'If it's Already Tough, Imagine for Me...' A Qualitative Perspective on Youth Out of School and Out of Work in Brazil

Drawing on in-depth interviews with young women and men in rural and urban Brazil, this qualitative research explores gender dimensions in the causes and consequences of being "out of work and out of school." A key conclusion from this research is that this term (or the Portuguese: "nem-nem" ) does not translate well the complex realities of this highly heterogeneous group. The paper develops inductively from the data a typology of these youth, who face different barriers along their trajectories: a) barriers to building aspirations and internal motivation to return to school or work, b) barriers to action, and c) external barriers. Participants' position along this spectrum is shaped by social context and gender norms that frame youth's trajectories and envisioned futures. These observed patterns are particularly strong in rural areas, where youth perceive fewer quality economic opportunities and stronger division of gender roles within the household and in farming activities, which keeps young women in lower paid or unpaid roles. Participants who have successful trajectories to technical schools, universities, or formal work demonstrate strong resilience, which seems to be built on their relationships with their families, peers, partners, and role models.


Introduction and Motivation
"Se já é difícil, imagina para mim..." ("If it is already tough, imagine for me"). The title of this paper is based on a quotation from a young woman who does not study or work and lives in a poor urban community in Pernambuco, northeastern Brazil.
by choice or lack of options. There is also a lack of specific data for rural populations and very few studies offer a dedicated gender analysis. Research that delves deeper into these issues can help deepen the understanding of the problem, the different groups affected and individuals' own attitudes and behaviors.
What are young individuals' motivations for not participating in work or school? The young woman quoted above refers to the general difficulty of searching for a job; and reflects throughout her interview how her (socio-economic) background and lack of qualifications make her an undesirable candidate for a good job opportunity. At the same time, the resignation she shares is built on her reiterated experiences of failure in finding a job or finishing school within a context of limited opportunities, leading to her perceived inability to affect change in her own life. She lacks the support systems to facilitate access to these opportunities and -as a woman -feels she faces an additional set of disadvantages to do so.
Drawing on qualitative data from Pernambuco, northeastern Brazil, this research explores gender dimensions in causes and consequences of being out of work and out of school for poor youth. The paper is based on 108 in-depth interviews with key informants and young males and females (18-25 years) 7 in rural and urban Pernambuco, Brazil. Interviews explored a range of themes including: family and community contexts; experiences and perceptions about work and education; aspirations, plans for the future and perceived barriers to implement them; and knowledge about public policies.
This research adds value to the existing literature in two main ways. First, the use of qualitative methods provides additional insights on the topic since most of the available literature relies on quantitative methodologies. Open-ended questions enable participants to express their views in their own words and to invoke themes that are most relevant to them. This allows researchers to explore new perspectives on the issue; and helps unpack subjective and intangible factors such as meanings, attitudes, beliefs and perceptions of youth and how their family and community contexts may have shaped their behaviors and choices.
Second, the specific focus on gender-related issues enables an understanding of the different underlying causes and consequences faced by young women and men. Girls and boys may have different experiences and perceptions about their roles in society. Gendered social norms can shape girls' and boys' aspirations. Motivators to drop out of school and labor markets are likely to differ across genders. Understanding gender dimensions of the "nem nem" phenomenon helps identify strategies that may break the intergenerational persistence of gender inequalities -particularly among the vulnerable populations.
A key conclusion from this research is that the terms "nem-nem" or "disengaged youth" -commonly used in the literature -do not translate well the complex reality of the youth to which the terms refer. First, the term obscures the fact that this group is highly heterogeneous. It is composed of young men and women of different backgrounds, family structure and age groups, with varying levels of participation in income generation activities. Their situation is spurred by a variety of reasons related to their social context; capacity to form aspirations related to work and school; access to opportunities; availability of financial resources and support systems, among others. Assuming homogeneity may lead to the design of policy measures that fail to address the different obstacles faced by these subgroups. Furthermore, by defining their situations negatively, the term embeds a normative notion that stigmatizes them as individuals who are to blame for not choosing the "regular track", while ignoring constraints that limit their access to opportunities throughout their individual journeys.
In addition, we derive four major findings from the data. First, the path towards participation in the labor market or the educational system seems to be divided in three dimensions: having aspirations related to work and school and the internal predisposition to act on them; taking sustained action to achieve these goals; and having the ability to surpass external barriers encountered (e.g., lack of local training centers). Participants' position along this path is strongly organized by gender. Second, we demonstrate how exposure to a context of deprivation and poverty, restrictive gender norms, and the lack of support systems diminish youth's agency (capacity to aspire and take sustained action) and access to economic and educational opportunities -particularly for young women. Third, these observed patterns are stronger in rural areas, where participants perceive fewer quality economic opportunities, especially for young women. Finally, urban youth from the same communities who were successful in entering the labor market or continuing formal education beyond secondary school (despite similar negative educational opportunities) demonstrate significantly different attitudes towards work and education. Their resilience seemed to be largely built upon their support systems: their relationships with families, peers and partners.
The structure of the paper is as follows. Section 2 outlines the objective, main questions and methodology of the paper. Section 3 presents the typology of youth out of work and out of school. Section 4 discusses the role of gender norms, the social context, the lack of support systems in shaping labor and educational outcomes and the risk of resignation among youth. Section 5 describes the patterns observed specifically in the rural sample. Section 6 outlines factors that contribute to resilience based on the "positive deviant" cases. Section 7 discusses the findings and presents the conclusions.

2 Methodology
This research explores the trajectories, attitudes and perceptions of youth out of school and out of work in Pernambuco, Brazil, focusing on the gendered dimensions of the causes and consequences of their situation. The following questions guided the research: i) What are the trajectories of youth out of work and out of school and which barriers did they experience to remaining active in work and/or the educational system? ii) What are their aspirations related to work and education, plans for the future and perceived barriers to implement them? iii) How may their different family, community, school and work experiences have shaped their mindsets and choices? iv) Do the answers to the above questions differ between male and females? If so, why?
The paper is based on 77 in-depth interviews with young men and women (18-25 years) out of work and out of school 8 in rural and urban Pernambuco. The interviews explored a range of themes, including family and community contexts; experiences and perceptions about work and education; aspirations, plans for the future and perceived barriers to implement them; and knowledge about public policies. In addition, a set of nine "positive deviance cases" (PD) 9 was collected from the same urban sites to serve as a benchmark, helping to investigate the factors and strategies that enabled some individuals to continue their education beyond high school and to successfully enter the job market, despite sharing the same community background. Finally, 18 key informant interviews were conducted with local community leaders, policy practitioners and academics in the areas of education and labor to inform the research design as well as to capture their perspectives on the issue (see annex 2 for extended methodology).
Certain themes and realities were not explored in-depth in this research, such as: participation in crime and violence; the situation of underage youth; and how sexuality, race and ethnicity might impact the situation of youth. 10 In addition, the sample is limited to low-income populations in certain locations of rural and urban Pernambuco who may face a reality different from that of youth in families with higher incomes or in other geographic locations.

