X Close

JDI Latin America and Caribbean Unit


Supporting research on crime and citizen security, and the professional development of policing in the LAC region.


Archive for the 'Insights' Category

The Relationship between Firearm Proliferation and Violence in Latin America

By Diego Sanjurjo, on 22 July 2020

The vast majority of Latin American countries share a similar trend of rising crime and violence. Most forms of violent crime have intensified, with property crime being the most notorious. Nonetheless, homicides are easier to account for. Murders experienced an 11 percent increase in the region between 2000 and 2010, resulting in more than 2.5 million killings since the turn of the century. Furthermore, homicides have arguably turned into a Latin American singularity, because it is the only region in the world in which murder rates increased during the first decades of the twenty-first century. According to the United Nations Office on Drugs and Crime, the region surpassed Africa for the first time and became the region with the highest number of murders globally, both in absolute numbers as in relation to its population.

Guns are fundamental in these dynamics and their use as murder weapons supposes a particularity of the Latin American homicide epidemic. Firearms were used to commit around 50 percent of all homicides worldwide between 2010 and 2015[i], but their impact in Latin America is even more pronounced. In Brazil and in Central America, for instance, gun homicides correspond to at least 70% of the total. Their use in non-lethal forms of crime has drastically increased as well in recent decades, while national reports show that rising insecurity and mistrust in state authorities appear to be linked to increase popular disposition to acquire them as instruments of self-defense. Hence, self-protection and criminal predatory behaviour fuel a rising demand for guns in a region that already possesses important surpluses dating back to the civil wars and military dictatorships of the 20th century.

As a result, the best and latest available estimates indicate that there are approximately 71 million small arms[ii] in Latin America. Among these, 14 percent belong to state security forces and paramilitary forces. The remaining 86 percent likely belong to civilians, resulting in a relatively low average country distribution of 9.87 civil guns per every 100 residents. Among independent countries and territories, national estimates of firearm ownership vary considerably, ranging between 2 firearms per every 100 inhabitants in Bolivia, Peru and Cuba, to 34.7 in Uruguay[iii]. Despite being on opposite ends of the gun-ownership spectrum, these countries present relatively low levels of violence. It is an indication that the relationship between civilian gun ownership and firearm homicides is not linear (Figure 01), and the correlation between gun proliferation and crime is usually low. This is not a distinctive feature of the region. The United States’ estimated gun ownership rate is 12 times higher than Latin America’s average, but its firearm homicide rate is 7 times lower.

Figure 01: Estimates of civilian guns (2017) and violent deaths by guns (2016).

Source: Own elaboration using estimations by Karp (2018) and Small Arms Survey (n.d.).

The influence of gun proliferation on violence and insecurity is more complex and difficult to predict than the regional literature usually suggests. Efforts to find evidence of causality are strongly contested and in-depth studies have not been carried out in the region. Ultimately, guns can be an effective instrument for personal defence against crime, as well as a lethal instrument to generate violence, commit crimes, and confront state security policies. However, their influence on regional violence presents distinctive particularities that go beyond such a linear relation. Guns do not suppose a problem in Latin America because of their sheer quantity, but due to their prominent incidence in crime and violence at an aggregate level. Indeed, lethal gun violence is not only disproportionally high, but national violent death rates also show a positive correlation with the percentage of violent deaths committed with firearms. In other words, countries in Latin America with high rates of violence also tend  to have high percentages of violent deaths committed with firearms (see Figure 02). This is the case in countries like Brazil, El Salvador, Honduras, Jamaica or Venezuela, for instance.

Figure 02: National violent death rates and percentage of violent deaths committed with firearms (2016)

Source: Own elaboration using data from Small Arms Survey (n.d.);

The relationship between gun availability and armed violence is not linear, and this phenomenon does not constitute evidence of causality between gun proliferation and homicides. However, it does suggest important correlations that cannot be ignored. The proliferation of small arms leads to different consequences in Argentina than in Guatemala. Even between territories, cities, and neighbourhoods, small arms trigger different processes. It is significantly more likely for a gun to be used to commit a homicide in Central America than in the Southern Cone of South America. Therefore, as important as the quantity of firearms and their capability to produce harms, it is also the circumstances that surround firearms that determine their lethality.

What this relationship suggests is that firearms seem to add fuel to the fire in violent places. If the presence of a gun during a fight between neighbors can make the difference between a broken nose and a murder, a similar dynamic occurs at the macro level, whereby easy access to firearms can contribute to extreme levels of violence. One way of looking at it is through an epidemiological approach and the variation of local risk factors. Accordingly, small arms represent one factor among many: social exclusion, weak institutional contexts, rapid and uncontrolled urbanization, a recent history of armed conflict, violent events in neighbouring countries, high percentages of young men who do not work nor study, the presence of organized criminal groups, and the expansion of illegal markets. A quick overview of the region suggests that all countries concentrate some of these factors to a certain extent. Most experience several risk factors, and some are severely affected by all of them.

Ultimately, such extended structural deficits make the proliferation of small arms especially pernicious in Latin America, as the easy access to guns triggers (and exponentially increases) what is already an explosive combination of risk factors. Under these circumstances, it seems reasonable to suggest that a more restricted access to firearms is an insufficient but rational policy step.

[i] Even though Latin America is the region with the highest proportion of gun homicides, the information on homicidal mechanisms in places such as Africa, Southern Asia, or Southeast Asia is very scarce, which may result in under-reporting (GDS 2015).

[ii] For the purposes of this post, I use the terms ‘small arms’, ‘firearms’ and ‘guns’ interchangeably and refer to the following items: revolvers and self-loading pistols; rifles and carbines; shotguns; sub-machine guns; and light and heavy machine guns.

[iii] The national firearm ownership rate informs us about the national firearm culture. Societies with the highest rates are those with (1) an important tradition of owning and using firearms, such as in Guyana, Paraguay, and Uruguay; or with (2) access to the powerful small arms market of the United States, such as in the cases of the Bahamas, Puerto Rico, Surinam, Mexico, and Panama.

Diego Sanjurjo holds a Ph.D. in political science from the Autonomous University of Madrid, he is currently the Coordinator of the Policing and Preventive Police Component of the Comprehensive Citizen Security Program of the Uruguayan Ministry of Interior. His main research areas include public policy and security, gun and drug policies. He is the author of “Gun Control Policies in Latin America” (Palgrave Macmillan, 2020).

The views expressed in this blog post are the author’s own and do not necessarily represent the views of UCL, the Department of Security and Crime Science or the JDI LAC Unit.

Another Approach to Crime Reduction in Colombia

By Alberto Nieto and Herve Borrion, on 5 June 2020

Following concerns about the increasing number of crimes reported in Colombia, President Ivan Duque ordered last year the creation of a plan to reduce crime. In the past weeks, officials from the National Police (NP) have revealed the contours of this strategy that aims to reduce crime by improving the criminal investigation process. But, is it such a good idea? Colombia faces a daunting scenario. New accounts show that 1,136 people were robbed on average per day during the first semester of 2019. Violence is also a significant concern. During 2018, 25 people per every 100,000 inhabitants were killed, situating Colombia as the 9th country with the highest homicide rate in Latin America and the Caribbean. The strategy’s underpinning objective is to increase the effectiveness of NP and the Prosecution Service in dealing with crimes as soon as they are committed. Although not explicit, there is hope that this would produce a general deterrent effect, by sending a message to offenders about the risk involved in committing crime.

Improving part of the criminal justice system might seem like a good idea. However, this is not as straightforward because the criminal justice system is a complex system with multiple agents (e.g., citizens, law enforcement agencies, judges, prosecutors, public defendants, prisons, social workers), multiple functions and processes, as well as forward and feedback loops. In this case, it is questionable whether improving the criminal investigation process is really an effective and efficient way to reduce crime.

