“Bad with it, worse without it”: the will for continuity of the UPPs (pacifying police units) beyond the Olympics

Abstract

This article uses the literature on community policing projects and logistic regression models to identify which variables explain the wish for the UPPs (pacifying police units) continuity after the Olympics. Analysis of a household and probabilistic survey carried out with 2,000 residents was conducted. In the first round, the survey was applied to 10 favelas that received UPPs in the initial phase of the program’s implementation and, in the next round, in another 10 favelas when the “pacification” process already indicated signs of crisis. The results indicate that young black and pardo individuals present a high degree of rejection to this mode of policing and that the police’s ability to rebuild the legitimacy of their actions in the communities causes individuals to endorse the continuity of UPPs, and the Olympic city project.

Keywords:
UPPs; community policing; Rio de Janeiro; legitimacy; public safety policy

Resumo

Neste trabalho utilizamos a literatura de avaliação de projetos de policiamento comunitário e modelos de regressão logística para identificar quais variáveis explicam o desejo de continuidade das UPPs após as Olimpíadas. Neste sentido, apresentamos a análise de um survey domiciliar e probabilístico realizado com 2 mil residentes. Na primeira rodada, aplicamos o survey em 10 favelas que receberam UPPs na fase inicial de implementação do programa e, na rodada seguinte, em outras 10 favelas quando o processo de “pacificação” já indicava sinais de crise. Os resultados indicam que jovens pretos e pardos apresentam elevado grau de rechaço a essa modalidade de policiamento e, por outro lado, que a capacidade da polícia em reconstruir a legitimidade de sua ação nas comunidades faz com que os indivíduos endossem a continuidade da UPP, desejando-a para além do projeto de cidade olímpica.

Palavras-chave:
UPP; policiamento comunitário; Rio de Janeiro; legitimidade; política de segurança pública.

Resumen

En este trabajo utilizamos la literatura de evaluación de proyectos de policiamiento comunitario y modelos de regresión logística para identificar qué variables explican el deseo de continuidad de las UPPs después de las Olimpiadas. En este sentido, presentamos el análisis de un survey domiciliar y probabilístico realizado con 2 mil residentes. En la primera ronda, aplicamos el survey en 10 favelas que recibieron UPPs en la fase inicial de implementación del programa y, en la ronda siguiente, en otras 10 favelas cuando el proceso de “pacificación” ya indicaba signos de crisis. Los resultados indican que jóvenes negros y pardos presentan un alto grado de rechazo a esta modalidad de policía y, por otro lado, que la capacidad de la policía para reconstruir la legitimidad de su acción en las comunidades hace que los individuos endosen la continuidad de la UPP, deseando más allá del proyecto de ciudad olímpica.

Palabras clave:
UPP; policía comunitaria; Rio de Janeiro; legitimidad; política de seguridad pública.

1. INTRODUCTION

During their existence as public policy, Pacifying Police Units (UPPs) in Rio de Janeiro experienced two very specific moments. Between 2008 and 2012 they soared with great approval from a good part of the population.¹ This turned sour when the population began to question the correctness of the police procedures, something that began after the disappearance of the mason Amarildo in 2013 and culminated in news reports that contributed to the end UPPs in 2017.²

As a proposal of public policy focused on the provision of security in territories traditionally marked by violence, the UPPs started to be implemented in the city of Rio de Janeiro in 2008. Initially presented as a new model of community policing (Riccio et al., 2013), which innovated by the connection between security and social services provision (Costa and Castilho, 2011), although under an essentially militarized logic (Fleury, 2012), the UPPs became an object of interest to sociologists, anthropologists and public managements. A Google Academic search indicates that the term “Pacifying Police Unit” indexes a myriad of articles in journals and specialized magazines.³

It is common in these studies the use of qualitative methods to identify the demands, problems and expectations surrounding the establishment, institutionalization and, in the last years, the possibility of extinction of the program. The few existing quantitative studies seek to correlate the arrival of UPPs with the reduction of crime rates in pacified communities (Oosterbaan and Van Wijk, 2015; Cano, Borges and Ribeiro, 2012); to measure the impact of policies on reducing police violence (Magaloni, Franco and Melo, 2015) and on the reorganization of the drug market (Cano and Ribeiro, 2016); and to understand what police think about this new format of security provision in favela areas (Soares et al., 2010; Musumeci et al., 2013). Almost no quantitative work is intended to understand how the residents of pacified areas evaluate the policy (Oliveira et al., 2014) and what their expectations for the project after the Olympics would be (Musumeci, 2017), since the UPPs have always been directly associated to mega-events based in the city (Ribeiro and Santos Junior, 2013).

This paper is based on the understanding that an indirect way of assessing whether the UPPs achieved the intended objectives is to verify whether or not the beneficiaries of this policy want it to be continued. The importance of this study for the contemporary debate on the effectiveness of the UPPs lies in its capacity to produce results that can be generalized, taking as a starting point the findings and the hypotheses constructed by previous qualitative studies — with an interpretive scope restricted to certain contexts. Therefore, this study innovates as it is based on a sample and domiciliary survey with residents of 20 Rio de Janeiro favelas, allowing to verify, from a statistical point of view, which contextual characteristics are more related to individuals’ support for the continuity of this new format of community policing.

2. UPP: THE THEORY AND THE PRACTICE OF COMMUNITY POLICING

The UPPs were established at the end of 2008, when a troop of the Rio de Janeiro State Military Police occupied the Santa Marta favela to reduce the lethality of a dispute between local drug traffickers (Rodrigues, 2014). The results of this intervention were so enlightening that the experience was initially extended to four other favelas in the city of Rio de Janeiro. Then, as a consequence of the choice for the city to host the World Cup and the Olympics, the experience of UPPs was replicated in 38areasby 2017 (figure 1).

The UPP project was an innovation for creating “a new situational configuration in which the police permanently occupy the locality” (Dias, 2016:16), from the diagnosis that the strong resistance to police action was expressed in the episodic incursions of the police in favela areas (Albernaz et al., 2017), and also in the corruption engendered because of the police forces’ absence in the territory (Arias and Barnes, 2016). For this reason, UPPs were seen as a public policy that reinforced “the discourse of the ‘absence of the State’ in the favelas as a cause of urban crime in Rio de Janeiro” (Dias, 2016:17).⁴

Figure 1
Distribuition of areas with UPP

The expectation that the permanent presence of the police, in areas marked by confrontations between the State and other armed groups, could change the scenario in terms of guarantees of rights to the population, meant that in the first years of the project, citizens of the favela and citizens of the ‘asphalt’ supported the establishment of UPPs (Cano, Borges and Ribeiro, 2012).⁵ Evidence of this expectation is the State Presence Perception Index (Ippe) of the Fundação Getulio Vargas (FGV), calculated in 2011, when 90% of the population living in the city supported the installation of the UPP in the favela Complexo do Alemão and 56% of residents in this area reinforced such endorsement.⁶ The high approval of the project was evident from the first opinion polls (Oliveira et al., 2014).

Between 2008 and 2011, the UPP was consolidated as a practice, and its implementation followed the same routine: announcement in the press that the UPP would be deployed in a certain place, occupation of the favela by the Special Operations Unit (Bope) and then installation of police officers who would be involved in community policing activity (Rodrigues, 2014). Among favela residents, mention of the use of the “community policing” to identify this project meant that “the police had come to stay” (Menezes, 2014).