3 Towards a Typology 11 of Youth Out of School and Out of Work
In this section, we present a typology of youth out of work and out of school derived inductively from the data. The typology certainly does not -and is not intended topaint an accurate picture of the multiple realities faced by youth, since individuals will not fit entirely into one of the types. Nevertheless, the typology simplifies the understanding of the range of issues interviewees face in sustaining their participation in the educational system and in the labor market.
We find that youth's path towards participation in the labor market or the educational system can be divided into three dimensions. First, an individual must have aspirations related to work or school and the internal predisposition to act on them. Second, consistent action must be taken with the objective of achieving these desired outcomes. These two dimensions we refer to as agency. 12 Finally, the individual's external structure must provide him or her with the minimal conditions to access these opportunities (e.g., the supply of local quality schools or the availability of economic opportunities).
We identify three types of individuals in the data -each type facing a different set of barriers to advancing along their path towards participation in school or work.
The first type consists in individuals who expressed no aspirations related to work or school during interviews. The second type are individuals who express the desire to engage in school or work but admit not taking any action to make it happen. The third type are those who are motivated and demonstrate they keep trying to access opportunities but are limited by external factors.
Although these dimensions -internal motivation, action, external structure -are presented here in linear format to simplify the argument, we emphasize their mutually reinforcing relationship. In this sense, we follow Bourdieu's approach (1977), according to which individual agency and social outcomes shape each other mutually. 13 In other words, an individual exercises agency in a given social context with certain 11 Note that De Hoyos et al (2016) also suggest avoiding generalizing the phenomenon, highlighting the heterogeneity among ninis in the region. 12 Amartya Sen (1985) defines agency as "what a person is free to do and achieve in pursuit of whatever goals or values he or she regards as important" (p. 203). An agent, according to Sen is "someone who acts and brings about change" (Sen 1999, p. 19). The relevance of the concept of agency to gender inequality has been highlighted in the World Development Report 2012. Here, agency figures as one of the three primary dimensions of gender equality and is understood as the process through which women and men can use their endowments -such as land, property, education, or health -and take advantage of opportunities to achieve their desired outcomes. As a prerequisite for achieving these desired outcomes, one must first take a decision and then consistently act upon that decision (World Bank 2012). 13 The structure-agency debate is a prominent debate in sociological literature. The range of positions in the field includes those theorists who interpret social structure as determinant of one's possibilities and outcomes, making it thus often impossible to exercise "agency". Others emphasize the individual and his/her potential and see social structure as the result and consequence of individual actions.
conditions, power relations and predominant values. At the same time, the sum of individual actions results in changes to that structure or reproduction of the structure itself.
As with all typologies, differences are fluid and sometimes difficult to distinguish because our primary data are participants' own voices. Gender norms may shape the way individuals express themselves during interviews. For instance, the pressure to be a "breadwinner" may lead young men to speak with greater excitement regarding their future work lives in an attempt to fulfill social expectations before the researcher. On the other hand, we argue that having a narrative of motivation is a necessary condition for effectively pursuing these opportunities. 14 In the next subsections, we discuss the typology and the key patterns observed in each type.

Barriers to Internal Motivation: Participants Who Expressed No Aspirations or Internal Predisposition to Return to School or Work
This group was composed predominantly by young women with small children and in unions 15 and it was more present in rural areas. Those participants did not engage in any sort of income generation activity and were economically dependent on their partners or other family members. A section from the interview with a 21-year old rural woman exemplifies the lack of motivation, drive and aspirations related to work and education that is characteristic of this group. She married at age 16 and has a small daughter. The aspirations of this group revolved around having financial security and happiness as a family. They demonstrated a tendency to see the family as their priority project and responsibility and attributed less meaning to their individual aspirations and projects.
Their daily lives were focused on the domestic sphere and the needs of the family. Many of these young women dropped out of school after falling in love, having children or forming a union during their teenage years, without having gained substantial work experience. Others formed families later in life but gradually disconnected from work life after marriage to be able to focus on their families. Most participants spent their days performing household duties or taking care of their children and other family members, talking with pride about their dedication to their family and children. In most cases, they were responsible for the work within the household while older members (or male members) were "working outside" and had the role of economic providers.
Regarding their future, these young women referred to themselves more as recipients than as agents of change, which shows a normalization of their economic dependence. For most, the detachment from paid jobs was conveyed with a sense of relief as they referred to their predominantly negative experiences during their short work lives. 17 Additionally, the perceived limited set of opportunities 18 and burden associated with deviating from the expected domestic roles (e.g. conflicts at home, judgment from neighbors, and facing their own self-esteem issues), particularly in rural areas, shape women's meaning of work and their ability to envision self-actualization through that path. At the same time, some participants expressed discomfort with their economic dependence. Particularly in urban areas, women expressed shame and shared uncomfortable experiences about having to depend on others to be able to survive. Nevertheless, individuals in this group seemed resigned with their decision to not consider work or study as possible routes for their future.

Barriers to Action: Participants Who Express Motivation to Return to School or Work but Don't Take Consistent Actions Towards these Goals
This group was composed predominantly of single males and females -most of them living with their parents and depending economically on another family member.
Although some of them -particularly men -engaged in informal work sporadically, they did not actively look for work. 19 These were largely individuals who completed high school but did not take consistent action to transition to higher education or work life and were particularly present among urban youth.
These participants expressed the desire to have a good job or pursue specific careers they found interesting, 20 but admitted to have previously passed up opportunities to pursue these paths. For example, several interviewees had enrolled in the national exam for entering university 21 at least once but did not attend the actual exam. Other interviewees distributed their resumes to potential employers but did not show up for interviews when they were pre-selected. In justifying their behaviors, participants often demonstrated self-blaming or dismissive attitudes, for example, saying they "were simply too lazy" or "too spontaneous" to follow through. 22 The meritocratic belief that success depends only on their willpower becomes a source of shame and low self-esteem as they interpret their situation as their own personal failure, leading to further paralysis. The lack of role models, information and mentoring appeared to be important factors shaping the behavior of this group. Their parents and peers generally had low educational attainment and high participation in informal work. Parents and other family members were not able to actively participate in youth's decisions regarding work or school. Because they were not exposed to examples of individuals who had successfully pursued career paths they showed interest in, they did not see their aspirations as a realistic possibility and lacked the information on how to move forward. Schools also failed to support them in these endeavors. Hence, students who are motivated to study or work in areas that are less common in their social environment end up unable to make realistic plans, define concrete steps and cope with the emotional challenges associated to the pursuit of a long-term goal. Finally, because they were mostly single and economically dependent on older family members, these youth were able to be more selective with opportunities they pursued.