With the new plan, we can expect more cases to be pushed to the next stage of the criminal justice system: courts and prisons. The logic would be that a higher number of offenders will be rehabilitated or deterred. In practice, however, those cases are unlikely to be processed effectively, and the sought effect may not materialise. Indeed, there are already more than 36,000 individuals in pretrial detention (33% of prison population), and prisons have an overcrowding rate of 55%, making Colombia the 5th country with the highest occupation level in South America and 50th worldwide.

Considering how congested the criminal justice system is, reducing the number of criminal opportunities would probably be a more promising strategy. Recent studies suggest that the crime drop observed in most developed countries in the West can be attributed to improved security measures used to increase the perceived risks of getting caught, reduce the rewards for the offenders, or increase the effort needed to commit an offence. These means are the tenets of the situational prevention approach to reduce crime.

For law enforcement agencies to take a leading role in the prevention of crimes requires a real shift in strategy, mindset, and incentives. One of the most successful approaches, problem-oriented policing introduced by Herman Goldstein in 1979, puts a premium on an in-depth understanding of the problems which law enforcement agencies need to address. POP, as this strategy is also known, has tried to change the standard, reactive policing model that deals with crime events individually, similar to the one envisioned by the NP’s new plan. This approach focuses on identifying recurrent crime and disorder problems and understanding and tackling the underlying causes to prevent them.

Police forces’ commitment and interest in POP has fluctuated since its introduction, but it is regaining interest. The challenge remains in how well law enforcement agencies define the problems and understanding the social, ecological, and ethical considerations of the interventions intended to reduce crime, as discussed in this article published in the British Journal of Criminology.

The opportunity for countries like Colombia to adopt POP seems to be remote. Delivering process outcomes —i.e., the number of people arrested and sentenced—continues to prevail in how public opinion holds accountable law enforcement agencies, and law enforcement agencies might use these outcomes to regain confidence in people when the number of crime reports increases. The difficulty in quantifying the number of crimes that can be prevented, therefore, reduces the importance of a preventive approach to crime.

There is merit in trying to improve the investigative process, and therefore in NP’s new plan. However, significant crime reduction should not be expected from it, for as long as prevention continues to be a second-order approach and law enforcement agencies solely rely on the criminal justice system to produce general deterrent effects. For this, partnerships between law enforcement agencies and other actors of society will need to be established to address recurrent crime problems. Complex system theory might come in handy to start shaping a new understanding of criminal opportunities and testing alternative strategies to improve security.

Alberto Nieto is a PhD candidate in the Department of Security and Crime Science. His research focuses on the evolution of co-offending networks.

Dr Hervé Borrion is Associate Professor at the Department of Security and Crime Science.

The views expressed in this blog post are the author’s own and do not necessarily represent the views of UCL, the Department of Security and Crime Science or the JDI LAC Unit.

Montevideo: Between Violence and Urban Fragmentation

By Gonzalo Croci, on 27 May 2020

Uruguay currently suffers from the highest levels of insecurity and crime in the country recent history. Data prior to the 2002 economic crisis shows that structural factors such as unemployment and poverty were closely associated with the increase in crime in the period 1986-2002. However, between 2004 and 2015, Uruguay has made progress in the social and economic dimensions and has reduced its levels of poverty and inequality measured in real terms. This, however, happened alongside an increase in crime rates. The link between structural variables and crime seems not to be as linear as it had been claimed and is questioned both by academics and policymakers. This article argues that other causes namely, drug trafficking and urban fragmentation have influenced the increase of crime, and specifically of homicides, in Uruguay.

