Social Representations of Homicide Investigations by Judges, Prosecutors and Police: A Case Study from the Metropolitan Area of Brasilia

This article examines homicide in the Metropolitan Area of Brasilia (MAB), analysing social representations from elites in the criminal justice system, including police chiefs, prosecutors and judges. It draws on the theory of social representations (TSR) to explore the imaginaries constructed around the criminal justice system’s inability to adequately investigate the rise in homicides. The representations from focus group participants highlight a lack of resources, infrastructure, equipment and human resources, as well as unsatisfactory working conditions. In seeking to understand and situate themselves in new realities and contexts, these elite criminal justice actors ultimately place themselves within the available reserve of knowledge, in which they claim that ‘nothing works’. Hence, this enables these powerful actors to justify themselves and blame others, while denying their inability to adequately investigate homicides. A hidden rationale emerges that represents the homicide victims of drug crimes and gang feuds as unworthy of


Introduction
Widespread violence in unequal countries is an obstacle to development (United Nations 2015;World Health Rankings n.d.). Haugen and Boutros (2014) argue that one of the worst disasters in developing countries is the collapse of justice systems. Due in part to the legacies of colonialism, justice systems are unable to curb overwhelming violence (Haugen and Boutros 2014) and cultures of corruption that allow it to flourish. Cifali and de Azevedo (2017: 46) have specifically examined 'why efforts to build an effective public security policy committed to upholding civil rights and improving the control of police and police activities have failed' during the period of re-democratisation in Brazil. The rise of penal populism across Latin America during this period has filled jails but failed to reduce violent crimes (Sozzo 2017). In this article, we examine why this is the case in the Metropolitan Area of Brasilia (MAB), drawing on original research conducted with nodes of elite actors in the criminal justice system-police, prosecutors and judges.
Contemporary research has questioned the claim that violence rates have fallen in developed and developing countries, suggesting that the trend is not universal (Weiss, Santos, Testa and Kumar 2016). A United Nations (2015) report has highlighted significant asymmetries concerning violence across the globe, such as the high homicide rates in the Americas. A recent study has examined the link between homicide rates and the Gini index in 39 countries, concluding that inequality and crime rates are correlated (Fajnzylber, Lederman and Loayza 2002).
Violence in Latin America-particularly in Brazil-is a key focal point in numerous fields of research (Caldeira and Holston 1999;Ciocchini 2017;Cruz 2016), with homicide rates evolving asymmetrically in different cities. With respect to Brazil, Waiselfisz (2017) indicates that 830,420 of the homicides committed between 1980 and 2014 involved firearms, representing about 70 per cent of all homicides recorded annually since 2001. Additionally, violence mostly affects the poorest neighbourhoods (Waiselfisz 2017). Despite Brazil's high incarceration rates, experts emphasise the widespread impunity for homicides-over 60,000 per year. As data for such cases are scarce, the authorities remain unaware of most homicides. The chief problem is the inadequate governance of public security, which concerns the Executive and the Judiciary (Lima, Sinhoretto and Bueno 2015).
In Brazil, the study of public security is not limited to criminological focuses and has expanded in recent years. There has been growing curiosity in certain initiatives, whereas the scope of interest had previously been circumscribed by the paradigm of social reaction. Meanwhile, in the social sciences, the role of criminal justice agencies is garnering attention.
As Brazil is one of the world's most violent countries, many studies stress the ineffectiveness of its criminal justice system (Costa, 2015;Machado, 2014;Sapori, 2007). Although specialists advance different interpretations, they agree that there is a lack of coordination between the different agencies involved. Some studies specifically analyse the different rationales of the police and Prosecution Office (Adorno and Pasinato 2010;Costa 2015;Machado 2015;Oliveira and Machado 2018). Accusations of disorder and incompetence feature prominently in the media and there are increasing accounts of inefficiency, lack of control and conflict between the actors that compose organisations such as the police, Prosecution Office and Judiciary (Machado 2014). 2 Moreover, the dearth of resources to gauge and combat urban violence has attracted criticism (Sapori 2007). Notwithstanding the considerable amount of research undertaken, this remains a vast field requiring further study.
This article approaches the issue first by examining research on homicides in Brasilia. The investigation concentrates on the following municipalities, due to their geographic proximity and high degree of dependence on the Federal District: Águas Lindas de Goiás, Cidade Ocidental, Formosa, Luziânia, Novo Gama, Planaltina de Goiás, Santo Antônio do Descoberto and Valparaíso IJCJ&SD 87 Online version via www.crimejusticejournal.com © 2019 8(1) de Goiás. It explores the social representations offered by police chiefs, prosecutors and judges, based on focus groups conducted between 2010 and 2011. The study devotes attention to statelevel procedures and practices in the State of Goiás, posing several important questions: how do nodes of actors operating within different organisations of the criminal justice system describe their roles? What is their assessment of the flow of homicide cases within the justice system? What is their appraisal of working conditions and how do they link to the low clearance rate of homicides in MAB? What is their evaluation of the introduction of public security policies in the different cities that make up the MAB? What are the hidden layers of discourse buried in their representations for the inadequacy of homicide investigations?

