“The right side of the wrong life”

“The right side of the wrong life”

Tiê Franco Brotto June 2019

Student number: 11754079

Supervisor: Dr. P.E. (Polly) Pallister-Wilkins Second reader: Dr. M. (Martijn) Dekker MSc Conflict Resolution and Governance Graduate School of Social Sciences Universiteit van Amsterdam

Table of Contents Introduction ________________________________________________________________ 4 Theoretical Framework _______________________________________________________ 7 Motivation _______________________________________________________________ 8 Alternative______________________________________________________________ 10 Choice _________________________________________________________________ 13 (Re)production __________________________________________________________ 15 Methodology ______________________________________________________________ 18 Research design _________________________________________________________ 18 Method ________________________________________________________________ 19 Corpus _________________________________________________________________ 20 Operationalisation ________________________________________________________ 22 Limitations _____________________________________________________________ 22 Ethics statement _________________________________________________________ 23 Analysis__________________________________________________________________ 25 Motivation ______________________________________________________________ 25 Alternative______________________________________________________________ 32 Choice _________________________________________________________________ 39 (Re)production __________________________________________________________ 43 Conclusion _______________________________________________________________ 45 Bibliography ______________________________________________________________ 47

List of Tables

Introduction

It is difficult to grasp the full extent of violence in Brazil. Statistics say that more than 550,000 people were murdered in the country in the last ten years (Cerqueira et al., 2019, 2018, p. 3). Homicide rates grew in the last decades from 11.7 murders per 100,000

inhabitants in 1980 to 31.6 in 2017 (Cerqueira et al., 2019; Waiselfisz, 2011). Compared to other countries, Brazil registered the highest absolute number of homicides in the world, totalling 65,602 homicides in 2017 (Cerqueira et al., 2019). In a ranking of the 50 deadliest cities worldwide, excluding conflict zones, Brazil is featured 17 times (CCSPJP, 2018). In 2017, a staggering 3,270 homicides were registered in the city of Fortaleza alone (CCSPJP, 2018), and 6,416 homicides were registered in the state of Rio de Janeiro, most of them in the metropolitan area of the capital (Cerqueira et al., 2019).

Most of this violence is associated with territorial disputes between organised crime factions over illegal drugs markets and trafficking routes and as well as direct combat with the police and the armed forces (Brazil, 2018; Carvalho, 2017; Zaluar et al., 1994). Brazilian police is one of the deadliest in the world. While it is difficult to account for police killings, Amnesty International estimates that about 8,466 people have been murdered by the police in the state of Rio de Janeiro between 2005 and 2014 (Groll, 2015), while the yearly national estimate is around 4,222 deaths per year, or 6.75% of the total murders (Cerqueira et al., 2018).

Compared to homicides, other crimes such as theft and robbery are even more challenging to measure (Cerqueira et al., 2019, pp. 88–96). For this reason, criminologists have historically relied upon homicides rates to measure and compare crime and delinquent behaviour. In Brazil, statistics for such crimes are not consistently centralised and verified by independent organisations. The statistics of cargo theft, one of the leading property crimes in the country, is an amalgamation of data from the Civil Police, the Military Police, and the Federal

Highway Police. In 2017, it is estimated that 25.950 cargos were robbed, adding up to the loss of more than R$ 2 bi (450 million euros)(Camporez, 2019). The latest nationwide

research dedicated to understanding crime carried by the Brazilian Institute of Geography and Statistics (IBGE), found that 7.3% of the population over the age of 10 had been a victim of theft or robbery in 2009, totalling 11.9 million people (IBGE, 2010). These numbers, despite being out-dated and under-representative, suggest that crime in Brazil is far from being restricted to homicides.

Another, more structural type of violence manifests itself as social inequality and social exclusion. Despite its relatively normal GDP per capita (World Bank, 2019), Brazil is one of the most unequal countries in the world. According to the UNDP Human Development Reports, Brazil ranks among the 11 most unequal countries in the world in three main

indexes, including the GINI coefficient (UNDP, 2018). In 2017, the average work income of self-declared black and brown Brazilians was 27.9% and 26.3% lower than the national average, respectively, while the rest of the population had a work income that was 29.2% higher than the average (IBGE, 2018). In 2017, 75.5% of the registered victims of homicide were self-declared black or brown (others being white, yellow, and indigenous). Correcting for the size of these populations in Brazil, self-declared black or brown people are 2.7 times more likely of being victims of lethal violence in the country (Cerqueira et al., 2019). Additionally, most of the reported victims are between 15 and 29 years old, and the rate of homicides against young people grew 6.7% between 2016 and 2017. From the 35.783 young people murdered in Brazil, 94.4% were male (Cerqueira et al., 2019).

The aim of this research is to investigate the possible connection between these two radical features of Brazil, crime and social inequality. Taking a systemic perspective, I consider these two features as sides of the same coin. Building on 22 original narrative interviews conducted in Brasília, Rio de Janeiro, and São Paulo with young people involved in crime in the present or the past, I am able to investigate the hermeneutics and phenomenology of the structuration processes that render these youths both victimisers and victims. My research question is:

To what extent does social inequality contribute to youth joining violent crime in Brasília, Rio de Janeiro, and São Paulo?

With this question, this research intends to provide additional substance to alternative policies aimed at combatting crime in Brazil that go beyond attacking the symptomatic issues like the aforementioned repressive and punitive policies. In order to adequately and effectively fight crime and violence, policies need to target the structural factors allowing for their

perpetuation. In this sense, this research aims to demonstrate that other dynamics are at play that explain social exclusion, which forms the fertile ground for crime to flourish in. Policies aiming to prevent organised crime growth and facilitate the exit of individuals from

organized crime should be informed by awareness and knowledge of these dynamics and resulting patterns of exclusion.

From an academic standpoint, this research is relevant as it mobilises concepts of sociology and criminology developed mainly in the United States and Europe in a different context, thereby assessing their applicability, flexibility, and legitimacy. Moreover, the research advances debates around public security, democracy, and legitimacy. The social relevance of researching this topic is contributing to an important discussion on the power dynamics engrained in social inequality. By gathering evidence of different ways in which social inequality aggravates crime, it gives an additional push policy changes in the direction of transforming the root causes of crime.

The organisation of this paper follows a very simple structure. The first chapter defines the theoretical framework used in the analysis. It follows one temporal structure of social action, focusing on the relationship between social inequality and crime in terms of motivation, action alternatives, and choice. The relationship between social inequality and crime is situated in different theories from sociology, starting by cultural criminology. Then strain theories gain emphasis, and are problematised along other relevant theories. Critical criminology adds the perspective of power and lastly structuration theory is used as an attempt to explain the reproduction of crime in Brazil.

