Are we really seeing improvements in crime control like the Mexican government under President Enrique Peña Nieto claims? A favourable combination of conditions for successful crime control includes fewer committed crimes and more reported crimes by citizens, coupled with a more efficient justice system. The good news is that improving Mexico’s justice system would result in more citizens reaching out to the police and, with time, in less crime; the bad news is that we are far from having a fair and transparent justice system.

Authorities point to declines in reported crimes submitted by the 32 public prosecutor’s offices nationwide as evidence that crime is dropping. Notwithstanding the unreliability of these statistics, which will be discussed shortly, a fall in reported crime is not necessarily a good thing. Victims could be reporting less crime as a result of more perceived distrust or fear of authorities. Similarly, people may be reporting less crime as the latter actually increases, a sign that both law enforcement and criminal justice are failing.

There could also be an actual decline in committed crimes, but this can only be estimated from reported crime data and comprehensive survey data on victimization. If the former is unreliable, the latter can bridge the gap between the current state of uncertainty and the desired state of knowledge on crime trends.

Problems with crime data sources

As in numerous other countries troubled by proliferating violent crime, corruption and weak institutions, Mexico’s crime figures are fallible and devious. The most widespread crimes are those prosecuted at the state-level (i.e. common crime), which account for 95% of crimes committed nationwide. However, states are known for their inefficient, crooked, and abusive police forces and judiciaries. Given the scant trust citizens have on their authorities, especially at the state and local levels, some types of common crime often go underreported.

Data on common crime are available online at the National Public Security System (SNSP), a national public security database in which crimes are disaggregated by type (e.g. violent / non-violent theft of households / businesses / vehicles; murder / man slaughter by gun / knife, etc.) and recorded at a state level on a monthly basis since 1997. Municipal-level data is only available since 2011, but is even less reputable. Crime figures submitted to the SNSP represent only reported crimes for which a pre-trial investigation has been opened up.

Although the technical aspects of data manipulation are slowly improving, the major problem with crime data is the data itself, which is unreliable and must therefore be complemented with survey data.

Mexico’s experience with victimisation surveys is mixed. In 2002 the Citizens Institute for the Study of Insecurity (Instituto Ciudadano de Estudios sobre la Inseguridad – ICESI) launched the first all-encompassing study on crime victimisation (Encuesta Nacional sobre la Inseguridad – ENSI). An autonomous and non-governmental organisation, the ICESI was considered the most authoritative source on crime survey data, up until 2010 when the SNSP tasked the national statistics agency (Instituto Nacional de Estadística y Geografía – Inegi) to carry out the seventh ENSI. ICESI saw this as an affront to the ENSI’s key purpose of providing impartial and accurate information on public perceptions of crime.

As of 2011, Inegi launches a yearly study on crime victimisation entitled ENVIPE (Encuesta Nacional de Victimización y Percepción sobre Seguridad Pública). The survey offers both an outlook of public perceptions on various common crimes and an estimate on the number of actual committed crimes for some felonies. The range of crimes covered by the ENVIPE includes fraud, extortion, different types of assault, and a miscellaneous category that bundles up crimes as varied as kidnapping, rape, and exhibitionism. The ENVIPE complements Inegi’s yearly homicide report that, unlike the SNSP, is based on death certificates submitted by civil registry offices around the country.

Yet a major drawback of the ENVIPE is that it fails to capture public perceptions on a wider range of crimes at the state level, as well as any survey data on crime at the municipal level. Moreover, it has yet to provide thorough data on kidnapping, one of the least reported crimes that most affect public perceptions of insecurity.

The ENVIPE is still a valuable tool for filling that knowledge gap on crime trends. But an overview of the survey’s crime data paints a disheartening picture of Mexico’s insecurity environment and an ominous prospect for citizen security.

What has happened in the past three years?

The national homicide rate reached its peak a year before President Peña Nieto took office and has been slowly dropping since. However other types of crime have risen and show no sign of levelling off.

Despite the official communication strategy understating security-related issues, the Peña Nieto administration has stressed that crime is falling whenever in need to publicly defend its security strategy. Its major campaign pledge on security was to cut violent crime, specifically homicide, kidnapping and extortion. But with mass media cutting its coverage on insecurity and other aspects of law and order, the public has been largely devoid of accurate information on crime trends.

So what has happened at the national level? On the most common indicator of violence, homicide, there is relative good news. After registering a record-low national homicide rate of 8 per 100,000 inhabitants in 2007, this figure trebled to 24 by 2011 and began to sluggishly drop in the following two years, with 2013 registering 19. That is still well above the epidemic rate of 10 homicides per 100,000 people.

Moreover, those figures conceal sub-national variations in homicide. In 2011, 23 states registered homicide rates above the epidemic level. Two years later there were 24 states with epidemic homicide rates. While there were indeed improvements in some states like Nuevo León, Coahuila and Chihuahua, homicide levels in those states are still significantly above the epidemic level. The states with the highest homicide rates in the past three years are practically the same; Guerrero toppled Chihuahua in 2013 and now holds the ill-fated position of having the highest homicide rate (63 vs. 59), followed by Sinaloa (42), Morelos (33) and Colima (32).