Methodology
Since the research on homicides in the MAB was part of a major financially supported project, it is important to explain the methodological stages undertaken. The starting point was a georeferential analysis of the deaths recorded in 2010, which made it possible to outline the concentration of homicides across different neighbourhoods. The report contrasted the reality observed in these cities with the infrastructure for public services in the Federal District, highlighting a lack of human and material resources for public security. Many municipalities only have a Civil Police Station or CIOPS-Integrated Center for Security Operations. For example, Luziânia has a Regional Expert-Forensic Police Center, which also serves the municipalities of Águas Lindas de Goiás, Cidade Ocidental, Novo Gama, Mimoso de Goiás, Santo Antônio do Descoberto and Valparaíso de Goiás (Costa, Soria Batista, Machado, Zackseski and Porto 2013).
The official data of the Goiás Secretariat of Public Security indicate that Valparaíso de Goiás, Luziânia and Águas Lindas de Goiás are among the most violent municipalities in the country. Significantly, they comprise 54 per cent of the population of the MAB and account for 62.8 per cent of violent deaths. Over the course of this century, the homicide rate in the MAB has increased by 59.3 per cent, making the region an outlier (Costa et al. 2013).
Based on this marker, after computing the 566 recorded deaths, the study then examined the flow of cases within the criminal justice system in three selected cities: Luziânia, Cidade Ocidental and Águas Lindas de Goiás. Analysing the research report, specifically the item addressing the flow of homicides reported in the MAB and the study's base year (2010), it emerged that the Prosecution Office brought charges in only 8.1 per cent of cases. Drawing on the data documented in the report, the focus shifted towards criminal procedures initiated in Águas Lindas de Goiás, Luziânia and Valparaíso de Goiás. The examination uncovered 30 records at the Court of Appeals of Goiás and we found that charges had not been brought in seven of these cases. All the relevant trials held during the period resulted in convictions (Costa et al. 2013). In addition, evaluating the sociourban evolution of the MAB complemented the research.
Finally, as part of the qualitative research, between 2010 and 2011, Costa et al. (2013)  homicides, the non-investigative police were excluded from the study. Judges at jury trialsperformed with seven members of the community-are competent to adjudicate in cases of 'voluntary crimes against life', including homicides, as defined in the Brazilian Penal Code.
To establish the focus groups, we invited representatives from several organisations with backgrounds in the investigation, prosecution and adjudication of homicides. This was complicated, since many of these public servants were reluctant to discuss their jobs and experiences. Thus, it was necessary to convince them of the importance of the research, financed by the Ministry of Justice, while guaranteeing their anonymity.
Initially, the intention was to form groups of five to six professionals, which proved impossible. Ultimately, the following cohorts were put together: one group of three experts; one with three police chiefs; one composed of three judges; and another with five prosecutors. Each focus group lasted 2-3 hours, according to guidelines agreed to by participants. The primary goal was to use their experiences and shared discourses about the culture and organisation of the criminal justice system in the MAB, in some instances to sustain ineffectual or even criminogenic practices (Laslett 2017). In several groups, participants described the poor urban infrastructure, widespread violence and subpar working conditions. During the conversations, some related their career trajectories, revealing differing representations of their duties, conflicts between actors and their salient views on the investigation, prosecution and adjudication of homicides.
In this article, the theory of social representations (TSR) is the guiding methodological framework for sociological analysis. Exploring fields of knowledge by analysing social representations entails the production of second degree-or 'second-hand'-knowledge. Rather than centring the analysis on raw data regarding the criminal justice system, this study examines the imaginaries constructed by certain social segments, in which language is an analytical device. The authors selected the narratives obtained as the raw material of analysis (Moscovici 1994;2003;2012;Porto 2009;2010). In line with this approach, values and beliefs constitute the raw material of sociological practice. This premise follows closely the Weberian tradition, according to which empirical knowledge is invariably linked to value 'judgments', establishing values and beliefs as part of sociological explanation (Moscovici 2001;Porto 2010).
Another significant factor explains the methodological preference for the TSR. The data for this study were acquired through distinct modalities that resemble Moscovici's (2012: 31) approaches to data collection, which are expounded in his research on psychoanalysis in which 'surveys aimed at populations of individuals, in addition to content analysis of the "population" of documents, are currently the most appropriate techniques in the academic examination of social and ideological representations. These techniques are simple enough and very flexible for the purposes of providing valid results on the particular points of interest'. In the context of this study, focus groups were used with three different networks of actors in the criminal justice system in MAB-police, prosecutors and judges. Based on the group discussions, rich material was extracted from the underpinnings of the actors' representations of the criminal justice system. The major themes to emerge from an analysis of the focus group data are discussed below.

Police chiefs' discourse
Several categories guided the analysis of the social representations made by police chiefs. However, two were central to understanding how these actors, informed by their representations, structure their work routines and practices. The first is the unequal value of life and death that shape how police come to routinely view some victims of homicide as valid victims and others as unworthy of police investigation. Our research explored how police chiefs define these concepts and apply them to the social actors with whom they deal during a homicide investigation. The second is the prejudice that underlies perceptions of certain behaviours. Fear, working conditions and the interaction between public security units are also important categories for understanding the police chiefs' representations and homicide investigation practices.
During the focus group held with the three police chiefs (one female and two males) from different cities in Brasilia, the accounts exposed the distinct banality surrounding violence. Participants frequently recounted events such as score-settling over non-payment of drugs and related conflicts, along with killings in bars resulting from trivial disagreements and a desire to exact revenge. The police chiefs depict an image of frenzied murder. Perpetrators of homicide are often the victims of previously attempted killings, as the female police chief explained: 'The perpetrator had previously been a victim, the witness had previously been a victim in the investigation and everybody was dead within a year' (Police Chief 1).
Social representations of local populations are laced with stereotypes, dispersed into conventional wisdom, for instance: 'Santo Antônio, Novo Gama, Valparaíso, even here, if you look at the origins of these people, they are migrants from the northeast; guys come here seeking a better life, an improvement, and they do not come alone-they come with their wives, with five, six children' (Police Chief 3).
The participants emphasised key problems affecting the area, including substandard urban infrastructure, a lack of basic sanitation and unemployment. The accounts invoke images of 'the other' or 'outsider', in which 'it is a matter of behaviour, of education; these people are generally rude, rough, and so that causes these situations … There's [usually] a knife involved' (Police Chief 3). Meanwhile, Police Chief 1 commented that this is 'typical of Northeasterners'. Police Chief 3 went on to say that they are 'really brutish'. These representations depict cities being invaded by 'a different people, full of vices'. The portrayals continue: 'drinking, drugs, prostitutes; this then leads to problems. It's always like this. And it's at this point in arguments, bar fights, that knives and machetes are pulled out' (Police Chief 3).
These remarks highlight a need to constantly invoke 'others' to accuse and blame, transforming them into individuals who succumb to crime, consistent with Misse's (2008) thesis. The comments distinguish the 'I' from the 'other', whereby the latter is accused of violence, brutish behaviour and inhumanity, which ultimately objectifies the individuals and devalues their lives as unworthy of investigation. Homicides involving drug trafficking are commonly motivated by an urgent need for money. In the images depicting the rationale for violent deaths, people kill to steal and protect their business. The poor education and public assistance offered to young people are also significant factors. Many children are left at home, alone, exposed and neglected (Police Chief 2).
Except for cases involving bar fights or domestic violence, our police participants were of the same view that other homicides are difficult to solve because there is an overwhelming fear of testifying. The witness protection program can be difficult to access, often requiring huge sacrifices. Rarely is it contemplated as an option. However, the greatest obstacle that emerges from the police chiefs' accounts is not fear but rather the lack of material and human resources to examine all the facts. Investigations are complex and lengthy and investigative resources are scarce. Police Chief 3 relays the pressure from the Prosecution Office calling for police action, saying 'we investigate many homicides at the police station following requests from the Prosecution Office; sometimes we have to stop everything, with all these files coming in that generate huge demands'.
The police chiefs regard expert evidence as fundamental in shedding light on the perpetration of criminal acts. Nevertheless, they condemn the inadequate infrastructure, with Police Chief 2 claiming that 'you can't completely count on our Forensic Police; (but) we can count on them to IJCJ&SD 90 Online version via www.crimejusticejournal.com © 2019 8(1) cover the crime scene'. Hence, very few cases are closed in an ideal fashion, through conclusive proof, both expert and testimonial. In practice, given the dearth of expert evidence, testimonial evidence is more prevalent.
In their interaction with prosecutors, police chiefs resent the demands made by the Prosecution Office, as well as the indifference towards the unsatisfactory conditions for conducting investigations. Routines require varied strategies and secrecy is highly cherished. One account reveals that not everything can be 'made explicit'.
The statements of the research subjects are consistent with the idea of using secrecy (Simmel 1991) as a strategy to concentrate power. Further, they illustrate the nexus between knowledge and power, as suggested by Foucault: These relations of 'power-knowledge' should not therefore be analysed by starting from a knowing subject who would be free or unfree with respect to the system of power; the subject that knows, the objects to be known and the modalities of knowledge must be regarded as so many effects of these fundamental implications of power-knowledge and their historical transformations. (1993: 30) The success of interactions varies depending on personal relations. Police Chief 1 holds the view that prosecutors do not appreciate the massive caseloads and difficulties faced by the Civil Police, yet Police Chief 2 maintains that the infrastructure at the Prosecution Office is adequate. However, they both believe little effort is made to improve working conditions for the police. Moreover, few seem to grasp the issues involved in homicide investigations, which require considerable expertise. It is also claimed that prosecutors do not understand the techniques employed to obtain confessions. On occasion, the Prosecution Office has described such 'techniques' as police torture. The police chiefs in the focus group disputed this interpretation and made a counter claim that a specialised Prosecution Office did not necessarily enhance homicide investigations. Demands for swift task completion, along with unrealistic expectations by the Prosecution Office displayed an ignorance regarding the complexity of investigations (Police Chief 1; Police Chief 2). This dispute between prosecutors and chief polices led to the suspicion that torture is used as an illegal investigative procedure. In the absence of adequate resources, it seems that extracting confessions under duress is sometimes accepted in police culture as an efficient way to 'solve' cases.
Given the high number of individuals arrested for various crimes, better infrastructure is deemed necessary. Further, police stations contend with competing priorities, notably people management versus the efficient coordination of investigations. Many case files end up at the bottom of the pile, waiting to be processed, while homicide arrests are relatively rare. To address the issues provoked by witnesses' fear of coming forward, one potential solution is a genuinely effective tip line. Based on the statements, distrust and hostility from the population are not the problem, although it seems that few calls are received from individuals offering up information freely.
The police chief accounts also highlight the state's failure and indifference concerning the expansion of the MAB. For the police chiefs, the state needs to invest in the Civil Police and expert police work, as this represents the only way to improve investigatory conditions and solve homicides. Moreover, they felt that further training and additional qualifications are not a practical solution because relevant programs are concentrated in Goiania, the capital of the State of Goiás. As Civil Police earn low monthly salaries of around R$2,900 (approximately US$800), this precludes their participation in additional training programs. The police chiefs claim that incentives for training are scarce, while the sacrifices are enormous.

Prosecutors' discourse
The low number of formally prosecuted complaints of homicide relative to the amount of officially recorded violent deaths (8.1 per cent) is a source of concern for the prosecutors who participated in the study. As with the police chiefs, several analytical categories guided our interpretation of the social representations advanced by prosecutors. Three are fundamental in understanding how this node of actors, informed by their representations and institutional cultures, adjust their practices: first, the poor working conditions across all the institutions of the criminal justice system undermined the quality of homicide investigations; second, prosecutors regarded the police chiefs as the primary culprits for ineffective investigations; and third, few judges wish to conduct jury trials because they are exhausting and require technical expertise. The prosecutors' accounts reveal social representations of heated power struggles between the different sectors of public security, entailing mutual accusations against the various agents. Additionally, the prosecutors' accounts suggest widespread incompetence and an inadequate work ethic among police who investigate homicides.
This focus group included five prosecutors who dealt primarily with voluntary crimes against life, including homicide. Of the prosecutors undertaking duties at jury trials held in cities surrounding the MAB, only one had previously lived in Brasilia. The rest were from different states. Living in Brasilia was attractive to agents posted in neighbouring cities, because the airport of Brasilia has routes to other Brazilian capitals. Career lengths varied from four to 14 years.
In the descriptions of the various cities, Águas Lindas is identified as a point of transit for many prosecutors-not because of an 'organisational desire'. Rather, the reason is precisely the traditional lack of interest in the post, which comes with an excessive workload and unsatisfactory working conditions. Some years ago, conditions were substandard in many cities, there was restrained demand and officials worked on cases implicating detained defendants. Judges accrued roles in different judicial districts and normally appeared for urgent hearingsthose involving detained defendants.
According to the prosecutors, the shortage of judges remains a major problem. Certain judicial districts apparently had no tenured judges, leading to excessive workloads while compromising the delivery of justice. Jury trials are unattractive to many judges because they are exhausting and require highly peculiar procedures. Some prosecutors suggested that demands from the National Council of Justice pressurise judges into conducting jury trials for old, static cases. Equally, the participants considered that jury trials lead to burdensome communication with the public as well as constant pressure, rendering the career unappealing to many prosecutors. Few people fit the profile required for the procedure and the work is physically and emotionally draining.
The prosecutors regard the external monitoring of police as a complex task due to the huge workload involved, particularly in cities where the Prosecution Office also oversees jury trials, which is the general rule, except in Luziânia. Monitoring procedures only occur contingently, in cases that warrant closer attention. Prosecutor 1 mentions that 'investigations often remain at police stations for months, without any monitoring or demands for investigations to be returned for follow-up'. This flatly contradicts the police chief discourse that the Prosecution Office make unrealistic demands for swift investigations.