The second chapter explains the methodological process employed on this research. Here, I detail the process of case selection, method, corpus, operationalisation, limitations, and finally the ethical considerations. The third chapter reflects on the interviews, based on the theoretical framework. Both the analysis and the theoretical framework are structured in a similar fashion to facilitated the flow and understanding. The conclusion brings back on the research question, summarises the key learnings from the analysis and suggests future research topics, based on the limitations of this one.

Theoretical Framework

Criminology has been successful in observing important correlations between crime and biological, psychological, and social conditions (Wikström, 2006, p. 67). Such correlations are observed from different angles, organised in different theories which serve to explain different causal relations. Since the 1980s, criminologists have been concerned with the limitations of the “million modest little studies” producing a “million tiny conflicting results” in criminology (Wolfgang, Figlio, and Thornberry 1978, p. 4 apud Bernard and Snipes, 1996, p. 302). As such, the challenge became the integration of theories into ‘general’ explanations of crime (Bernard and Snipes, 1996), based on the assumption that integration was possible so long theories did not contradict each other.

Today we know that most delinquents are not pathological (Bunge, 2006, p. 22). As a result, the most influent theories currently focus on the psychological and social levels (Agnew, 1992). The main general theories explain crime through either strain, control, or learning (Bernard and Snipes, 1996). Using abroad simplification, strain theories focus on the way in which individuals respond to either the absence of positive stimuli or the presence of noxious or negative stimuli. Control theories focus on the absence of self and institutional control, either through sanctioning, deterrence, moral appeal, or otherwise. Learning theories focus on the cultural and ecological positive relationships that model delinquent behaviour (Agnew, 1992). These general theories can offer plausible explanations to the same occurrence, from different angles (Agnew, 1992). On top of these, a number of other theories have been developed to explain crime, from critical criminology (Carrier, 2011; Quinney, 1977) to cultural criminology (Hayward, 2004), and the new realism (Walklate, 1992).

It is beyond the scope of this research to review each of these theories in depth. Instead, the below selectively discusses elements of the three theories that best align with, and support the research question. In other words, how these ideas help explain the relationship between social inequality and crime in terms of motivation, action alternatives, and choice. The first, addresses the effect of social inequality on motivation, from general theories of sociology and cultural criminology. The second focuses on strain theories and the role of social inequality to selecting action alternatives, problematising certain aspects of strain theories and

commenting on other relevant theories. The third, offers a reflection on social inequality on crime from a critical perspective of power in criminology. Lastly, the fourth segment engages with the elements raised throughout the framework, and puts forward a final theory to explain

the reproduction of crime in Brazil, informed by structuration theory. In the next chapter, the operationalisation of the concepts and theories discussed here will be elaborated on.

Finally, social inequality in this research is understood not as poverty, but as the relative exclusion of a group or an individual “from at least one of the main social systems: economy, polity, and culture” (Bunge, 2006, p. 23). This definition reflects not only the concepts of early theorists of stratification, for whom social inequality was mostly resumed to a class and therefore material argument (Leahy, 1983), but also the relation between class and urban divisions (Lefebvre, 1968; Wacquant, 2016, 1995), cultural power dynamics (Bordieu, 1979), and concepts of gender, race, and ethnicity, as well as their intersectionality (Collins, 2002). For the sake of simplicity, I will refer to what some of these theorists would call ‘lower class’, ‘underclass’, ‘oppressed’, ‘marginalised’, only as ‘socially excluded’ from now on, unless there is a conceptual need to make the distinction.

Motivation

Considering the way that the research question is put forward, the emphasis is put on the effects of the structure (social inequality) to the object (youth in crime). It follows that the theoretical framework should, to better accommodate the research question, focus on the structural explanations of crime. Fortunately, most of the established theories in criminology go in this same direction. The first segment is centred on the motivations which will drive the actions of young individuals. In particular, one fundamental motivation is presented and reflected upon with the main concepts of strain theories. In the next segments, this motivation will be contextualised within a pool of action alternatives, and finally within the process of choice. Inequality is the main thread connecting these elements.

According to Hirschi and Gottfredson (1983), crime—excluding offenses which require high occupational roles—is committed mostly by 15-25-year olds males. Conversely, the victims of violent crime are also disproportionally young, at least in Brazil (Cerqueira et al., 2019). While the exact age and profile of the population interviewed for this study will be detailed in the methodology, for the purpose of the theoretical framework, youth should remain the somewhat vague period of maturing, between infancy and the social and familiar

responsibilities of adulthood. A period of a person’s life characterised by an intense search for independence and identity. Paraphrasing Arendt, a person incapable of action, incapable of revealing one’s unique distinctness, has ceased to be a person. It is only through action that

we insert ourselves into the world, that we experience ‘natality’ (Arendt, 1998, p. 176). Every person, irrespective of their background, and especially the youth, is motivated by the pursuit of an identity: “In acting and speaking, men show who they are, reveal actively their unique personal identities and thus make their appearance in the human world” (Arendt, 1998, p. 179).

Common institutions that support identity development are the family and profession. The vast majority of people will consider these options, including those who later on engage in crime. In fact, research has shown that criminals “value education and aspire to obtain legal employment” (Dickson-Gomez et al., 2017, p. 562). In other words, the pursuit of identity or the struggle for recognition, as universal motivations, do not necessarily lead to crime. It is the impossibility to achieve such goals that generates conflict. When asked about separatist movements, McLuhan said “all forms of violence are quests for identity. When you live out on the frontier, you have no identity. You are a nobody. Therefore, you get very tough. You have to prove that you are somebody” (McLuhan and McManus, 1977 emphasis added). The idea had already appeared in his earlier work, observing the effects of globalisation in culture and multimedia (McLuhan and Fiore, 1968).

In a similar vein, Honneth writes about the ‘struggle for recognition’. Continuing the critical theory tradition of his predecessors at the Frankfurt School, notably Habermas, Honneth takes the inverse perspective to social conflict and proposes that they originate from a perceived attack to an individual or collective identity. The recognition is at the same time, personal, social, and legal (Honneth, 1996). These two are examples of instances when social

exclusion, by denying someone’s identity, is associated to a form of violence or conflict, and crime.

Another form of pursuit of identity, and being recognised, is shaped by culture (Bordieu, 1979). Currently, culture is heavily influenced by the consumer society (Baudrillard, 1970; Bauman, 2005). Thus, the pursuit of identity in the context of the consumer society, is the pursuit of consumption, to exist as a person is to consume. The interplay of social inequality and consumerism seems, at first, paradoxical. Socially excluded individuals ‘over-identify’ with this consumer culture “in an attempt to create a sense of identity” (Hayward, 2004, p. 181). It is paradoxical because these people “experience tremendous feelings of alienation and exclusion from traditional employment and educational opportunities” (Hayward, 2004, p. 181), at the same time they are seduced by the consumer culture, which is amplified by

mass media (McLuhan and Fiore, 1968). This tension between the consumer ideal and the means of achieving it, fuelled by social inequality, is well captured by the strain theorists, and will be developed more in detail in the next segment.

Institutional anomie theory, puts forward a similar understanding, in which an institutional arrangement that highlights the economy in detriment to other institutional spheres, such as family, will be more criminogenic (Messner and Rosenfeld, 2012). Considering identity one general motivation, the question addressed here was the explanation of how social inequality affects the pursuit of identity and potentially leads to socially unapproved means, defined by law as crimes. Next, I will explore how social inequality affects the action alternatives to attain it, mainly employing strain and social learning theories.

Alternative

Influenced by the materialism of the early theorists of stratification, Merton focused on explaining crime as a product of class structures (Carrabine et al., 2009, p. 79). When a significant discrepancy exists between internalised socially approved goals and the perceived socially approved means to attain said goals, the subject group or individual is frustrated and angry (Agnew, 1992). This relates to the feeling of ‘revolt’, described by Zaluar as a refusal to participate in the social game when its rules are fixed or unjust (Zaluar, 1985, p. 130). In practise, they may either reject the first and adopt different goals, or reject the second and adopt alternative means to cope or attain such goals, some of which are illegal, and therefore criminal (Merton, 1938).

Cohen built on Merton’s work, suggesting that people engage in crime not necessarily

motivated by economic goals, but as a way to oppose middle-class values and means (Cohen, 1955). Although this opposition goes arguably without questioning culture. Zaluar observed that most offenders in Rio de Janeiro, for example, continue to be consumers, not developing an alternative to the ideals of the consumer society (Zaluar 1985:165). Another important contribution associated with Merton’s work, is the illegitimate opportunity theory. Expanding from the basic idea of strain theory, that crime is a possible manifestation of the perceived incapacity to attain internalised socially approved goals by socially approved means, Cloward and Ohlin suggest that the ease to access illegitimate opportunities is also relevant (Cloward and Ohlin, 1960, pp. 144–160). In other words, an important factor for engagement in

criminal behaviour is simply the availability of means to engage in such behaviours. Individuals will favour the means that are closer and most familiar to them.

Most recently, Agnew expanded strain theory beyond the failure to achieve such socially approved goals, to explain crime as the inability to legally escape from painful situations, varying from child abuse and physical or verbal violence, to “noise, heat, air pollution, personal space violations” (Agnew, 1992, pp. 58–60) From his perspective, strain may lead individuals to three different coping mechanisms: “(1) make use of illegitimate channels of goal achievement, (2) attack or escape from the source of their adversity, and/or (3) manage their negative affect through the use of illicit drugs” (Agnew, 1992, p. 49). Applying this theory to the impossibility of consuming would suggest that socially excluded people eventually have to choose between these three options.

Because of their focus on social discrepancies, strain theories are normally the most closely related to explaining crime in terms of social inequality. On the flip side, this is commonly indicated as a limitation to explain crime in general. This criticism was particularly strong around the 1970’s, when a number of large self-report studies concluded that class

differences were statistically non-significant, and even questioning the relationship in the first place (Farrington, 1989; Tittle et al., 1978). A comprehensive review of such studies

demonstrates that the majority of people have, at some point in their young life, committed a crime, irrespective of their background. Later, such studies have since been criticised, mainly because “massive class and race differences for officially recorded offenses of high

reportability such as homicide and car theft cannot be explained away” (Braithwaite, 1989, p. 49).

Despite their criticism, these self-reporting studies contributed to the strengthening and developing of alternative theories worth exploring. In one study, 95.6% of mostly white, working-class, British males reported having committed at least one offense between the ages of 10 and 32, ranging from burglary, car theft, assault, vandalism, and drug use, among others (Farrington, 1989). Since the majority of their sample has, at some point in their young life, committed a crime, a moral assessment on the basis of a particular background was contested. That is because, either individuals are generally capable of rationally justifying or

temporarily suspending their moral judgements a priori, much like neutralisation theory suggests (Sykes and Matza, 1957), or moral assessments are not a primary impediment for individuals engaging in criminal activities. Given the criticism that neutralisation theory

received for its focus on rationality (Hamlin, 1988), it would be more appropriate to conclude that engaging in crime is, to a significant extent, informed by emotion rather than a moral assessment.

Another way to interpret these self-reporting studies is through control theories, which assume that everyone is a potential criminal and a combination of self- and social control is necessary to refrain individuals from engaging in criminal activities. This self- and social control can be exercised by either morality or deterrence (Carrabine et al., 2009, pp. 82–84; Laub and Sampson, 1991). However, there are three reasons why this theory, despite its potential to explaining other cases in criminology, falls short in addressing this research question. The first is that control theories as Bunge suggests, morals and self-control are only regulators: “one must look elsewhere for crime mechanisms, because people act from habit, need, or desire, not from restraint. The analogy with the car is obvious: what moves the car is the engine, not the brakes” (Bunge, 2006, p. 22).

A final contribution of these studies is that they suggest that the historical difference between self-reported data and official records can be explained as “biases in the criminal justice process which now have been corrected” (Tittle et al., 1978, p. 654). A biased criminal justice, giving different treatment to particular individuals, promotes another form of social inequality and social exclusion. According to labelling theorists, stigmatisation of

determinate segments of society, in this case the socially excluded, perpetuates a

‘self-fulfilling prophecy’ by selectively policing and punishing (Becker, 2008; Mead, 1934). More broadly, this differentiation is internalised by socially excluded individuals in two main ways. The first is, as elaborated in labelling theory, the embodiment of a criminal caricature. The second is perhaps part of a larger phenomenon in social learning, where individuals’ perception of their self-efficacy, or capacity to succeed, is diminished (Bandura, 1982). In other words, it is the internalisation of failure by socially excluded individuals, that

increasingly keeps them from aiming for particular goals, which they start to consider above their capacity to succeed. A critical reflection of this bias will be explored more in depth next, with the elaboration on the concept of crime.

Social learning, more broadly, is also a powerful theory in criminology. In the early 1930s, Sutherland developed his theory of different association, which was widely accepted at the time (Brookman, 2005). Integrating the early work by Sutherland on different association, social learning theory developed with the focus on the learning process of direct or indirect

associations and interaction with others, in terms of behaviour, norms, and values. Primary groups such as family and friends are given more importance, but also secondary and reference groups are included (Akers, 2009).

Recapitulating, as proposed in the previous segment, social inequality is a selecting factor to the universal motivation for recognition and identity, that often results in overemphasising consumerism. Here, the role of social inequality was emphasised in determining the

alternatives to engaging in such consumerism, and the particular situations where it can lead to crime. Strain theories were favoured are the most closely related to explaining crime in terms of social inequality. However, the problematisation of such theories opens up space for considering other important theories in criminology, such as neutralisation, control, labelling, and social learning theories. Combined, they suggest socially approved means of achieving recognition and developing an identity are lowered by social inequality for socially excluded individuals, leaving criminal means as an alternative. Next, the focus is on the evaluation and choice of engaging with this criminal alternative.

Choice

As it has been constantly stressed here, action is always situated. Taking the choice as one of these actions, this segment offers a reflection on the background against which individuals consider entering crime. The reflection is mainly informed by critical criminology, which offers important elements for the analysis. On the other hand, it lacks sufficient theories to explain why individuals engage in crime. This, rather than a limitation, should be taken as an opportunity to explore unfamiliar ways, addressed in the following segment.

Early theories in criminology used to adopt a rather limited concept of crime, mostly based on legal definitions, “which leave out the multiple components of the crime phenomenon” (Henry and Lanier, 1998, p. 610). Some still choose to define crime as “an act of breaking a moral rule defined in criminal law” (Wikström, 2006, p. 63). Despite recognising that criminal law is only one of the many existing moral rulebooks, such theories take a unitary system to explain deviation. Others have proposed a more nuanced view, trying to account for the interaction and collision between different sets of moral rules (Agnew, 2011, pp. 12– 43). More broadly, conflict theorists argue that the definition of crime comes largely from power. Crime can be seen as a selection of particular harmful behaviours, that depends on the power to enforce such definitions (Henry and Lanier, 1998, p. 612). Because of its

particularity, the definition of crime can also ignore behaviour that is only harmful to others than the ones who define it. Sutherland was particularly vocal about ‘white-collar’ crime almost a century ago, and critical theorists have since expanded this idea to other forms of harm such as “imperialism, racism, sexism, poverty, and other denials of human rights” (see Henry and Lanier, 1998, p. 612).

Understanding power is an important perspective to take when studying criminology, and especially the relation that crime has with social inequality. The apparently black and white definition of crime, “where the boundaries between licit and illicit activities appear to be clear cut” (Aas, 2013, p. 125) is to a large extent a feat of modern states. Looking back a few centuries, the distinction between state authorities and criminals was far from clear (Aas, 2013, p. 125). Tilly’s famous analogy of the state as “quintessential protection rackets with the advantage of legitimacy” (Tilly, 1985, p. 169), is intriguing because of its historic resonance. The provocation is based on the work of early social contract theorists, notably Hobbes, and the anarchic setting upon which different parties disputed the right to use violence legitimately. “To the extent that the threats against which a given government

protects its citizens are imaginary or are consequences of its own activities,” Tilly writes, “the government has organized a protection racket” (Tilly, 1985, p. 171). Even though such

historical, and particularly European, observations on state making might not apply today (Leander, 2003), the critical observation of power and crime reveals important structures of domination.

When a party has established the legitimate monopoly of violence, all actions that endanger such legitimacy will be deemed illegal, and ultimately a crime. Quinney, for example, states that “when a society cannot solve the social problems of its own creation, policies for the control of the population must be devised and implemented” (Quinney, 1977, p. v), with the purpose of maintaining the status-quo and the legitimacy of its institutions. Any biased interpretation of crime, to the extent that is based on a power social inequality, fuels social exclusion. In such cases, there is a differentiation between crimes of domination and crimes of accommodation and resistance (Quinney, 1977). To the extent that power is the

legitimising element, this tension provokes a crisis of legitimacy (Carrier, 2011, p. 332), as the dominant party continuously employs violence at the expense of power (Arendt, 1970). Inequality, as social exclusion, contributes to a crisis of legitimacy that increases the chance of actions occurring that will be deemed criminal, by the traditional definition based on

criminal law, and by other forms of harm such as those pointed out by critical theorists. Taking social inequality as the root cause of crime, “there can be no positive peace without social justice” (Quinney, 1995, p. 155). While social inequality exists, it will demand a violent dispute between crimes of domination and crimes of accommodation and resistance, to use Quinney’s words. It is against this background that individuals consider engaging in criminal activities. Again, while focusing on power dynamics, critical criminology does not offer established theories to explain why individuals engage in crime. This, in turn, sets an opportunity to explore unfamiliar ways, explored in the following segment.

(Re)production

Despite the contributions from critical criminology on the political dynamics of crime, the fundamental premise of the theory favours, much like strain theories do, one-sided diagnosis of crime, in which the structure is highlighted. According to Bunge, criminology needs to take an integrative, systemic, and multi-level approach, and understand offenders as both victim and victimiser at the same time (Bunge, 2006, p. 24). The new realism theories attempt to overcome the “tendency to talk about victims and offenders as if they were

hermetically sealed, rather than overlapping and blurred, categories” (Walklate, 1992). In her early observations during the 1980’s, Zaluar questions whether the phenomena she observed should be seen as a social movement or as organised crime. In other words, whether

‘criminals’ should be regarded as victimisers or victim. She seems to tend for the first option, while raising important observations. Paraphrasing, Zaluar concludes that we should not idealise, even more than some local residents do, these people involved in illegal activities. They are not romantic heroes of a social movement, alike Hobsbawnian social bandits. Despite the social injustice, they are attracted by the economic gains of drug trafficking and their life style is not contestatory. “They are neither reformists, nor revolutionaries,” they are consumers, she concludes (Zaluar, 1985, p. 166).

What Zaluar observes is to certain extent an extrapolation of social learning theory.

According to her, criminals employ the same repressive techniques which they experience as members of the underclass, perpetrated by the repressive apparatus of the state (Zaluar, 1985, p. 167). They not only ‘imitate’, as Akers would put (Akers, 2009), but they ‘reproduce’ violent domination (Zaluar, 1985, p. 167). On top of that, criminals are not imitating or reproducing something learned predominantly within primary social groups, such as family

and school, but broad societal institutions such as the repressive apparatus of the state, the police. Going back to Arendt:

“Although everybody started his life by inserting himself into the human world through action and speech, nobody is the author or producer of his own life story. In other words, the stories, the results of action and speech, reveal an agent, but this agent is not an author or producer. Somebody began it and is its subject in the twofold sense of the word, namely, its actor and sufferer, but nobody is its author” (Arendt, 1998, p. 184). A number of studies in criminology have considered the concept proposed by Giddens of structuration (Bruce et al., 1998; Densley and Stevens, 2015; Farrall and Bowling, 1999). The theory of structuration, in Giddens own words, is about how “we should see social life not just as ‘society’ out there or just the product of ‘the individual’ here, but as a series of

ongoing activities and practices that people carry on, which at the same time reproduce larger institutions” (Giddens and Pierson, 1998, p. 76). From there three ideas relevant to this study emerge. The first is the duality of structure, a dynamic where “social action of any kind is always situated in time and space but it also gives substance to time and space” (Giddens and Pierson, 1998, p. 13, emphasis by the author). The second is the condition of

knowledgeability, in which agents always and inevitably draw upon their knowledge of their structural context when they engage in any kind of purposeful action (Stones, 2005, p. 17), without necessarily achieving the intended results. The third is the idea of homeostatic loops, defined as “causal factors which have a feedback effect in system reproduction, where that feedback is largely the outcome of unintended consequences” (Giddens, 1984, p. 375) Taking these three ideas, and structuration theory in general, it is possible to understand the contribution of social inequality to crime, explored in the previous segments, in terms of a mutual production and sometimes unintended reproduction by structure and individual (Giddens, 1984, p. 25). In practise, knowledgeability of the youth engaging in unlawful behaviour is situated, as presented above, a culture of consumerism and domination.

Considering that the violence perpetrated by law enforcement, widespread corruption in the state, and its overall absence from marginalised communities (Zaluar, 2007), the legitimacy of state’s law is damaged, while on the other hand, negative incentives like fear, and even sometimes positive incentives like basic service provision, from criminal groups increases their own local legitimacy. For people in such environments, where the criminal rule overlaps the state law, conforming to one or another is an inevitable decision. Morality, despite its

theoretical relevance, is set to the background when violence is all options. “The act of stealing or killing is not deemed necessarily as bad, negative, or criminal” (Zaluar, 1985, p. 164).

For their violent coercive means, multiple organised crime factions in Brazil are able to exercise a function similar to the state, albeit in a much smaller level (Santos et al., 2017). In the three decades since then, these organisations evolved in different ways. The largest faction of the country, the First Command of the Capital (PCC), which effectively detains the monopoly of drug trafficking in its home state of São Paulo, is notably able to regulate violence in multiple areas of the state (Santos et al., 2017). Similar dynamics were observed in other drug trafficking organisations in Medellin, Colombia during the 2000’s (Garzon, 2016, pp. 9–10) and other countries in Latin America (Durán-Martínez, 2015). In Rio de Janeiro, the same centralisation did not happen and the illegal drugs market is divided between three main factions. The largest, the Red Command (CV), is confronted by the Pure Third Command (TCP) and the Friends of Friends (ADA), which also clash between each other.

Again, the desire for control, as well as to consume, is not particular, as Zaluar highlights, but different people will find different means which they are able to attain it, as suggested by strain theories. Inequality shapes the knowledgeability of many youths seeking to escape the domination which they are subject to, and crime offers them the possibility of becoming dominators themselves. The unintended consequence is that instead of resisting social inequality, they reinforce violence amongst themselves and the wider society, which feeds into more repressive measures and the domination. Although beyond the scope of this

research, it is important to highlight that a similar dynamic of unintended consequence can be said to be existing taking the state, and especially repressive law enforcement, as the actor (Bruce et al., 1998; Densley and Stevens, 2015, p. 114). This way, building on Hallsworth and Silverstone, crime can be seen as a “self-destructive response” to social inequality, marginalisation, and social exclusion (Hallsworth and Silverstone, 2009, p. 374).

Methodology

Studies in criminology have employed a multitude of research designs and methods. This research is a case study based on interviews. In this chapter, I detail the process of case selection, method, corpus, operationalisation, limitations, and finally the ethical

considerations. Research design

This research was motivated to a great extent by my own interest in understanding one of the growing challenges of my home country, Brazil. The alarming violence in the country, resulting from territorial disputes between organised crime factions, different types of robberies, and police brutality, combined with structural violence, led to soaring social inequality and social exclusion (Brazil, 2018; Carvalho, 2017; Zaluar et al., 1994). The majority of the victims and a significant part of victimisers are young men (Cerqueira et al., 2019). The research question derived from this interest was formalised as:

To what extent does social inequality contribute to youth joining violent crime in Brasília, Rio de Janeiro, and São Paulo?

Taking Brazil to research the relation between social inequality and crime is an opportunity to observe a scenario where both are highly present. The research is designed as a case study, focusing on young individuals from three major cities in Brazil. The benefit of this is having some comparability, given that at least 7 interviews were done on each of these cities. This is done at the expense of going more in depth in one location. The decision was also informed by preliminary contacts made with social workers, which suggested that gathering more interviews in one location would be difficult given the available time on both the time of the researcher and the participants.

The three cities were selected to represent three dynamics of crime. São Paulo is home to the largest and most organised crime in the country, the First Capital Command (PCC). Rio de Janeiro is characterised by constant territorial disputes of large factions. Brasília is disputed by local and larger factions, mainly the PCC. Despite a modest basis for comparability, as mentioned above, considering the modest size of this research, limited by time constraints and other resources, it is difficult to generalise its conclusions to contexts particularly

different than the cities studied. The intention is to gather evidence of existing dynamics, which can then be further explored in future studies.

The academic relevance of this research is first to expand criminology geographically. Criminology has traditionally focused on relatively wealthier societies, and while a

considerable body of literature exists on organised crime in Latin America, only a portion of case studies focus on the relation between social inequality and crime. Despite the smaller number, Brazilian criminology produced a number of relevant case studies (Meirelles and Gomez, 2009; Rolim, 2016; Silva, 2015; Spagnol, 2008; Zaluar, 1985; Zaluar et al., 1994) and ethnographic studies (Biondi, 2018, 2010; Dias, 2013, 2011; Feltran, 2018; Manso and Dias, 2018). This research adds to this existing body, complementing it with current

discussions in western criminology, while making Brazilian criminology more accessible to an international audience.

The social relevance of researching this topic goes beyond the most obvious, to lessen or even eradicate crime. This is not just for the 65,000 fatal victims every year. Rather, this dissertation continues an important discussion on the power dynamics engrained in social inequality and its not-so-hidden costs. The paper gathers evidence of how social inequality aggravates the unintended reproduction of crime, which can be added to other important studies that support policy changes.

Method

Surveys and interviews lie at the heart of criminology studies. Considering the quality of information necessary to provide insightful evidence of different phenomena, rather than trying to confirm a specific hypothesis, the method employed in this case study is interviews. This is the most appropriate method to capture influences, experiences, circumstances, issues, themes and lessons (Brinkmann, 2018). The interviews were semi-structured initially, but quickly transformed into narrative interviews to accommodate the openness of different participants. I started with a small list of questions, divided in three main areas: childhood, environment, and the individual’s participation in crime. However, I realised that this

structure did not allow for the exploration of much more complex stories that were presented to me. As such, I shifted my approach and let participants prioritise topics and themes. The main focus of each interview varies, in consequence. Yet, the participants all addressed the core questions of this research albeit through different routes. They all have in common that

they cover different accounts of criminal activity, discourses and motivations for participating in crime, goals in life and relationships with family, friends, and state institutions, including school.

