The annual race to chart Mexico’s homicide rates has begun, but an examination of the claim that the drug-related killings have levelled off shows the inadequacy of this measure.
Some news outlets have kicked off by reporting that drug-related homicides in Mexico may have plateaued. However, tracing the trail of such data shows the futility of trying to use this metric to understand Mexico’s crime problem and ignores the more important question of how to lower the actual murder rate.
Homicide stats in Mexico are a tricky and politically-charged business. Media, analysts and politicians play around with big numbers at their own peril. The most recent example was Alma Guillermoprieto, whose New York Review of Books blog, “Day of the 40,000 Dead,” drew ire from Mexico watchers.
“I would point out that continuing to repeat the 40,000 number is wrong,” wrote Molly Molloy, who compiles the influential and very informative Frontera listserv. “The Mexican government’s official tally of ‘drug war homicides’ [as they define it] was 34,612 at the end of 2010. To think that only 5,388 people have been added to that toll in nearly 11 months is ludicrous.”
The problem begins with the various sources used to determine drug-related homicides. In her post, Molloy cites several different government sources.
But other sources are more like a breadcrumb trail. The recent reports, for example, about murders reaching a plateau come from Sign On San Diego and the Latin American Herald Tribune. They used preliminary data from Trans-Border Institute (TBI) that suggests drug-related murders are increasing at a slower rate than in previous years.
However, in an email exchange with InSight Crime, a TBI representative explained that its 2011 mid-year data that led to the articles is based on La Reforma newspaper’s “Ejecutometro,” a running tally of publicly reported killings.
TBI is also comparing this newspaper’s data with government data that shows a 2010 end-of-year figure of 15,273 drug-related murders.
In sum, the recent news reports are based on a combination of newspaper and government data that is channeled through a university but is both preliminary and sketchy (as noted by the same university that is gathering these tallies).
To be sure, news organizations are filling an important hole in this debate. As the murder rate related to organized crime rose in 2007, the government did not provide statistics, which left news organizations scrambling to quantify the violence.
But as the homicides have increased, so have the differences in news organizations’ tallies. Zeta magazine, for example, estimated that there were 19,546 organized crime-related killings in 2010, much higher than the estimates of La Reforma (11,583) or the government. Expect similar differences between media again this year.
There are also wide discrepancies within the government’s own data. Last year, as InSight Crime noted, the Attorney General’s Office released data that was very different from the presidency’s estimates of drug-related murders.
At the heart of these inconsistencies is a difference in methodology. Zeta uses the number of intentional homicides committed by firearm but also includes data on intentional homicide by “other” or “unknown” means, which would include the many victims of violence that have been beheaded, disemboweled, hung from bridges or found buried in mass graves, but not shot by a firearm.
La Reforma’s classification depends on the type of weapon and the way it was employed (e.g., execution style); how many dead; whether there are markings on the body or if it is mutilated; whether there are signs or symbols near the dead; official reports indicating the connection to organized crime; the presence of weapons and/or high volumes of cash at the crime scene.
The presidency uses three categories (download methodology here): executions, confrontations, and aggressions. The vast majority of these murders are classified as executions, which the presidency says have certain characteristics such as multiple victims, a message from a rival drug trafficking organization, and signs of execution-style killings. But even this is somewhat arbitrary.
As Molloy points out in her post, the only part of this equation that seems clear is that the total number of homicides have increased. Classifying these homicides without proper judicial investigations and records is pure guesswork.
A bigger problem may be the use of these statistics as the main barometer of criminal activity in Mexico. Indeed, the optimistic assessment from these media sources also failed to consider other forms of violent crime and ignored how violence has spread to new areas, making some parts of Mexico more dangerous than ever.
Homicides are obviously not the only form of drug related violence in Mexico. President Felipe Calderon’s approach to fighting drugs has changed the dynamics of organized crime, fragmenting the cartels. The new, smaller groups are increasingly involved in other criminal activities, like extortion and kidnapping, and have pushed crime into new parts of Mexico.
According to the Citizen’s National Observatory (ONC), the number of kidnappings, extortions and violent robberies is on the rise in 2011. The increase in these high-impact crimes has been felt across Mexico, but particularly in areas that saw very little violence until recently.
The worst part of this debate may be its futility. Media, analysts, academics and government institutions should spend more time trying to understand what causes violence to fluctuate.
They could measure, for instance, the cumulative effect of years of international investment in the security sector and domestic funding for federal and state police forces, and whether this has increased Mexico’s police investigative capacity.
They could assess whether the implementation of major judicial reforms is improving the ability of prosecutors to convict criminals, using forensic evidence gathered by police investigators, and thus reducing Mexico’s high rate of impunity for criminals.
They could also explore questions that have caused controversy in the United States when assessing its plummeting homicide rates, namely, incarceration rates and demographics. Specifically, Steven Levitt, of Freakonomics fame, argues that four factors account for the dramatic and sustained decrease in violence in the U.S. over the past several decades (pdf version of Levitt’s 2004 article available here).
The declining popularity of crack cocaine, the legalization of abortion in the 1970s, the increase in the number of police officers on U.S. streets and the enormous growth in prison populations due to much harsher penalties for drug offenses and mandatory minimum sentencing.
It would appear that Mexico is applying this logic to its own problems, at least in part. Mexico’s armed forces and police have arrested thousands of suspected criminals in recent years, and jails are close to 20 percent over capacity (pdf report on prison population available here).
However, two things are worth pointing out here. First, there is a great deal of debate over the social costs of mass imprisonment and whether it is a cost-effective and sustainable response to high levels of crime. Second, because of an underresourced and overwhelmed judicial system, only 42 percent of the nearly 220,000 people incarcerated in Mexico have even been sentenced for a criminal act, which suggests that, even if the logic behind mass incarceration was correct, it may not work for Mexico.
So, just as mass imprisonment in the U.S. may have hastened the end of the crack-related crime wave of the 1980s, Mexico watchers should ask themselves if incarcerating thousands of Mexico’s young men can lower levels of violent crime over the next decade. Using preliminary data that is both questionable and varies from source to source, the answer appears to be no.