External Barriers that Limit even the Motivated Youth
The third group identified in the sample is composed of individuals who demonstrated consistent efforts to continue their education or return to work but were constrained by several external factors. Most of them had incomplete secondary 20 These included medicine, nursing, business administration and arts. 21 The National High School Exam (Exame Nacional do Ensino Médio -ENEM) is Brazil's tertiary-level entrance exam at the end of high school. While general secondary education has a modern system of student assessment, national exams like Enem are not compulsory for all students completing upper-secondary education (Silva, Almeida, and Strokova 2015). 22 This mismatch between intention and action serve as an example of what the World Development Report on Behavior, Mind, Society (World Bank 2015) refers to as "intention-action divide": the tendency to make decisions that favor the present over their future, even if individuals cognitively understand the implications of their action. school, engaged in sporadic informal jobs but looked for more stable opportunities. A distinctive feature of this group in comparison to the other two is their supportive partners, peers or family members, who seemed to act as enablers for action (this will be explored further in the "positive deviance" cases). Males who were married and had children were the majority in this group.
Although they wanted to continue their education, most youth in this group saw finding a "trabalho com carteira assinada" ("formal fixed job") as a necessary first step to consider returning to school or enrolling in university. School materials, foregone wages, fees, the need to pay for private care support and transportation require resources they did not have, leading them to prioritize work over education. The financial constraint was particularly strong for young males, who were more reluctant to seek support from other family members as they saw financial dependence as a source of shame. Participants with incomplete secondary schooling had very negative perceptions of adult education programs, 23 sharing their several failed attempts to enroll or complete them in the past. Key issues were the lack of vacancies and the inadequacy of curricula and teaching methodologies. Many perceive these programs as geared towards retired individuals; and found their structure did not take into consideration the physically demanding work many students completed before attending classes (paid jobs and taking care of small children). Hence, participants were only able to sustain the "double work-study" routine for a short amount of time, dropping out of programs after some time and continually trying to re-engage.
Although some participants demonstrated interest in attending technical schools, 24 they felt excluded from these opportunities. Enrollment criteria often required 23 In Brazil, adult education programs are available for people 18 years or older. These are a secondchance or adult education for youth and adults with low educational attainment for their age (an approach called Educação de Jovens e Adultos [EJA], also referred to as "Supplementary Education"). Because only individuals over 18 years of age were recruited for this study, completing secondary education necessarily means attending such programs. When someone turns 18 years old, they should be automatically forwarded to EJA classes instead of attending regular ones. 24 PRONATEC is an umbrella program coordinating a variety of existing and new vocational education and training policies, including courses in two modalities: Technical Education (TEC) and Initial and Continuing Training Programs ("Formação Inicial e Continuada"-FIC). Technical vocational education and training (TVET) delivery in Brazil involves multiple providers. FIC are short-term vocational training programs usually directed at improving existing workers' qualifications and not tied to the formal education system. The Sistema S (S System) network of providers has a key role in the delivery of these courses and other providers include federal institutes, state and municipal complete secondary education, leaving them ineligible to benefit. The small number of vacancies and high levels of competitiveness to enter marginalize participants with lower levels of educational attainment. The mismatch between the courses offered and their interests was another issue raised. Similarly, youth in this group who had completed their secondary education recognized the importance of governmental programs facilitating access to private and public universities, 25 but felt left out. One of the explanations given is that public schools do not prepare students well enough to compete in the national exam, leaving them unable to pass or benefit from quota systems 26 in public universities. As a result, their only realistic option was private universities, which were considered too expensive for this group. Even though Brazil has programs such as PROUNI 27 and FIES 28 to facilitate access to private higher education to those who cannot afford it, interviewees mentioned that payment installments are still prohibitively expensive and access to these opportunities is too competitive. These constraints reinforce their perception that they must have a stable job first to be able to invest in education.
Women with small children mentioned additional barriers related to education such as: inadequate times at which educational programs were offered; large distances and lack of safe transportation. Adult education programs are normally offered at night to allow the combination of study and a regular work schedule. Some female participants with small children considered the evening unviable because it is the time their children return from school. Others mentioned they could only find vacancies in night schools distant from their houses, which raised safety concerns. Women without school networks, and private providers (INEP 2013;Almeida, Amaral, and Felicio 2015;Brazil, Court of Accounts 2015). TEC courses are longer-term technical education programs that do form part of the formal education system. They are mostly offered at the secondary and postsecondary levels (Silva, Almeida, and Strokova, p. 69, 2015). 25 In Brazil, public universities are tuition-free. However, there are limited places and entry is determined by highly competitive exams, thereby often excluding those who have not had a highquality secondary education or attended an expensive preparatory course. 26 Law 12.711/ 2012 or "Lei de Cotas" (Law of Quotas) established that federal universities must reserve 50 percent of their vacancies to students from public high schools until 2016, taking into consideration social and racial criteria. Regulated by Decree 7.824 / 2012, the law proposed the allocation of 25 percent of vacancies for students from public schools with family incomes equal to or less than 1.5 minimum salary, 25 percent for candidates with family incomes equal or higher to 1.5 minimum wages and a percentage for self-declared blacks, pardos (mixture of white Brazilian, Afro-Brazilian and Native Brazilian) and indigenous people. 27 The University for All Program (PROUNI) is a federal program that provides scholarships to poorer students to study in private universities. 28 Financing of Higher Education Students (FIES) is a federal program that provides financing to students who want to pursue a higher-education degree at a private university.
supportive partners or family members were more burdened because they lacked financial resources to cover schooling and care expenses and they did not have anyone to accompany them to school in the evening.
In terms of barriers to access labor markets, the economic crisis and a constrained demand for jobs was the key issue mentioned. 29 The combination of financial crisis and the greater pool of individuals with higher levels of qualification or experience applying to the same jobs put them particularly in disadvantage in accessing economic opportunities. A few interviewees also referred to Jovem Aprendiz, 30 a program designed to facilitate access of inexperienced youth (14-24 years) to labor markets. The program aimed to provide youth with specific skills and experience that could help them start their professional careers. Although interviewees saw the program as positive, they highlighted that coverage was too low, only benefiting a small number of individuals in their communities. 31 Another obstacle for women's access to jobs was the fact that most available opportunities involved long work hours and included night shifts. The lack of affordable care infrastructure makes it difficult for women to combine their working lives with their care responsibilities. Either women must face a strenuous double burden, or they will have to negotiate the allocation of household tasks at home. 32 As a result, many partners were unsupportive of their partner's plans to complete their 29 After growing around 4 percent between 2003 and 2010, Brazil's economy gradually lost steam in 2012-13, was stagnant in 2014 and went into recession in 2015. The economic slump took a toll on the labor market, as the unemployment rate gradually rose and payroll indicators pointed to job destruction (World Bank 2016). 30 The Apprentice Law (2000) to facilitate development of programs for the inclusion of youth in the labor market. It requires all medium-size and large companies to hire youth between the ages 14 and 24 years while they undergo technical and professional training in a special employment contract of up to two years. 31 This perception is in line with the analysis put forward by Marra et al (2015), indicating firm compliance to be overwhelmingly low. For example, in 2013, only 23 percent of potential apprentice slots were filled. 32 Studies from Latin America and the OECD countries show that public provision of childcare results in high take-up, but not necessarily in higher maternal labor force participation rates. Rather, it leads to a substitution of existing arrangements (switching from informal to formal care) and improved quality of labor force participation (Chioda et al 2016). The report shows that demand for formal childcare appears to come from women who have already made the decision to participate in the labor market and who just readjust their existing childcare arrangements in favor of formal care. education or to take on a job. Families ended up prioritizing men's work. 33 This arrangement is reinforced not only by men's socially expected role as providers, but also with participants' expectation that men will earn higher wages. 34 It shows that, even when women value having a paid job, power dynamics within households and negotiating the organization of domestic work is still a major barrier to equal participation in the work force.
Additionally, women with small children mentioned experiences of discrimination.
Women reported being asked about whether they have a partner and small children during job interviews and not being considered for the positions after their positive responses. While it remains a speculation, the level of detail to which the interviewee, in the following quotation, was inquired about her childcare responsibilities suggests that discrimination based on gender and care responsibilities may account for some of young women's difficulties in accessing jobs.

How Poverty, Gender Norms and Lack of Support Systems Can Limit Youth's Participation in the Educational Systems and Labor Market
In this section, we show how the combination of a social context of poverty and deprivation, a set of restrictive gender norms, and the lack of support systems may contribute to limit youths' outcomes related to work and school. However, it is important to point out that, although we are portraying major patterns, there were clearly deviations from those. The actual extent to which these patterns will determine individual outcomes can vary depending on the specific combination of contextual factors related to family, institutions and community as well as the individual's capacity to exercise agency.
Individuals' capacity to form aspirations 35 in relation to work or education is adaptive to their social contexts and framed within a context of reference and sense of belonging. 36 As discussed by Duflo (2012), in contexts of poverty and social exclusion, hopelessness and the perceived lack of opportunity for life transformation may undermine one's willingness or ability to make choices, causing individuals to deliberately "hold back". Hope can fuel an exit from the poverty trap; and, conversely, the deficit of hope can create a poverty trap, hindering the development of other capabilities and limiting how much individuals can achieve.
A recurrent theme from the interviews is the demotivation participants felt after failing several times to access work or educational opportunities, and how these reiterated failures led them to slowly give up on a desired course of action. A significant share among these youths did aspire to return to work, school or to engage in tertiary education. Nevertheless, the constant failed attempts can generate frustration and resignation in a context where services are of bad quality or where opportunities are unavailable (frustration/ lack of opportunities effect). For example, a young woman who is motivated and keeps trying to find a good job may eventually give up. After failing several times to be selected for a job, she seems to have accepted her unemployment as insoluble and expresses having "perdido as esperanças" ("lost hope"). She begins perceiving her efforts as useless, causing her to stop trying to look for a job. Gender norms also frame young men's and women's experiences and references of the trajectories that are possible and desirable for their future -including in the realm of work and education. As noted by the World Development Report 2015 Mind, Society and Behavior (World Bank 2015), rather than considering all possible costs and benefits from a self-interested perspective, people are malleable and emotional actors whose decision making is influenced by contextual cues, local social networks and social norms, and shared mental models. These play a role in determining what individuals perceive as desirable, possible, or even "thinkable" for their lives. Most women interviewed had female family members and peers who formed families during adolescence. This shapes their views of what is normal, acceptable and socially desirable. If young women perceive their roles as that of being wives, mothers, and not to be working for an income unless there is physical need for it, they will likely not aspire to do something different. The same issues were found in rural areas, but they seemed to be even more intense and result in barriers even more difficult to overcome (see figure 1).