The UPP was institutionalized as a public policy a few years after it began as a practice, based on the regulations of the Decree-Law no. 42787, of 6 January 2011 (Cano, Borges and Ribeiro, 2012), reformulated by Decree no. 45186 of 17 March 2015. This legislation established the nuances of this new form of public security provision: (i) emphasis on the resumption of territorial control of the favelas, by reducing the armed power of drug traffickers, even if this did not mean elimination of drug trafficking (Cano and Ribeiro, 2016); (ii) exclusive use of police officers in front-line activities, to avoid police corruption (Corrêa et al., 2016); and (iii) a concern to increase security and guarantee peace through the reduction of shootings and lethal violence (Magaloni, Franco and Melo, 2015).

In order to carry out the activities, the philosophy adopted was the community policing (Riccio et al., 2013), a name given to the different strategies implemented by US police forces for decades in order to rebuild their legitimacy in poor communities (Skogan, 2008). Community policing is understood as a police reform strategy that seeks to replace the traditional approach that the police officer only acts after the crime occurs (to register or repress it) by a more proactive practice aimed at the presence of a police officer for community action in the area, concerned with understanding how the dynamics of sociability contributes to criminal actions (Weisburg and Neyround, 2011). From this knowledge, with the support of those who live in the area, the police officer tries to intervene in the context, with a view to improving the quality of life of all citizens (Eck and Maguire, 2000).

In the US, the focus of community policing programs was on neighborhoods with concentrated disadvantages, based on an attempt to reconstitute the legitimacy of police practice, which is always seen as violent, brutal and inadequate (Braga, 2015). Based on the idea of building a new form of interaction between police and society, through respect for human rights and less use of violence (Skogan, 2008), this program was implemented by police officers who would take care of preventing the occurrence of the crime, guaranteeing respect and security to the local population (Beato, 2008). Therefore, the adoption of community policing was a commitment to build a new type of interaction between police and society, which would lead to greater legitimacy of the police institution (Weisburg and Neyround, 2011).

Legitimacy, according to Zanetic (2017:5), is a word related to the correctness of the procedures of the police and, therefore, depends on three main dimensions: (i) form and adequacy of the procedures adopted (which means respect for the local population); (ii) quality of decisions made, measured by effectiveness indicators (in terms of crime reduction and increased perception of security); and (iii) quality of interpersonal treatment (implying non-victimization of the individual by the police). Therefore, a successful community policing program has legitimacy only if it can guarantee the correctness of police procedures in the eyes of those who interact with the police.

Community policing programs became very popular in Latin America during the redemocratization between the 1980s and 1990s, when police institutions sought to rebuild their legitimacy by leaving behind the notion of a police that serves the government and adopting the notion of a police that serves the citizens, as a citizenship institution (Dammert, 2007). It was in this period of transition that models capable of helping the police to focus on the citizens well-being, especially with regard to guaranteeing their civil rights, were imported (Frühling, 2007). In an attempt to bring about this transmutation demanded by regime change, these institutions resorted to the model of community policing, whose implementation attempts occurred between the decades of 1990 and 2000. The failures in the true institutionalization of a community policing program have made this category rhetoric (Carrión, Pontón and Armijos, 2009). Analyses of the meanings of community policing indicate that, despite the discourse of democratic legitimacy, some police practices were far from being considered strategies for community approximation (Dammert, 2007).

In the specific case of UPPs, some elements allow them to be considered as community policing while other elements lead to deny this idea peremptorily. Perhaps the most defining element of this distance between the meaning and practice of community policing is the purpose of policy: the resumption of favela territory, dominated by drug traffickers (Vaz, 2014), to the detriment of increasing the legitimacy of police action through approximation with the community. It is, therefore, a form of police action that has an eminently repressive dimension, even if this does not mean reducing drug trafficking (Cano and Ribeiro, 2016).

On the other hand, there is an approximation between the concept of community policing and the practice of UPPs, due to the continuous presence of the police in the favela in order to change the sociability and contribute to the reduction of shootings (Rodrigues, 2014). The basic idea of community policing is that police must act before crime occurs and must therefore always be present in areas with high rates of violence (Riccio et al., 2013), mediating conflicts which may result in violent crimes (Mourão and Strozemberg, 2016) and preventing the occurrence of other crimes from the deterrent effect of their permanent presence (Beato, 2008).

Therefore, UPP as a theory of community policing should promote a complete reform of the Military Police of the State of Rio de Janeiro (PMERJ), reconstituting the bases of its legitimacy in favela areas, substantially reducing crime and violence rates. The UPP as a practice of community policing means the permanent occupation of some of the favelas of Rio de Janeiro, changing the dynamics of sociability and reducing, even temporarily, the shootings that cause panic to local population.⁷ Due to the dual nature of the implementation process of UPPs, as a theory and practice of community policing, the purpose of this article is to unveil the determinants of favela residents’ desire for the continuity of this public policy.

3. WHY SHOULD WE LISTEN TO THE POPULATION LIVING IN AREAS WITH UPP TO UNDERSTAND PACIFICATION?

The evaluation of a social program has the purpose of determining if the objective of a governmental policy was achieved by using the planned means. Examining a social program means establishing the cause-effect relation between the government intervention and the results achieved, considering in the process the profile of the beneficiaries, the changes occurred in addition to those planned by the government and the perceptions of beneficiaries regarding the effects of the public policy. Assessing UPP as a public policy through the beneficiaries’ perception is a way of verifying the extent to which and why citizens have become attached to this new form of governmental action.

In Latin America, in order to understand if the implemented community policing programs comply with what they promise, surveys were conducted in order to listen to what citizens considered the police did and how (Frühling, 2007), as well as to measure perceptions about correctness of police procedures, expectations and changes in the lives of residents (Mohor, 2007). In addition, listening to the community would be a way of forcing the police to position themselves in the face of the needs of this public and to review possible courses of action that impact the legitimacy of the policy, which could contribute to the improvement of police practice (Pinto and Garay, 2014).

In the case of UPPs, community policing would be the philosophy that support front-line police action (Riccio et al., 2013). Although the evaluations comparing crime rates and police violence before and after the introduction of the UPP indicated that this program was effective — at least for a while — in reducing these indicators (Cano, Borges and Ribeiro, 2012; Melo, 2015), few studies have attempted to understand whether the UPP was efficient in improving the relationship between police and community (Oliveira et al., 2014).⁸ This scenario contrasts with the various studies that sought to understand, from an ethnographic perspective, how the interactions between these two publics (police and community) were built and rebuilt, resulting in approximations and distancing (Pacheco de Oliveira, 2014).

In this context, far from advocating for the continuity of the UPPs, this article intends to present the determinants of the desire for continuity of this policy by the population living in areas served by the policy. The aim is to understand the policy of pacification under the logic of its users.⁹ We want to point out that, despite the criticisms presented by the interviewees regarding the correctness of the police procedures, there is a demand for permanence of the police service, as a way to avoid a return to the logic of war that, in the eyes of the favela residents, is even worse (Musumeci, 2017). Thus, the intention is to understand to what extent individuals endorse the action of the UPP and, by legitimizing it, begin to desire the policy continuity after the Olympics.

4. METHODOLOGY

The data used in this article were collected in the research Dimensionamento dos impactossociais das UPPs em favelas cariocas (Measuring the social impacts of UPPs in favelas of Rio de Janeiro), conducted by FGV-Opinão in partnership with the Center for Studies on Crime and Public Security (Crisp) of the Federal University of Minas Gerais. The study was carried out in two rounds: the first was funded by CNPq (2014) and the second by Fundação Getulio Vargas (2015-2016) and allowed to understand the dynamics of pacification, its challenges and problems. In both stages, the main objective of the research was the assessment of pacification process by its beneficiaries. Additionally, the study sought to understand the maturation effects of the security policy. Considering that the policy needs time to accommodate the expectations of both groups involved (police officers and local population) and thus shape the interaction between them, what would happen when the UPP was no longer a novelty? How would individuals rate it?

