Human Rights

Explaining High Murder Rates in Latin America: It’s Not Drugs

By Elaine Denny and Barbara F. Walter

In a background paper for the World Bank on homicide rates James D. Fearon found that in the last two decades murder rates around the world have either stayed steady or declined, except in Latin America and the Caribbean where they have significantly increased. This increase is across the board; as H. Hugo Frühling and Joseph Tulchin note in Crime and Violence in Latin America: Citizen Security, Democracy, and the State all homicide rates have been rising over the last 25 years regardless of whether a country had low or high homicide levels to begin with.

Most people might assume that this trend is due to higher levels of drug trafficking and the violence that surrounds competing drug cartels. This is wrong.  Although Fearon found “a slight tendency for higher homicide levels in drug producing or trafficking countries,” this effect was not very large, and the increase in homicides can be found throughout the region whether drugs are present or not. Fearon’s analysis also found that trends in violence were not correlated with a country’s history of civil conflict.

What then might explain the rise in homicide rates in Latin America compared to the rest of the world? In a recent Global Study on Homicide, the United Nations highlighted a mix of factors that might be associated with a country’s murder rate. As complied by Rodrigo Soares and Joana Naritomi, these include high income inequality, the availability of guns, gang activity, the drug trade, alcohol abuse, unemployment, poor public education, a high youth population, as well as low incarceration rates and small police forces.

Of these factors, a number stand out as being uniquely present in Latin America. Compared to other regions of the world, citizens in Latin America do experience relatively higher levels of inequality, easier access to guns, and greater gang violence. But this doesn’t tell us why this causes individuals – especially young men – to kill each other.  Here are three possible explanations.

Inequality, Individuals, and Institutions

Inequality appears to be key to explaining violence. Statistical studies have consistently found that income inequality predicts homicide rates better than poverty does. Moreover, other countries with similarly high levels of inequality, such as South Africa, have been found to have comparably high murder rates. We think this relationship exists for at least two reasons. First, large disparities in wealth create greater competition within large youth populations facing high unemployment and limited upward mobility. From an individual perspective, murder might be the consequence of young people driven to extreme measures – people who turn to violent crime or gang involvement as the easiest path for increasing wealth or status.

Second, income inequality also creates two disparate groups competing for public goods from the state – the rich and the poor – with the rich successfully lobbying for a disproportionate share of services (or for a service’s elimination). Latin America’s economic elite has strong political ties and their interests do not necessarily coincide with the interests of those living in poorer neighborhoods. Whereas poorer citizens might push for greater law enforcement and better policing in their neighborhoods, wealthier citizens might have no interest in public policing that does not directly benefit them. The result is the sub-standard provision of security in neighborhoods that need it the most. This argument would also apply to more indirect policies that benefit one group more than another such as job creation, public education, and affordable healthcare.  Promotion of elite interests, therefore, not only calcifies the inequality gap, but siphons money away from law enforcement in poorer neighborhoods.

The Big Neighbor to the North

One notable difference between Latin America and the rest of the world is its proximity to the United States; proximity that results in a relatively easy southward flow of weapons, a northward flow of drugs , and more meddling by a powerful neighbor. Latin America’s relations with the hemispheric superpower arguably have often resulted in destabilizing political intervention and lackluster growth initiatives, which in turn may contribute to the inequality described above.

Furthermore, in a region where over 70 percent of murders are committed with firearms, the availability of guns plays a major role in violence levels. Consider Jamaica – a country that consistently has some of the highest homicide rates in the world. In Jamaica the majority of guns seized in the last decade can be traced back to Florida.  Initial studies also suggest that the region’s access to guns – including semiautomatic weapons – increased after the US Federal Assault Weapons Ban expired in 2004. The expiration of this law has been shown responsible for at least 16.4% of the increase in Mexican homicides between 2004 and 2008.

Taken together, these factors provide a motive and the means to commit murder, on top of a legacy of insecurity when national interests or citizen mandates have conflicted with US interests.

Gang violence and suboptimal equilibria

Still, these factors do not necessarily explain why murder rates are rising. We think this increase has something to do with the particular form of violence – gang-on-gang violence – that is happening in Latin America.

