Reality Check: What Kind of Violence Kills Mexicans?

The Foundry, a conservative blog from the Heritage Foundation, contains an account of Mexico as a land of total lawlessness and hopelessness where: "Americans in general face a deadly threat close to home."

Although many readers will doubt the ethical bases of these views, many others will indeed agree that they serve as indications of the image that this country has recently gained overseas and the damage in the international reputation that the war against organized crime has caused.

This article was republished from Geocrimen, the crime and geography blog of Mexico's investigative research institution CIDE. See original post here.

One disclaimer, however, before proceeding: The following discussion is not concerned with proving or disproving the capacity of the Mexican State to provide safety and peace, which is a topic that consistently dominates these kind of discussions, nor is it interested in complaining about the failure of the Calderon strategy against organized crime which is also another dominating topic. The discussion at hand is only interested in informing of what kind of violence actually kills Mexicans. This discussion is about the empirical truth of the last twenty years.

I will explain now the method that informs and guides this (very) brief discussion. Unlike the Foundry, I use valid data to check reality. The Mexican National Institute of Statistics (INEGI) provides recorded mortality data all way back to 1893. These data is divided into four main sections: General, maternal, infant, and accidental and violent deaths. The latter section makes a distinction of three main types of accidental and violent deaths: Accidental, homicide (i.e. intentional homicide) or suicide. Some deaths are recorded as either "ignored cause" or "not specified". The original sources of these data are either the Police and/or the Civil Register (Ministerio Publico and Registro Civil).

SEE ALSO: Mexico News and Profiles

The period of my observation runs from 1990 to 2011, which I believe is the history period most readers will be interested in. In this respect, INEGI reports that a total number of 1.2 million violent deaths occurred in this period of 22 years (See Table 1). It is clear that most violent deaths have been accidents. Nearly 64% of total violent deaths are related to an accident. An "accidental" violent death is defined as any death due to external causes and unintended circumstances such as injuries, poisoning or any other adverse causes. Homicides and suicides cannot obviously be classified as such.

It is convenient now to calculate some basic rates from the data to make comparisons and to discern patterns. First, in the last row of Table 2, I show the arithmetic mean of the rates per type of violent death for the entire period. The historical rate of accidental deaths (38.0) is 2.6 times the rate of deaths by homicide (14.7) and 5.6 times the rate of "other" violent deaths (6.8). Between 1990 and 2007, the rate of violent deaths decreased steadily to become a half of what it was at the beginning of the period (See Chart 1). Year 2004 is the year in the history of this country when we finally entered the one digit rate of homicide. And 2007 was the year with the lowest homicide rate ever recorded (including before 1990). However, a breaking point can be seen in 2008 at which this trend of peace abruptly changed direction. The cause of this change is clear: This breaking point can be associated to the beginning of the war against organized crime. (And this break in the trend is of course statistically significant at p < 0.01. In other words, it was unlikely, and very unexpected by pure chance.)

Chart 1. Rates of violent deaths per type, 1990 to 2011

Data source: Sistema Municipal de Bases de Datos (SIMBAD). Defunciones Accidentales y Violentas. INEGI. Yearly rates are per 100 thousand inhabitants. Own calculations.

Notice two more things. First that the trend of violent deaths by accident has not changed direction: it has been slowly decreasing (almost every year) since 1990. Second, the "other" category of violent deaths, seen over time (again most of these deaths can be associated to suicide), is an entirely different case from the previous two. Here data show an increasing trend (in fact are the increases in suicides what are driving this trend).

Thanks to the data at hand it is possible to understand, even for loyal readers of the Foundry, two really simple facts. First, homicide is not the leading cause of violent death in Mexico. This is not the case, by far. Second, if it had not been for the war on organized crime, ceteris paribus, the homicide rate would have fallen to the 4.7 per 100 thousand inhabitants mark by 2011. (Based on a linear prediction approach.( This is a figure similar to the U.S. homicide rate of 4.8 per 100 thousand in 2012, according to United Nations numbers.

But let me be quite clear too: These data also provides an example of the fragility of peace and progress. It clearly shows too that political decisions (or the absence of) play a key role in explaining how quickly a society in the path of peace can return to levels of violence not seen in many years. Briefly put: Mexican political elites were not successful in maintaining the advances made against violence. In just two years (2008 and 2009) we lost what we had advanced in twenty. But a very different thing is to say that our American neighbors face a deadly threat in Mexico. Not even us living here expect this deadly threat. That argument is intellectually unstructured. It is simply not true when one examines reality.

 Table 1. Total number of violent deaths per type, 1990 to 2011

 Data source: Sistema Municipal de Bases de Datos (SIMBAD). Defunciones Accidentales y Violentas. INEGI.
*After 1995 more than half of these deaths are related to suicide.

 Table 2. Rates of violent deaths per type, 1990 to 2011

Data source: Sistema Municipal de Bases de Datos (SIMBAD). Defunciones Accidentales y Violentas. INEGI. Yearly rates are per 100 thousand inhabitants. Own calculations.
After 1995 more than half of these deaths are related to suicide.

*Carlos Vilalta is a professor at the Centro de Investigacion y Docencia Economicas (CIDE), a Mexico City-based research institution and university. This article was posted on CIDE's blog Geocrimen. See original here.