Lecture: Studying and Reducing Police Violence in Latin America

The Political Science Department hosted Professor Hernán Flom of Trinity College in Hartford, Connecticut on Wednesday, November 2. Flom presented his preliminary work for an ongoing project in a presentation titled “Police, Politics and Violence in Latin America.”

Flom’s presentation was the very beginnings of a long term project he is starting, in which he is hoping to determine the causes of police violence in Latin America in both cross-national and subnational spheres, as well as changes over time. According to Flom, while Latin America accounts for 13% of the worldwide population, it makes up an astonishing 37% of homicides, and 8 in 10 of the most violent police forces reside in the region. Overall, 10-15% of all homicides in Latin America are committed by law enforcement officers. 

To illustrate his motivation for comparing rates of police violence within the subdivisions of Latin American nations, he showed that the Brazilian state of Rio de Janeiro has not only experienced some of the highest rates of police violence in Brazil, but also the most dramatic changes in violence rates over the decades, compared with the state of São Paulo, whose police violence rates have remained stable over the last 40 or so years. 

Flom attributed this to one major factor: São Paulo has experienced decades of political continuity, resulting in the entrenchment of its successful violence reduction programs and laws. Rio de Janeiro, on the other hand, has changed hands at the political level much more frequently, resulting in wild swings in police policy, and thus dramatic shifts in violence levels over time. Flom argues that this example can be extrapolated to many regions and nations throughout Latin America. Within Brazil, the subnational variations in violence rates are stark, with the Distrito Federal, containing the capital city of Brasília, experiencing the lowest rates of violence, and the poor, far northern state of Amapá home to the highest rates of per-capita violence. Nationwide, Brazil has seen a tripling of police violence between 2012 and 2020. 

Flom also attempted to showcase the stark differences between different countries, many of which are infamous for their rates of crime and violence. According to 2017 data, Colombia and Mexico both saw a rate of 0.3 police killings per 100,000 people, compared with Venezuela, which saw an astonishing rate of 15.9 killings per 100,000. Flom saw Colombia as an interesting outlier, with lower rates of violence despite attributes that typically predict high rates, such as histories of civil war and right-wing governments. Flom noted that Venezuela’s status as an authoritarian nation also placed it somewhat outside the realms of his study.

Flom summarized three primary arguments among the circulating academic literature as to why Latin American nations experience high rates of violence. The first is top-down influence, in which politicians implement repressive policies like “mano dura” under the guise of being tough on crime. The second explanation is bottom-up influence, in which public perceptions of crime, as well as protest against police violence, influence the eventual policies. Finally, Flom noted that there has been comparably little research done examining the effects of internal factors, or the independent policies of the police forces themselves, which often have little political oversight among national and local governments.

Flom summarized his overall argument as such: political fragmentation, added with internal police structures, create the eventual policies and practices that lead to violence. He stressed the importance of also looking at incidents of off-duty violence, in which acts of vigilantism and the formation of militias among police officers creates instances of violence that are not conducted by the state or by law enforcement agencies.

He then suggested a handful of broad reforms that have been proven to have an impact, if implemented properly. There were “marginal” reforms, like mandating body cameras and demilitarization, as well as “structural” reforms, like changing tactics, strategies, the recruitment and promotion process, and changes in leadership. He noted that all of these things take willpower among both politicians and police departments. Politicians have a choice to restrain police violence, tolerate it, or incite it. Police, on the other hand, have the choice to comply with mandated reforms, or deviate, and not enforce the intended changes successfully. Flom again used Rio de Janeiro as an example for this. In the 1980s, the newly elected governor Leonel Brizola manded new reforms, but the execution was so poor, it resulted in a restrain/deviate politician-police relationship. In the 1980s, the “faroeste” policy was an incite/comply relationship. And finally, beginning in the early 2010s, the short lived policy of Pacifying Police Units (UPPs), was a huge success, and the first instance of a restrain/comply relationship.

All of these points helped Flom plan his study. He plans to study violence rates in the Brazilian states of Rio de Janeiro and Minas Gerais, two places with drastically different rates of violence. He plans to study violence in Cali, Colombia, a city with frequent changes in leadership similar to Rio de Janeiro. He also plans to study violence in El Salvador, to determine why violence in that country is so much higher than in Colombia. All three of these cases account for the subnational, international, and temporal factors he introduced at the start of his presentation. To conduct his study, Flom plans to obtain statistics on police violence, interview both political and law enforcement leaders, as well as on-duty officers, and visit both low and high-violence neighborhoods. His research may provide a newly comprehensive view of the cases of police violence in Latin America, and the most successful reform that could be implemented in order to reduce it.    

Notify of

Inline Feedbacks
View all comments

Related Stories


We would love your thoughts, please comment!x
%d bloggers like this: