Report No. 32310-BR Brazil Youth at Risk in Brazil (In Two Volumes) Volume I: Policy Briefing May 10, 2007 Brazil Country Management Unit Human Development Management Unit Latin America and the Caribbean Region Document of the World Bank Youth at RiskinBrazil Volume I:PolicyBriefing TABLEOFCONTENTS I Introduction.................................................................................................................. 1 . II Mainfindings............................................................................................................... . 3 Youth at Riskare a SubstantialPopulation inBrazil ...................................................... 3 BrazilianYoung People are a HeterogeneousGroup that RequiresHeterogeneous Policies............................................................................................................................. 7 FindingWays to Reduce Risky Behavior i s a Core Component of Youth Development ....................................................................................................................................... 11 Brazil's Current Investment Strategy does not Focus on At-risk Youth....................... 13 14 III ConclusionsandPolicyConsiderations................................................................. The Cost of Not InvestingEarly is Billions ofReais Per Youth Cohort....................... Basethe Youth LnvestmentStrategy on EmpiricalEvidence........................................ . 17 17 Get the Environment Right to ReduceRisky Behavior................................................. 19 Provide Remedial Programs.......................................................................................... 21 Advantages .................................................................................................................... Assign andCoordinate Institutional Responsibilities according to Comparative 24 References.......................................................................................................................... 29 Annex I: The YouthWelfare Index................................................................................... 32 Tables Table 1: Youth at RiskinBrazil andinComparable Countries. age 15-24unless otherwise noted.................................................................................................................... 4 Table 2: Young People versus Adults and Heterogeneity among the Young ..................... 7 Table 3: How Demographic andSocioeconomic FactorsAffect the Propensity of Young Table 4: Defining"Youth" inTerms of Transitions from Childhood to Adulthood.........11 People from Marginalized Neighborhoods to EngageinRisky Behavior .......................... 8 Table 5: Key FactorsCorrelatedwith Risky Behavior amongBrazilianYouth ...............13 Table 6: Costs of Risky Behavior by Brazilian Young People, 2002 (millions of reais)..15 Comparative Advantages................................................................................................... Table 7: ProposedRole of EachActor inthe Youth Development Strategy basedon their 27 Figures Figure 2: Current and Optimal Social Investment Strategy over the Lifecycle ................10 Figure 1:Brazilian Youth Welfare Index. by State ........................................................... 14 Boxes Box 1:Adolescent BrainDevelopment ............................................................................. 11 15 Box 3: Sources of Information for EvaluatedPrograms ................................................... Box 2: The UnquantifiableCosts of EarlySchool Leaving .............................................. 18 Box 5: The LatinAmerica Jdvenes Models ...................................................................... Box 4: The Perry PreschoolProgram................................................................................ 21 24 iii Box 6: Youth Policy by andfor Young People................................................................. 25 Box 8: Putting Young People at the Center of Policy ....................................................... Box 7: CoordinationbetweenCity and State GovernmentinRio de Janeiro...................26 28 iv PREFACE This report was produced by the World Bank as a contribution to increasing knowledge of youth at risk in Brazil today. This group was selected as the subject of the study since the general topic of youth in Brazil was already being analyzed in several excellent studies that were ongoing when this report was being prepared. This focus on at-risk youth also fits in with the goal of the World Bank to work with the Government of Brazil to define strategies to invest in the poorest and most vulnerable Brazilians to improve their lives. The report uses more rigorous statistical analysis than i s usually used in studies on young people in order to add a new dimension to the youth development debate. The main contributions madeby the report are: It summarizes in one report the incidence of the various kinds of risky behavior that young people try relative to young people elsewhere and relative to adults. 0 It develops a methodology for estimating in reais the cost that youth at-risk impose on themselves and onBrazil. 0 It collects and analyzes data to identify environmental factors -atthelevelof the individual, home, community, country, and general society - that predisposeyoung people to engage inrisky behavior. Itusespaneldatato map out the pathfrom childhood to adulthood. 0 It draws lessons from the Bank's work in other countries and highlights programs that have been shown empirically to be effective at preventing or mitigatingrisky behavior andthat may be applicable to anduseful inBrazil. This report is the first volume of a two volume series. The second volume is a technical report that presents the statistical results indetail. This first volume summarizes the 100- page second volume. It gives a brief review of the report's empirical findings and of the implications that the analytical results have for youth development policy and programming in Brazil, drawing on experiences from other countries to underpin the discussion. The report does not cover several themes. It does not include a comprehensive review of the governmental andnon-governmental organizations working with youth or of the issue of young people's citizenship and participation in society. These issues are fully covered inother reports, andwe redirect the readerto those studies. Thisreport does not lay out a specific program for the Government of Brazil, preferring instead to contribute to the more general discussion of youth issues. V ACKNOWLEDGEMENTS The report team includedWendy Cunningham (Team Leader), Andrea Guedes, Peter Holland, Lorena Cohen, Sandra Rosenhouse, Lerick Kebeck, Ceres Prates, and Lucas Medeiros. Background papers were prepared by Silvia Koller, J'aims Ribeiro, Elder Cerqueira-Santos, NormandaAraujo de Morais, Maycoln LeoniTeodoro, Juliano AssunGiio, Leandro Carvalho, Ceres Prates, Maria Beatriz B.Nogueira, Erik Alda, Debora Dalbosco Dell'Aglio, Vicente CasseppBorges, Joana Severo Leon, Bernard0 Kipnis,JosCFerreira BelisArio, ClaudiaWerneck, PatriciaMoreira, andClaudia Maia . Peer review was provided by Dr.Robert Blum(Johns Hopkins University), Gary Barker (Instituto Promundo, Rio de Janeiro), Andrew Morrison, Maria-Valeria Pena (World Bank), AndrC Noblat, Nara Kohlsdorf, Reinald Chaves Gomes, Josbertini Virginio Clementino (Universidade da Juventude/Conselheiro Nacional de Juventude), andMaize Soares. Additional inputswere providedby Alberto Rodriguez, Kathy Lindert, Bernice von Bronkhurst, Zeze Weiss, Juliana Barbosa, Yasuhiko Matsuda, Lucas Siga, Maria CaridadAraujo, Pablo Lavada, Leonard0 Lucchetti, Fernando Landa, Mariano Bosch, and Ariel Fiszbein. Special thanks to all the partners, friends, andteachers inBrazil who generously shared with the team their deep knowledge of youth inBrazil, inparticular, Miriam Abromavy (UNESCO), Paulo Vanunuchi, Pedro Paulo Martoni Branco, Helena Abramo (Instituto Cidadania), