Public Security, Criminal Policy and Sentencing in Brazil during the Lula and Dilma Governments, 2003-2014: Changes and Continuities

Since mid‐1980s crime rates in Brazil started to increase with the transition from military dictatorship to democracy, a period marked by the rise of poverty and hyperinflation. Given levels of police corruption and protection of criminal gangs, trust in the criminal justice system was low, leading to a heightened dependence on private security, gated communities and the politicisation of law and order politics. An impressive punitive turn began at in the 1990s. This paper analysed penal policies under successive left‐wing Lula and Dilma governments from 2003 to 2014. During this period the rise of ‘postneoliberal’ and ‘postneoconservative’ politics, led by the Workers Party, developed governmental strategies and initiatives that formulated and legitimized rich and complex expressions of traditionally left‐wing politics. Despite the simultaneous implementation of distributive policies, the increase of human development levels throughout the country, the reduction of social inequalities and the redirection of the official discourse of Federal Government security policies towards crime prevention, incarceration rates continued to increase. This article examines 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.

The following Figure 1 represents the homicide rates in Brazil from 1980 to 2012. A growth trend can be identified from 11 homicides per 100,000 in 1980 to 29 homicides per 100,000 inhabitants by 2012. 3 A downward trend and stabilisation can be perceived from 2003 onward, when some important States like São Paulo and Rio de Janeiro experienced a significant drop in homicide rates. This did not have a wider impact by virtue of the southern and north-eastern States experiencing increasing rates in the same period. However, from that time onward, a clear shift in the homicide curve can be noted. This has been attributed by researchers to changes in policies and social dynamics, one of which was the implementation of the Statute of Disarmament, which established an effective arms control policy and a ban on the carrying of weapons by the civilian population, as well as the consolidation of the criminal group 'del Primer Comando de la Capital' (PCC) in São Paulo. PCC emerged in prisons and had a monopoly on the illegal drug trade. The policy thus contributed to the reduction of homicides caused by disputes over territory. 4 The year 2012 was marked by a new record in homicides, reaching the highest level since the start of the measurement of statistics. This objective component is accompanied by a 'subjective' component, or 'sense of insecurity', which can be understood as a change in expectations and feelings of the inhabitants of large-and medium-sized cities in relation to the possibility of becoming a crime victim. This unease, in turn, is expressed by innumerable self-protection and avoidance tactics, which have a considerable impact on behaviour in everyday life and constitutes a new component of the insecurity crisis.
These objective and subjective components of the insecurity crisis have become recurring themes in news and media debates, to the point that they have now become one of the fundamental forms of sense-making by the media. Additionally, it is viewed as a sign of State failure by Brazil's surrounding countries, due to the fact that security policies and criminal justice are traditionally arranged by State-level governance (police, judiciary and the penitentiary system). As a result, even beyond the growth of insecurity, there has been a growing 'politicization' of the issue, which increasingly means that the 'political' arms of State structures are seen as part of the problem, or as the one that, together with experts (for example, lawyers and criminologists), must make key decisions that act in the area, while moving the administrative part to a secondary position.
Although many argue that something has to be done about crime, the trust in the criminal justice system remains very low. It is viewed as slow, ineffective and biased in favour of the wealthy and powerful. 5 Brazilians therefore rarely involve the justice system in dealing with their conflicts but, instead, rely on their existing interpersonal networks to resolve issues. One consequence is that the middle-and upper-classes isolate themselves by moving into closed compounds with electric gates, walls and private security. This is reflected by the fact that the private security industry employs more people than the public security system in Brazil. In several regions of the country it is not unusual to hire a professional hitman in order to eliminate political opponents or suspected assailants. Additionally, in recent years urban 'militias' emerged in the favelas of Rio de Janeiro. These militias are controlled by the police and charge residents money in exchange for protection of the community from drug traffickers and other assaults.
Efforts to build a public security policy committed to guarantee civil rights and improvement of the control of police and police activities have failed since the 1980s. This occurred because they were blocked by the ruling punitive populism at the time and the inability of the government to confront often corrupt and violent resistance by corporate police structures. During the military dictatorship, political murders were secret acts by the government that were never publicly admitted. Under popularly elected governments, homicides committed by the police during violent and unprofessional raids in the urban periphery -ones that often remained unpunished -gained popular support. This was with the apparent aim of controlling crime within a situation where there was no trust in the justice system. In practice, however, they served as a currency agreement between the police and trafficking groups, characterizing what Misse (2008) termed 'political goods'.
Due to the growing perception of a public security crisis, the Federal Government was forced in the mid-1990s to assume a stronger leadership role in the development and induction of security policies in the States. It did so by creating the National Secretariat of Public Safety (in 1996)  Despite the complexity of a political system that is made up out of 30 parties and that depends on the construction of coalitions in a rather heterogeneous parliament that often is based on the exchange of political favours and patronage, the fact remains that, since the stabilization of the currency due to the 'Real Plan' (Plano Real) in 1995, there was a steady growth of social expenditures by the Federal Government. Between 1995 and 2010 the federal social expenditure more than doubled, from 234 million Reales in 1995 to 638 million in 2010, an increase of 172 per cent over 16 years. In relation to the country's Gross Domestic Product, state social expenditure increased from 11.24 per cent in 1995 to 15.54 per cent in 2010 (Castro 2012).
During Fernando Henrique Cardoso's first term (1995)(1996)(1997)(1998), the Gini index of income inequality (where 0 corresponds with perfect equality and 1 with perfect inequality) remained virtually unchanged, and then fell slightly from early on in his second term (1998)(1999)(2000)(2001)(2002)   The return of economic growth to the country in 2004 had beneficial effects on the labour market, such as the increase in employment and the rise of wages. The reduction of unemployment and the increase of GDP could be verified by a heightened demand in unskilled labour and the increased income of those occupying the base layer of the so-called social pyramid. This is reflected in a decrease of 5.7 per cent in the Gini index from 2002 to 2007, with an observed acceleration in equality in 2004 (IPEA 2008). However, this trend does not significantly alter the historical picture of social inequality, which remains the main challenge to be faced by current and future governments and depends on the unlikely continuing stability of the economic growth rates experienced in recent years.  526 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 IJCJ&SD 150 Online version via www.crimejusticejournal.com © 2017 6(1)

Criminal policy and legal reforms
To analyse the movements of criminal policy and its impact on incarceration rates in Brazil involves addressing the complexity of crime control. The National Congress has the sole power to legislate in criminal matters in Brazil. The Congress, however, does not have the autonomy to define criminalized behaviour beyond what has been established by federal law; nor is it able to create procedural rules. It is important to note that one of the characteristics that marks the Brazilian post-dictatorial political system, especially after the end of the Collor de Mello government in 1992, is the consolidation of a presidential system based on the development of coalitions. These coalitions are made up of parties that exist in the broad political landscape and need to create partnerships that reflect the majority in parliament. In this sense, the governmental system adopted is 'presidentialism', in which governments resort to coalitions in order to garner support for their initiatives (Limongi 2006). In fact, this is the modus operandi of most contemporary democracies. In this system minority governments face serious governance issues, while governments with majority coalitions are able to mitigate such difficulties. In Brazil, the coalitions in this sense act as a basic operating budget of functioning presidentialism, in which the government is more successful adopting legislation when it can consistently count on the support of a broad coalition of parties. Therefore, members of the parties that are part of the coalition supporting the president generally follow the vote of the government's leader (Limongi 2006).
Based on empirical studies Limongi (2006) showed that the Lula government, in his first term, received an average support of 89.1 per cent of 164 votes from the deputies of its supporting parties, thus ensuring the maintenance of governance (Limongi 2006). In addition, it appears that the losses are not the result of indiscipline of the parliamentary base of support, but rather its internal division. In other words, losses occurred in situations in which at least one party did not follow the direction of the government leader (Limongi 2006).
It thus becomes evident that whether proposals of the Executive Power become successful is not the result of case-by-case negotiations. Rather, it is the ability of the government to control legislative production, which in turn is based on the interaction between the Executive of the party in power to produce an agenda whilst maintaining majority support (Limongi 2006). In other words, because the process of legislative decision favours the Executive Power, objections of the Congress become more predictable. In order to structure and preserve support, it is important to focus on plenary debates that will most likely be accepted and avoid those that might present a risk to the coalition (Limongi 2006).
General research in terms of legal reforms in the area of criminal justice has pointed towards the existence of a hypertrophy or inflation in criminal rules, which invade areas of social life that previously were not covered by criminal sanctions. It is a criminal remedy used by political decision-makers in response to almost all types of conflicts and social problems. The penal response is the symbolic response by the state to the growing demands of security and punishment by society, which are expressed by the media. However, the response occurs without measuring its instrumental effectiveness as a means of reducing crime, or forecasting the potential direct impact on the prison system, or assessing any other possible side-effects that might result from the use of force.
As for the coordination of public security legislation in the country, the landscape in Brazil was marked by a rigid division of powers, in which States were granted great autonomy in designing IJCJ&SD 151 Online version via www.crimejusticejournal.com © 2017 6(1) and implementing their own policies and initiatives. According to Sá e Silva (2012), the Federal Government action was summed up by the mobilization of the Federal Police (responsible for the investigation of crimes within the jurisdiction of the Federal Court) and the Federal Transit Police. The second notable characteristic was the centrality of extensive action of the State Police, backed -or, rather, ignored -by the Federal Government.
The political landscape starts to change with the start of the Fernando Henrique Cardoso government (1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002). The National Secretariat of Public Security (SENASP) was created under the Ministry of Justice and launched the first National Plan of Public Security in 2000. In addition the National Public Security Fund (FNSP) was formed, which allowed greater power induction and ensured systematic initiatives in the Federal Government sector.

The National Public Security Plans during the Lula-(2003-2010) and Dilma Governments (2011-2014) 6
After three previous attempts (in 1989, 1994 and 1998), Luiz Inacio Lula da Silva was elected President of the Republic in 2002. With a career built on the union struggle in the 1970s, which culminated in the founding of the Workers Party in the early 1980s, Lula became president in a coalition that united the parties of the so-called democratic and popular bloc. The coalition was mainly comprised of the Brazilian Socialist Party and the Communist Party of Brazil and was located to the left of the national political spectrum. These parties prioritised the formation of a coalition that aimed to implement policies to combat poverty and to reduce the huge social and regional inequalities. In addition it increased social spending, implemented state-organised income distribution, and directed international policy towards greater autonomy while prioritising stronger relationships with Latin American countries and other so-called emerging countries. These elements influenced the government to be considered left-wing. This characteristic was not substantially altered despite the broadening of partnerships in terms of the Congress, which resulted in further ambiguities and difficulty to implement more profound reforms.
Not coincidentally, public safety was one of the priorities identified in the electoral process, with a focus on a greater role for the Federal Government in reducing violence. Lula had gone through the experience of governing some important States before his victory in the presidential elections, and had faced difficulties in managing the police apparatus, particularly in Rio Grande do Sul. This was due to left-wing programmatic weaknesses on the issue of security and to the boycott of the police forces by important sectors that were committed to the defence of human rights and historical allies of social movements. Lula therefore proposed a National Plan of Public Security during his electoral candidacy, in which the following was specifically expressed: The Brazilian people are dominated by a general feeling of insecurity and, for that reason, our government will seek to establish a Public Security System on a national level. Social exclusion, with unemployment as its main expression, affects millions of men and women and results in daily despair for many people, if not crime. The statistics show that firearms as the causa mortis of youth. And the impunity with which organized crime acts, threatens to compromise the often infiltrated functioning of democratic institutions. The same impunity can be observed in the hundreds of crimes committed against rural workers, unionists, lawyers and religious people who fight for the Agrarian Reform. (Workers Party 2002) It became clear that the commitment to arms control policy was inherently related to the fight against the impunity of crimes committed by criminal organizations who were opposed to social movements and human rights. The plan also recognized that: The lack of human and material preparation of the police apparatus and the slow judicial system encourage violence and aggravate crime, which in turn reproduces itself and broadens due to the absurd prison system. The impunity of the powerful and brutal misery of much of the population, which is in stark contrast with the continued appeal of consumerism, causes a crisis of moral values that fuels violence. Both the rich and poor are frightened and locked in their neighborhoods and homes. Forms of sociability for Brazilians becomes increasingly restricted. The poor are stigmatized as criminals and civil coexistence is threatened. The organizations responsible for national defense are put to the test by the increasingly more insolent acts by organized crime. (Workers Party 2002) Therefore, the National Public Security Plan encouraged, on the one hand, social policies that allowed for the reduction of inequality and contributed to social appeasement and, on the other hand, improved the performance of public security organs, in particular that of the Federal Police but also of the State Police. The policies would be carried out with the use of new technologies and the creation of a database that would allow for the coordination of State forces in efforts to fight organized crime, with a special focus on drug trafficking and arms smuggling.
The national public security policy was assessed to lack planning and management capacity due to fragmenting actions by several States in the area. The proposed model therefore involved increasing States' understanding of the actual and current status of violence and crime, in order to promote preventive, strategic and targeted actions, which would permanently be monitored by actors of public security and in the criminal justice system.
The actual implementation of the proposal, however, was only partially carried out. Moreover, the organizational reforms of police structures was completely removed from the agenda with the fall of the National Secretary of Public Security in the second year of Lula's term. This was due to pressure from groups that were highly involved in public security organizations, particularly the Federal Police, and that wanted to maintain their central role in policy coordination in the area.
The withdrawal of support by the Ministry of Justice, led by lawyer Marcio Tomaz Bastos from San Paulo, and the concomitant fall of Luiz Eduardo Soares who was in charge of public safety, meant that the development of knowledge, the structuring of research networks and the systematic production of data were brought to a halt. However, this was taken into account only to a limited extent when informing strategic management. The first agreements between the Federal Government and the States to implement prevention programs and training for public safety professionals were seen as verification of the progress of prevention policy. Even though the plan was still situated in the police schools, they received an important contribution to establish the National Network for Advanced Studies in Public Safety (RENAESP). This facilitated more permanent contact between police, Brazilian universities and other safety operator in order to enable consolidation between, or creation of, several research groups in the area.
During the electoral process which led up to Lula's second term, the development of the Government Security Plan was coordinated by sociologist José Vicente Tavares dos Santos. Tavares dos Santos was a professor and researcher at the Federal University of Rio Grande do Sul with a proven track record in the field of public safety and the sociology of violence. The proposal finally confronted major problems such as the question of obsolescence of the State Police; the failure of the penitential model; the fragmentation of the public security system; the low success rate of State agency interventions and service provision; the developing instrumentalization of police intelligence; the lack of local public power to deal with the topic; and the lack of interaction and involvement in the discussion, evaluation and development of public policies in the area. (Tavares dos Santos 2006 The proposal was articulated around the concept of public and citizenry safety, with initiatives aimed to strengthen State security and criminal justice institutions and, especially, dismantle organized crime in Brazilian territory. It was also aimed at stimulating new patterns of coexistence between people; at strengthening and improving participation in democratization; and at promoting human rights; while at the same time developing prevention policies to counter crime and (physical and symbolic) violence, in order to ensure peace as a fundamental value in Brazilian society. In order to do this, three basic policy guidelines were presented: first, democratization and participation of both society and the state in the management of safety in an articulated manner; second, the integration of public security institutions; and third, defining public security policies in scientific information and research systems.
The proposal was developed in 2007 following the broader trend of introducing State performance into the area of public safety, which had begun with the establishment of the SENASP and was soon followed by a proposed Single Public Security System (SUS). It was developed by the Ministry of Justice, which was led at the time by Tarso Genro, a lawyer and political figure in the Workers Party who had been twice elected as the mayor of the city of Porto Alegre in Rio Grande do Sul. Genro launched the National Program for Public Security with Citizenship (PRONASCI) with the use of the previously developed plans, at the same time bearing in mind the complexity of the phenomenon of violence, and with greater emphasis on the socio-cultural roots and underlying dynamics of violence and criminality. The program proved to have an understanding of the inherent embeddedness of security issues within other social, cultural and environmental factors such as education, health, culture and infrastructure services.
The PRONASCI was based on two major principles: structural measures and local programs. Among the first structural projects developed by PRONASCI were the improvement of professional public security and criminal justice system; the restructuring of the penitential system; and the combatting of police corruption and organized crime and the communities' involvement in it. The local programs that are worth mentioning are Territory of Peace, Women for Peace, the PROTEJO and the Community Justice programs. In return access to justice and the recovery of public spaces were guaranteed through the revitalization and redevelopment measures.
Despite advances in the design of the plan and linking the proposals and programs to resources for their implementation, the results were quite fragmented and dispersed. The reviews showed that there were issues with the limited room to question the proposed solutions, the pure and simple accession of the States involved, the lack of adequate mechanisms to monitor the implemented policies, and the abandoning of the planned structural reforms of public safety organizations. 7 In 2011, Rousseff won the second round of presidential elections over Jose Serra, a PSDB candidate with conservative views on administrative management who also spoke out in favour of moral issues such as the criminalization of abortion and criminal hardening. Dilma Rousseff subsequently took office with 373 partnered deputies (against 111 for the opposition) and 62 senators (against 18 for the opposition). However there was no unity in political programs which made it difficult to carry out the planned reforms.
The area of public safety did not receive much attention during the election period and the National Public Security Plan was not even formulated and released until Rousseff's second year in office. When it did appear, it turned out to be much more discreet than previous proposals and had been transformed from an integrated whole to a number of points to be implemented according to the conveniences of the partnering State governments. Even though Dilma Rousseff's election campaign had supported proposals on public security issues that closely resembled those of the Lula government, the fact was that, once elected, the plans were redirected to focus on other priorities.
IJCJ&SD 154 Online version via www.crimejusticejournal.com © 2017 6(1) As president of the Republic Dilma Rousseff argued for a more reluctant role of the Federal Government in public safety based on Article 144 of the Federal Constitution, which attributed the responsibility of Public Safety to State governments with the use of civil and military police. The PRONASCI was discontinued, due to the withdrawal of Federal Government support for municipalities to implement crime prevention. As its replacement the government presented the 'A Safer Brazil Plan', which was never implemented, even as a pilot study in the State of Alagoas where homicide rates reached 92 per 100,000 inhabitants in 2010. Additionally, resources were redirected towards a focus on border control policies, by establishing partnerships between the armed forces and State Police and Federal Police. At the same time policies were prioritised that assured the security conditions in the 12 State capitals during the 2014 FIFA World Cup by developing command and control centres coordinated and occupied by the armed forces and the Federal and State Police.

Incarceration rates and penitential policy
Despite the simultaneous implementation of distributive policies, the increase of human development levels throughout the country, the reduction of social inequalities and the redirection of the official discourse of Federal Government security policies towards crime prevention, it is noteworthy to mention that incarceration rates continued to increase throughout the analysed period.
Whereas the average monthly incarceration in 1990 was 61 prisoners per 100,000, by 2014 this monthly average (to July) had increased to 300 prisoners per 100,000 inhabitants. As shown in Figure 3 this growth trend was virtually uninterrupted over the entire period.  The prison population from 1990 to 2014 in absolute numbers as measured in the month of June grew by around 575 per cent (from 90 per 100,000 inhabitants to 607 per 100,000 inhabitants) (Figure 4), while population growth over the same period was only slightly greater than 30 per cent.

Figure 4: Growth of incarceration in Brazil, in absolute numbers (per 100,000 inhabitants), June 1990-June 2014
Source: DEPEN (2015) By 2005, after a period of relative stability from the start of millennium when the prison population was around 230,000 prisoners, incarceration rates started to grow significantly in absolute numbers. This increase of approximately 8 per cent per annum remained relatively consistent, reaching a total of 574,000 prisoners in June 2013. By July 2014 the rates had increased to 300 inmates per 100,000 inhabitants. These rates varied greatly by State.
Among the States with above national average incarceration rates were: Acre with 521 inmates per 100,000 inhabitants; Mato Grosso do Sul with 519; São Paulo with 503; Rondônia with 495; and the Federal District with 476 inmates per 100,000 inhabitants ( Figure 5). Conversely, the State of Piaui ranked among the States with below national incarceration rates with 95 prisoners per 100.000 inhabitants. In addition Maranhão and Bahia had comparatively low average rates with, respectively, 99 and113 prisoners per 100,000 inhabitants.
Of the total number of prisoners in Brazil in June 2013, 56.2 per cent was serving their sentences. Prisoners with an interim status, some even without a conviction, represented 43.8 per cent of the total, which was an increase of 2 per cent compared to December 2012. The constant growth in the number of provisional prisoners over the last decade proves the ineffectiveness of the new law relating to precautionary measures in criminal proceedings (2011, art. 12.403). The law was devised to give the judiciary a series of new possibilities to ensure the advancement of the process without the need to imprison the accused. One of these measures was electronic monitoring of the accused, a method that is still little used either due to resistance of the judges or due to the lack of available State structures.  Crime against property is the number one type of crime that generates imprisonment in Brazil, representing 49.1 per cent of all convictions in 2012. Crimes included in this category are simple and aggravated theft, simple and aggravated robbery, larceny, extortion, extortion by kidnapping, wrongful appropriation, embezzlement, and intentional and unintentional receiving and handling of stolen goods. The next largest group of prisoners were those convicted of drug related crime, which represented 25.3 per cent of all crimes in 2012. This number has been growing since the implementation of Law 11.343/2006, which increased the minimum sentencing for drug trafficking-related offenses. Beyond the increased sentencing, the law had an unexpected side effect which was the option to charge under drug trafficking situations that were previously typified as personal drug use. Because Brazil has no clearly defined criteria for the amount of drugs that represents personal use, it is the profile of the accused that gets weighted. This reinforces the selective nature of the system and contributes to the overcrowding of the prison system. The third highest ranked type of crime is crimes against a person, accounting for 11.9 per cent of all sentenced prisoners in 2012. Crimes against a person include manslaughter, (premeditated) murder, kidnapping and false imprisonment. The so-called crimes against customs, which include rape, violent attempted assault, corruption of minors, and international and domestic human trafficking, accounted for 3.9 per cent of the total. Finally, it is noteworthy to comment on the seemingly high percentage of prisoners ( According to the data by DEPEN, the increasing decision to opt for imprisonment in Brazil (as can be observed in Figure 6) is linked to two main discourses: the steady increase in preventive incarceration and the increasing criminalization of the drug market. These discourses have not been accompanied by improved prison conditions, which in turn contribute to prison violence, the spread of diseases and the growth of criminal organizations. In 2011, the deficit between number of available places and actual numbers incarcerated stood at 175,841 places. By 2012 this number had risen to 211,741, an increase of 20% in the short period of one year, to a national average of 1.7 prisoners per place available in the prison system. In some States the situation was even more acute and the proportions of inmates per bed reached over 2, as in Bahia (2.2), Rio Grande do Norte (2.3), Amapá (2.4), Pernambuco (2.5), Amazonas (2.6) and Alagoas with the record of 3.7 prisoners per place. With no guarantees of places in the prison system and the growing number of prisoners each year, it seems clear that prisons in Brazil end up assuming a criminogenic role. The prison system becomes a vehicle by which crime becomes worth the risk and in which the State's own performance is delegitimized. The responsibility for this phenomenon can be shared by the Federal Government and State governments, which have the duty to guarantee places in prisons, and by the Congress, that is unable to advance criminal law reforms and define a more rational and judiciary criminal policy. Further, delays and selective actions end up worsening the high provisional imprisonment rates.

Where are Brazil's penal policies heading?
The current Brazilian public security system, which was developed at the start of the Federal Constitution in 1988, established individual and collective security as a legal commitment. However, since then Brazil has not been able to confront its major structural issues in the public security and criminal justice system. military and civilian police dividing extensive policing and criminal investigation activities, strengthens militarism and bureaucracy. Both concepts are characteristics of the two State police forces and have remained unchanged. As a result, violence and corruption remain trademarks of the Brazilian police. The external mechanisms of control to manage police activity are ineffective in altering this picture. Additionally, criminal investigation is a highly inefficient and bureaucratised process with a low success rate for serious cases -such as murder -as a result. Additionally, the judiciary is slow and inefficient in guaranteeing due process for vulnerable social groups who often are the targets of police selectivity, which, despite the steady increase in incarceration rates, results in a general sense of impunity in Brazilian society. Prisons therefore are overcrowded, dominated by criminal organizations, and largely occupied by provisional prisoners and young drug dealers.
An enormous increase can be noted the incarceration rates over the last decade, which has brought the results close to the historically high mark of 600,000 prisoners, compared to a prison population of around 150,000 inmates in the late 1990s. The general data for the prison system show that especially preventive detention -that is, people who are in prison without a sentence -has increased and now accounts for 40 per cent of all prisoners in the country. Increased incarceration rates, resulting from a punitive demand supported by the parliament (primary targeting) through the performance of public security and criminal justice organs (secondary targeting), did not have the expected effect of decreased crime. This can be explained by the selective nature of the criminal justice system which only attacks the lower levels of the criminal chain and imprisons individuals that are easy prey for the criminal organizations that rule the illegal market in and outside of prisons, due to their social vulnerability (Oliveira and Azevedo 2012).
Despite a growth in recent years of the number of cases against white-collar criminals, the increase has not produced a significant change in the profile of the prison population. This profile is still characterized by individuals with low incomes and limited education and who, in the vast majority of cases, are imprisoned for crimes against property (theft) or for drug trafficking. Because they are within the penitentiary system, they will most likely become permanently ingrained in the network of illegalities.
It is important to analyse the diversity and ambiguity of current governmental policy. According to Matthews (2013: 14): 'there are mixed messages and competing imperatives emanating from official sources'. It is therefore possible, at one point, to be talking about alternative measures and 'de-incarceration' and, at another, about more severe measures to prevent impunity. Official criminology thus becomes increasingly dualistic, polarized and ambivalent, a fact that does not help the comprehensiveness of some aspects of the current situation of newly emerging public policies that are distinct in their fundamental guidelines. Therefore, 'it is not the exclusive punishing that is the defining characteristic of this period, but rather the ambivalence' (Garland 1999: 75).