To answer these questions, we conducted a survey with 2,000 individuals in 20 Rio de Janeiro favelas. A structured questionnaire was used with 38 questions, coordinated with other research already done on (i) police action and forms of interaction with the beneficiary communities (Riccio et al., 2013); (ii) local population’s perceptions of the social impacts of UPPs (Burgos et al., 2013); (iii) valorization of the presence of the State in the communities (Cano, Borges and Ribeiro, 2012); and (iv) the desire for continuity of this policy (Musumeci, 2017).

In order to guarantee the representativeness of the results, a probabilistic sample was constructed with a hundred interviewees from 20 favelas, totaling 2,000 interviews. The interviewees’ selection consisted of three stages: first, ten census tracts were selected in each favela, followed by randomly selecting ten households within each tract, and finally, in each household, one respondent was randomly selected to answer the questionnaire.

Regarding the research design, the first round of the survey consisted of interviewing the residents of the first favelas that received the UPPs, established between 2008 and 2010, during which time this policy enjoyed a high degree of public support. In the second stage, however, the limits of the process of accelerated expansion of the UPPs were already more than visible, a fact exacerbated by the set of favelas selected by the government, among them Complexo do Alemão, Vila Cruzeiro, Manguinhos and Rocinha.¹⁰ The research design, the favelas surveyed, as well as the time of the interviews, are presented in box 1.

Box 1
Distribution of favelas researched, per survey year and establishment of the UPP

After the two rounds of the survey, a final database was created that reflects different moments of public policy and covers heterogeneous favelas. The interpretation of the data should be done considering this overlapping of the maturation effect and the heterogeneous set of favelas selected by the public security policy and reflected in the two rounds of the survey. Therefore, the analyzed information will consider this difference, in order to verify the impact of these moments on the desire for continuity of the UPP after the Olympics.

5. THE ANALYTICAL MODEL

The constitution of the UPPs as a public policy was part of a government effort to prepare the city for the mega-events that it would host. The policy was planned to last until the end of the 2016 Olympics. It was, therefore, a policy to guarantee a “tourist city, integrated city, safe city” (Araújo, 2012:115). At various times the UPP was announced as a strategy of occupying so-called “conflagrated” territories with a view to “forming a ‘security belt’ around the mega-events” (Araújo, 2012:116).

It was hoped, however, that the UPPs would be able to take root in order to transform the relationship between police and community, giving legitimacy to the actions undertaken by the police officers (Rodrigues, 2014), which would guarantee their permanence beyond the mega-events (Cano, Borges and Ribeiro, 2012).

In the early years of implementation of the UPPs, the results were so surprising that they were seen as the way out of the issue of urban violence, with Rousseff promising to expand this form of community policing to all regions of the country in her presidency campaign in 2010.¹¹ However, the atmosphere collapsed with the crisis after the World Cup, with several reports highlighting the deaths in the locations and abuse by police.¹² The biggest landmark in this regard was the disappearance of the mason Amarildo in 2013.¹³

In the meantime, the academia followed the development of the policy, initially, with some enthusiasm (Cano, Borges and Ribeiro, 2012) and then, with a great deal of pessimism, criticizing the rules that were imposed on those residing in these areas (Silva, 2016). This research sought to discuss this knowledge accumulated, aiming to understand the phrase that often appeared in the interviews with residents of pacified areas about what the UPP’s permanence in these localities meant: “Bad with it, worse without it” (Ribeiro and Vilarouca, 2018).

Two questions about the future of the UPPs after the 2016 Olympics were included in the survey. One related to the respondents’ expectations regarding the continuity of this public policy beyond the mega-events that led to its creation: “Do you think that the UPP will end or continue after the Olympics?”. The other referred to the rules and how the interviewee felt about the public policy implemented in their place of residence once the project of Rio as Olympic city ended: Should the UPP end or continue after the Olympics?”

For 43.4% of respondents, the UPP would end after the Olympics, indicating how peacemaking was a project designed to ensure “the safety of major sporting events and would not endure after them” (Musumeci, 2017:7). However, for 75.8% of the interviewees, the UPP should continue, which indicates how respondents recognize the relevance of this public policy and the difficulty of returning to the previous status quo (Cano, Borges and Ribeiro, 2012). The most interesting result is the finding that among those who thought that the UPPs would end after the Olympics, 69.7% affirmed their desire for continuity (table 1). It can be seen, therefore, that despite the disbelief in the UPPs as public policy, there was a certain desire for them to continue.

Table 1
Crossing between variables related to expectations and the desire for continuity of UPPs after the Olympics

Interviewees’ opinion whether UPPs will end or continue after the Olympics	Interviewees’ opinion whether UPPs must end or continue after the Olympics
	They must end		They must continue		Don’t know/No response		Total
	N	%	N	%	N	%	N	%
Theywillend	203	23.4%	605	69.7%	60	6.9%	868	100%
Theywill continue	67	8.1%	736	88.8%	26	3.1%	829	100%
Don’t know/No response	22	7.3%	175	57.8%	106	35.0%	303	100%
Total	292	14.6%	1516	75.8%	192	9.6%	2000	100%

• Chi-square: 362,886 (0,000)14Source: Research data.

The belief in the discontinuity of the UPP as a public policy may be related to the residents’ understanding of the social dynamics in the city of Rio de Janeiro, which is directly connected to the memory of other community policing programs already developed by PMERJ (Cano and Ribeiro, 2016). On the other hand, the desire for UPPs continuity is based on the experience of this community policing in all its limits and possibilities and, therefore, it is associated with police practice and, of course, its effects on the life in the favela (Oliveira et al. 2014).

The dependent variable chosen for this study, the desire for continuity of UPPs, was expressed by 75.8% of the interviewees. The assumption is that the desire for continuity of this public policy means that, in the residents’ evaluation, the UPPs have presented positive results, which means that there is some level of legitimacy attributed to this modality of policing. It does not necessarily mean that the interviewees accept pacification fully, since there is resistance, criticism and complaint (Musumeci, 2017). The desire for continuity seems to indicate the aspiration that policing continues to be part of the local population’ reality.

In order to clarify these various assumptions, the study tested to what extent this feeling is related to the characteristics of the individuals residing in these areas, to the legitimacy they give to pacification, and also to the improvement of the quality of life after the arrival of police. The independent variables — those that may explain the desire for continuity of the UPPs after the Olympics — were organized into three distinct groups shown in figure 2.

Figure 2
Model built to understand the determinants of the desire for continuity of UPPs

The first group of variables refers to the personal characteristics of the respondents (sex, age and race), with a focus on the young and black, who are the most susceptible to police actions and, therefore, the most critical of pacification (Corrêa et al., 2016).

The second concerns the performance of the public policy in terms of legitimacy of police action (Zanetic, 2017) and therefore includes (i) the adequacy of police procedures to the concept of community policing (measured by the degree of trust placed in UPPs police officers); (ii) the quality of the decisions made (related to the fulfillment of the highest promise made by the UPP, i.e., reduction of shootings); and (iii) the correctness of the operational procedures (which means no victimization — aggression, extortion or verbal abuse — by the UPP).

The third group, related to the improvement to quality of life, is composed by the variables that measure perceptions about the feeling of security and the odds of suffering violence, as well as the perception of better life conditions after the establishment of the UPP, parameters pointed out in qualitative studies as public goods generated through the pacification policy (Cano, Borges and Ribeiro, 2012). In addition to this information, the study included a variable identifying the year of the survey, which clarifies the distinct moments of pacification and the related public perception (Menezes, 2017), considering that the moment can have direct effects on affection from the individual toward the policy.