The prosecutors in this study implied that homicide investigations are virtually non-existent. As a rule, inquiries are not pursued if no arrest is made or if the perpetrator is not identified soon after the crime is committed. Expert evidence is unreliable because experts are insufficient in number and they operate in such a vast and demanding field. Some focus group members explained that they sometimes must request exams in Goiania, leading to long delays. However, the inadequacy of expert evidence is not the only obstacle. The prosecutors also cite the problem of improperly managed police investigations, where police chiefs conclude procedures with the testimony of the Military Police officers who detained the perpetrators. The prosecutors criticised the Civil Police for prematurely concluding many investigations without eyewitnesses, which ultimately weakens the evidence presented at the evidentiary stage and undermines the arguments submitted at the jury trial. Fear of retribution also explains the absence of eye witnesses. Some prosecutors recounted cases in which witnesses retracted their statements in court, often motivated by doubt and a lack of confidence in state protection (Prosecutor 3; Prosecutor 5).
The participants propose various and recurring reasons to explain the challenges faced in elucidating cases. For instance, homicides involving gangs or score-settling require intricate investigations, making it difficult to address the identity of the offenders. Such cases usually concern score-settling, revenge and retaliation; therefore, they are imbued with a climate of fear and worry. Witnesses tend not to come forward, even if there are leads on the identity of the perpetrators or relevant suspicions due to past events (Prosecutor 1; Prosecutor 4). The criminal records of victims are decisive in investigations, since 'the police seem to show less interest in elucidating (cases) when the victim is apparently involved in the world of crime' (Prosecutor 1).
In describing the interaction with police chiefs, Prosecutor 4 hints at difficulties: 'there are police chiefs who do not want to work'. In practice, professional good will is crucial to the success of operations. Prosecutor 1 explains that 'there are committed police officers'. Notwithstanding these exceptions, the comments display general criticism concerning a perceived lack of professionalism and substandard training of police chiefs and officers. From his experience in other cities of the MAB, Prosecutor 1 outlines the utility of having specialised homicide units at police stations. Greater specialisation enhances the quality of work performed, in addition to the team's commitment. Though this may be insufficient to meet the pressing needs, the prosecutor suggests that the shortcomings of investigations are nonetheless rectifiable.
The participants regret the bad relations between police chiefs and officers. In their representations of police routines, the prosecutors relate that chiefs undermine officers by taking the credit for their work, although it is the officers who undertake investigations (Prosecutor 3; Prosecutor 5). As Prosecutor 2 mentions, officers, thus, feel devalued and neglected. Meanwhile, some prosecutors disregard the final reports from police chiefs-the files in which the Civil Police submit a legal analysis of the case. Occasionally, these reports are deemed useful, especially in complex cases. However, in simpler cases, prosecutors typically ignore them (Prosecutor 1; Prosecutor 2; Prosecutor 5). Such narratives, recurring in recent studies, highlight the tensions between the two groups, as the conclusions of police chiefs can be vastly different from those of prosecutors, who ultimately decide which cases to pursue (Costa 2015;Machado 2014;. Moreover, the prosecutors believed that poor relations between chiefs and officers led to discriminatory practices. One account provided by Prosecutor 3 describes an occasion in which an officer allegedly approached the prosecutor directly, requesting that 'measures be taken', given the chief's inertia and refusal to move forward with the investigation. In the power struggle between officers and chiefs, the former liaise directly with the Prosecution Office (Prosecutor 3). According to the discourse of the prosecutors, officers portray police chiefs as outsiders (Elias and Scotson 1965), as '[they are] sat on a chair, not out on the field [yet] still lap up the credit for the investigation' (Prosecutor 1).
In recent years, the frequent rotation of police chiefs, prosecutors and judges from different cities has contrasted with the regular presence of officers posted at stations for long periods. The prosecutors described a context of mutual distrust. Some prosecutors stated openly that the Civil Police do not investigate certain instances of police abuse (Prosecutor 2; Prosecutor 4). Although death squads are uncommon, they have undoubtedly existed and are difficult to penetrate. As stated, even when prosecutors unravel the facts, it is difficult to punish the members since the prevalent view is that 'they killed a bandit. The police officer killed a bandit because the justice IJCJ&SD 93 Online version via www.crimejusticejournal.com © 2019 8(1) system isn't doing anything' (Prosecutor 4). When the victim has a criminal record, the recurring representation is that 'they did society a favour. This is the stereotype and we note that what underpins the jurors' decision is this image of the vigilante police officer; they killed a bandit' (Prosecutor 5). In the prosecutors' depictions, the criminal-not the crime-becomes the 'object' to be eliminated.
Widespread fear makes the tip line an important instrument for enabling the population to contact the police. It has led to the execution of several important practices, including on-the-spot arrests. Nevertheless, it is generally difficult to obtain witnesses, as emphasised during the focus group. Tips alone are not always sufficient to establish findings. Consequently, many judges deny provisional measures if the grounds advanced rely exclusively on information from the tip line.
In the prosecutors' representations of the difficulties in conducting investigations, they agree that success often depends on participation from the victim's family and direct pressure on the police. Given the abundance of other, often similar, matters, pressure seems to be a key factor in determining whether cases are pursued or not. The witness protection program is also difficult to implement owing to the constraints imposed on witnesses. Victims only decide to participate it they either have no choice or wish to see the offender held liable.
