral. For example, in a war of succession over the Tijuana Cartel, El Ingeniero formed an alliance with the Zetas who then used this as an opening to expand their influence in Tijuana.

These maps show the number of criminal groups across Mexico’s states, ten years apart: between November 2009 and October 2010, and between November 2019 and October 2020.

The places where criminal groups operate in Mexico have shifted considerably over the last decade, reflecting the growing diversification of illicit revenue streams. 

In 2009-2010, groups operated in 11 per cent of municipalities. By 2019-2020, this had increased to 29 per cent. In part, this change reflects how organisations grew to sustain themselves through a far wider array of activities. For example, by 2014 the Zetas were involved in at least 25 different industries. Illegal logging, poaching, opioid production and distribution, iron ore extraction and theft, online crimes, kidnapping, extortion of small businesses and human trafficking have allowed groups to sustain operations across more territory. As Crisis Group has recently documented, the rise of fuel theft has led to greater fragmentation and violence in areas with gas pipelines.

Criminal group fragmentation and proliferation into new territories are concerns in large part because of the link to worsening violence. Crisis Group’s data show a clear correlation between the number of groups present in a municipality and the average per capita murder rate, with violence being by far the lowest in regions with no known criminal presence. Violence continues to increase broadly in step with the number of additional groups – up to a certain point, that is, when average per capita homicides flatten out. The data suggest this flattening is probably because higher levels of fragmentation signal the greater presence of minor, local criminal organisations. The presence of larger groups, and particularly the nine major drug-trafficking organisations identified by the DEA, appear to underpin higher levels of violence.

Behind this correlation lies a complex mix of factors, which has been described in previous reporting by Crisis Group and other bodies. For example, speaking of the southern state of Guerrero, Crisis Group noted that: “Guerrero is a welter of local feuds. Each quarrel is distinct, based on the structure of the groups involved as well as the resources over which they are fighting”. A UN Office on Drugs and Crime study cited “disorganisation” – meaning the growing complexity of the criminal landscape – as a major factor in the murder rate.

The link between fragmentation and violence also reflects the fact that certain regions are more valuable for criminal groups to control. Areas that offer the greatest economic returns are more likely to attract new groups and more likely to be prone to violence as shown in Crisis Group’s research on fuel theft and the sources of violence in Michoacán state.

In late 2006, Felipe Calderón sent the military into Michoacán to eradicate drug trafficking and dismantle criminal organisations, marking the beginning of the most intensive chapter in the militarisation of Mexico’s “war on drugs”. Supported by the U.S. Mérida Initiative since 2008, this militarisation has been widely blamed for Mexico’s rocketing homicide rates in the years that followed. 

Despite criticisms of Calderón’s approach, subsequent presidents have largely upheld his policies. Enrique Peña Nieto’s administration (2012-2018) promised to focus on violence reduction rather than the dismantling of crime groups, but continued to rely heavily on the military and a new national police force to break up criminal operations.

Peña Nieto’s administration criticised the kingpin strategy for making groups “more violent and much more dangerous”, estimating that dozens of smaller outfits had splintered from larger organisations. Nevertheless, it continued to place significant emphasis on arresting criminal leaders. Peña Nieto voiced regret in his final weeks in office that his administration had reduced the power of large criminal organisations, but lacked the local capacity to address the emergence of smaller groups.

Mexico’s current president, Andrés Manuel López Obrador (elected in 2018), has publicly broken with the kingpin strategy, declaring that “no capos have been arrested because that is not our main purpose. … Officially, there is no war”. Partly because Mexico does not have a clear definition of what a kingpin is, however, it is difficult to substantiate this claim. 

There are also some dissonant facts. For example, the government did arrest one of El Chapo’s sons, though authorities released him following a major confrontation with security forces in 2019. The U.S. bounty for all four sons came two years later. Security forces also detained El Marro, head of the fuel theft-focused Santa Rosa de Lima Cartel, in August 2020. In 2022, the security services arrested the leader of a Sinaloa Cartel faction. Moreover, although López Obrador ran on a policy focused on “hugs not bullets”, critics have pointed out that his newly created National Guard, established in 2019, looks like a continuation of his predecessors’ highly militarised strategies. 

Critics of the kingpin strategy argue that the policy weakened large criminal structures, but the power vacuums left behind were quickly filled. Loss of leadership may breed succession contests, reduce criminal groups’ oversight of local cells, leading to factionalisation and splintering, and encourage new groups to enter into formerly tightly controlled areas. For example, the extradition of El Chapo to the U.S. led to a conflict over succession between los Chapitos and his former lieutenant. Violence in Sonora exploded as local groups, Sinaloa factions and other outfits fought to take over a region once firmly viewed as under the control of El Chapo.

The link between the kingpin strategy and fragmentation is demonstrable from available data, including from the U.S. government, which has worked hand in glove with the Mexican authorities to shape and implement the policy. Important information on kingpins is available from the U.S. Treasury Office of Foreign Assets Control’s Specially Designated Nationals List, which includes individuals sanctioned under the Foreign Narcotics Kingpin Designation Act. The data are not exhaustive, however: U.S. Treasury data are limited to individuals linked to groups in some way of interest to the United States. The data thus cover major organisations, like the Sinaloa Cartel, but do not include kingpins from groups that operate largely domestically within Mexico, including the Santa Rosa de Lima Cartel. Standardised data on kingpins were provided to Crisis Group by the Center for Advanced Defense Studies and linked by Crisis Group researchers to arrests and deaths. 