Statistics on homicides are widely accepted as a metric for security, especially violence. Further, homicides are commonly held to be a good index of quantitative criminal tendencies, on the assumption that police and judicial authorities are more diligent in reporting and prosecuting homicides, while other crimes such as theft can escape the attention of public institutions. Uruguay’s homicide rates have risen continuously since 2011. In that year, the homicide rate stood at 5.8 homicides per 100,000 inhabitants. By 2018, the government reported a homicide rate of 11.8 homicides, double the world average. A particularly high spike of homicides was registered in 2017, when the total number of homicides increased by 45.8% compared to the previous year. The latest data confirms that the trend continued in 2019 and seems to stabilise in 2020.


The sudden increase of homicides in recent years seems to be closely linked to confrontations among gangs over the control of drug traffic routes and territories. Data from the Ministry of Interior shows that 47% of homicides were due to conflicts among criminal groups, drug trade, and hired killers. Foreign traffickers, mostly from Brazil and Colombia, are working with local gangs and taking advantage of the countrys porous borders with Argentina and Brazil; they are now using Montevideo as a base for drug transit operations. Although drug movements in the port of Montevideo are far less substantial compared to those in the ports of Rosario, Santos or Rio, it is becoming a centre for Europe-bound cocaine, smuggled from the Paraguay and Parana rivers, as well as other land routes. As such, cocaine seizures have constantly increased: 134 kg of cocaine were seized in 2015, 148 in 2016, 144 in 2017, 586 in 2018, and 12.042 kg in 2019. There was a noticeable spike in seizures in late 2019, when several large cocaine shipments were seized in Uruguay or after coming out of the country. One among them, a cargo ship seized in Hamburg, was a particularly eye-opening event: the ship contained four tons of cocaine with an estimated value of € 1 billion, the largest-ever cocaine shipment seizure in Germany. Moreover, recent activities of the Primeiro Comando da Capital(First Command of the Capital, PCC), Brazil’s largest criminal organisation, seem to be reinforcing the problem. It is unknown how much power the organisation has in the country, but it is clear that PCC’s interest in Uruguay is increasing. In December 2019, a PCC cell was dismantled, and the police claimed that the group had recruited 84 criminals in Uruguay.

Drug trafficking and clashes among criminal organisations may well be the short-term cause of violence in the country; however, there are also long-term underlying causes that facilitate violence. Drug trafficking organisations use violence to appropriate the territory, clientele, routes, and other assets of their rivals. Generally, this type of violence occurs among criminal organisations that are active in the same territory, where simultaneously the state plays a limited role. Studies have shown that the absence of the state and of clear and effective norms facilitate the increase of violence. Hence, phenomena such as homicides related to organised crime, drug markets, and gangs can be understood as a consequence of a weak or non-existent state monopoly of power in specific territories.

Ungoverned spaces provide a safe haven for lethal criminal groups that pose a constant threat to individual security. These areas are not uncommon in Latin America: favelas and villa miserias are broadly uncontrolled and unregulated urban areas that exist in several countries in the region. As formal state structures do not provide public goods to certain segments of the population, other actors involved in illicit activities step in and take on the functionally equivalent role of the state by monopolising violence, asserting authority and identity, providing parallel justice mechanisms, and establishing informal economic opportunities.


The metropolitan area of Montevideo includes 62% of the total population of the country on a surface that amounts to 10% of the national territory. It is in Montevideo where the biggest urban problems are, constituting the phenomena of urban violence as one of the most worrying social processes of contemporary Uruguay. With a homicide rate of 16.1 homicides per 100,000 inhabitants in 2018, approximately 54% of all homicides in the country occur in its capital. Additionally, and confirming previous research on crime patterns, homicides are not dispersed proportionally throughout the territory. Violence and crime often tend to cluster around a small number of places, people, and behaviours and Montevideo is no exception. According to official data, homicides are concentrated in specific areas within the city, with 44.4% of homicides occurring in only three sectional districts out of 25: Sectional 17 (Casavalle), Sectional 24 (Casabo-Cerro) and Sectional 18 (Piedras Blancas-Punta de Rieles). These districts have homicide rates higher than 30 homicides per 100,000 inhabitants. The consequence of high levels of violence not only carries a human and economic cost but also tarnishes basic state legitimacy, which in turn foments more citizens to act outside the law. When institutions are perceived to be unjust, weak, or ineffective, citizens withdraw their support, contributing to weakened social control mechanisms and higher levels of violence.