A Downward Spiral of Gradual Detachment from School
Participants, particularly in the urban sample, had negative perspectives on their school experiences, portraying school as spaces that felt demotivating, oppressive and chaotic. The often-mentioned description of schools as "an anarchy" referred to poor infrastructure, cancelled classes and absent teachers and the social environment among peers. There were several reports of bullying, drug use, trafficking and violence within school and its surroundings. Teachers and school staff were often described as dismissive or emotionally abusive, leading them to believe that school was not a place where their voice mattered. Most of the youth who were interviewed did not find school content and teaching methodologies engaging. 38 Many were unable to state their preferred subjects or the usefulness of school content to their lives. Those who were able to usually refer to teachers with whom they developed positive relationships or who inspired them to think for themselves, which shows how teachers' positive attitudes towards students seemed to play a vital role on how youth perceived the value and importance of their own education. The perception that friends who had finished school are "in the same place in life" leads to an evaluation that efforts will not pay off, as good work prospects for them are limited anyways. Many participants pointed out how peers who had finished school were in the same place as them or even had a more difficult life. For example, one urban young woman explained her friend had finished school and secured a fixed job, but that this job was exhausting, with low pay, long hours and no leisure time. In a context of high burden, lack of support systems and low perceived incentives to continue, youth start skipping classes or repeating grades. Many youth come from mono-parental families with working parents who had little time to be very present in their daily lives. In addition, parents mostly did not have successful trajectories in school and may not feel prepared to support their children in their studies. As they progress in school, they watch their peers gradually dropping out, which leaves them questioning the rationale for staying. Hence, youth lacked parental figures or mentors who would support and monitor their school performance. As the school is often unable to fulfill that mentorship role, youth end up without the motivation to keep surpassing the challenges they face. For example, a young individual struggling with school content and lacking the support to improve their learning capabilities may feel like they lack the capacity to continue in school. The final decision to drop out is usually taken in critical moments, such as the transition into high school or family shocks. When families move to other locations, youth may fail to enroll at the school at the new location. The sickness of family members may also force them to prioritize paid work or confine them in the role of caregivers. Parental separation or the introduction of step-parents can lead to conflicts over decision-making and division of work within their households, triggering a desire to become independent.
Critical moments may trigger youth's attempts to transition into adulthood, which are largely shaped by traditional gender roles reinforced early on at the family and community levels. For young men, the pressure to become financially independent leads them to assert their masculinity through their ability to navigate independently in the professional world and display symbols of financial success. As they see their peers start engaging in informal work while still in school, they may start doing the same. In some cases, the family exercises pressure for them to work, which creates conditions in which they feel they are unable to continue with both education and work. For young women, the transition is often expressed in their desire to form their own families, which may lead them to drop out after teenage unions and pregnancies. 40 Many young women interviewed had mothers and many peers who married or got pregnant at a very young age. Although focusing one's priority and efforts on family building and childrearing can be a source of fulfillment and accomplishment for women, it is important to note that this is a choice rather than an imposed path. For some women, this option must be understood in a context of a limited set of 39 See for instance Branson et al (2014) who show that overage students in South Africa are more likely to drop out because they started late or because they repeated. 40 The idea that motherhood is the predominant way poor women in Brazil imagine their transition to adulthood is also emphasized in Heilborn and Cabral (2011). The authors argue that in a context of lack of female empowerment and lack of other opportunities, young women do not see other transitions to adulthood other than by becoming a mother. On the other hand, education levels are highly correlated with a woman's degree of sexual autonomy and with a delay in child bearing and marriage (World Bank 2012). Hence, on the one hand their immediate decision-making capacity may increase with these important life events (marriage, union, pregnancy) because they are associated with adulthood and not adolescence. At the same time, the decisions they may take at this point are often a product of short-term thinking, lack of thorough planning or consideration of the consequences they will face.

How Dropping out of School Contributes to Labor Market Outcomes
After dropping out, social norms around marriage and the pressure to fulfill the roles of the "good wife" and "good mother" will tend to shift women's focus towards their domestic and family lives while also implying economic dependence on a partner. 43 A common theme in the trajectories in married life for women is the 41 Attitudes and expectations about the future can influence the probability of teenage motherhood (Plotnick 1992(Plotnick , 1993(Plotnick , 2007. When young women are poor and face a lack of opportunities, teenage childbearing and motherhood can be perceived as a rational choice. In other cases, given the specific circumstances, the incentives to avoid a pregnancy may not be high enough (Azevedo et al. 2013;Cater and Coleman 2006). 42 For the concrete case of Brazil, Heilborn and Cabral (2011) emphasize the differences in the perception of ways to transition to adulthood based on social class. The authors clearly link these observed differences to availability of opportunities or the absence thereof: "This ideal conception of passage to adult life (including the completion of an education and the building of a professional career) ignores that the availability of social opportunities is not offered in equal conditions for youth of different social classes" (pg. 42).
perceived need to give up on their independence and life projects for the good of building the family. This reflects how the "good wife" and "good mother" must be available for and attentive to the family and the needs of the partner. Marriage, union and childbearing (which often follows dropping out of school) deepens their lack of mobility even further, reducing their exposure to networks and finally perpetuating their status as economic dependents. 44 Q17: P: Like...for work, many times there has been an opportunity for me to work and he is like "no, don't go to work", these things…He said "I already work, you don't need to work and etc.", these things… IR: And how is that for you? P: It's bad. But I talk to him a lot and I believe he is changing… Woman, 24 years, rural Family and community attitudes about women's mobility and work shape their meanings of work as a means for self-actualization. In most families, young girls were responsible for household chores either by themselves or sharing these responsibilities with another woman in their house. Particularly in urban areas, women tended to be more confined to the private space. While for young men, work becomes a way to confirm their status in the household and the community, this may be the opposite for women. Working may involve burdens such as having to deal with conflicts with jealous or unsupportive husbands, having to face the double burden of paid and unpaid work or facing judgement from members of their community about being a "good mother". For example, many young women perceived the public space as a threatening space or a space of social surveillance. In this context, women will tend to live more domestic lives and are less exposed to other women who may be pursuing less traditional paths involving careers. On the other hand, after dropping out, because young males are expected to provide for themselves or for other members of the household, they are drawn into precarious work situations due to their low educational attainment. The "bicos" may have served well as a short-term strategy in a certain moment in their lives; but once was described quite uniformly as "an obedient, caring, and respectful mate to the good husband" (pg. 37), responsible for domestic chores and for caring for other family members. 44 Heilborn and Cabral (2011) also document that girls who became mothers in adolescence had been more engaged in domestic work prior to having their baby. At the same time, the social isolation found prior to early childbearing in these women often becomes exacerbated after child birth. The same authors (Heilborn and Cabral 2011), summarizing the GRAVAD study in Brazil, state that young women become more confined to the house after giving birth. Importantly, they also show that this is not necessarily true for middle-class women who do go out more, receive more support for maintaining social relations, and do not experience confinement as radically as young women in the lower socioeconomic segment.
they become providers of a household, the instability of their work situation puts a strain on their emotional health and self-image. Given the low educational attainment, lack of experience and qualification, youth will mostly find opportunities in temporary, low stability and informal jobs with low salaries. 45

Observed Patterns are Stronger in Rural Areas 46
In this section, we explore more in-depth the data collected from rural sites. Rural sites consisted in "agrovilas" 47 located in the semiarid region of the state of Pernambuco. As we will demonstrate, rural youth perceived similar -albeit seemingly intensified -constraints to their participation in the educational system and labor markets as poor urban youth: more traditional gender norms and sexual division of labor; fewer opportunities; and more stringent external barriers.