I invited the participant to collaborate, admitting my lack of experience and giving them voice while sharing some control of the interview. At the same time, I did not ignore the implicit power relations tied to the very act of interviewing someone. I was aware of how I presented myself and how that could interfere in one way or another with the interviews. I also did not try to hide who I am and where I come from, in many interviews the participants asked me questions about my life and I answered them honestly. I was decided that if I wanted them to trust me, I had to be honest and also trust them. Despite speaking the same language, accents and slangs largely differ, constantly reminding the participants and myself of the fact that we come from different places, geographically and socioeconomically. During one of the interviews, while talking about the possibility of getting a new job, one person told me that he felt very insecure because he “did not speak the language of the workers world”. My challenge was to find a healthy and productive mix between on the one hand empathising with their stories and on the other one, keeping some distance to allow for objective analysis. I also learned a lot from talking to other people in the same environments where the

interviews happened. These conversations and other observations contributed to the descriptions of the participants’ world.

Corpus

The criminal justice system in Brazil directs offending minors to two types of facilities. Serious offenses usually are sentenced to at least 45 days in closed regime while lighter offenses are given much lighter sentences in a conditionally open regime. The closed regime is usually centralised by the state government while the open regime is carried in partnership with different institutions and organisations, which differs from state to state. For this

research, I contacted two of those open regime organisations, one in Brasília and one in São Paulo, aiming to interview people who had already completed their sentences. In Rio de Janeiro I had several contacts with social workers, NGOs, and other people working with youth involved crime, that helped me to arrange the interviews. The participants were selected by the intermediaries, according to the availability of the participants. The only requirement was fitting the following profile: “Between 18 and 30 years old, past or present involvement with crime, not serving a sentence.” In Rio de Janeiro many interviews were

cancelled last minute because of security issues, and I decided to extend the age limit to 39 years old.

Table 1. Participants

# Initials Age City Date

1 JN 19 Brasília 09-Apr

2 ML 18 Brasília 09-Apr

3 EL 18 Brasília 10-Apr

4 JV 18 Brasília 10-Apr

5 RI* 19 Brasília 10-Apr

6 IN 22 Brasília 11-Apr

7 IE 18 Brasília 11-Apr

8 LS 25 Rio de Janeiro 24-Apr

9 RO 39 Rio de Janeiro 24-Apr

10 VS 28 Rio de Janeiro 24-Apr

11 VR 29 Rio de Janeiro 24-Apr

12 EN 39 Rio de Janeiro 25-Apr

13 RT 21 Rio de Janeiro 26-Apr

14 RN 27 Rio de Janeiro 26-Apr

15 WN 19 São Paulo 29-Apr

16 IL 18 São Paulo 29-Apr

17 NY 18 São Paulo 29-Apr

18 JP 19 São Paulo 30-Apr

19 AX 19 São Paulo 30-Apr

20 RD 19 São Paulo 02-May

21 LO* 19 São Paulo 02-May

22 MV 18 São Paulo 02-May

Between April 9th _{and May 2}nd, _{I conducted 22 interviews with people who have been directly}

involved with crime, some of which still are involved. From these, 20 self-identify as men, and 2 as women. Their ages range from 18 to 39, with an average of 22 years old. I did not inquiry beforehand about what crimes the participants were accused to. Instead, I preferred to learn this information from their perspective, in case they did not feel comfortable sharing particular details. From what they told me, 3 only participating in robberies, 4 mentioned participating in robberies and drug trafficking, 14 only mentioned participating in drug trafficking, and 1 mentioned drug trafficking and murder. At least 7 people were interviewed in each of the three cities. To protect their identities, names have been changed by acronyms. Table 1 summarises this information. All interviews were recorded in audio format except

from two, marked below in asterisks, in which the participant preferred that I took written notes instead. In all interviews, the relationship between my primary contact and the participant is fundamental to establish a minimal level of trust for the interview. The daily trips from where I was staying, wealthier areas of these cities, to the poorer areas, also became part of my observations.

Operationalisation

Before going to Brazil, I had an initial understanding of the main theories of criminology and knew of a handful of case studies done in Brazil, notably in São Paulo and Rio de Janeiro. After employing an inductive model for the interviews, I returned to the theoretical

framework I had drafted and made the adjustments necessary. To a significant extent, these adjustments reflected some of my observations during the interviews. From that point, the analysis was conducted in most part from a deductive model. At first, operationalisation was a challenge because the combination of different theories did not point to an obvious format. On top of that, one of the conducting lines of the analysis, structuration theory, has been criticised for focusing overwhelmingly on ontology and neglecting epistemology and methodology (Stones, 2005).

Given that challenge, I decided to code the interviews, each one averaging one hour, according to the main concepts of the theoretical framework, while allowing for the emergence of other patterns. From the 20 hours of interviews, I coded 373 quotes into 43 different categories, ranging from education, social acceptance, and violence, to status, drugs, and goals. From these, about 90 were selected to compose the analysis. In the references to dialogues, I refer to myself as FB (Franco Brotto).

Limitations

Especially in a qualitative research like this, where discourse is analysed, it is important to be aware that any person carries a set of values that can interfere in one way or another with the results. In this case, I was born in Brazil, and it is impossible to entirely suspend my opinions on the challenges which the country faces. On the one hand this allows me to understand the situation from a perspective that an outsider would not be able to; understand the nuances of the language and the context. I attempted to address this by having a transparent methodology and analysis, where the interpretation of the discourse is clear and open for contestation. For instance, I use a great number of direct quotes in the analysis, so as to allow readers to draw

their own conclusions and evaluate and possibly criticise my own interpretation of the participants’ narratives.

Another possible limitation could be the validity of self-reports, given the topic of this inquiry is fundamentally illegal behaviour. This will also be taken into consideration when preparing and coding the interviews, but research has demonstrated that “disclosure rates of gang members did not differ significantly from those of non-gang members” (Webb et al., 2006), suggesting that self-reporting is a valid measurement technique even when inquiring about illegal behaviour.

Ethics statement

This research does not include children because it is understood that children cannot fully grasp the consequences of their actions in this context and therefore cannot knowingly consent to being interviewed. The role of intermediaries in this research is also central. All participants were contacted by social workers or members of the civil society who explained the goal of the research and invited them to an interview. They were informed that

participation is voluntary, and that they can withdraw at any time. A number of individuals decided to end the interview after a few minutes, and were not included in the research. Others confirmed the interview but did not show up. For those who participated in an interview, they have the right to remain anonymous by the use of a pseudonym. They were also asked permission to record the interview in audio, and those who did give permission have their consent included in the recording. Two individuals did not give permission to record the audio, when written notes were taken. The digital and written records will not be shared with anyone unless necessary to confirm validity of this research.

Participants were in liberty during the time of the interview, in order to minimise bias and harm from the expectation of reward or punishment from a public institution. In keeping with the principle of ‘do no harm’, the security of the interviewer as well as the security of the participant were given special attention. For the interviewer, entering particular

neighbourhood unannounced could have caused problems. For the participants, speaking to strangers in public places can also be harmful because they might be mistaken as informants, which can put them in danger with organised crime. In both cases, the intermediary played a role in minimising risks, introducing the researcher to the necessary people and suggesting places where the interview could take place that were safe for all parties. At the end of every

interview, I reserved some time for additional stories or questions from the participants. Finally, the participants were not promised and not given any sort of compensation that I know of, for participating in this research.