Gang violence has a specific dynamic that makes it self-perpetuating.  The initial cause might be competition, anger, an initiation rite, or even a mistake. But once the initial killing takes place, the code of conduct surrounding gangs demands retaliation. Retaliation is not only a way to exact revenge, but is also a way to establish one’s status. The effect, however, is that one killing begets another killing. Once someone is killed, retaliation becomes a self-perpetuating equilibrium that is impossible to escape without losing standing in the game.

How do you stop this cycle? An event that occurred in El Salvador this past March offers insight. In March 2012 a truce was brokered by the Catholic Church between El Salvador’s two major gangs. Since the truce, homicide rates in the country – which previously had hovered around the third highest in the world – plummeted by almost two-thirds. Murders are now down from an average of fourteen murders per day to five, with a decrease of over 50% from the same period last year.

Why was the Catholic Church so effective despite no real means to enforce the truce? The Church’s intervention offered gangs a face-saving way to break the cycle of reprisals they were stuck in. Suddenly, a gang’s decision NOT to retaliate represented taking the moral high ground rather than showing weakness.

Prior work has applied game theory’s Stag Hunt to post-crisis situations, showing how a disintegration of social order can result in a suboptimal equilibrium of violence or lawlessness. Perhaps Latin America’s economic disparities and institutional realities have catalyzed such destabilization in slow motion, leading to a status quo where young men are killing each other in the streets with little incentive to stop. Such a theory is consistent with Fajnzylber, Lederman, and Loayza, who find that homicides are caused by inequality and are self-perpetuating. The violence then increases over time as a culture of payback creates a world where one murder builds on another.

While the problem of high homicide rates is complex, it’s not intractable. This discussion highlights different possible points of entry for policy to reduce murder rates, from motive (institutionalized inequality and its frustrations), to the means (illicit arms trade), to the mechanism (retaliatory and status-driven violence). Of course, other factors, such as strained social fabric or the effectiveness of police and judicial institutions, are at play as well. We invite comments on the causal pathways driving homicide trends in Latin America, as well as the most promising domestic and foreign policy approaches to stem the region’s rising murder rates.


  • This is a very interesting post that makes some good points. I see a few problems, however.

    First of all, “Latin America” does not have an unusually high murder rate. Argentina, Bolivia, Chile, Costa Rica, Peru, and Uruguay have lower murder rates than the US. In South America, the countries with high homicide rates are Brazil, Colombia, and Venezuela.

    Second, in the case of Central American countries with high homicide rates, Fearon’s claim that “trends in violence were not correlated with a country’s history of civil conflict” is fairly meaningless. The countries with the highest homicide rates–El Salvador (57 homicides/100,000 pop.), and Guatemala (30)–are the countries that experienced the most severe political violence during the proxy Cold War of the 1980s. The gang violence in El Salvador is directly related to the legacy of their civil war. There might not be a correlation between the “trend” in those countries, because their levels of violence were already high to begin with.

    Next, there are obvious cases, such as Colombia and the 50+year civil war that have no relation to the sources of violence in other countries. Finally, while I appreciate that the main point of the post is to correct the false impression that drugs is the main cause of violence in Latin America and to call attention to questions of inequality, the militarization of the US-led “war on drugs” is complexly related to a range of other factors, all of which vary to large degrees depending on the particular country.

  • This is an EXCELLENT post–well-documented and thorough. While there is variation in homicide levels within the region, it is impossible to ignore the larger picture that Fearon points out: Latin America escapes the downward trend in lethal violence found in most of the rest of the world. And in the Southern Cone countries where violence is notably low, the reasons for this are highly predictable including lower inequality, stronger democratic institutions, smaller (or no) youth bulge, and shorter shadow of the U.S. both in terms of its policies and its small arms market. For an explanation of the mechanisms of inequality I highly recommend Wilkinson and Pickett’s “The Spirit Level: Why Greater Equality Makes Societies Stronger” (2009) which advances an argument that I utilize in explaining gang participation and violence in my book “Homies and Hermanos: God and Gangs in Central America” (2011).
    My only (mild) criticism of this post regards the argument that the drug market does not contribute significantly to violence levels. In the Fearon addendum, he buttresses this claim with evidence from data in which Honduras is categorized as a country with “NO” drug involvement. The miscategorization of that country alone–the most violent in the hemisphere last year–is enough to skew the data significantly. While it is certainly true that not all violence in Latin America is drug related, anyone who has been reading the national newspapers in Central America knows that a great deal of the lethal violence there has been the result of warring between local transportistas and the Mexican cartels.

