Seven years ago, I drove around Mexico’s most dangerous towns. In January of 2018, I may need to do so again, albeit for different reasons. So the topic of violence is on my mind, even if Coyoacán is still as safe as ever.
After that trip, I noted that Mexico spent shockingly little on law and order, defined broadly. Total spending on police, courts and defense came to only 1.1% of GDP, less than a third of the level in the United States. Seven years later, as violence once more swings upwards, Mexico spends ... all of 1.4% of GDP on law and order.
Spending data come from page 20 of this new report; figures divided by GDP from here.
The report digs in a little further with international comparisons. Consider how much Mexico spends on police (seguridad interior, as opposed to justicia or seguridad nacional.) Well, 1.4% of all federal spending goes for that ... compared 4.7% for most OECD nations and 6.2% in Italy ... and Italy, unlike Mexico, is a country which has managed to contain its serious problem with organized crime.
Since I was not sure how Ethos made its calculations, I dug up the OECD numbers on general government expenditure by function. (General government includes subnational governments.) For comparability, I converted everything into a percentage of GDP.
What has the money been spent on? The number of federal police has almost quadrupled between 2006 and 2015, to 43,724 from 11,663. Real average salaries only rose 9% between 2010 and 2015. (Page 34, deflated by the INPC.) You might wonder how such a mighty expansion in payroll was sustained on such a small increase in spending, but note that the federal police payroll takes up only 5% of all security spending.
So more cops, but becoming a cop did not become more attractive (quite the reverse, given the rising danger) and the quality of police officers did not rise. Now, to be fair, federal federal police officers are well-paid by Mexican standards; approximately US$15,400 per year. But state and local police are paid much less: US$7,800 on average. The highest paid are in Sinaloa, reaching barely US$10,900.
And by other measures, the Mexican criminal justice system is a mess. The U.N. Office on Drugs and Crime (UNODC) collects figures on prosecutors and judges. Per 100,000 people, Mexico had 6.8 prosecutors and 4.2 judges in 2013. Call it 11 per 100,000. That same year, crime-ridden Scotland mustered 9.9 prosecutors and 4.9 judges ... 15 per 100,000. And Italy enjoyed only 3.1 prosecutors but 17.1 judges ... 20 per 100,000. (In the Italian system prosecutors are considered judges; it is not clear how UNODC made the distinction.)
Combine that with badly-trained cops, and you have a mess. 85% of all criminal cases in Mexico were commited en flagrancia; meaning that the cops caught the perp while he or she was committing the crime. And in the rest, most cases have been resolved by beating confessions out of whomever the police pick up. In theory, new reforms have made confessions inadmissible save in the presence of a defense attorney (page 20). In practice, it is not clear that anything has changed, especially given as the ability of the police to investigate crimes is practically nil and bribery still reigns supreme.
I still tend to think that Mexico can solve its crime problem. I also tend to think that while throwing money at it is not a sufficient condition for doing so, it is a necessary one. 2.6% of GDP, like Colombia? Five percent, like nowhere? I do not know. But more seems clearly necessary. Am I wrong?
Estimate police elasticity! Chalfin and McCrary claim that U.S. cities are underpoliced.
Posted by: Leticia Arroyo Abad | August 12, 2017 at 08:00 PM
Is the Mexican data good enough to replicate their cost of crime variable?
By God, that's an elegant old-school paper.
Posted by: Noel Maurer | August 12, 2017 at 08:38 PM