Analysis

The analysis is informed by theoretical framework, and the coded data from the interviews. In trying to simplify the structure of the analysis, it has been divided in four segments that loosely coincide with the four segments of the theoretical framework. To reinforce the narratives of the participants, these four segments are composed an elusive linear narrative. Motivation

The starting point of any action is, as discussed in the theoretical framework, the motivation to become, to be recognised, to assume an identity. In the interviews, every participant responded to several broad questions that are relevant to this topic. The first is about goals, the second about family, and the third about school. When asked about personal goals, what they wanted to be when they were younger, and what they want to be now. In short, the answers reveal simple goals, normal for any child. The answers also reveal details of their environment and of their families, which are continuously marked by obstacles:

“Since I was little, I wanted to be a football player. Modesty aside, I’m a good player. I went to some tests, was accepted, but because of the financial difficulties it was hard to go to the trainings. I wanted to be a football player, but as I’m older now I can’t. I would like to maybe get a degree in physical education, and work with football. I don’t know, training in one of these schools, be a football coach” (RN, p. 32).

In this, as many other answers, the motivation comes promptly tied with its impossibility. This type of answer was repeated in many interviews. The impossibilities are nearly always related to economic constraints. This is pushed to a point that the association between motivations and impossibility is gradually internalised:

“I really wanted to take a drawing course […] There was an ad on television. Then my mother went there to get information but the materials were too expensive. […] So, like, this was something that slowly faded away, and other things appeared. Not that I forgot about it, but it slowly lost the importance” (LS, p. 22)

This suggests a first and very direct impact of social inequality to the lives of these few people I interviewed. Bridging to the theoretical framework, these stories align with one of the main elements of strain theories; “strain as the failure to achieve positively valued goals”

such strain. Instead of abandoning socially approved means to achieve their goals, the participants reveal that they simply abandon socially approved goals. The strain between these socially approved motivations of becoming a football player or an artist and the impossibility of achieving such goals does not lead to crime directly. Rather, this is more indicative of a process of social learning, of internalised failure, that Bandura refers more broadly to as ‘self-efficacy’ (Bandura, 1982). A good example of this is when EL talks about the possibility of attending university:

“You think we don’t want to go to university? […] We are not stupid like that. I wanted… when I was younger, I wanted to be an engineer, man. Only because I saw a video on YouTube. […] But I will get a degree in physical education […] because it’s easier. […] Who doesn’t want to be a lawyer? Everybody wants. But who has the patience?” (EL, pp. 45-46).

Other participants would provide different, less elaborate answers. It is unclear if those answers are in some way related to the internalised failure hypothesis, because they interpreted the question in a less ambitious or more near-future-focused way, or simply because they naturally had different goals in life. The difference of these responses was that they were not accompanied by a contrasting view. IL said that studying mathematics is important because he wanted to work at the supermarket as a cashier (p. 10), and JN said that if he could convince his mother to go with him, he would like to open a pizzeria or a small shop in another state, where they have family (p. 18). For ML, talking about the future was more difficult:

“They asked me this so many times when I was here, but I still can’t answer. I wish I had this kind of thinking, and say what I want for my life. […] My dad told me a few things… gave me a few ideas but… Talked about a course of radiology […] and it looks like they make good money” (ML, p. 8).

The reticence present in many of the accounts similar to this last one suggests the lack of visible alternatives with which the participants identify. That can be either because, as suggested before, particular visible options are perceived to be too difficult or simply not meant for them, or because there are too few options available. Not one of the participants mentioned their family members’ occupations when talking about motivations. Applying strain theories here seems difficult because the participants did not demonstrate a strong

identification with socially defined goals when talking about their motivations to become or be something. However, what does seem interesting when looked at in the context of strain theories, is the apparent lack of positive input. Ironically, AX would only find out what he wanted to be in a juvenile correction facility:

“FB: Where did you get this idea from?

AX: From when I was at the Fundação Casa. I read a book, I started reading books because when I was there this was my thing. There was a TV and stuff but my thing was just reading books. I went to the library every day to read books. Stayed there reading books. I read a few books about personal management, and then I saw how it was, how it’s done, and I liked it […] So I thought, trying won’t hurt, I’m not sure I can do it, but if we don’t try things we don’t get anything in this life” (AX, p. 10).

These stories about the participants’ motivations all seem to include aspects of social inequality, either as the impossibility to achieve known goals, or the absence of positive influences that inform such goals in the first place. One way or another, the result of social inequality here is limiting the options of becoming someone, being recognised, assuming an identity.

Another common source of recognition and identity formation is the family. When asked about their families, the majority of the participants reported being raised by a single mother. Only two participants grew up and were still living with both parents. The other 20 (91%) either have an absent father, or has divorced parents. According to the IBGE, the divorce rate in Brazil is a little over 30% (IBGE, 2017). The lack of a father figure is a common element in their stories. Sometimes this would only become evident when they talked about their own children:

“When it was fathers’ day at the church, I cried… or at school, I cried… When we had to write letters for our dads. This is the minimum, even just taking my son, like ‘let’s go to the park, daddy can’t buy anything, but I can give you a hug, this hug is worth more than money can pay. I wanted a hug like this when I was your age.’ But of course, I will want to give more” (JV, p. 38).

This is not to indicate that in their stories, their mother was represented as a negative figure. As JV also explained, “there is a space for each, the father can’t fill in the mother’s and vice

versa. Because my mother did everything for me” (JV, p. 39). Despite the common element of abandonment, mother figures were regularly referred to in positive ways: “Like my mother used to say, ‘I’m not gonna visit you in prison.’ But then when I was arrested, the first person to be there was her” (AX, p. 21). Their family stories are most about abandonment,

sometimes simply because their mother was the main or single provider in the household and spends most of the day working. In some cases, this resulted in constant moving:

“Like, you have to leave your mother’s house because you mother doesn’t want you there, and you go to your grandpa. Your grandpa doesn’t want you do you go to your dad. Your dad doesn’t want you, and you go to your grandma, to your aunt. I was like that” (JP, p. 17).

Another common negatively valued reference was mother’s new partners, often in relation to some kind of abuse. Relating back to strain theories, in particular to the additions of Agnew (date) to general strain theory, those cases represent, on top of the removal of positively valued stimuli, the presentation of a negative stimuli. The lack of the father figure being the removal of positive and the presence of an abusive new partner being the negatively valued stimuli. The story of VR illustrates exactly this point:

“For me it was hard because I grew up without a father figure at home. My stepfather, that my mother had a long relationship with, wasn’t someone who offered a reference, you know? He was an aggressive person and abused of drugs, he was always drunken at home” (VR, p. 3).

