There are many theories out there about why we have seen a huge uptick in violence in Mexico – now running close to 25,000 homicides a year. An interesting academic paper by Melissa Dell, PhD candidate at the Massachusetts Institute of Technology (MIT), tests one particular theory – elaborated by Eduardo Guerrero among others — that the policies spearheaded by Calderón and the PAN more generally have actually caused the increase in violence. To do so she uses statistical models to examine how PAN victories in close mayoral elections affect violence locally, and whether they have “spillover effects”, causing traffickers to divert their routes to neighboring municipalities.
She finds that when a new PAN mayor comes in after a close election, homicides become 9 percent more likely, and drug traffickers are much more prone to have confrontations with the police. The movement of drugs also shifts to nearby towns — causing an increase in violence there — confirming the so-called cucaracha, or cockroach, effect. Dell argues that government’s policy is behind these statistically significant differences, and specifically that the PAN’s decisions — from top to bottom — to take on drug traffickers more aggressively than other parties is behind the surge.
This rigorous analysis is extremely helpful, and is the type of work that academics should be sharing with policymakers on both sides of the border. Yet we should also be mindful of the limitations. For one, Dell only considers locally produced drugs – marijuana, heroin, meth – leaving out the biggest cash cow, cocaine. Her analysis also exclusively focuses on drugs and not organized criminal groups’ other businesses such as extortion, kidnapping and human trafficking (she does nod to these, but finds no adequate dataset to use). As the business model has changed, so too have the targets, bringing these criminal groups much closer to the general population –as customers and as prey.
This leads to the third limitation – the assumption that “more than 85 percent of the [drug] violence consists of people involved in the drug trade killing each other,” a figure repeated a number of times without any footnotes. Though this has albeen the mantra of the federal government over the last five years, so far neither the Mexican government nor outside sources have provided any proof that this is true. Of the nearly 50,000 drug trade-related deaths since 2006, the Attorney General’s office has investigated less than 1,000 (and solved less than 350). Given the shifting commercial interests of the criminals (bringing them closer to innocent civilians), it seems doubtful that the deaths are still almost all between the gangsters themselves, or that the percentage of bad guys killing bad guys hasn’t changed. Indeed, as a recent Human Rights Watch report points out, there are many cases of misclassification, where the authorities presume that murder victims are linked to drug traffickers until proven otherwise (which they rarely are, since the Attorney General’s office investigates less than 2 percent of the killings). The rise in extrajudicial killings by the military, also laid out in detail by Human Rights Watch, further questions these claims.
Finally Dell makes the assumption – repeated in the press and elsewhere – that drug-related violence picked up with Calderón and his “war against narcotraffickers.” But the data show that the uptick started earlier, under president Fox, increasing some 40 percent from 2004 to 2005, and another 25 percent from 2005-2006. This doesn’t necessarily disqualify a PAN-ista effect (given both Fox and Calderón hail from the same party), but it needs to be explored more, as the security policies of the two differed in some respects.
The paper provides some policy suggestions, particularly regarding how to best use scarce law enforcement resources (for starters, don’t set up roadblocks). But the other more ominous implication is that if drug traffickers are rational economic actors, and PAN victories are so costly for them (in terms of relocating their routes or bringing in competitors), it makes sense for them to invest up front – and buy more local elections. As we head into 2012, all should be worried about this conclusion.