I have been in Mexico only twice, in 1969 and in 2002, and the increased fear of crime in my second visit was palpable. On my first visit, I felt as safe as in the United States; on my second visit, although I did not feel quite as constrained as Becker describes his recent visit (maybe the crime situation has worsened in the last four year), there were constant reminders of the crime menace (such as the metal detectors at the entrance to the Four Seasons Hotel in Mexico City) and my hosts insisted on driving my wife and me everywhere. As Becker points out, despite the absence of good statistics there is no doubt that the Mexican crime rate is very high. Moreover, an increase in a crime rate greatly understates the increase in the demand for crime that generates the increased rate. The reason is that a rising crime rate induces defensive measures (alarm systems, private guards, gated communities, etc.) that, though they nominally reduce crime, actually just transform the costs of crime to crime victims into costs of crime avoidance by potential crime victims.
The increase in the Mexican crime rate is mysterious. It is true that income inequality in Mexico has increased, in major part it seems because of liberal economic policies that, as in the United States, have increased the demand for skilled workers by subjecting employers to greater competition, both foreign and domestic. And increased income inequality, even when it is not associated with an increase in poverty, increases the gains from crime by increasing the number and wealth of people worth robbing, kidnapping, and defrauding. Yet it seems unlikely that the increase in income inequality in Mexico has been great enough to explain the increase in crime, especially since poverty in Mexico has declined dramatically since the mid-1990s, though the crime rate seems to have continued to grow--at least it has not declined.
Becker rightly stresses the corruption and incompetence of the Mexican police as an important factor in the crime rate. This too is mysterious. Mexico is not a poor country by international standards. Its per capita GDP calculated on a purchasing power parity basis is slightly over $10,000, which places it in the upper third of the world's nations and ahead of such countries as Bulgaria, Romania, Turkey, and China, and within spitting distance of placid Costa Rica. Furthermore, because Mexico has a large population, that per capita figure translates into a total Gross Domestic Product of more than $750 billion. Mexico can afford an efficient police force in its capital!
In 2003 Rudoph Giuliani was hired by Mexico City to advise on how to reduce the crime rate. He made 146 recommendations, based in part on his successful campaign to reduce crime in New York City when he had been mayor and in part on the obvious need to increase the salaries of Mexico City's police. Mexico City's police chief announced that he had accepted all of Giuliani's recommendations. But there was no implementation. I find that baffling.
Although the underlying cause of Mexico's astronomical crime rate may be income inequality, it doesn't follow that reducing inequality is the the most efficient way to reduce the crime rate. It is a fallacy to think that the only sound solutions to social problems are those that remove the underlying causes of the problems. If crime can be repressed by improved law enforcement at lower cost than by attacking inequality, then improving law enforcement is the superior strategy. Efforts to reduce Mexican income inequality would probably either be totally ineffectual or stifle economic growth. Certainly the reform of law enforcement is the place to start.
So the task for the think tanks is not to study the Mexican (or broader Latin American) crime problem as such; for the problem is obvious, and so is the solution (better law enforcement). What they should study is why the Mexican government and other Latin American governments are incapable of implementing the obvious solution.