6. THE VARIABLES¹⁵

In terms of individual characteristics, 46% of the interviewees were male, 35% were young (up to 35 years of age); 75% described themselves as black or pardo, characteristics that make up the profile of those most susceptible to police violence and approaches (Ramos and Musumeci, 2005; Lima and Sinhoretto, 2015).

In the second axis, regarding the elements that underpin the legitimacy of public policy, we found that 40% of the interviewees trusted the UPP police officers and 71% said that the UPP officers respect the resident; requirements for the construction of any community policing program.¹⁶ In addition, 75% said that with the UPP, the shootings are no longer routine in the community, which does not eliminate the fact that for 25% this violence is still part of favela life. Most concerning is that 10% indicated that they suffered victimization due to verbal offense, aggression and/or extortion by UPP police officers,¹⁷ which hampers any strategy of closer proximity between police and community. How are the residents supposed to trust a police officer if, on some occasions, they are treated disrespectfully?

In the third axis, concerning the elements that measure the improvement of the quality of life after pacification, 51% stated that the UPP meant more security, 50% affirmed that the police presence made life in the favela better and another 21% perceived that there are high odds of suffering some kind of violence in the favela. Despite this percentage of individuals who believe that theft, aggression and robberies are part of the pacification dynamics being relatively low, the same cannot be said of the rest. The fact that only 50% of survey respondents believe that the UPP improved safety and quality of life is an understated result of the promises of this public policy.

In the opinion of the residents interviewed, police presence alone does not mean an improvement in the quality of life of those living in the favela, given the lack of adequate provision of services such as health, education and social service, which, although promised by the UPP, did not take effect in most cases. Consequentially, only half of the respondents said that the implementation of this public policy has contributed to improve community life.

Finally, it is important to highlight that half of the interviewees were in the favelas that received the UPP at the beginning of the project, when this public policy was seen as a form of “crime transformation” in these areas (Riccio et al., 2013) and the other favelas pacified after were the target of intervention when some studies indicated signs of crisis and identified problems that needed to be remedied so that the idea of community policing could effectively be institutionalized (Leite, 2014). In addition to the immense sociodemographic diversity and the degree of heterogeneity, this study grouped the favelas into two groups in order to analyze the effects of the time of implementation of the UPPs.

Table 2
Statistics of the variables that will be included in the model (Rio de Janeiro, 2014-2016)

Variables		Minimum	Maximum	N. cases	Average	Standard-Deviation
Individual characteristics	Sex	0 — Female	1 — Male	2,000	0.46	0.50
	Youth	0 — Non youth (over 36)	1 — Youth (18 to 35)	2,000	0.35	0.48
	Race/ethnicity	0 —White	1 —Black (including pardos)	1,918	0.74	0.44
Legitimacy, based on a policy of keeping promises	Trust	0 — Does not trust UPP police officers	1 — Trusts UPP police officers	1,881	0.40	0.49
	Respect	0 — UPP Police officers do not respect residents	1 — UPP Police officers respect residents	1,860	0.71	0.46
	Police_violence	0 — Was not victim of verbal offense, aggression and extortion by UPP police officers	1 — Was victim of verbal offense, aggression and extortion by UPP police officers	2,000	0.10	0.31
	Shooting	0 — With the UPP, daily shootings did not stop in the community	1 — With UPP, daily shootings stopped in the community	1,854	0.75	0.43
Improving the quality of life	Security	0 — Feels unsafe after the establishment of the UPP	1 — Feels safe after the establishment of the UPP	1,887	0.51	0.50
	Life_after_UPP	0 — Worse or the same	1 — Better	1,913	0.50	0.50
	Odds_violence	0 — Low odds of suffering violence in the community (aggression, robbery, theft)	1 — High odds of suffering violence in the community (aggression, robbery, theft)	1,937	0.21	0.41
Dependent variable	UPP_must	0 — UPP must not continue after the Olympics	1 — UPP must continue after the Olympics	2,000	0.76	0.43
Local	Pacification	0 — Old pacification	1 — Recent pacification	2,000	0.50	0.50

• Source: Research data.

When comparing the dimensions related to police performance between the two geographic groups (old and recent areas of pacification) surveyed (table 3), all of them show the average number of individuals who trust the police or believe that the UPP police officers respect residents, is less among recent pacifications.

When considering the dependent variable, there are marked inequalities, with more support for pacification from the interviewees in the first round of the survey than in the second. Two indicators did not present a significant statistical difference: police violence and the odds of suffering violence, dimensions that seem to be equally distributed between the two territories.

Taken together, the data summarized in this section indicate that the UPP is not a public policy accepted by all equally, which is why the indicators of police legitimacy are very different and are related in part to the moment when pacification was conducted. Moreover, the arrival of this type of policing did not imply a complete transformation of the community, and positive externalities seem to be more visible among those interviewed in 2014 than among those who participated in the second round of the survey. Finally, the dimensions where all interviewees agree are those related to violence (whether committed by individuals within or outside the community, or by the police officers). Considering these results, it would be important to understand the dimensions that explain the desire for the continuity of this policy.

Table 3
Comparation of variables averages with police performance in old and recent UPPs (Rio de Janeiro, 2014-2016)

Pacification	Trusts UPP police officers	UPP Police officers respect residents	Was a victim of verbal offense, aggression and extortion by UPP police officers	With UPP, daily shootings stopped in the community	Peace	Feels safe after the establishment of the UPP	Life after UPP is better	High odds of suffering violence in the community (aggression, robbery, theft)	UPP must continue after the Olympics
Old pacification	43%	73%	10%	90%	86%	61%	59%	20%	87%
Recent pacification	38%	68%	11%	59%	60%	40%	42%	23%	64%
Total	40%	71%	10%	75%	73%	51%	50%	21%	76%
Anova	5.78	6.24	1.74	286.36	176.68	79.99	53.87	3.56	158.17
Significance	0.016	0.013	0.188	0.000	0.000	0.000	0.000	0.059	0.000

• Source: Research data.

7. MODELS

In this section the four models of binomial logistic regression are presented. They explain which factors affect the desire for continuity of the UPP by the interviewees. The first model tested to what extent the individual characteristics determine the desire for the continuity of the policy. The second analyzes the question from the point of view of legitimacy. The third model verifies if the perception of improvement of the quality of life with the establishment of the UPP means acceptance of the pacification. Finally, an integrated model is built the aim of verifying whether the time of pacification (which expresses the institutionalization and accommodation of the policy), controlled by all other dimensions, is the variable that most explains the desire for continuity of UPPs.

To model the data collected with the survey, the technique chosen was the binomial logistic regression, recommended when the response variable is categorized as zero (pacification should not continue) and one (pacification should continue). As the response-variable is binary — the desire versus the non-desire for continuity of the UPP after the Olympics — the models were estimated according to a binomial logistic function, according to the following equation. In this equation, the response variable (Θ) assumes value one, if the individual wants the UPP to continue beyond the Olympics, and zero, if the individual rejects the policy. This result is generated depending on the β values of each of the dependent variables.

$\theta ={\beta }_0+{\beta }_{1x1}+{\beta }_{2x2}+\cdots {\beta }_{nxn}$

As argued by Oliveira (2016:145), β is a vector of parameters to be estimated. It measures the impact of variations on the explanatory variables at xi’ on the probabilities of the respondent to say that they want the continuity of the UPP, and this function of cumulative distribution is restricted to the interval {0,1}. Therefore, the binomial logistic regression model is the one that best fits the data, because each coefficient informs the “impact” of a variation occurred in the explanatory variables, on the response variable. Based on the β of any independent variable, we obtain the marginal effect of this variable on the probability that the individual wants the UPP.