In the statements on the exercise of discretion in daily practice, the police were portrayed as the archetypal discretionary actors. In the prosecutors' discourse, police discretion was considered unbefitting of proper police practice as it lacks transparency and accountability. When information is presented at the police station, the response 'this is not a crime' often signifies that a case is closed and no record is taken. In addition, the prosecutors' lack of access to the electronic database of statements and investigations undermines the access and authority of the Prosecution Office. The prosecutors lambasted the database of the National Institute of Investigation, which they view as unreliable (Prosecutor 1; Prosecutor 3; Prosecutor 5).
Prosecutors also expressed reservations regarding collective endeavours to perform jury trials, normally conducted in small rooms, where defendants are physically close to jurors without the due formalities or assurances of safety. Specially appointed, overwhelmed prosecutors and judges typically undertake the proceedings without the necessary precision (Prosecutor 2; Prosecutor 5). For the prosecutors, this transmits a negative image to victims and family members (Prosecutor 5). They claimed that the policy may have a merely functional rationale that satisfies the prescriptions of the National Council of Justice yet fails to deliver justice. Further, the prosecutors levelled criticism at the task forces involved in concluding investigations. According to one prosecutor, who monitored the teams, they concentrate on old cases, dating back prior to 2007. The archiving of cases often demands immense exertion and large-scale production of documents. This jeopardises more recent cases requiring dedication and special investigative skills.
Perceptions of the reasons behind homicides also differed among the prosecutors. Unlike the police chiefs, they acknowledge that many jury trials relate to bar fights, domestic violence or homicides implicating known individuals. Cases involving drugs are rare for prosecutors, but they arise at the level of the police. Indeed, as the prosecutors explained, these are precisely the types of cases that the police, who are familiar with the investigations, must manage (Prosecutor 3; Prosecutor 4). However, few cases are ever solved, which explains the differences in the assessments of the submissions presented to the Judiciary.

Judges' discourse
Two specific analytical categories guided the examination of the social representations of judges: First, the inadequate working conditions, especially at the Judiciary and among the Civil Police; and second, the contention that both police chiefs and prosecutors are partially responsible for the ineffective investigation of homicides. The Judiciary of the State of Goiás-unlike the Prosecution Office, which has become specialised in different fields-still struggles because of a shortage of judges and public servants to cover the innumerable jurisdictions created in recent years. In the MAB, judges perform diverse roles in the criminal domain, beyond sitting on cases of voluntary crimes against life.
Due to urban expansion and population growth, the state has created new judicial districts over the last 10 years. Nevertheless, the growing workload means that court structures are not entirely suited to delivering justice. Among the criminal judges in the cities studied, women head between five and seven criminal courts. In Luziânia, one court has acquired the jurisdiction for voluntary crimes against life and for enforcing sentences. However, it has been vacant for over a year. Initial reports point to a deficit of over 100 judges, as well as 154 unoccupied judicial units.
The focus group with judges was composed of two females and one male. In choosing a city in the Metropolitan Region, the proximity to Brasilia was an important factor. One female judge recounted the strain of travelling weekly to serve other judicial districts, resulting from a lack of available judges.
In assessing performance in homicide cases, Judge 1, with 21 years of experience at the Judiciary, explained that important changes have occurred. Given the dearth of judges, cases awaiting trial used to pile up. The Court of Appeals of Goiás has addressed this deficiency through joint endeavours, by bringing in judges from other judicial districts. However, the cases usually concern detained defendants, who represent a priority within the operation of the jurisdiction. In the past, Judge 2 explained that it was common to receive cases involving bar fights and vengeance. Nowadays, it is apparently different. According to Judge 1, 'they [the police] arrive and if there are three or four people in the group, all four die. It is impossible to even know who the target was … Nobody wants to testify, there are no witnesses, because in reality if a person testifies they will also die, so there are no witnesses. So, police investigations are accumulating homicides with no [identifiable] perpetrators'.
Further, the judges cast aspersions on the investigations and facilities at police stations. The lines of inquiry adopted are seldom aimed at identifying the perpetrators. Many arrests are for theft, drug trafficking and domestic violence, but rarely homicide. Hence, investigations pile up and are numerous and interminable. The judges explain that they refer many cases back to police stations for further investigation, where they remain for years, never to emerge again. Judge 2 revealed: I requested that they call the police station and check on the investigation. The officer who responded said he did not know where the file was, in other words, that the investigation had not been pursued. These investigations that are not pursued are investigations that do not have a [identified] perpetrator.
The National Council of Justice sets specific targets, resulting in an enhanced commitment to closing cases, with the help of taskforces. Nevertheless, the identification of perpetrators is rarely broached. The judges believed that the situation has worsened in recent years. They stated that material conditions are awful owing to a shortage of basic equipment, printers and cars. Judge 3 assessed the Civil Police as follows: 'The police in Goiás does not exist anymore; they catch people in the act and pass it [the case] on to the Judiciary, that's all'.