The population of these neighbourhoods not only suffers the consequences of high levels of violence but also copes with extremely precarious living standards. Social and territorial fragmentation and economic deprivation take on an unusual strength in these neighbourhoods.Social disorganisation theory argues that an extreme concentration of social disadvantages in neighbourhoods creates a distinctly different social and structural environment that facilitates crime. Data from the National Institute of Statistics (INE) shows that these neighbourhoods are some of the poorest in the city, with low income levels, low levels of education, little presence of health services, high levels of unsatisfied basic needs, significant levels of overcrowding, and limited provision of basic services, such as public lighting and sanitation. In short, these neighbourhoods have limited presence of the state and have characteristics that enable criminal organisations to prosper.

According to O’Donnell (1993), state theories are often based on the assumption that there is a high degree of homogeneity in both the territorial and functional domains of the state and of the social order that it provides and supports. As such, it is assumed that public policies have a similar impact and effectiveness throughout its population. In Montevideo, instead, deep economic and social segregation has generated urban fragmentation, thus creating territories where the Uruguayan state is no longer a guarantor of order and security.Marcuse (1989) introduced the concept of the a “quartered city”, which refers to fragmented cities that mirror a deep sense of social division. Urban fragmentation facilitates the appearance of conflicts and violence as parts of the city suffer from a reduction of private investments and from neglected public services. These areas thereby become spaces of stigmatisation, which in turn encourages new waves of desertion of the middle classes abandoning those areas and deepening the social gap.

In Uruguay, most of the public policies that have been carried out to raise the welfare of the urban poor have neglected the problems of their integration with the rest of society. Policymakers assumed that, by improving the living conditions of certain segments of the population, citizens would be able to establish (or re-establish) meaningful links across existing social barriers. That, however, does not seem to be the case.Previous studies have shown how the socio-economic gap between wealthy and poor neighbourhoods has increased over the past 15 years: for example, the gap between the most and least educated households was higher in 2017 than in 2007. It would seem that Uruguay’s economic and social improvements have been more significant in neighbourhoods that were already better positioned, which implies a deepened gap between neighbourhoods during the same period of time when violence increased. The signs of deterioration are undeniable: Montevideo is witnessing not only a process of increasing violence and criminality, but also of increasing urban fragmentation – two developments whose link to each other we need to better understand to reduce violence.


Structural disadvantages, urban fragmentation, disparities in public services, and the increase in violence have a common denominator: a weak and ineffective state. To solve these issues, the state must recover its lost territory in specific urban areas, not only by improving its security services and imposing order, but also by improving the most basic services and reintegrating fragmented neighbourhoods into the rest of society. Strong institutions and effective policies are needed to neutralise the effects of rising violent crime. This strategy presupposes, above all, a significant increase of the presence of the state in the most vulnerable areas.

Each city and, indeed, each neighbourhood within a city has its own specific security risk factors that need to be addressed through careful policy-making. Local development entails the design of concrete, data-driven, and measurable policies by municipal and central government agencies, in addition to an effective coordination between the different levels of governance. For the Uruguayan state to be once again a guarantor of order and security, it is essential that innovative policy-makers in public security understand the cross-cutting nature of security problems and therefore focus their efforts on the development of holistic policies.

This article was first published in the Urban Violence Research Network: https://urbanviolence.org/montevideo-between-violence-and-urban-fragmentation/

Gonzalo is currently a PhD candidate at the UCL Department of Security and Crime Science

Please note that views or opinions expressed in this blog are the author’s own and do not necessarily represent the views of UCL, the Department of Security and Crime Science or the JDI LAC Unit.