7.1 Model 1 — individual characteristics

Of all the variables inserted in the model, only the variable ‘sex’ was not statistically significant. Color and age contribute to the rejection of the UPP. In other words, being young decreases the chance of desire for continuity by 0.499 times, whereas being black or pardo decreases chances by 0.636 times.

These results are in part consistent with qualitative analyzes, since youngsters, blacks and pardos have traditionally made up the suspicious element of the police and are therefore the most susceptible to police brutality and lethality, and those most critical of pacification, given their understanding that it excessively limits the rights and rules of favela life with values that are typical from population living in the ‘asphalt’, disregarding the current culture in these areas (Silva, 2016). The most important result of this model is to indicate how the interviewee’s physical characteristics influence their sympathy for the public policy because such elements are determinant of how police interact with them (Ramos and Musumeci, 2015).

7.2 Model 2 — Legitimacy of police action

All the variables included in the model were statistically significant, and the trust in the UPP police officers and the perception that these professionals treat the residents with respect contributed to increase the odds of the interviewee to desire the continuity of the policy by 2.612 and by 2.864 times, respectively. However, the variable that most contributes to the UPP’s legitimacy status is the reduction of shootings, a variable that increases by 3.307 times the odds of the individual wanting the project to continue. On the other hand, victimization by a UPP police officer for verbal offense, extortion or aggression decreases the chance of the individual accepting this type of police action by 0.533 times.

In a way, these results were anticipated by qualitative studies, especially those that highlighted how police violence erodes the legitimacy of this institution, generating repulsion to pacification as a public policy (Cunha, 2015). After all, the great promise of the UPP was the change in police practice in favelas, which would result in a true revolution of procedures within the Military Police of the State of Rio de Janeiro (Cano, Borges and Ribeiro, 2012). Local violence, especially of those perpetrated by the police officers, causes the UPP to be perceived as a policy incapable of fulfilling the promises enunciated and, therefore, needing to be reformulated, since it reifies characteristics traditionally associated with the PMERJ, seen as highly inefficient and lethal (Cano and Ribeiro, 2016).

Therefore, to victimize the individual that the institution should protect is to demonstrate that the old police continue to exist even within this new program, provoking rejection by those living in the favela.

7.3 Model 3 — Improving the quality of life

In this model, believing that the UPP improves the quality of life of those living in the favela was the most important variable, increasing by 3.817 times the odds of the individual desiring pacification to last after the Olympics. Secondly, the perception that there is more security with the police in the favela, increases the chance of the desire for continuity by 3.436 times. On the other hand, the understanding that the chance of suffering violence in the community is high, reduces by 0.753 times the possibility of the subject supporting the UPP after the mega-events.

Again, the quantitative results reinforce the findings of other studies on the subject, which portray how support for peacemaking results from the ability of police officers to transform life in the favela to a guarantee of a better life (Burgos et al., 2012) as well as the reduction of typical security problems, such as aggression, robbery and theft (Cano, Borges and Ribeiro, 2012). In short, to become attached to the UPP, one must feel that this policy has altered the quality of life in several dimensions, making the UPP areas safer and with less criminal activity, elements that the UPP as public policy is often not able to deliver.

7.4 Model 4 — integrated perspective

The last model, which integrates all variables previously analyzed, seeks to indicate to what extent the variables are better understood when analyzed with each other rather than in separate models, which test only the effect of individual characteristics, perception of legitimacy of the police and of improvement with the implementation of the policy.

Among the variables inserted in the model, two were not statistically significant: sex and the odds of suffering violence. These variables are, however, essential to adjust the model, indicating that they measure something that others do not, although they do not present themselves as dimensions that directly affect the desire for continuity of the UPP. In other words, there are phenomena that overlap being a man and believing that the chance of suffering violence is high when the proposal is to understand the individual’s idea for the future in terms of policing patterns in the area in which they live.

Table 4
Odds ratio (Exp(B)) of independent variables included in the logistic regression models that estimate odds that UPPs must continue in the residents of pacified areas’ point of view (Rio de Janeiro, 2014-2016)

Variable	Description	Model 1		Model 2		Model 3		Model 4
		Exp(B)	Sig.	Exp(B)	Sig.	Exp(B)	Sig.	Exp(B)	Sig.
Sex	Male	1.176	0.135					1.120	0.440
	Reference: Female
Youth	Youth (18 to 35)	0.499	0.000					0.609	0.001
	Reference: Non youth (over 36)
Race	Black (including pardo)	0.636	0.001					0.647	0.018
	Reference: White
Trust	Trust in UPP police officers			2.612	0.000			1.854	0.001
	Reference: No trust in UPP police officers
Respect	UPP police officers respect residents			2.864	0.000			2.269	0.000
	Reference: UPP police officers respect residents
Police violence	Victim of verbal offense, aggression and extortion by UPP police officers			0.533	0.001			0.557	0.008
	Reference: Was not victim of verbal offense, aggression and extortion by UPP police officers
Shooting	With UPP, daily shooting stopped in the community			3.307	0.000			1.619	0.004
	Reference: with UPP, shootings did not stop in the community
Safety	Feel safe after the establishment of the UPP					3.436	0.000	1.883	0.000
	Reference: Feel unsafe after the establishment of the UPP
Life after UPP	Better					3.817	0.000	2.470	0.000
	Reference: Worse or the same
Chance of violence	High odds of suffering violence in the community (aggression, robbery, theft)					0.753	0.042	0.839	0.284
	Reference: Low odds of suffering violence in the community (aggression, robbery, theft)
Pacification	Recent pacification							0.379	0.000
	Reference: old pacification
Control		1.237	0.063	0.715	0.008	1.373	0.000	2.334	0.002

• Source: Research data.

Controlling personal characteristics, we see that the main variables affecting the desire for continuity of UPPs are the improvement of the quality of life with the arrival of pacification and the perception that the police officers treat the residents with respect (increase the odds of supporting the policy by 2.470 and 2.269 times, respectively). Beyond these dimensions, feeling safer after the arrival of this model of “community policing” and having confidence in the work done by these professionals increases the chance by 1.883 and 1.854 times (respectively) that the individual desires the continuity of the public policy. Finally, to believe that the greatest promise of the UPPs — the reduction of the shootings — was deferred, means 1.619 times more chances of supporting this intervention after the Olympics. Indeed, the binomial logistic regression model points to improving the legitimacy of the police and the quality of life in the favela as dimensions that contribute to a greater sympathy for pacification and, therefore, support the continuity of this policy after the Olympics.

As expected, the individuals who declared themselves to be black and pardo, the youth and those who have already suffered violence by the UPP police officers are the most resistant to the UPPs. The rejection of the continuity of politics is very large among this public, which points to the need for public policy to be rethought in order to guarantee a police practice less biased by the stereotype of the “suspicious element” (Ramos and Musumeci, 2005), of training more in keeping with the principles of human rights within the favela, since this is one of the pillars of support for the continuity of pacification.

Finally, living in areas of recent peace means not supporting the continuity of the UPP compared to those residing in areas of old pacification (the odds ratio was 0.379). In a way, this result was also expected, given the conclusions of the qualitative studies carried out in these areas, about the constant clash between residents and police, the numerous cases of violence and lethality, and the continuation of the shootings, which would indicate the impossibility of expanding the project beyond those communities pacified between 2008 and 2010. Ultimately, perhaps this result helps to understand the constant protests of those residing in favelas of the Complexo do Alemão against police action in the area. Apparently, the UPP is far from guaranteeing its promises in the 10 favelas surveyed in the second round of the survey, held in 2015 and 2016.