Experts were also considered unreliable. The facilities are unsatisfactory for such a vast area, given the lack of material and human infrastructure to undertake basic exams. The judges singled out the Legal Medical Institutes and the infrastructure for expert evidence in Luziânia and Formosa. The latter city serves up to 30 municipalities, along with the entire northeast region of Goiás, despite having only one court-appointed expert. In addition, the Forensic Police reports submitted as evidence are regarded as inadequate.
Routinely, judges only receive reports at the end of the evidentiary stage and the committal stage. Jury trials sometimes take place without the appropriate documents, including forensic assessments in cases of attempted crimes, or other routine exam reports. Thus, 'of the 100 jury trials I have held, if I'm not mistaken, one or two had crime scene reports, [while] one or two had no ballistics comparison report. It is also common for jury trials to happen without an autopsy report' (Judge 3).
The judges portrayed the Military Police in a more positive light than the Civil Police, describing them as more professional and devoted to their job. One judge recalled that, years ago, he noted numerous instances of abuse by the Military Police. The more experienced, female judge describes how she '… also picked up on the idea that the Military Police used to batter [people]' (Judge 1). However, abuse now appears to be far less common. Another judge stressed the numerous successful on-the-spot arrests made, including homicide arrests, which are largely due to the prowess of the Military Police.
The absence of appropriate levels of state support for social infrastructure was a recurring theme in judges' accounts of the reasons for homicide. While the population has grown substantially, urban infrastructure remains substandard. Cities lack schools and basic public services and children grow up on the streets. Participants associate these local realities with a universe that is far removed from their own. As the judges reside in Brasilia and travel to cities with serious social and public security problems, they naturally identify little with the social context in which they operate. The social conditions of the way children and families live in the barrios also create what Judge 3 called the 'banality of violence': 'We always joke that this is linked to cachaça [Brazilian liquor] and women. In general, drugs and alcohol are involved … Lots of banality … We, people of the middle class, we hold values [that are built up] via our daily lives, our relations with society, and we need to understand that these values are not the same as theirs'.
The judges also directed opprobrium at the Prosecution Office for being too intimidated by the Police Chiefs. Judge 3 was of the view a that prosecutors should not accept the negligence of the State of Goiás regarding public security and, especially, the debilitated Civil Police. The situation is so critical as to warrant institutional action. Judge 3 argues that 'it's up to the Prosecution Office to act … the Prosecution Office leadership should activate the Prosecution Office so that it sorts out the police, except we know things don't work this way'.
As unveiled in the assessments, the peculiarity of certain cases often complicates investigations. Many concern drugs and battles between rival gangs. The police often know or suspect who the offender is but there are rarely any witnesses. Indeed, 'people do not feel safe and fear is widespread' (Judge 1; Judge 2). It is also difficult to implement the witness protection program. The judges stated that few people agree to participate because of fear of reprisal and the inherent problems faced by witnesses under 'protection'.
The judges also lamented the low percentage of investigations converted into criminal proceedings. Nonetheless, there is a high conviction rate for cases that reach trial, reflecting an association with the 'banality of the violence'. Judge 3 explains: 'we feel society is also dissatisfied. Sometimes, in situations where you yourself may have doubts, you take [the case] to jury trial and the person is convicted'. Meanwhile, Judge 2, who practices in a city with a high crime rate, attributed recent acquittals in her court to intimidated jury members who feared persecution. Perceptions of impunity are widely held among judges, jury members, police, prosecutors and the public.
The conditions for enforcing sentences were a matter of further concern for the three judges. The state's indifference is such that the Judiciary interdicted the prison establishment of Planaltina. They also condemned the relaxation of prison sentences under prison regimes that offer IJCJ&SD 96 Online version via www.crimejusticejournal.com © 2019 8(1) reduction in sentences as incentives for good behaviour and the conditions of the prison regimes in the MAB, which mix criminal offenders with the mentally ill-criticised as a common situation in many Brazilian prisons.
Despite the appalling prison conditions that we verified in the fieldwork, the judges nevertheless justified the need for tougher prison sentences. For instance, Judge 3 regarded lenient prison sentences as an incentive for crime. For this judge, prison functions 'to exclude the citizen from society'. Another judge defended the application of harsh sentences, even if appellate courts reduce them, arguing that ' … since the sentence is not only about rehabilitation, it also needs to punish [criminal] conduct for the sake of members of society' (Judge 3). The idea that punishment of the offender has a social function is consistent with Durkheim's (1984) theory that punishment serves as a symbolic mechanism to reinforce the collective conscience. Judge 2 suggests that 'if you also remove people from social interaction, at least you know that they won't commit crimes during this period'. According to Judge 3, in cases where the offender is believed to be a psychopath, the fear is that they will kill again, thereby making it necessary to remove them from social interaction via tougher sentencing.

