Has Rio's 'pacification' simply pushed violence to city limits?

Yes, homicides are increasing in the municipalities that surround Rio de Janeiro, but these areas have traditionally witnessed higher rates of violence than the city itself.

• InSight Crime researches, analyzes, and investigates organized crime in the Americas. Opinions are the author's own.

Officials in the Rio de Janeiro metropolitan area are increasingly complaining that the so-called "pacification" of favelas is displacing violence to the city's periphery. The hard evidence for this claim, however, is scarce.

Rio de Janeiro's security strategy has made some impressive gains since the state government first rolled out the pacification program in 2008. In the neighborhoods where Police Pacification Units (UPPs) have been installed, local criminal control has been broken and murder rates have dropped considerably.

While allegations of abuse and the number of residents killed in police crossfire have drawn attention to its shortcomings, pacification is generally held up as a success story. At the same time, however, it has drawn criticism from some who say that rather than solving Rio's crime problem, the program is pushing violent drug gangs outside the city and into surrounding municipalities.

This narrative is commonly invoked in media coverage of crime in the Rio area, with local papers reporting on the apparent "migration" of gangs to the city's periphery.

It has also become popular among officials in the Baixada Fluminense region, a section of the northern Rio metropolitan area that includes smaller cities like Duque de Caxias and Nova Iguaçu. Luiz Antunes, Nova Iguaçu's security secretary, recently told journalists that, "Violence has invaded the Baixada Fluminense."

Duque de Caxias Mayor Alexandre Cardoso has repeatedly referred to an uptick in crime in his city as the result of Rio's pacification. Lindberg Farias, a candidate for Rio state governor in October's elections, has made the same criticism of the UPPs.

The specifics of the argument vary, but one of the most common claims is that the increased police presence of Rio's favelas has pushed gang leaders to relocate to slums outside the city.

Supporters of the idea point to the fact that while violent crime has reduced in the city, in the Baixada it has increased drastically in recent years. In Nova Iguaçu, for example, officials registered 310 murders from January to May 2014, almost twice the 177 seen in the same period in 2008.

InSight Crime Analysis

At least partially in response to this perception, the state government is taking steps to expand the pacification strategy beyond the city borders of Rio de Janeiro. In July, authorities set up the first UPP in the Baixada Fluminense – and the first one outside of the city – in the Duque de Caxias neighborhood of Complexo da Mangueirinha. Another UPP, in the bordering municipality of Belford Roxo, will be established in October.

Current Governor Luiz Fernando Pezão has said that if he emerges victorious from October's election, he intends to open up several new pacification units outside of Rio. On the campaign trail, he has promised to not only establish UPPs in every municipality in the Baixada Fluminense, but also in the neighboring city of Niteroi.

While the media and some officials in the metropolitan area have accepted the claim that pacification has spread violent crime to Rio's surrounding municipalities, the hard proof for it simply isn't there. For one thing, the criminal "migration" theory has been flatly rejected by the office of the Rio de Janeiro State Secretary of Security (SESEG). According to a November 2013 SESEG report, the vast majority of suspects arrested on drug trafficking-related charges in the Baixada from January to August of last year were locals. Just 5.6 percent of suspects were residents of the city of Rio de Janeiro.

What's more, asserting that the violence in the Baixada was "pushed in" ignores the fact that it has never been particularly peaceful in the first place.

In reality, the area has long been home to some of Rio's most infamous criminal groups, like the Red Command (Comando Vermelho) and Pure Third Command (Terceiro Comando Puro). As security expert Ignacio Cano of the Violence Analysis Laboratory told newspaper Extra when asked to comment on the issue: "The displacement would only be meaningful if criminals were going to areas where there weren't any gangs."

The fact is, the municipalities that ring Rio de Janeiro have traditionally witnessed higher rates of violence than the city itself. From 2002 to 2006, before the word "pacification" had even entered Brazil's political discourse, official data shows the average homicide rate in Duque de Caxias was 81.5 per 100,000 residents. Nova Iguaçu's was 64.2, and Belford Roxo's was 55.2. All of these were higher than the average homicide rate in Rio for the same period, which was 44.8.

It is true, however, that homicides are increasing in the municipalities of the metropolitan area (excluding Rio) after a period of decline from 2008 to 2012. Brazilian researcher Joana Monteiro points this out in a very useful analysis of crime statistics released by the Rio de Janeiro Public Security Institute.

Nevertheless, the data doesn't support the idea that crime is shifting to the city's periphery. As Monteiro notes, the number of so-called "crimes against property" (a category which includes extortion, theft and muggings) has consistently been far higher in the city than in the outer municipalities since 2007, though such crimes appear to be on the rise in both.

So if pacification hasn't caused some kind of mass criminal exodus from Rio, why is violence in the Baixada Fluminense on the rise?

The answer likely has to do with the simple lack of police presence. In 2012 the 13 municipalities that make up the region had, between them, just 2,910 military police officers operating there. At the time, this was 20 percent less than the number of police stationed in UPPs in Rio. As R7 reported, this breaks down to a police-to-resident ratio of one officer for every 1,254 people, compared to one to 478 in the city itself.