As statistical analysis based on this data show, fragmentation increases in municipalities affected by kingpin captures – those areas where a group operated in previous years and experienced an arrest or capture of a leader. On average, relative to other municipalities, one additional group operates in affected regions following a kingpin capture.

Different types of groups enter these territories. Powerful organisations, splinter groups, and other medium- or small-sized groups may all increase following a kingpin capture. In other words, the relationship between the kingpin strategy and fragmentation appears to reflect both the breakdown of existing outfits and widening opportunities for new groups to enter a territory.

These findings likely reflect a combination of causal factors, including kingpin captures weakening groups and weakening groups seeing their kingpins captured. Fragmentation may also in part result from the decomposition of a group’s hierarchy, a direct consequence of the arrest or death of leadership. For example, Los Rojos, Los Ardillos and Los Guerreros Unidos squared off against one another in Guerrero state after splintering from the Beltrán Leyva Organisation, following the killing or capture of its top bosses – which at the time President Calderón called an “important success for the Mexican people”. As one high-ranking officer observed, “When the big one [Arturo Beltrán Leyva] wasn’t around anymore, nobody accepted the other’s leadership and they started betraying and fighting each other. That’s the situation we’re in today”. But the very fact that a kingpin was captured may also show groups were already weaker. Often, the application of the kingpin strategy followed financial and other military operations that laid the groundwork for capture.

Similarly, after La Familia Michoacana started to split up in late 2010 as a federal offensive aggravated internal disagreements, a number of splinter groups began to operate more independently over the following years. This is illustrated in the figure above which shows La Familia and its offshoots (weighted by the number of appearances in Borderland Beat, used as a proxy for the importance of a group). The largest splinter group, the Knights Templar, remained a major drug-trafficking organisation until being weakened by the emergence of self-defence groups in the state from 2013, and by the death and arrest of key leaders in 2014 and 2015 respectively. 

The weakening of major, targeted criminal structures was, in part, a goal of the kingpin strategy. But there is no evidence that the kingpin strategy reduced violence. To the contrary, major networks like the CJNG and Sinaloa Cartels continue to operate widely. Homicides have risen since the militarisation of the “war on drugs” and remain high. Major groups have extended their presence across Mexico, from operating in 268 municipalities in 2010 to 531 in 2021. They are also significantly more likely to face competition: in 2010, the nine major groups operated unchallenged in 47 per cent of the municipalities where they were present, compared to 29 per cent in 2020. At the same time, there has been a proliferation of smaller and midsized groups, making targeting leadership an increasingly futile task.

There is no quick or easy fix to Mexico’s conflict-level criminal violence. As noted, the kingpin strategy has exacerbated the situation. Fragmentation aggravates the violence caused by competition among groups for territory and resources. It opens up more fronts of conflict and spawns more criminal contenders, too often catching civilians in the crossfire. Widespread links between state officials and criminal groups hinder progress. 

A new strategy is warranted. Given the frustrating results of the kingpin strategy, a more promising way forward would see resources concentrated in the most violent parts of the country through tailored regional action plans which would have the goal of addressing the specific regional characteristics of conflicts and their causes. 

One element of such an approach might involve stepped up state and federal support for the efforts of local intermediaries facilitating agreements between non-state actors – as Crisis Group has previously recommended – which some mediators believe have contributed to reducing violence within Guerrero. Criminal groups and state security forces already routinely engage in “conversations” and “agreements” as members of criminal groups, as well as military commanders, have phrased it in Crisis Group interviews. But existing channels of communication could be used to encourage better protection of civilians – increasingly victims of extortion, displacement, execution and disappearance – as well as to lower homicides by introducing minimal rules of behaviour. In parallel to communicating these rules to violent actors, authorities could concentrate on intelligence-led law enforcement efforts that hone in on the most predatory groups involved in homicides, kidnapping for ransom and extortion of licit businesses. Doing so, however, would require state authorities that have both the will and capability to enforce such rules and do not side with any preferred criminal group. 

The heart of the regional action plans would accordingly be a focused effort to sever the links between state officials and criminal groups through the use of dedicated, highly vetted prosecution and police units, with the aim of sharply reducing sky-high impunity rates for serious crimes. Unlike under the kingpin strategy, a priority would be to tackle the official collusion and corruption that give rise to new generations of criminal groups and undercut state responses to them. Regional action plans would also bring resources to bear on bolstering legal economies, to provide locals with an alternative to recruitment by criminal groups, and on setting up pathways back to civilian life for those already involved in criminal groups. Aside from some local initiatives, significant efforts of this ilk have been sorely lacking until now in Mexico. 

Meanwhile, international cooperation is needed to starve Mexico’s internal armed groups of weaponry and funding. Criminal groups get their arms and ammunition predominantly from abroad, chiefly the U.S. and Europe. At the same time, more and more of the fighting in Mexico is now over originally licit markets, including agricultural produce and natural resources being exported to markets in North America, Europe and Asia. Foreign governments should help by offering financial and technical support for the above-referenced regional plans rather than continuing to support counterproductive kingpin strategies.

A sustained state effort to reduce violence and crime in Mexico should not and need not rely on methods that end up raising murder rates and multiplying criminal groups. Understanding and avoiding the undesired effects of the kingpin policy is a crucial step in a more promising direction.