In short, attachment to the UPP is more prevalent in the older projects, where pacification was implemented before the mega-events. In addition, the individuals who embrace this policy are those who have felt changes in the police practice implemented by this project compared to the traditional actions that characterize the PMERJ and therefore believe that life is better with the UPPs.

8. FINAL CONSIDERATIONS

The purpose of this paper was to carry out an evaluation of the Pacifying Police Units from two rounds of survey conducted with individuals residing in UPP areas in the years 2014 and 2015-2016. The assumption was that if the police practice implemented through this project is considered as positive, this would result in the desire for its continuity. In the case the practice is understood by residents as negative, that would lead to the repulsion toward the policy. The results of the logistic regression models show relationships similar to those indicated by the qualitative studies on the constitution and effects of UPPs in the city of Rio de Janeiro.

However, the two rounds of survey reflect very different realities and moments in terms of the relationship between the society of Rio de Janeiro and the Military Police. In the first round the first established UPPs were researched, those considered consolidated (Rodrigues, 2014) and at a time when the questioning of the pacification project was not so intense. A very different scenario was found in the second round, where the areas surveyed were marked by intense conflict between the police and the communities. In addition, a constant questioning of the legitimacy of pacification practices was observed, which would once again put favela residents “under siege”, thus representing only a substitution of the abuses carried out before by local drug traffickers (Machado da Silva, 2015).

The last model of logistic regression indicates that this duality, between old and new pacifications, is visible in the desire for UPP continuity. In the communities that experienced the first years (or the golden age) of the policy, residents strongly wanted continued stay of the UPPs. But 1/3 of the interviewees of those who received the project when the situation was already in crisis (Cunha, 2015), believe that they can live without them. As argued by Zanetic (2017), police legitimacy is essentially contextual, and although the UPP is often treated as an indivisible whole, the results of this study demonstrate Zanetic’s argument for the importance of the context by offering evidence that the moment and date of pacification greatly interfere in the affection that the individual develops for this project.

Perhaps this result is directly related to the capacity of pacification to constitute a community policing program in a less rhetorical way in certain areas (Riccio et al., 2013), which apparently does not happen in all favelas. Ensuring security, treating local population with respect and building trust between police and society are indispensable elements for the recovery of police legitimacy in favela areas (Cano, Borges and Ribeiro, 2012), this is also one of the explicit objectives of the program in its first years of existence. When the population perceives that the UPP has these characteristics, it is possible to observe sympathy for the project and a general defense for its permanence after the mega-events in the city. This confirms the assumption of the more theoretical perspectives on community policing (Braga, 2015) and other studies on the positive effects of these programs in Latin America (Carrión, Pontón and Armijos, 2009; Frühling, 2007).

On the other side of the spectrum are the problems traditionally questioned in the practice of PMERJ, which is excessively abusive to young, black and pardo citizens, and residents of favela areas (Machado da Silva, 2015). If the rejection of police action by this part of the population is not new (Holmes, 2016), it is noteworthy that the UPP was not able to change this scenario, pointing out a police practice outside the expected parameters in terms of procedural fairness (Zanetic, 2017). It is clear, therefore, how the police failed to advance in terms of community relations, because in almost a decade of UPPs this dimension has not been adjusted. This is why, for this public, the favela is better off without the police than with it. Thus, the presence of police officers in the areas where these young people reside makes the community a worse place, where there is no possibility of freedom and, least of all, leisure spaces in which they can develop their own culture (Silva, 2016).

Those who want the UPP to continue after the Olympics are those who believe that life in the favela is “bad with it, but worse without it.” They understand that the pacification measures have transformed the routine of sociability in the area, making life better, especially by reducing violence. As stated by one of the interviewees, the UPP “brought some peace when it comes to the shooting. The amount of shots reduced. They go up and down doing an inspection. This shows that they value the community, a moral value; they have more respect for the community.” In other words, those who want the continuity of pacification are those who perceive in police practice quality in interpersonal treatment and ability to comply with rules in an essentially professional organization (Zanetic, 2017), as police action should be in a democratic society (Braga, 2015).

Therefore, instead of enacting the death of the project,¹⁸ the Rio de Janeiro State Secretary of Public Security might have to look for ways to change police practice, making it better suited to the principles that govern a democratic society, as a way of improving life in the favelas, guaranteeing the basic right of any individual, guaranteeing the life. Without a police action that emphasizes citizens’ survival, hardly any community policing program, new or old, will be desired beyond the time it takes for the novelty of its announcement to wear off.