During the first election of Luiz Inacio Lula da Silva in 2002, his government's program prioritised public safety and advocated a greater role for the Federal Government to reduce violence. Lula's candidacy represented a commitment to arms control policy, by focusing on combatting the impunity of crimes committed by criminal organizations, and against social movements and human rights. These commitments on the one hand encouraged social policies that would allow for the reduction of inequalities and contribute to social pacification and, on the other hand, would improve the performance of public security organs, especially the Federal Police but also the State Police. These aims would be carried out with the use of new technologies and the creation of a database that would allow for the coordination of efforts to fight organized crime, with a special focus on drug trafficking and arms smuggling. Based on the diagnosis that the national public security policy lacked planning and management capacity -this due to the fragmentation of actions by the States in the field -the proposed model involved increasing the IJCJ&SD 159 Online version via www.crimejusticejournal.com © 2017 6(1) knowledge on the reality of violence and crime. The aim was to fuel preventive, targeted and strategic actions that were to be constantly monitored by actors in the public security and criminal justice system, as well as necessary reforms of the police forces.
The implementation of the proposal, however, occurred only in part and the issue of structural reforms of police organizations was quickly removed from the agenda. However, there were important advances in knowledge management with the development of research networks and a more systematic production of data, which culminated in the establishment of the National System of Information for Public Security, Prisons, and Drugs (SINESP) during the Dilma government. But that was hardly taken into account when informing the originally proposed strategic management. Advances in prevention policies were assumed on the basis of the first agreements between the Federal Government and municipalities to implement prevention programs and training for public safety professionals. Even though the policies were still centred in police schools they received an important contribution by the establishment of the National Network for Advanced Studies in Public Safety (RENAESP), which facilitated a more permanent tie between the police and other public safety operators and Brazilian universities.
In his second term, and now with Tarso Genro leading the Ministry of Justice, Lula implement the National Program of Security and Citizenship. It was the first program that presented a clear link between the budget and specific State-and municipality-implemented programs that addressed violence, such as training community leaders, youth and witness protection, community justice programs focused on conflict mediation, and the expansion of mechanisms that addressed domestic violence against women. In addition the program sought to qualify and valorise public safety officials through training courses, scholarships, equipment investments and the restructuring of the police force.
As a result of this attempt to implement a new paradigm for the reduction of violence in Brazil, the government mobilized civil society to carry out a National Public Security Conference in 2009, seeking to source and build social support that was necessary to implement the policy. At the end of Lula's second term, the proposal still hadn't been consolidated and much had to happen to structure a National Public Security System. But the innovative left-wing approach left an important mark in the area, laying the foundation from which the succeeding government could work, and even the difficult fight against homicide rates had a partially successful outcome with stabilizing and reducing levels in the early 2000s.
In the following election, when Dilma Rousseff was launched as governmental candidate, the public safety issue was relegated to second place in the program and in the electoral debate. This occurred largely by virtue of the relative success achieved in the previous period when the growth in homicide rates had stalled, especially in the South East. The National Plan of Public Security, which wasn't developed and released until Rousseff's second year in office, was much more discreet than the previous disclosure. Rather than an implemented approach, the plan presented a series of points that would be implemented according to the convenience of the governments of the partnered States.
By the end of her first term and with the advent of a new electoral dispute, the Dilma government lacked favourable indicators in the area of security. This was influenced by the regrowth of homicide rates and the government followed the momentum to create a resurgence of the law and order discourse, lowering the age of criminal responsibility, and increasing penalties and other measures linked to punitive populism. If Dilma's proposal has the merit of recognizing the need for the effective participation of the Federal Government in the management of security, one must recognize that little changes in this regard. Accordingly, the implementation of effective control mechanisms of policing, police reforms and the necessary competence improvements of municipalities in this area are often being omitted from proposals. In fact, the police reform proposal might even counteract and end up reinforcing a logic that does not question the issue of police militarization but rather establishes a new model of police structures. Neither is the need to incorporate popular participation in security management promoted, an idea that developed in the Cabinets of Integrated Management which disseminated from PRONASCI. Finally, and perhaps most importantly, nothing is said about the necessary construction of alternative pathways to the hardening of criminal law, in order to revise drug policy and implement effective mechanisms to manage everyday conflicts in large and medium urban centers. Conversely, the governments that were led by the Workers Party left a significant mark on attempts to address the punitive culture and reduction of violence with the approval of the Disarmament Statute in 2003. This fact has led them to be the subject, even today, of attacks from free arms trade advocates in the country, who are often linked to the arms industry and the punitive discourse. At the same time, the Workers Party governments contributed to the passing of tougher laws to combat corruption and encouraged crime prevention policies by the States and municipalities.
Additionally, the Dilma government passed an important legal reform which allowed for the expansion of precautionary alternatives in criminal proceedings in an attempt to reduce the use of remand imprisonment, as well as legislation and criminalization of the use of torture by State agents.
