On the Distributive Costs of Drug-Related Homicides

Nicolas Azjenman, Sebastian Galiani, and Enrique Seira , January 2014.

Violent crime is a serious obstacle to human development and economic growth in many nations throughout the world. It can have dramatic consequences for life expectancy, social capital, and human capital.  In Mexico, a country in which violent crime is relatively prevalent, some have suggested that its economic costs may be as high as 1 to 2 percent of GDP per year. Using more rigorous methods, Azjenman, Galiani, and Seira study the effect of drug-related homicides on house prices in Mexico.  Their finding of a large negative effect that is borne predominantly by poor households implies that lowering the homicide rate would substantially increase the well-being of the least well off in affected communities.

Although violent crime has been a longstanding problem in Mexico, the number of drug-related homicides rose sharply after 2006. This increase in homicides is believed to have been  spurred by President Calderón's declaration of a war on drugs and the ensuing fragmentation of many of the major drug cartels, who then began to fight among themselves.  Although widespread, the timing and the magnitude of the increase in drug-related violence varied across areas for reasons largely unrelated to local economic conditions. 

Using detailed, community-level statistics on drug-related homicides together with similarly detailed data on house prices gathered from over 1.3 million home appraisals connected to mortgage applications during the 2008 to 2011 period, Azjenman et al study the effect of the sharp increase in drug-related homicides after 2006 on home values. They find that the increase in violence had significant effects on house prices, but only for low quality housing. More specifically, they find that a one standard deviation increase in homicides caused a 3% to 4% decrease in the price of poor quality homes, with persistent increases in violence being associated with a 36% greater decrease in value. This implies that the burden of increasing crime on home prices was borne primarily by the poor in Mexico and thus that the growth in crime had a regressive redistributive effect. Given that houses represent a very large share of the assets possessed by poor families, this 3-4% decrease in value amounted to a substantial loss of wealth. Azjenman et al estimate that the poor in Mexico would be willing to pay approximately 3.2% of their annual household income to roll back the increase in violent crime.  To put that figure in context, it is a sum that would total 11% of the cost of Oportunidades, Mexico's largest social program to help poor households.

The effect of violent crime on house prices thus is another channel through which violent crime disproportionately affects the poor. Policy makers in the unfortunate situation of facing high violent crime rates should be aware of how particularly detrimental crime is for low-income households. As such, it is important to remember that, in addition to the direct benefits of reducing the homicide rate, policies that reduce violent crime also may double as anti-poverty policy.

Links to Researchers