Exclusion from Opportunities and Diminished Value of Education
According to the rural youth interviewed, unless one migrates or has access to private transportation they are unable to access educational opportunities and other economic opportunities outside of farming. 48 Most individuals made a living from agriculture by taking on small daily jobs -known as "diárias" 49 in larger farms, planting on their own family plots or both. In addition, a small number of families have small businesses such as bars, food stands or mini markets. Nevertheless, agrovilas are situated over 60 km (and connected by precarious unpaved roads) from the nearest urban hub and, from the late afternoon onwards, there is no public transportation available to these areas. 50 45 This is consistent with findings in De Hoyos et al (2016) and Monteiro (2013). 46 On average in the Latin America region, the share of youth out of work and out of school is higher in rural areas (21 percent of youth) than in urban areas (17 percent . 47 Agrovilas are typically planned villages in which houses are grouped together in one area rather than on each farmer's parcel of land, which are located at further distance. This way of grouping houses is meant to facilitate the provision of basic services to households such as electricity and potable water (Zibechi 2007). 48 Agrovilas had access to irrigation systems, that function as key enablers of farming activities. Such systems are not commonly available to family farmers in other semiarid regions of Brazil, which can suffer from intense drought periods. Upon the formation of these agrovilas, families who were relocated there were each given a house, with access to water, a plot to plant with irrigation systems, and a living stipend. 49 Informal daily work for women will generally involve laying seeds, removing fruits and vegetables from trees ("despencando o amendoim"), cleaning and boxing them, while males will do the heavier work such as sowing the land and applying fertilizers and pesticides ("evenenar"). 50 Costa and Ulyssea (2014) point to the lack of opportunities and distances as major obstacles for young people to enter the labor market and access education opportunities in rural areas in Brazil.
At the same time, migration was not seen as a viable or desirable option to the group interviewed. Important factors discouraging migration were: the perceived precarious nature of jobs available to them in cities; the fear of the "unknown" (particularly for girls); the need to "know someone" in the city to have a place to stay temporarily and to secure a good job; the desire to be close to their families or attachment to the community; and the need for rural succession of their family properties. According to key informants interviewed, several productive inclusion policies that were implemented in recent years shifted the perception that to be a farmer was lack of option to more positive interpretations. This has had implications for the aspirations of young women and men living in rural areas: Now, youth were fighting for their right to remain on their land, which has meant greater demand for opportunities and policies to be brought to rural areas to enable rural succession for youth and less willingness to migrate.
In this context of exclusion from opportunities, school seemed to have even less meaning to rural youth in comparison to urban sites. First, youth must deal with the double burden of work and school from even younger ages. Rural youth's relationship with farming is more "naturalized" as compared to the relationship urban youth have with work. Girls will share the responsibility for the domestic work with their mothers and other female family members from a young age. Boys are expected to start working early on the family plot under the guidance of older family members. 51 This work is physically demanding and even more challenging to combine with a school routine.

Q19: P: Because we have a profession since we are born. Who is born in the rural area, is already born with a profession, that is [the profession of] farmer. The academic formation is a process that comes later, after which I may or may not have another profession...But I am already born with one in that place. Key informant, rural
Second, school curricula seem even more disconnected to their realities and perceived possibilities. Differently from urban youth, when rural youth described their experiences in school, most spoke positively about the school quality and infrastructure. Nevertheless, most were unable to explain what they learned in school that was useful for their lives. Family members and peers have even lower educational attainment in comparison to urban areas. Finally, there was a perceived lack of opportunities to apply or derive returns from formal education in the absence of jobs Third, young males perceived family farming as a high-risk activity which requires very specific and practical knowledge, which was provided in their social networks and not in schools. This was a recurrent concept mentioned by male interviewees in farming as a "game" or a "lottery." 53 In this context, informal, male-dominated learning networks (fathers, uncles, and other family members) often prepare young men to take over family farming activities and partly work as substitutes for formal education, as this type of knowledge is not acquired in schools. In this sense, male youth not identified as currently working or studying may be in a learning phase, preparing to make the transition into working their own land.

Man, 18 years, rural
Consequently, although schools in the visited rural communities were not of bad quality, 54 they still failed to retain youth. Overall, participants report positive or neutral experiences at school in terms of the quality of supply and the infrastructure. Nevertheless, drop outs occur even earlier than in urban areas, typically around middle school in 4 th or 5 th grade. The predominant triggers are the same as in urban areas: the 53 The image of a 'game' is explained as follows: First, one must invest a substantial amount of resources to prepare the land without guaranteed returns. This initial investment requires either having a large amount of savings -which most do not have -or acquiring loans in-kind with local shops selling inputs such as fertilizers and seeds. To acquire such loans, it is important for farmers to be in good standing with shop owners. Second, after securing the inputs, a farmer needs to "saber trabalhar bem" ("know how to work well"). This refers to the fact that lack of experience may lead to employing the wrong combination or quantities of inputs and therefore having only limited output. In addition, one must know how to administer the entire activity to succeed: This involves taking only the necessary loans, spending your money carefully and saving when you make a profit for your future investments. Even so, this "game" has no guarantees as climate factors may cause a farmer to lose their entire investments. 54 Interviewees had mostly good perceptions of local schools, pointing to the fact that teacher absence was uncommon, transportation systems were functional and school uniform and meals were regularly available. Nevertheless, some also reported poor teacher-student relationships and classes combining too different age groups as issues.
start of work life, unions and pregnancies -but those tend to happen earlier than in the poor urban communities included in the sample.

More Stringent Gender Norms Impact Access to Opportunities to Some Groups
Strong occupational segregation and division of gender roles in the households results in even scarcer and lower quality paid work opportunities for women. Social norms regarding the division of labor within households seem to be more stringent than in urban areas. The family operates also as a production unit and work inside households is divided according to gender and generation. Although some informants mentioned that arrangements were changing recently, most expressed the view that "trabalho de casa é trabalho de mulher" ("domestic work is women's work"), 55 "roça é trabalho de homem" ("farming is men's work"). 56 Fewer women interviewed worked outside the home in comparison to urban sites, especially those married. Women in families who had their own property often did not participate in farming activities, which were operated by males. In addition, women who do engage in informal daily jobs perform activities that pay less in comparison to the work done by men. In addition, limited access to land, information, financial resources, care support and negative attitudes towards women's mobility also limit women's access to educational and economic opportunities in rural areas. Rural succession usually takes place within the family, with the father choosing a child or dividing the farm between his children -usually the men -who will be responsible for the property and care of parents in old age (Brumer 2014). Hence, it is common for many children, but especially 55 For most women in unions, a typical day will involve waking up early to prepare the breakfast for the husband before he goes to work on the farm, cleaning and tidying up the house in the morning, preparing a meal for the husband when he returns home for a mid-day break, and cleaning and watching TV the rest of the day. Typically, in rural areas, gendered segregation of work begins very early in life: for the male children, work involves farm labor, usually conducted by the father and his male children (Paulo 2016), while domestic activities are carried out by the mother with the involvement of her daughters (Brumer 2007). 56 Although, in some locations in the northeast, women may perform farming activities for selfconsumption, such as growing greens and herbs and raising small animals, this was not found to be the case in the agrovilas visited, where such activities were also considered men's work. 57 Very few women could work in services, such as working as domestic workers or working in small local markets. Few women also worked as public servants in the local social service units as teachers, nurses, cleaners, etc. Nevertheless, according to the stories shared, access to these jobs usually required knowing someone who either appointed them to these jobs or let them know beforehand that these vacancies would be open, which gave them the lead. However, most of these positions are already filled by older women, limiting the opportunities for younger individuals.
for women not to inherit the land (Ferreira and Alves 2009). No women interviewees owned, took decisions and worked on their own land. 58 Furthermore, male-dominated learning networks contribute to constraining women's possibilities to conduct farming activities autonomously. In addition, restrictive family and community attitudes towards women's mobility posed an even severer barrier to women's access to education opportunities in the city. Finally, the issue of sexual identity and its contribution to social exclusion also emerged in rural areas. Homosexual men interviewed mentioned the difficulties faced in accessing work opportunities due to prejudice. Some "recruiters" will refuse to speak to them or will only offer typically "women's work". Interestingly, the view that they would not be apt to perform certain types of work due to their sexual identity seemed to be internalized. A participant who identified himself as gay explained why he only had access to certain types of work: "Since I am gay, it is heavier for us, you know…that is more men's work". This demonstrates the importance of working on gender norms in communities to enable the inclusion of individuals who face overlapping constraints.

Less Access to and Information about Public Policies and Programs
Most interviewees shared the impression that they feel "forgotten" by the government and demonstrated very little knowledge about public policies in the areas of education and labor. This speaks to the notion that resources are generally concentrated in urban areas, and that if poor urban youth face difficulties in accessing these opportunities, the struggle of rural youth is even more pervasive. A key informant from the education sector explained that, although in the past decade there have been several advances in the promotion of policies to improve the well-being of youth in Brazil, reaching out to the population is still a challenge, especially in the case of rural youth. 59 The best known governmental intervention by rural youth was ENEM. 60 Nevertheless, they had very little knowledge about the specificities of the program and how to access it. In the following quotation, a young rural man discusses the extent of the sense of "disconnectedness" faced by rural youth: The disconnectedness goes beyond simple access to information on educational or training opportunities. It is also accompanied by a fear of the unknown and an inability to envision themselves navigating through the necessary steps one must go through to access such opportunities.