In the area of penal reform, during the Lula government, the measure that was definitely of an opposing direction was undoubtedly the new drug law in 2006. The parliamentary debate that led to the adoption of the new law showed no clear perspective on the role of the criminalization of the drug market in the 'hyper-incarceration' nor an ability to develop a proposal that objectively distinguished between users and dealers. In other words, it showed the limited capacity of left-wing Brazilian politics to deal with this issue and perhaps therefore fell into a trap.
In return for the decriminalization of personal drug use -which, it has to be noted, did not lead to incarceration before the adoption of the new law -a reform was adopted that hardened measures against small drug dealer by increasing the minimum sentencing. However, the reform lacked any objective criteria to define offenses, leaving the fate of small drug dealers up to the police and judges, which led to increasingly overcrowded Brazilian prisons, especially in São Paulo.
To understand the paradox between the attempts to contain punitive power but at the same time reduce the all-time high incarceration rates that played out at a federal level during the Lula and Dilma governments, one must consider the weight of the São Paulo State in these results, amongst other factors. Throughout the 12 years that the Federal Government had a political orientation that was inspired by a tradition of Latin American left-wing politics, the government of São Paulo was under the control of the conservative PSDB party. This is a political party that, both discursively and in their management practices, adopted policies related to the law and order movements, which argued in favour of harsh imprisonment, especially for drug-related crimes. This resulted in the discrepancy of incarceration data of São Paulo compared to other States and reflects the orientation of the Executive Power which is responsible for the coordination of the civil and military police.
The current practices of police forces have crossed the line from a legality that is considered oldfashioned and inquisitorial if not authoritarian, to a type of 'practical illegality', a logic justified to improve efficiency. If criminal investigations become an obstacle, alternative practices are created to increase efficiency, in order to deal with what otherwise would be an overwhelming demand for police. However, this 'efficient illegality' diminishes the guarantees of the rights of the IJCJ&SD 161 Online version via www.crimejusticejournal.com © 2017 6(1) accused and has not shown to increase the investigatory capacity of the police. Michel Misse (2008) proposed the concept of criminal submission to address the phenomenon, which, although not new, is reinforced by contemporary social dynamics.
To understand criminalization processes and the incarceration geography of Brazil, it is necessary to look beyond the importance of the provincial State governments and their influence on police action and take the performance of the other two other punitive control agencies into account: the Public Ministry and the Judiciary. In relation to the Public Prosecutors Office, which is in charge of criminal proceedings, investigations into the conceptions of criminal policy advocates and public prosecutors (Azevedo and Vasconcellos 2012;Azevedo 2005) indicate a growing identification with zero tolerance proposals for crime as a way to combat high crime rates. The current inalienable rights, guided by the idea that the prosecution cannot violate the constitutional rights of the accused, has fewer and fewer supporters within the institution.
Entrusted with law enforcement and the prosecution of specific cases, judges are central actors in the criminal justice system. Magistrates therefore became the objects of study with respect to the content and reasons for their decisions. Research on professional behaviour, the mentality of judges and their social background exposed the fallacy of the myth of neutrality, which led to the opening of inquiries into the real motivations of judicial decisions. The overall findings point to the existence of a twofold of selective judicial activity. Firstly, there is a selectivity in law enforcement, in which social groups that are economically and culturally disadvantaged are more likely to be sentenced as opposed to favouritism for the upper classes. Secondly, there is selectivity in the interpretation of the law, in which judges use their discretionary power to benefit their political and ideological options.
For a number of reasons -including the various corruption scandals that occurred over the last 12 years and the lack of a clearer direction about the decarceration program which, at the same time, addressed the social demand for the reduction of violence -governments led by the Partido dos Trabalhadores (Workers Party) in Brazil were losing the ability to build an effective hegemony around public security in relation to the affirmation of rights and the adequate and Republican functioning of the government organs responsible for the police persecutions. They also lost their political effect on other areas of government, such as the States and municipalities, in addition to other institutional dimensions, such as the legislative and judicial powers. These effects are crucial in order to be able to change the direction of judicial decisions and expand a legal system that is able to restructure law enforcement agencies and implement effective control mechanisms. This is how we arrive at the start of a new government, still with a leftist perspective, but with a continuous and uninterrupted growth in incarceration rates, higher levels of violent crime, and the resurgence of the law and order movement, which threaten the few achievements that have been accomplished and throw a shadow of doubt on public policy in criminal matters in Brazil in the near future.