  • Sorry, but I have studied criminal violence for 40 years, and the obvious Marxist “flavor” of this author’s “research” comes through loud and clear – all we need is massive wealth redistribution and voila! violence fizzles out. The reality is that even within a country like the United States, homicide rates very dramatically depending on where you are standing. But the real “elephant in the room” that nobody (especially left-wing socialists) want to admit is that the biggest predictor of violent behavior (in the U.S.) is not income, education, or “opportunity” – no, the best predictors are age, sex, and race, and the single biggest predictor is…race. For example, poor white males have a lower homicide offender rate than black males from middle income families. But THAT is not an issue that they want to delve into to any significant degree.

    • First, you have obviously not spent much time in Latin America. If you had you would know that North and South America cannot be compared. And race is much less of a factor in Latin America than the US. Guns, however, namely those that make their way from the US to Mexico, are a significant problem. I could continue here, however, I have a feeling that no matter what I say you already know better.

      • The private ownership of firearms in Chile is rather high, when considering both legal and illegal ownership. Yet firearms violence and homicide are decidedly low, much lower than can be predicted by mere “firearms availability” and the factors pointing to high drug use and drug trafficking, along with weak conviction and sentencing practices. So there is something else at work that should be honestly explored. And that something else likely has to do with the prevailing culture.

    • You obviously seem to feel that by calling an analysis “Marxist” you have dealt a death blow but you haven´t! What all you reactionary propaganda sufferers fail to realize is that Marx was a genius whose analysis changed world history. He is NOT the devil. You should take a look at the quality of life in Europe to get a clue about the effectiveness of his views! A further point would be that as a propaganda driven American you immediately trot out racial statistics from NORTH AMERICA to counter an argument about LATIN AMERICAN violence. Your slip is showing.

    • your analysis is flawed .. you cite race … and you just proved the point of the author, that inequality and the poverty are the leading factors right here in the good ol’ U.S. of A. … you are typical denier of the facts that make Blacks the leading committer of crime in this country. When you put whites in the same circumstances the crime rate is the same .. nice try on trying to blame race.

    • i thought along those lines so time ago. My family is half latin- and I suppose that explains the in-fighting. But I read (20 years ago) that Spain had the lowest murder rate.

      I think it is best explained by redrawing the map of america. America up top shitting and south central where all the shit goes – guns amo, destabilisation. America pump a trillion dollars into the evil gang networks— but trades in real terms so much less. If America need to get serious about its health as a nation– a nation of sex and drug addicted- then voila it will witness a flourishing democracy of mexico and wonder family life- the bulwark against tyranny.

  • Okay first of all, I disagree with the its not drugs. Hispanic cartels from all over and to the Island control those drugs. Facts.

  • The homicidal rate in Costa Rica is two times higher than in th US. h
    Lack of death penalty, alcohol and drugs consumption are the real triggers of violent crime.

  • The notion of “inequalty breeds homicide” – – while “politically correct” — falls apart when comparisons of comparable countries are made. Compare Argentina and neighbouring Chile, with similar cultures. Chile has high inequality and low Gini numbers, while Argentina has high Gini and lower inequality. Yet Argentina’s urban homicide rate is significantly higher, and its robbery rate is outrageously higher, both compared to Chile. The Peronist socio-political model for creating equality in Argentina has had the effect of increasing both types of crime, lowering worker productivity, and augmenting drug-related violence problems.

    • Estanislao

      An analysis based on a single comparison is inevitably flawed. The study does obviously not apply to every single case.
      And no, Chile’s Gini coefficient is actually higher than Argentina’s. Get a clue.

Leave a Comment