Building Resilience: Support Systems and the Sense of Possibility 61
This final section presents the results from the analysis of the data collected through in-depth interviews with individuals who achieved successful school completion and entry into the job market despite their origin in the same urban neighborhoods in urban Pernambuco. It will explore facilitating factors and strategies that enabled these individuals to move forward in their educational and work trajectories. 59 This "distance" between rural youth and governmental resources was reflected in interviews with both youth and local key informants from the education and social sectors. Furthermore, it was also observed when visiting a social service unit targeting youth in the area where the fieldwork was conducted when unsuccessfully trying to explore more in-depth the condition of youth in the area through this entry-point. This may be interpreted as a reciprocal process: Youth are not reached by these services -because they are either not aware of such programs and institutions or do not see the value added from reaching out and services themselves fail to reach out to youth. 60 Although Brazil has governmental programs to support family farming, such as PRONATEC Campo (rural technical courses), Technical Assistance and Rural Extension Services (ATER), Program to Stimulate Rural Productive Activities (Programa de Fomento às Atividades Produtivas Rurais) and Seeds Program (Programa Sementes), they were not mentioned by participants during interviews. This suggests they are either not available locally or that the local population is not aware of their existence. 61 Five women and four men were recruited from the same urban communities used in the first phase of data collection. They were divided among two distinct groups: individuals who were attending or had concluded a technical course or a university degree; and individuals who had attended a technical course or a university degree and managed to secure a "trabalho fixo com carteira assinada" (fixed job) upon completion.
Although this group shared similar negative school 62 and work experiences compared to the first group interviewed, they demonstrated very different attitudes and meanings of these experiences. These young men and women attributed positive meanings to these experiences, seeing them as "stepping stones" to the achievement of their longer-term goals and speaking with pride about the need to persevere to achieve better living conditions. They mostly aspired to specific and high paying careers -such as business administration and engineering -and demonstrated structured plans with clear steps and realistic expectations of the difficulties they would face along the way. Their greater capacity for resilience, higher levels of selfconfidence and self-reliability seem to be largely built upon their relationships with their families, peers and partners.

Man, 21 years, from urban Brasilit
They also demonstrated clear and personal motivations to continue their educational and work trajectories. Interviewees who were out of work and out of school often conveyed aspirations in terms of their future work lives in a vague manner 63 and seemed unable to pinpoint the reasons for which certain jobs seemed attractive to them or how finishing high school could make a difference in their lives. In contrast, interviewees from this group demonstrated clear and personal motivations for pursuing a career or enrolling in a technical or university course. These included: the recognition of higher financial returns they could derive from more years of schooling; the intrinsic value of learning; the desire "to be somebody" or "to be successful in life"; the desire to contribute to the wellbeing of their families and to make other family members proud.

Support Systems -How Families, Friends and Schools Can Make a Difference
The most striking difference between youth in this group and youth who were out of work and school was the level of parental involvement in youth's decisions about education, work or life in general. Almost all the interviewees were single, had no children and still lived with both parents. This is different from the setting of other interviewees, in which the incidence of mono-parental families was higher and unions and pregnancies happened early in their life. Youth in this group demonstrated more positive family relationships, using words such as "diálogo" (dialogue), "apoio" (support) and expressions that denote the family as a team. Parents and older family members acted as role models and motivating figures for them, monitoring their school activities and holding youth accountable for school outcomes. This is rather different from other groups interviewed, in which parents were less involved in adolescents' school lives and were not able to enforce their attendance in school. These interviewees referred to their parents' involvement positively, as protecting them against the harms of life and poor decision-making. This greater involvement provided them emotional support and encouragement in critical moments in their work or school lives. Transitions to high school or from high school to university or work life were described as emotionally and psychologically distressing experiences, when they had to navigate through an environment with new peers and rules of engagement. Participants shared several moments in which they considered giving up and when their parents' encouragement was crucial to move forward. This demonstrates how a parental or mentor figure can provide reassurance, guidance in managing tasks, as well as reassure the young adult that his or her reaction is not uncommon.
In addition, family members of this group also provided them with practical and financial support, protecting their children from the burden of early work. In some cases, parents actively searched for opportunities for their children or accompanied them to inquire more details about specific opportunities. This contrasts sharply with the experiences shared by other groups interviewed, who reported not knowing what to do, being discouraged by the "fear of the unknown" and the complex process required to enroll in a university or take a new job. This support seems to have the important effect of reducing the costs of searching for youth by creating a sense of shared risk and "team effort". In addition, most families did not require children to contribute financially to the household, and they protected them from excessive household chores. This is not trivial as the evidence base showing that negative effects of child labor on schooling outcomes is quite strong. 64 By instructing their children to prioritize school as their key responsibility, parents not only signal an important symbolic meaning to the children's education for the family as a whole, but also allow their children to dedicate their full time and energy in learning.
Also, from the young women's perspective, their relationships with partners seem more supportive and equal. Although most female interviewees from this group were single individuals living with their parents, some had relationships, others have had them in the past. When they refer to their relationships, they use words such as "mutual incentive" and "respect". Nevertheless, some young women in this group were still exposed to controlling behaviors from their partners. With clear aspirations and sense of priorities in their lives as well as a solid support system at home, these women demonstrate higher levels of agency and an increased capacity to make their own decisions and to exercise their power in relationships.

The Sense of 'Possibility': More Role Models and Networks -Particularly for Women
Individuals in this group also had more positive perceptions of their friendships and were more likely to have an extended network of friends outside the community. While they typically had a few long-term friends from the community, most of their current friends were from outside their community. Some described the process of having lost friends in the community throughout their lives as following "caminhos diferentes" ("different paths"). Their social mobility impacts their social networks (and vice versa), gradually detaching them from their local peers. This process is not easypeers from their original networks show resentment which can create emotional distress. On the other hand, ties with peers who had similar interests gave them additional capacity to persevere as they found support in sharing the same struggles.
They had more cases of siblings, other family members and peers who had completed high school or had successfully entered the job market or university.
When probed about what facilitated their access to study or work opportunities, most of them mentioned accessing information through a family member or a friend, who could pass on the information about the opportunity, suggest their names to prospective employers and help prepare them for job interviews or examinations. In addition, knowing people who had successfully graduated from technical schools or university in their social circle also gave them a greater sense of possibility for their aspirations and a means to emulate successful strategies to achievement them.

Woman, 23 years, from urban Caranguejo-Tabaiares
The existence of role models was particularly important for young women.
Interviewees in this group were significantly more exposed to working women in their lives than other interviewees. In addition, these young women were also more likely to identify working women as persons of reference in their lives precisely because of their perseverance or hard work. This shows how positive role models for girls are important for helping girls build aspirations related to work and school.

Discussion and Conclusions
This research explored the experiences, meanings, attitudes and perceptions of young men and women in Northeast Brazil who do not fully participate in the labor market nor in the education system. Interviews explored a range of themes related to individual, family and community contexts as well as experiences, perceptions and aspirations related to work and education.
We find essentially three different groups of young individuals who are not at work or at school. The first is composed of mostly women in unions or with small children who do not express any aspirations to return to school or work. These women have negative attitudes towards women's participation in the workplace and have their lives focused on the domestic sphere. The second group is composed of young men and women, predominantly single and with no children. They are not attracted to the lowskilled jobs they see available. They aspire to go back to school, to return to university or pursue specific careers, but do not take measures to achieve these goals. Their inaction seems to be related to the absence of role models in their close lives who have attempted to engage in similar tracks, which limits their access to information and instills a sense of not belonging. Finally, the third group is composed by young men and women who aspire to return to work or school and continually try but are faced with external barriers that they are unable to surpass. In terms of work, these barriers are associated with their lack of access to opportunities due to the perceived economic crisis and their lack of qualification and experience. In terms of education, they face difficulties in combining work and study, and their lack of financial resources leads them to prioritize work over education. Long distances, inadequate hours in which most education and work opportunities are offered, and discrimination from employers are other factors that emerge and that specifically constrain women's ability to combine employment or study with their care responsibilities.
One of the key findings is that the term "nem-nem" does not capture well the complex reality of these youth. The term seems to hide the fact that this is a highly heterogeneous group -individuals are participating in neither the labor market nor education for a variety of reasons. By treating them as a homogeneous group, policy makers may fail to provide effective policy measures to help change their circumstances.
The application of the term "nem-nem" in current research may also favor a mismeasurement of the problem. As the term often relies on a strict definition of "work," 65 it may fail to differentiate between complete inactivity, informality and engagement in unpaid work. Few of the individuals interviewed for this research were completely inactive. Participants demonstrated informal relationships to several forms of work. When asked whether they had worked in the previous week, participants answered negatively, but upon further questioning, they would describe having performed several informal sporadic jobs. They explained that, for them, "work" meant "having a fixed job" or "working for a boss". In addition, although many women did not engage in paid work, most fulfilled some type of caregiving role that involved long working hours at home. Given the differences in values and meanings attributed to work by society and participants themselves, studies may not always capture adequately the work conducted by these young people, who end up categorized as non-working individuals.
Second, by defining their life situations negatively, the term embeds a normative notion that stigmatizes youth while not recognizing their social exclusion is largely created by systemic policy failures. It portrays the picture of individuals who are to blame for not choosing a more "positive track of action", while ignoring the internal and external constraints that limit their access to opportunities throughout their individual journeys. This seems associated with the misconceived popular image of the "nem-nems" in Latin America as either middle-class individuals who choose not to work or study, or as young men who engage in crime and violence. 66 A better understanding of the social context reveals that many young women and men interviewed are following the tracks perceived as possibilities given the social reality they live in. While it may be socially expected that a "middle-class" young woman will finish high school, go to university and engage in the formal workforce, this is not the case for most of the youth we encountered. Most youth interviewed have parents and peers who did not finish secondary school and who dropped out of school to engage in informal work. The results show how, in the absence of role models of individuals who achieved success in their careers in their proximity, young adults may not aspire to achieving anything beyond what they have at the moment (social context/ poverty and deprivation effect).
As the paper shows, even youth who are actively seeking to pursue new lives for themselves and their families were often defeated by the limited set of opportunities. A significant share among these youths did aspire to return to work, school or to engage in tertiary education. Nevertheless, the constant failed attempts can generate frustration and resignation in a context where services are of bad quality or where opportunities are unavailable (frustration/ lack of opportunities effect).
Besides the social context of poverty and lack of opportunities, gender norms are also powerful in keeping girls from aspiring to economic activity, which may create a vicious circle. Most women out of work and out of school who were interviewed also had female family members and peers who formed families during adolescence. This shapes their views of what is normal, acceptable and socially desirable. If young women perceive their roles as that of being wives, mothers, and not to be working for an income unless there is physical need for it, they will likely not aspire to do something different (gender effect). However, economic opportunities are important catalyzers for agency. 67 Not only will these women have lower aspirations related to employment and paid work, but they will also not accumulate the necessary endowments to succeed in the labor markets. This may have important negative effects also on their children, who will grow up in similarly vulnerable situations and conditions of poverty and gender inequality, framing their preferences and aspirations based on those.
The heterogeneity among youth out of work and out of school calls for a multipronged approach to reach these different groups. First, a challenge among all groups is the external barriers faced to access economic and educational opportunities: limited supply of governmental interventions and too stringent eligibility criteria (e.g., ENEM, Jovem Aprendiz, PROUNI, etc); inadequate times and methodologies of adult education and technical courses offered; lack of financial support; lack of secure and safe transportation; lack of care services; among others. Nevertheless, there is the need to go beyond these external factors to effectively reach youth.
Second, a vital step is building youth's agency -their capacity to aspire and take consistent action -to support them in creating and pursuing life projects that go beyond their limited perceptions of what is achievable. In other words, it is likely insufficient to increase the supply of technical schools with the objective of facilitating youth's participation in labor markets if this is not associated with interventions that: facilitate access to information about opportunities and how they can change their lives; instill a sense of belonging and preparedness for youth who feel the opportunities available are not for them; offer support or mentorship arrangements that help them cope with the difficulties associated the pursuit of a goal that requires challenging their mental frames of possibility. The paper shows that the promotion of aspirations related to work and education among youth, particularly women, is an important entry-point for programs and policies. It may be important to also offer young women and men higher-order cognitive and socio-emotional skills training. Programs which seek to improve higher-order cognitive skills build on basic cognitive competencies of literacy and math, and include skills such as critical thinking, decision-making, organization, efficiency, problem solving, written and oral communications (Cunningham and Villaseñor 2014). Socio-emotional skill components training can equip young people with the capacity to think critically about social and gender norms and to avoid behaviors that perpetuate gender inequalities (Gimenez et al. 2015). Evidence shows that providing young people with these skills has positive impacts along many dimensions including education, health, and the labor market, both in the short and long term (Ibarraran et al. 2012, Bandiera et al. 2012, Cunningham and Villaseñor 2014. Third, it is important to address gender norms in families and community, and at the individual level. Many women interviewed are unable to envision lives for themselves beyond the role of the caregiver and specific challenges in attempting to change their lives. It is important that interventions: expose them to other female role models and mentors that can instill a sense of possibility and support them in the challenges associated with the process of looking for a job or studying; support women's negotiation of household responsibilities; offer positive coping strategies to deal with negative community and family attitudes towards their work, education, and mobility. One channel to address norms is the media: In Brazil, La Ferrara et al (2012) show that exposure to soap operas, showing women with smaller families reduced fertility and increased divorce. Jensen and Oster (2009) provide similar evidence for India, showing that exposure to soap operas portraying urban women reduced fertility and domestic violence as well as changed beliefs about women's autonomy in rural India. A recent review of the existing evidence found that enhancing mass media programs with discussion groups at the community level can turn into particularly effective transmission channels for new values and norms, especially in rural sites (Marcus et al, 2014).
Finally, specific interventions are needed for rural areas. The findings from the rural sample indicate a need to: a) address gender roles and gendered division of labor; b) link education content with the reality of rural women and men while also allowing them to think of possibilities beyond farming; c) promote rural youth's capacity to aspire and create autonomy, especially among girls; d) connect youth with opportunities by ensuring affordable transportation and mobility to city hubs. Finally, because rural youth have less access to information, it is important that these interventions are combined with awareness raising on existing governmental youth programs, active search and mentoring schemes, which can be implemented through the school system. Data collected. The paper is based on 77 in-depth interviews with young men and women (18-25 years) 69 who were out of work and out of school in rural and urban Pernambuco, Brazil. Interviews explored a range of themes including: family and community contexts; experiences and perceptions about work and education; aspirations and plans for the future, and perceived barriers to implement them; and knowledge about public policies. A set of nine "positive deviance cases" (PD) was also collected from urban sites to investigate the factors and strategies that enabled some individuals to complete high school and enroll in technical schools or university or to enter the job market. In addition, 18 key informant interviews were conducted with local community leaders, policy practitioners and academics in the areas of education and labor to inform the research design as well as to capture their perspectives on the issue.