REFERÊNCIAS

• ALBERNAZ, Elizabete R.; CARUSO, Haydée; PATRÍCIO, Luciane. Tensões e desafios de um policiamento comunitário em favelas do Rio de Janeiro. São Paulo em Perspectiva, v. 21, n. 2, p. 39-52, 2017.
• ARAÚJO, Marcella C. S. Rio em forma olímpica: a construção social da pacificação na cidade do Rio de Janeiro. In: MISSE, Michel; WERNECK, Alexandre. Conflitos de (grande) interesse: estudos sobre crimes, violências e outras disputas conflituosas. Rio de Janeiro: Garamond; Faperj, 2012. p. 117-130.
• ARIAS, Enrique Desmond; BARNES, Nicholas. Crime and plural orders in Rio de Janeiro, Brazil. Current Sociology, v. 65, n. 3, p. 448-465, 2016.
• BEATO, Claudio. Compreendendo e avaliando: projetos de segurança pública. Belo Horizonte: UFMG, 2008.
• BRAGA, Anthony Allan. Crime and policing revisited Washington, DC: US Department of Justice, 2015.
• BURGOS, Marcelo Baumann et al. O efeito UPP na percepção dos moradores das favelas. Desigualdade & Diversidade, v. 11, p. 49-97, 2012.
• CANO, Ignacio; BORGES, Doriam; RIBEIRO, Eduardo (Org.). Os donos do morro: uma avaliação exploratória do impacto das Unidades de Polícia Pacificadora (UPPs) no Rio de Janeiro. São Paulo: Fórum Brasileiro de Segurança Pública, 2012.
• CANO, Ignacio; RIBEIRO, Eduardo. Oldstrategiesand new approaches towards policing drug markets in Rio de Janeiro. Police Practice and Research, v. 17, n. 4, p. 364-375, 2016.
• CARRIÓN, Fernando; PONTÓN, Jenny; ARMIJOS, Blanca. 120 estrategias y 36 experiencias de seguridad ciudadana Quito: Flacso-Sede Ecuador, 2009.
• CORRÊA, Juliana et al. Poor youths and ‘pacification’: dilemmas between discourse and practice from the perspective of young people about policing in Rio de Janeiro’s favelas. International Sociology, v. 31, n. 1, p. 100-120, 2016.
• COSTA, Marcelo H. da; CASTILHO, Sérgio R. R. Desenvolvimento econômico solidário e segurança pública na cidade do Rio de Janeiro: estratégias para a ação. Boletim Mercado de Trabalho, n. 46, fev. 2011.
• CUNHA, Christina V. O medo do retorno do medo: um ponto de inflexão no programa das UPPs. Dilemas-Revista de Estudos de Conflito e Controle Social, v. 8, n. 1, p. 41-62, 2015.
• DAMMERT, Lucía. Dilemas de la reforma policial en América Latina Policia, Estado y sociedad: prácticas y saberes latinoamericanos. Rio de Janeiro: Publit, 2007.
• DIAS, Ariley P. R. Para ver o futuro: utopia, risco e paz no Jardim Batan. 2016. Dissertação (mestrado em sociologia e antropologia) - Universidade Federal do Rio de Janeiro, Rio de Janeiro, 2016 .
• ECK, John E.; MAGUIRE, Edward R. Have changes in policing reduced violent crime? An assessment of the evidence. The Crime Drop in America, v. 207, p. 207-265, 2000.
• FLEURY, Sonia. Militarização do social como estratégia de integração - o caso da UPP do Santa Marta. Sociologias, v. 14, n. 30, 2012. p. 194-222.
• FRÜHLING, Hugo. The impact of international models of policing in Latin America: the case of community policing. Police Practice and Research, v. 8, n. 2, p. 125-144, 2007.
• HOLMES, Tori. Ethical dilemmas in studying blogging by favela residents in Brazil. In: PURI, Shalini; CASTILLO, Debra A. Theorizing fieldwork in the humanities: methods, reflections, and approaches to the global south. Nova York: Palgrave Macmillan, 2016. p. 131-149.
• LEITE, Márcia Pereira. Entre a ‘guerra’ e a ‘paz’: Unidades de Polícia Pacificadora e gestão dos territórios de favela no Rio de Janeiro. Dilemas-Revista de Estudos de Conflito e Controle Social, v. 7, n. 4, p. 625-642, 2014.
• LIMA, Renato S.; SINHORETTO, Jacqueline. Narrativa autoritária e pressões democráticas na segurança pública e no controle do crime. Revista Semestral do Departamento e do Programa de Pós-Graduação em Sociologia da UFSCar, v. 5, n. 1, p. 119-141, 2015.
• MACHADO DA SILVA Luiz Antonio. A experiência das UPPs: uma tomada de posição. Dilemas: Revista de Estudos de Conflito e Controle Social, v. 8, n. 1, p. 7-24, 2015.
• MAGALONI, Beatriz; FRANCO, Edgar; MELO, Vanessa. Killing in the slums: an impact evaluation of police reform in Rio de Janeiro. Stanford: Stanford University, 2015. (CDDRL Working Paper, n. 556).
• MENEZES, Palloma V. Os rumores da ‘pacificação’: a chegada da UPP e as mudanças nos problemas públicos no Santa Marta e na Cidade de Deus. Dilemas: Revista de Estudos de Conflito e Controle Social, v. 7, n. 4, p. 665-684, 2014.
• MENEZES, Palloma V. Será que estaremos aqui quando as Olimpíadas chegarem? Novas oportunidades e preocupações pós-UPP na “favela modelo”. Trama: Indústria Criativa em Revista, v. 1, n. 1, 2015. p. 42-58.
• MOHOR, Alejandra. Uso de indicadores para evaluar el funcionamiento policial Documento de Trabajo n. 2. Santiago: Universidad Catolica de Chile, 2007. Disponível em: <Disponível em: http://www.cesc.uchile.cl/publicaciones/op_07_documentodos.pdf >. Acesso em: 2 set. 2017.
  » http://www.cesc.uchile.cl/publicaciones/op_07_documentodos.pdf
• MONET, Jean-Claude. Polícias e sociedades na Europa São Paulo: Edusp, 2001.
• MOURÃO, Bárbara M.; STROZEMBERG, Pedro (Org.). Mediação de conflitos nas UPPs: notícias de uma experiência. Rio de Janeiro: CESeC/Iser, 2016. Disponível em: <Disponível em: http://www.ucamcesec.com.br/livro/mediacao-de-conflitos-nas-upps-noticias-de-uma-experiencia/ >. Acesso em: 2 set. 2017.
  » http://www.ucamcesec.com.br/livro/mediacao-de-conflitos-nas-upps-noticias-de-uma-experiencia/
• MUSUMECI, Leonarda. UPP: última chamada Visões e expectativas dos moradores de favelas ocupadas pela Polícia Militar na cidade do Rio de Janeiro. Relatório de pesquisa. Rio de Janeiro: Centro de Estudos de Segurança e Cidadania (CESeC), 2017. Disponível em: <Disponível em: http://www.ucamcesec.com.br/wp-content/uploads/2016/03/UPP-moradores-Relatório_3.pdf >. Acesso em: 2 set. 2017.
  » http://www.ucamcesec.com.br/wp-content/uploads/2016/03/UPP-moradores-Relatório_3.pdf
• MUSUMECI, Leonarda; et al. Ser policial de UPP: aproximações e resistências. Boletim Segurança e Cidadania, v. 14, p. 1-28, 2013.
• OLIVEIRA, Carolina V. Prática policial e ordem moral: um estudo da relação moradores-polícia em uma comunidade em processo de pacificação. 2016. Dissertação (mestrado em letras/estudos da linguagem) - Pontifícia Universidade Católica do Rio de Janeiro, Rio de Janeiro, 2016 .
• OLIVEIRA, Fabiana Luci de et al. Cidadania, justiça e pacificação em favelas cariocas Rio de Janeiro: Editora FGV, 2014.
• OOSTERBAAN, Sarah; VAN WIJK, Joris. Pacifying and integrating the favelas of Rio de Janeiro: an evaluation of the impact of the UPP program on favela residents. International Journal of Comparative and Applied Criminal Justice, v. 39, n. 3, p. 179-198, 2015.
• PACHECO DE OLIVEIRA, João. Pacificação e tutela militar na gestão de populações e territórios. Mana, v. 20, n. 1, p. 125-161, 2014.
• PINTO, Alejandro F.; GARAY, María S. de. Community policing in Mexico. The framework of resistance and conditions of possibility. Stability: International Journal of Security and Development, v. 3, n. 1, p. 1-17, 2014.
• RAMOS, Silvia, MUSUMECI, Bárbara. Elemento suspeito: abordagem policial e discriminação na cidade do Rio de Janeiro. Rio de Janeiro: Civilização Brasileira, 2005.
• RIBEIRO, Ludmila; VILAROUCA, Márcio. The meanings of Pacifying Police Units (UPPs) in the ‘marvelouscity’. Police Practice and Research, 2018. no prelo.
• RIBEIRO, Luiz Cesar de Q.; SANTOS JUNIOR, Orlando A. dos. Mega sporting events in Brazil: transformation and marcantilisation of cities. Território, v. 64, 2013. p. 54-60.
• RICCIO, Vicente et al. Community policing in the favelas of Rio de Janeiro. Police Practice and Research, v. 14, n. 4, p. 308-318, 2013.
• RODRIGUES, Robson. The dilemmas of pacification: news of war and peace in the ‘marvelous city’. Stability: International Journal of Security and Development, v. 3, n. 1, p. 1-16, 2014.
• SAWAYA, Ana Lydia. Políticas públicas: pontos de método e experiências. Estudos Avançados, v. 20, n. 56, p. 131-148, 2006.
• SILVA, Bruno Marques. Reformar a polícia e pensar a cidade: o policiamento comunitário e a segurança pública pedetista no Rio de Janeiro (1983-95). Libertas, v. 15, n. 2, p. 189-214, ago./dez. 2015.
• SILVA, Luciane S. Baile Funk, missão civilizatória e UPP: cultura e segurança pública na cidade do Rio de Janeiro. Brasiliana-Journal for Brazilian Studies, v. 4, n. 2, p. 318-342, 2016.
• SKOGAN, Wesley G. An overview of community policing: origins, concepts and implementation. In: WILLIAMSON, Tom (Org.). The handbook of knowledge-based policing: current conceptions and future directions. Chicago: John Willey & Sons, 2008. p. 43-57.
• SOARES, Barbara et al. O que pensam os policiais das UPPs? Ciência Hoje, v. 49, n. 294, p. 34-39, 2010.
• SOUZA, Luís Antonio Francisco; BATTIBUGLI, Thaís. O difícil caminho da reforma: a polícia e os limites do processo de reforma pós-redemocratização. Dilemas-Revista de Estudos de Conflito e Controle Social, v. 7, n. 2, p. 293-319, 2014.
• VARGAS, João H. C. Taking back the land: police operations and sport megaevents in Rio de Janeiro. Souls, v. 15, n. 4, p. 275-303, 2013.
• VAZ, Bruno O. E. Três ensaios em microeconometria sobre crime, política e migração. 2014. Tese (doutorado em economia) - Pontifícia Universidade Católica do Rio de Janeiro, Rio de Janeiro.
• WEISBURG, David, NEYROUND, Peter. Police science: toward a new paradigm. In: NEW perspectives in policing. Washington, DC: U.S. Department of Justice, Office of Justice Programs, National Institute of Justice, 2011.
• ZANETIC, André. Ação institucional, confiança na polícia e legitimidade em São Paulo. Revista Brasileira de Ciências Sociais, v. 32, n. 95, p. 1-19, 2017.