It is important to point out that there were a few exceptions to this rule. VS for instance, stressed that his parents were supportive throughout his difficulties, but then again, other members of the extended family were not. Given the limitations of this research, it is difficult to verify the extent to which social inequality influenced their family situation. Statistically, divorce or absence of a father is nearly three time more common within the participants than the average of the population of Brazil, as indicated above. However, the relationship

between this factor and social inequality is unclear. What can be assumed is that, given their family situation, the participants’ mothers were likely to assume the role of sole provider, which reduces the time to develop their relationship. This is, however, a vastly generalised assumption that would need more attention in further studies. Setting aside the implications

of social inequality to their family configuration, nearly all of the stories account for a reduced time spent with family, resulting in weaker social bonds.

Another way to understand this factor within a criminological perspective is by using control theories. It is logical that weaker social bonds reduce the ability of parents and family in general to transmit moral rules, reinforce self-control, and act as a form of deterrence and social control. This is not contested here, but as Bunge suggests, morals and self-control are only regulators: “one must look elsewhere for crime mechanisms, because people act from habit, need, or desire, not from restraint. The analogy with the car is obvious: what moves the car is the engine, not the brakes” (Bunge, 2006, p. 22). In this respect, what these stories suggest, as well as diminished control, is another form of removing the means through which individuals find recognition and develop their identities. The result of fragile families is visible in the difficulty to self-express, even in the most literal sense. Two straightforward examples are: “I never managed to talk to anyone about what was happening” (VR, p. 4) and “I went to mental institutions, I had psychiatrist, psychologist, but no one could understand me” (JN, p. 14). Although, the ability to self-express was slowly developed with the help of social workers in some cases:

“I used to keep everything to myself. You know, when your mind is just accumulating stuff? I just kept thinking and after a while… To be honest, not even with my mother I used to express myself. It was here that I managed to express, with [social worker’s name]. I started to talk, like, I don’t know what happened, but I felt safer” (AX, p. 26). “I trust the people here more [in the NGO], people who come to talk to us, like you even, I trust because if I keep it in my heart, like, later this will be in my head, sad, crying, thinking about my family. Like, the evil pushing from one side and the good from another. So, I let it out for you because then I feel lighter” (JP, p. 32)

Similar experiences happened at school, albeit less often. This was visible in both positive and negative experiences at school. “Only one teacher, who I never saw again, this teacher. She asked ‘how are things at home?’, because she knew how things were in my house,” said JN (p. 3). When asked if he enjoyed talking to this teacher, he replied “Yes, it was a way to vent. I was younger, wasn’t involved in crime still. I vented” (JN, p. 3). A similar story was told by LS, “School was there. The directors tried, they called my mother, tried to contact. There was a teacher that even came to my house” (LS, p. 23). On the other hand, WN story

illustrates the same exclusion and abandonment that many of them experienced in other social groups:

“I was a little yob at school. Because they called a bunch of teachers, a bunch of students, and started to say ‘this person wants to be expelled’ and then the teachers talked. Then when it was my turn there wasn’t a single teacher that didn’t raise their hands. They all raised their hands. So, I left. But they… instead of saying ‘take this paper and go to another school,’ explain that they were transferring me… they only said that I was being expelled. They just said ‘you’ve been expelled, bye’ (WN, p. 6).

It is important to stress that only a few had extreme experiences like that in school. Most of them were just not engaged, like JN and LS. Like the family, these stories are not definitively positive or negative, but they reveal the difficulties and limitations of the two most important social groups to young individual’s growth, family and school. Inequality might have a significant albeit sometimes indirect impact on the development of their motivations. First, and more indirectly, by affecting their perceptions of social acceptance within primary social groups such as the family and the school. And second, and more directly, by limiting their knowledgeability, to use the structuration term (Giddens, 1984), when (a) they internalise the impossibility to accomplish their goals, and (b) the lack of role models increasingly

diminishes their options. The result is that the perceived structure upon which these individuals, as agents, draw to engage in purposeful action—in this case, to gain

independence and become a person with a socially recognised identity—is reduced by social inequality.

In that vacuum, other ways of being recognised and assuming an identity are stressed, as brought forward in the theoretical framework. From here onward, the argument summarised by Hayward, that socially excluded individuals ‘over-identify’ with consumerism “in an attempt to create a sense of identity” becomes apparent (Hayward, 2004, p. 181). The first instance where this comes up is culture in general, and more specifically music. In this respect, genres that relate to their own group identity, mainly of exclusion, are predominant. This relation, is far from unidirectional. It is from within socially excluded groups similar to the ones where they live that virtually all culture comes from that the participants access. The preferred music genres are two, rap/hip-hop and Brazilian funk. The first one is older and historically included socially criticism in their lyrics. The main group, ‘Racionais MCs’, is

still very relevant in São Paulo, where it originated. Funk is a more dispersed, commercial genre, which often abuses of highly sexual and overly simplistic lyrics. Like current North American hip-hop, both genres are also known for emphasising consumerism as a successful lifestyle. These distinctions are somewhat common sense to Brazilians in general, who are used to listening to these genres. The participants confirmed this: “Man, because [Tupac, famous American Rapper] is like… the king of gangsters, they say it, you know. Because his music is crazy, it’s thug” (EL, p. 17). “Sabotage [similar to Racionais MCs] is an idol to me, man, really… I always liked. I never liked funk, you know. [Sabotage] talks about the reality, what is really happening” (JN, p. 3). “Only funk. Only fornication. Only sex. Only music about sex” (RT, p. 4).

It became clear how music and their environment stimulated an association with

consumerism to compensate for their exclusion of other spheres. This is aligned with the idea of institutional anomie, put forward by Messner and Rosenfeld, in which market institutions are granted predominance over other noneconomic institutions like family and school (Messner and Rosenfeld, 2012). As it will become clear in the next segments, this form of more detailed strain offers a better explanation of the respondents’ trajectory into crime. When asked about the ways which they use or invest their money, participants gave clues to items or experiences that they value, offering an insight into their consumption ideals. The answers reveal relatively minor and yet vague extravagances, mostly involving clothes, food, and drugs.

A few examples are: “To live well, having money to buy clothes, eat well, go to different places […] Sometimes you just want to eat a roasted chicken, steak with fries, something sweet on the street. Go to McDonald’s [laughs]” (RN, p. 27). “To buy weed, buy clothes, to eat. I like to eat well, you know. Street food in the middle of the night. Nike, Lacoste, Oakley” (RT, p. 9). “You always want to have a nice motorbike, a nice car. You learn from those who have” (EL, p. 27). “Back then I remember that the advertisement was that the phone would open like that, and the screen shined, blue and red [laughs]. Then I said, well I need to have a phone life this” (LS, p. 31).

Consumerism is internalised in different ways. One interesting instance is when talking about raising their children: “When I have a son… he won’t have the same childhood I had, man. He will have everything he wants; the boy is going to walk around wearing Armani” (EL, p. 41). References to music are often mixed in their accounts as well: “I want to make a million,