Conclusion
In analysing the performance of organisations in the criminal justice system, one of the central issues is to interrogate the distinct rationales of organisational culture that emerge at a local level (Machado 2014). Understanding the local contexts, typified by spiralling violence, calls for an examination of several factors. The information made available by the Secretariat of Public Security of Goiás illustrates that the increase in homicides over the last decade has coincided with the disorderly urban growth of the cities composing the MAB. This growth has not been matched with state-supported social infrastructure, such as schools, hospitals and roads. This was the context in which large numbers of homicides were not adequately investigated by the local police. The lack of investigation frustrated the Prosecution Office, who complained of police ineptitude and bungling. In response, police complained that they had neither the resources nor the willing eye witnesses to adequately investigate homicides, especially where they feared reprisal. Hence, they rationalised the use of police discretion-even the use of duress and force-in extracting confessions as these are the easiest homicide cases to resolve when insufficient investigative resources exist. The judges complained that the Prosecution Office was too intimidated by the police and needed to be more assertive in demanding proper homicide investigations.
Importantly, the focus groups exposed the challenges involved in performing functions effectively, in addition to the suboptimal working conditions at all the institutions of the criminal justice system in MAB. Moreover, the narratives from prosecutors, police chiefs and judges create important social representations about homicide. Without ignoring the obstacles, it is nonetheless appropriate to discuss these narratives as social representations.
As Moscovici (2001) argues, the function of representations is to familiarise the unfamiliar. Beyond what appears in the accounts there is a layer of unstated assumption. It is important to question what lies beneath the social representations that emphasise the actors' troubles to carry out their duties. Thus, participants' representations cannot be imputed solely to the substandard working conditions or the lack of resources, infrastructure, equipment and human resources. They raise points that are implied, unstated and perhaps even forbidden and provide insight into the cultures that frame homicide investigation. The narratives evoke the perspectives of those who, in seeking to understand and situate themselves in new or unfamiliar realities and contexts-particularly the growth of violence and criminality-ultimately place themselves within the available reserve of knowledge, thereby making these realities and contexts normal within the organisational culture. Under this reserve of knowledge, the commonplace, the familiar and the normal is predicated on the idea that 'nothing works', enabling actors in these institutional networks to justify themselves despite their inability to follow through with procedures. The social representations assemble devices that conceal the underlying essence of the matter, namely that this is how the system functions (Vargas 2014). Alternatively, it is difficult to treat all the organisations that handle investigations, criminal prosecution and trials as components of the system because they derive from different trajectories with their own directions and rationalities (Machado 2014).
Significantly, in the agents' representations, this loosely articulated dynamic is portrayed as a sign that 'nothing works'. The dynamic plays an essential, functional role within the system, as research shows. Taking this further, Garland (2001, pp. 61-63) describes a wave of demoralisation that undermines the credibility of key institutions of crime control and, at least for a period, the entire criminal justice system. Notably, in competing for a monopoly over procedures, the operating agents themselves are often important elements in these processes of disillusionment.
The notion that 'nothing works' is, or may be, a 'concealment' that purports-albeit subconsciously-to overshadow the fact that control over crime and the functioning of the system exists in consonance with certain prerogatives, interests and strategies. Another possible interpretation is that the idea that 'nothing works' alludes to an entity operating according to 'certain interests'. The actors' representations of inertia, conveyed across the testimonies of judges, prosecutors and police, certainly suggest an attempt to 'explain' the impotence of the system in investigating homicides in MAB.
However, they effectively hint at a potential form of interchange between the various bodies that compete for power and a monopoly over expertise in criminal procedures. The capacity to read the 'invisible' in social representations enables analysts to understand the meaning behind practices. Moreover, it allows them to appreciate the distinct levels of tension that imbue interaction and lack of interaction, as well as the explicit or latent conflicts between the different actors that compose the criminal justice system. Although the interviewees offered divergent interpretations of the inadequacies of the police and the prosecution in investigating homicides, treating each other as outsiders (Elias and Scotson 1965), there was a common view that most unsolved or inadequately investigated homicides in Brasilia involved turf disputes or drug wars between rival gangs. As homicide victims are portrayed as protagonists in a war over illicit markets, there was an implicit suggestion that 'those homicides are not very significant to society'. In the social representations of the judges, prosecutors and police, those killed in drug wars and rival gang fights were victims unworthy of investigation. They were represented in this discourse as having deserved their victimisation and sealed their fate. Consequently, this discursive representation was used to justify the priority accorded to the investigation of homicides regarded as more worthwhile, since the justice system does not have sufficient resources at its disposal.  (40): 294-325. We thank both reviewers and Kerry Carrington for the detailed comments and positive suggestions. 2 Studies also show that the investigation and prosecution of crimes have a debatable influence on crime rates (Costa 2004: 47-49). 3 In this item, we reproduced much of our report, following the interpretation advanced in the focus groups (Costa et al. 2013).