• {Translated version} Note: All quotes in English translated by this article’s translator
• 1
  Available at: <http://memoria.oglobo.globo.com/jornalismo/premios-jornalisticos/haacute-democracia-nas-favelas-9216055>. Acessed on: 2 Sept. 2017.
• 2
  Available at: <www1.folha.uol.com.br/colunas/marco-aurelio-canonico/2017/08/1912524-fim-das-upps-acentua-fracasso-das-megaoperacoes-no-rio.shtml>. Acessed on: 2 Sept. 2017.
• 3
  Research conducted on Google academic on 2 September 2017 with the term “Pacifying Police Unit” returned 1,240 results. The term in the plural resulted in 1,910 references. We do not use the acronym because it also indexes medical articles related to “Pressure Sores”.
• 4
  This association is extensively worked on in policing patterns. Monet (2001:12), one of the greatest experts on this subject, states that “the excluded population are led to reject political authority, whether abstaining and being indifferent, or refusing to submit to law and regulation, or by plunging into open violence. In this case, the police are in grave danger of adopting a repressive attitude, of taking coercive measures. But this kind of reaction increases the precautions against the police, further separating this population from the rest of society.”
• 5
  The use of the designations ‘hills’, for favelas, which are usually on mountain slopes, and ‘asphalt’ meaning the ‘rest’ of the city, represented by well urbanized and prosperous neighborhoods, is an allusion to the work of Zuenir Ventura, “Cidade Partida” (Split City). This clarification explains the frequent references to the ‘rise’ of the police or the ‘fantasy world’ of asphalt. In addition, it determines the proximity of the two social worlds.
• 6
  For more information, see: <http://cpdoc.fgv.br/fgvopiniao/pesquisaspublicas#id_1>. Accessed on: 29 Aug. 2017.
• 7
  Musumeci (2017:4) points out that “less than a decade after the inauguration of the first UPP and after a few years of blatant success of police occupation in favelas, it became recurrent to declare the ‘collapse’ or ‘failure’ of this model. Violent crime rates have risen again, and shootings, ‘stray bullets’, police killings and police death have increased again.”
• 8
  It is important to emphasize that when writing this article, in the first half of 2017, the city of Rio de Janeiro suffered from an increase in the rate of violent crimes and also an increase in cases of police lethality, a situation that challenges the results of previous research on the subject. For more information, see: <www.ispdados.rj.gov.br/#>. Accessed on: 7 June 2017.
• 9
  The term ‘pacification’ is highly controversial. As described by João Pacheco de Oliveira (2014), in Brazilian historiography, it is a category used as a synonym for civilizing and including, and therefore, problematic for a public security policy, since it is equivalent to treating residents in favelas (22% of the inhabitants of Rio de Janeiro) as uncivilized individuals or completely excluded from the city. As throughout the national history the pacification policy covered up actions of genocide and ethnic discrimination, its use in the scope of police practice can be interpreted as a kind of control and exclusion of this part of the population. This article does not ignore the debate, but uses ‘pacification’ as a native category, since this is the term from which several interviewees identify the arrival and permanence of police officers of the UPP project in favela areas.
• 10
  There are two exceptions that do not fitt his classification. In 2014, due to the death of dancer Douglas Rafael da Silva Pereira, we did not conduct the survey in the favelas Pavão-Pavãozinho and Cantagalo, and ended up including those favelas in the subsequent round, held in 2015/2016. In the first round of the survey, replacing Pavão-Pavãozinho and Cantagalo, we included the Vidigal favela, whose UPP was inaugurated in January 2012.
• 11
  Find out more at: <http://politica.estadao.com.br/noticias/geral,upp-vira-marca-da-campanha-de-dilma-e-cabral-no-rio,593466>. Accessed on: 5 Sept. 2017.
• 12
  Find out more at: <https://oglobo.globo.com/rio/escalada-no-numero-de-autos-de-resistencia-revela-tendencia-de-alta-na-letalidade-policial-21276182>. Accessed on: 6 Sept. 2017.
• 13
  Musumeci (2017:10) describes the Amarildo case as follows: “Charged with drug trafficking, Amarildo was kidnapped, tortured and killed by police at the UPP’s headquarters in Rocinha, and his body has still not been found today. This occurred at the moment of the great street demonstrations all over Brazil and gained an extraordinary visibility, giving rise to the slogan ‘Cadê o Amarildo?’ (Where is Amarildo?), featured on posters, leaflets, t-shirts and banners during the protests, in addition to being widely spread on social networks”.
• 14
  The chi-square test measures the independence of variables to see if there is an association between two pieces of information. However, a causal relationship cannot be established, as this is not an appropriate test. That is, it can be said that the phenomena are related, but not that one necessarily causes the other.
• 15
  To facilitate the interpretation of the results, all variables were re-coded to 0 (meaning absence of the phenomenon) and 1 (which means presence of the phenomenon). This transformation of variables is done by recoding and joining categories such as “strongly agree” and “partly agree”; and “strongly disagree” and “partly disagree”. Or “always” and “almost always” and “never” or “almost never”; and so on, assigning the values 1 and 0 for each of the groupings of variables (Zanetic, 2017:10). With this, we hope to verify which are the variables that increase or decrease the odds of the individual to desire the continuity of the UPP after the Olympics (value one) versus wanting the end of the policy (value zero).
• 16
  This percentage is higher than that found when the individual was asked about the degree of trust in non-UPP police officers, which is around 39.8%. It is interesting to note that the “institution” in which the interviewees most trust is the neighbors (71.1% said they trust). Only 55% said they trusted, to some extent, community associations in the locality. The less reliable subjects are politicians: only 16.5% said they trusted them to some extent. Therefore, UPP police officers seem to enjoy greater reliability than other institutions operating in the favela. A similar result was found by Musumeci (2017:33).
• 17
  According to data released by the Brazilian Forum of Public Security (2018), 6% of the population of the city has already been victim of police violence. In the Brazilian population, 3.5% were already victims of violence by Military Police. Find out more at: <www.forumseguranca.org.br/wpcontent/uploads/2018/04/FBSP_Rio_sob_Intervencao_2018_infografico.pdf>. Using this source as a basis, it is possible to infer that the percentage of police victimization among residents in the UPP areas is almost three times higher than that of the general population.
• 18
  Find out more at: <https://g1.globo.com/rio-de-janeiro/noticia/secretario-de-seguranca-do-rio-diz-que-politica-nao-e-de-confronto.ghtml>. Accessed on: 13 Sept. 2017.

Publication Dates

• Publication in this collection
  Nov-Dec 2018

History

• Received
  13 Sept 2017
• Accepted
  07 May 2018