Mexico’s Unreliable Crime Data Calls Govt’s Claims Into Question

A new report indicates that one of Mexico’s principal systems for collecting crime data is riddled with errors, raising questions about the reliability of the conclusions regarding the nation’s recent drop in murders.

As reported by Animal Politico, during an 18-month period in 2013 and 2014, more than 70 percent of Mexico’s 32 states altered local crime data after it had already been submitted to the National Public Security System (the SNSP). 

In some cases, the changes in the data after its initial submission were enormous. Veracruz neglected to submit 299 murders committed inside the state during 2013, out of a total of 863 ultimately reported. It also retroactively added 360 cases of extortion, out of a total of just 461. 

In 2014 Michoacan erased nearly 3,000 crimes from its records after having already submitted them to the SNSP. This included nearly 400 cases of murder and manslaughter, and scores of extortion and kidnapping cases. The state’s explanation was that the crimes had been “reclassified,” but it has offered no additional details. Similar examples abound.

The changes were made without much fanfare — there is a page within the SNSP site that lists modifications for each state — and the reasons for them remain unclear.

The SNSP, which relies on the state governments to record the instances of a wide range of crimes, is one of the only resources for comprehensive and recent information on Mexican crime rates, and its data is widely used as a barometer of Mexican security. The fact that the data was so frequently and drastically mistaken suggests that any conclusions based on the SNSP should be taken with a grain of salt. 

Such errors and uncertainty over Mexican government data are unfortunately quite common, even beyond the time frame under consideration in the Animal Politico report. Diego Valle-Jones, a statistician who posts drug war analysis on his blog, has highlighted several inexplicable lapses in government crime reporting in recent years, whether in health records in the southern state of Chiapas or the more than 1,000 homicides missing from Chihuahua’s records. 

The mistakes don’t appear to be anything more sinister than bureaucratic errors, but it does reflect a grave shortcoming for Mexican authorities. The fact that such dramatic mistakes could be made in the collection of basic crime stats, and that the alterations have been made without a full accounting of the source of the errors, calls into question the depth of the security gains made during the nearly three years of Enrique Peña Nieto’s presidency. 

From 2012 to 2014, the SNSP’s tally of murders nationally dropped from 21,736 to 15,653. This nearly 30 percent decline has become a foundation stone of Peña Nieto’s claims to have turned the page on the chaos of the Felipe Calderon era, but if the initial reporting errors are hiding much larger lapses that have not been corrected, the improvement may have actually been much milder. 

SEE ALSO: Mexico News and Profiles

The inability to reliably tabulate murders, where there is typically ample and indisputable evidence of a crime, raises even greater doubts about the statistics on crimes where uncertainty is built in, such as extortion and kidnapping. Such crimes, when successfully executed, leave no evidence of their existence for authorities, and both victim or perpetrator have a powerful incentive to not report them. That natural barrier to reliable reporting of kidnapping and extortion, coupled with the errors outlined above, suggests that the SNSP figures on both of these crimes are little better than worthless. 

The statistical errors are symbolic of a larger problem plaguing Mexico, in that they suggest the civil service is not fully competent at multiple levels of government. A reliable cohort of government employees at every level, from the agency heads to the low-level staffers responsible for implementing a given policy, are vital for any successful government strategy, including in security matters. Counting crimes is hardly the most complicated element of Mexico’s security problems, and a state that cannot be relied upon to arrive at a trustworthy sum is unlikely to be capable of the far more complicated tasks affecting security, such as building self-perpetuating mechanisms to root out corruption and tackle the anarchic state of the prison system. 

As a result, strategies that aim to create an idealized Mexico without addressing the basic competency issues at all levels of government are unlikely to succeed.