Fieldwork.
Fieldwork was conducted in the period between March and November of 2016. The team included four field researchers from diverse backgrounds. Fieldwork was conducted in three phases: first with key informants and youth out of work and out of school in urban sites, then with "positive deviance cases" in the same urban sites, and finally with youth out of work and out of school in rural sites.
Criteria for selecting field sites. First, communities were locations in which field researchers had prior experience to facilitate participant recruitment and to secure safe entry for the field team. Second, communities were characterized as areas with moderate to high levels of poverty. 70 Third, communities did not have strong presence of NGOs. 71 In the case of rural communities, field sites should also have economies based primarily on agriculture and family farming. 72 69 Adolescent males and females were not recruited because of ethical concerns and respective research procedures. For example, Brazilian regulations require that informed consent is obtained both from the minor and their guardian, which would make recruitment substantially harder. 70 Data were collected from small communities composed of mostly informal settlements with low access to infrastructure such as water and sanitation or pavement and services. In the absence of official poverty data for these small communities, we relied on these characteristics as proxy for levels of poverty as well as on the assessment provided by fieldworkers. 71  Criteria for selecting participants. Two different groups of participants who were not at work or at school were recruited: 1) youth who self-reported as "not having studied or worked in the last year"; and 2) youth who "did not study but had informal temporary work (less than 20 hours) in the month prior to the interview." 73 The PD cases were also divided into two groups: 1) youth who were currently attending or had concluded a technical course or a university degree; 2) youth who had concluded a technical course or a university degree and managed to secure a "trabalho fixo com carteira assinada" (formal fixed job) 74 upon completion. 75 Recruitment strategy. A mix of recruitment strategies were adopted to accommodate the features of each site. Overall, participants were identified through local key 73 The literature review conducted for this study showed the heterogeneity in definitions applied in different studies to describe "ninis" or "nem-nems". The main differences were related to whether or not youth who are "unemployed" or "looking for work" should be included or not, or whether merely the "inactive/ not currently looking for a job" should be included. Similarly, the age range considered also varied between different studies as well as the consideration of whether nonremunerated work was accounted for as work or not. 74 We define "fixed job" as a formal, remunerated, personal, non-eventual job that is performed under subordination of a stable employer. The Consolidation of the Labor Laws (CLT) of 1943 instituted the use of the "carteira assinada" (formal contract), which records a worker employment history. The CLT also included a large body of laws that determine the individual and collective rights of workers such as: maximum hours of work per week; maximum overtime hours per week; the extent of paid leave; the duration of paid maternity leave; the minimum working conditions; the minimum notice period for dismissal; and so on. It also determines who is a formal worker and who is not, through the obligation to use the "carteira" (Gill 2002). 75 There are two reasons for using these parameters for defining "positive deviance" cases. First, a recurrent concept mentioned by the first two groups interviewed (youth out of work or school) when discussing their aspirations was their desire to have a fixed job with carteira assinada. Hence, we use this as a definition of success derived inductively from interviewees themselves. Second, because the non-PD sample is composed of those who do not work, work in infrequent and instable jobs and do not study, we can define that a parameter for success would be those who study or have formal fixed jobs.
The initial criterion for selecting all participants was stricter than the one applied in the very end: youth (18-25 years) who did not study or work (either formally or informally) within the past year. However, during fieldwork, recruiting male participants became nearly unfeasible. First, males were more likely to be out of the house and to refuse participation. Second, almost all males reported having engaged -although infrequently -in some sort of informal income generation activity in the last year. Third, many potential interviewees who local key informants identified as matches to the criteria did not self-identify as such. Their hesitation appeared to stem from feelings of shame and inferiority imparted by the social pressure to be economically active. Finally, even those who self-identified as matches to the stricter criterion were likely to reveal during the interview that they engaged in infrequent income generation activities in the last year. We interpret this as a demonstration of participants' informal and naturalized relationships to several forms of informal work, especially in rural areas. To adapt to this difficulty, we adopted broader criteria by dividing our target sample into the two groups mentioned above. Nevertheless, recruiting males remained difficult. As a result, most of our male participants belong in "group 2". All of them reported not having any fixed income.

Box 1: Research Design Adaptations Based on Realities Encountered in the Field
informants' suggestions, random "door-to-door knocking", snowballing 76 or a combination of those. Fieldwork was conducted in three phases in the period between March and November 2016. 77 Table 1 summarizes the data collected, recruitment strategies and criteria used in each phase. Research Instruments. The interview guide for phase 1 was developed based on the main research questions and taking into consideration topics that emerged from a selective review of literature 78 collected using the terms "nem-nems" and "youth disengagement" and focused on Brazil and Latin American countries. These emerging themes were used to develop interview guides and are broadly presented in table 2. They contained open-ended questions organized in the following sections: family and community contexts, relationships and daily lives; experiences and perceptions about education; experiences and perceptions about work; aspirations, plans for the future and perceived barriers to implement them; and a final section on public policies. 79 Phase 2 instruments were similar but with greater emphasis on facilitating factors to 76 Snowballing is a non-probability method used when the desired sample characteristic is rare. Respondents are selected on the criteria of their ability to provide certain information that may be relevant to the study. 77 Three female researchers and one male researcher interviewed women and men, respectively, aiming to ease participants' comfort when discussing sensitive topics. 78 Some traditional approaches to qualitative research resist formal literature reviews prior to the onset of collecting field data due to the belief that qualitative studies should attempt to capture the "meaning" of events, including their unique time, place, and distinct historical moment. Following the approach suggested by Yin (2015), the main purpose of the selective review was to refine our preliminary considerations regarding the specific array of previous studies directly related to the topic of study, method, and data source. 79 Interviewers were instructed to cover key questions laid out in the instrument while also probing for unforeseen themes brought by participants. remaining in school or at work. Phase 3 instruments had the same structure as the initial one but emphasized family organization of work and production. Data Storage. All interviews were audio recorded and fully transcribed. Audios, signed consent forms and transcriptions were stored under unique identifiers under the responsibility of the lead researcher to protect participants' confidentiality.
Data Analysis. The analysis process was organized in the following steps. (i) Prior to categorizing the data, two workshops with the four field researchers and two research assistants were held at the end of phases 1 and 3, in which each team member shared their experience and preliminary assessment of key themes emerging from the data.
(ii) An in-depth analysis of selected transcripts was conducted separately and simultaneously by the lead researcher and an independent research firm. 80 This involved reading each data document several times and identifying the key categories and patterns among them. (iii) Based on the previous two steps, a list of codes was created. All transcripts were then coded by two team members using the software Atlas.ti. (iv) Data organized under each category were read and analyzed according to participant characteristics (e.g., location, gender, marital status). 81 This allowed the team to better understand the heterogeneity of themes emerging for different groups in the sample. And (v) a comprehensive literature review was conducted to understand how the study's emerging findings dialogued with existing literature on several topics. This approach to data analysis allowed the team to balance the need for a policy focus, young men who are not "breadwinners" may have caused a self-selection of interviewees who felt less pressure to be economically active.
The sample was limited to low-income populations in certain locations of rural and urban Pernambuco who may face a reality different from that of youth in families with higher incomes or in other geographic locations. For example, families with higher incomes may have more resources to support their young members in coping with difficulties in school or when transitioning out of work. Different socioeconomic realities may shape youth's experiences with education systems 87 and their access to the labor market. Finally, different political contexts in other geographical locations may affect the implementation and availability of the public policies discussed here.
Qualitative research is not representative of the whole population, but that does not compromise its validity. Rather, findings from qualitative research should be understood and used in ways that are different from quantitative findings. The objective is not to test causalities and generalize findings. It is precisely by studying the conditions, meanings and practices found in specific contexts that qualitative research can provide a deeper insight into the heterogeneity within a generalized picture. Doing this also allows for the collection of data that can help in refining theories, but also alerts for the caution needed when applying those theories to specific contexts.

Description of Field Sites
This subsection provides a brief description of field sites-based data from observations, key informant interviews and data collected from official statistics when available.
Urban Communities: Caranguejo-Tabaiares, Brasilit and Córrego do Jenipapo, Recife (Pernambuco) The three urban field sites consist in informal urban settlements in Recife, the capital of Pernambuco. According to key informants, these areas were progressively occupied from the early 1900s by low-income populations due to rural migration and land price hikes. Without proper urban planning, they became densely populated while lacking basic services such as paving, water and sanitation and garbage collection. All households in these sites have access to electricity.

The areas have relatively good access to labor markets and educational services.
Some are embedded in neighborhoods with higher levels of income per capita, which facilitates the access to quality social services. 88 All sites have access to a variety of transport alternatives including several bus lines and, in some cases, metro. Also, labor constraints do matter in this context though: Black women in Brazil (when controlling for household wealth) have a higher probability to become a nem-nem compared to white women (Simoes et al 2013). 87 For example, in Brazil, youth from richer families will tend to attend private schools with significantly better infrastructure than public schools. 88 For example, Brasilit is located within the middle-class neighborhood of Várzea, and local youth make use of schools, health units, parks and leisure areas of that neighborhood. markets and formal employment opportunities in the industrial and commercial sectors are within feasible reach. 89 Neighborhoods in these areas are highly commercial, with several bars, markets, bakeries, hair salons and restaurants. According to key informants, the commercial sector, street vending and domestic work are common jobs for the local population.
Although all areas have access to public schools, their infrastructure is generally considered poor by informants. 90 In addition, two of the communities have access to public childcare, but vacancies are scarce. In all communities, educational attainment is low. 91 Crime and violence and drug abuse are present problems in these areas, especially among youth. Participants shared a general sense of deterioration of the security situation in all communities. This was attributed to drug and alcohol consumption as well as the increase in the number of gangs and drug trafficking.
Houses are typically very small and primarily made of masonry and stilt, and are usually built wall-to-wall. It is common for multiple nuclear families to live in the same household 92 or for extended family members to live in the neighboring houses. Streets in the communities are narrow and cars are not able to pass through several of them. Social life within the communities is very lively, with people typically staying in front of their houses talking and children playing in the street. The several bars located in the communities also contribute to social life.