Beyond homicides, ENVIPE’s overview of overall crime trends between 2011 and 2013 points to a troubling trend. Not only did both the number of committed crimes and victims expand, but also the rate of unreported crime (i.e. dark figure). The national rate of unreported crime expanded from 91.6% in 2011 to 92.1% in 2012 and 93.8% in 2013. In other words, of the 33,090,263 million crimes committed last year, 31,039,713 were not reported to the police.

These grim indicators however are impractical to the extent we cluster all crime under the same category.

A look at extortion

Extortion is one of the most prevalent yet least reported crimes, and is among the most distressing for citizens and businesses alike. The ENVIPE shows that extortion cases have spiked significantly countrywide in the past two years, growing 37% between 2011 and 2012, and 30% between 2012 and 2013.

More worrying still, only 6.8% of extortion cases were reported to the police in 2011, and just 3.4% were turned in for pre-trial investigations. This only worsened in 2012 and 2013, as merely 5% and 3.7% of crimes were reported respectively, while barely 2.2% and 1.5% were sent to pre-trial investigation. Furthermore the proportion of unreported crime for this felony is startling: 96.6% in 2011, 97.8% in 2012, and 98.5% in 2013.

Variations in extortion figures exist among Mexico’s states, too. Considering the three states with the highest homicide rates between 2011 and 2013 – Chihuahua, Guerrero, and Sinaloa – extortion was the most common crime during that three-year period (except in 2011 in Chihuahua, where assault in public transport or in the street was more frequent). In all three states, the extortion rate went up between 2011 and 2012, but fell between 2012 and 2013.

However, unreported extortion is even more staggering than at the national level. In 2011-2012, only Chihuahua saw a drop – albeit insignificant (from 98.8% to 98.1%), while upticks ensued in both Guerrero (from 99.3% to 99.9%) and Sinaloa (from 96.8% to 97.5%). Between 2012 and 2013, Chihuahua experienced the only meaningful fall in unreported extortion (from 98.1% to 92.2%), and Guerrero experienced a paltry decline from 99.9% to 99.5%. Once again, Sinaloa saw a rise from 97.5% to 98.2%.

Now, if we look at the least violent states in terms of homicides – Aguascalientes, Queretaro and Yucatán – the extortion rate is not much different than that in the most violent states, with the exception of Guerrero in 2012 and 2013. In fact, Chihuahua had a smaller extortion rate than Queretaro in 2011 (3,418 vs. 3,882 extortion cases per 100,000 people), when the Chihuahua’s homicide rate of 126 was among the highest worldwide (Queretaro’s was 6).

The rate of extortion for these more peaceful states also followed a similar upward tendency in 2011-2012, and in the next period all but Aguascalientes likewise saw a continuing upward trend. What’s more telling is that the rate of unreported crime is as overwhelming as in the more violent states: the lowest (94.2%) was recorded in Aguascalientes in 2013, while the highest (98.2%) was registered one year before in that same state.

This overall portrait of extortion confirms the worse possible trend we could expect: more committed crimes, less crimes reported to the police, and less reported crimes turning to pre-trial investigations (i.e. only the second link in the long chain of the criminal justice process).

What is the key issue?

The relative good news that homicides are slowly falling nationwide is misleading once we look at the other forms of insecurity that still plagues many parts of Mexico, such as deteriorating developments in extortion, impunity and public perceptions.

What’s worse, the recently uncovered mass graves in the state of Guerrero point to the not-so-surprising fact that the underworld is adapting to changing circumstances. Criminals are abstaining from frequently displaying mutilated corpses since the government communicated that it would prioritize targeting those who resort to extreme violence.

Similarly in the case of El Salvador, where there was a sudden drop in registered homicides following a gang truce in March 2012, the discovery of mass graves revealed that gangs continued to kill while obtaining concessions from the government. Uncovered mass graves in Mexico are also revealing the extent of criminal violence – much of it aided and abetted by the state – in which disappeared victims are plentiful and mostly disregarded.

As we dig deeper into the crime figures in Mexico, what we unearth is mostly more crime and more impunity. Improvements can certainly be observed in some crimes and in some states, but the overall trend is not at all promising.

It is evident there’s an overall collapse of the justice system, ranging from the police to the prosecutors to the courts. A reform replacing the inquisitorial system with a more transparent adversarial system seeks to break this vicious cycle, but its delayed implementation across states shows how difficult it will be to overturn these negative trends.

Instead of asking whether crime figures are up or down, shouldn’t we be asking if our criminal justice system is improving? It’s harder to achieve and measure, but it’s more practical in the sense that it shows citizens a real commitment in addressing insecurity in all its forms. This would in turn boost citizen confidence in authorities and help upend the vicious cycle of impunity that feeds crime in the first place.