Effectiveness versus legitimacy: Use of force and police training in Latin America

Recent news of police misconduct in the United States has led to a revival of old tensions between effective policing services and public trust in the institution. It is thought that effective policing requires being tough, but tough often means compromising civilian protection. In fact, President Obama recently convened a special task force for police reform, seeking to strike the right balance.[1] In Latin America, on the other hand, with a few exceptions, this debate about effectiveness versus legitimacy is non-existent—though needed.

Recent surveys show very low levels of trust in the police in most countries in the region. According to a 2014 report by CAF (Development Bank of Latin America), less than 20 percent of the population trusts the police in Buenos Aires, less than 10 percent in Lima and La Paz, and around 15 percent in Caracas. These numbers stand in contrast with the 65 percent of the population reporting to trust the police in the United Kingdom (IPSOS-Mori) and the 53 percent trusting the police in the U.S. (Gallup)—even amid its recent misconduct scandals.

Some might suggest this is not a problem insofar as crime is kept in check and “law-abiding” citizens feel safe. The problem is that weak legitimacy of public institutions makes this harder: i) crime reporting suffers, so authorities don’t have good crime diagnostics; ii) citizens stop demanding that the police to do a good job (weak accountability), which hurts performance; and, much worse, iii) criminal organizations may replace the state/police in providing security services, as we’ve seen in Sao Paulo with the Primeiro Comando da Capita or in Medellin during the era of Pablo Escobar. In addition, weak legitimacy boosts private sector—non-criminal—provision of security, which means people spend money on services for which they’ve already paid taxes and that those who can’t afford it are excluded from protection, which reinforces other perverse exclusion mechanisms.

The most pressing citizen security problem facing Latin America is the weak legitimacy of its criminal justice system, including the prosecution, judiciary, penitentiary, and especially the police. Indeed, while the best place to begin building trust and legitimacy in the system is not readily apparent, police training would likely be part of any formula. The role of the police should be to guarantee a safe, democratic society (where individual rights are respected), and achieving that goal requires balancing effectiveness with legitimacy. The extent to which these goals permeate the curriculum of police academies across the region will, to a degree, determine the commitment of street patrol police to not only seek and implement the most effective crime control and prevention strategies, but to do it in a way that increases, not lessens, their legitimacy as providers of security for all members of society.

Excessive use of force by the police and their involvement in unnecessary shootings may undermine people’s trust in the police, so Argentina’s experience with police training in the rational use of force is important to this discussion in all of Latin America. Beginning in April 2013, the Ministry of Security in Argentina implemented a 45-hour (5-day) training program (Centro de Entrenamiento y Doctrina Policial, or CEDOP) on the rational use of force and firearms, directed towards their most junior patrol officers from the Argentinean Federal Police. The program sought to promote the critical assessment of potentially violent situations that a police officer might face and for him to visualize the alternative use-of-force responses and choose the best one. The program accomplished this goal through the reinforcement of verbal and physical techniques for handling crisis situations as well as the rational and progressive use of force. It also provided training in conflict resolution, basic legal information, and first aid.

An impact evaluation of this training program (study is in Spanish) was undertaken as part of collaboration between the Security Ministry and CAF. Training was prioritized for police officers of the lowest rank—agents—but who were up for promotion, from all dependencies across the city of Buenos Aires. There were 2,913 eligible agents as of April 2013, and they were all expected to attend the program in alphabetical order in groups of approximately 40 per week. By March 2014, 1,170 agents had been called to the program—last names A through K. Since it is unlikely that the first letter of a person’s last name is related to his probability of being involved in a firearm event, this assignment was as good as random, so by comparing the incidence of events between April 2013 and March 2014 for those assigned to training in that year with those who would be assigned from April 2014 onwards (control group), we were able to estimate the causal impact of the training assignment.

Using administrative records of police involvement in firearm events, we found that treated agents were 8 percent less likely to be involved in a shooting compared to non-trained agents. However, this difference rose to 72.8 percent when we focused on non-operative dependencies, essentially agents who do not have regular exposure to confrontations with suspects and who are mostly dedicated to administrative work. Just over 67 percent of all firearm events in the metropolitan area of Buenos Aires occur while officers are off duty and outside their jurisdiction (near their homes), and more than half report that it occurred while attempting to avoid being robbed. A survey designed to elicit truthful responses to sensitive questions (“list experiment” method) revealed that thanks to the training program, officers felt more aware of the risks around them and increased their perceived valuation of verbal communication with a suspect. It seems that the program helps agents with less “street experience” handle themselves better when facing a suspect or a potentially perilous situation when off duty. This finding should be related to the way police handles risk situations while on duty as well, although we lack the data to show this precisely.

How can Latin America begin to move in the direction of a more professional, effective, and legitimate police service? Although this is but one very specific training program and evaluation, it does provide promise for other similar training efforts, since potential impacts on the number of firearm events can be large and materialize in a relatively short period of time. Training efforts within police academies and special programs for active police might be a stepping stone for more significant reform efforts that may seek not only to put the right incentives in place to reduce violence without compromising effectiveness, but also to ultimately end a culture of violence that erodes the democratic foundations of our society. The way forward, however, needs to be informed by science. Rigorous impact evaluation of pilot initiatives can provide a stronger hand than the lead-weight of larger budgets.

[1] See the discussion in Lum and Nagin (2015) “Reinventing American Policing: A Seven-Point Blueprint for the 21st Century”. George Mason and Carnegie Mellon Universities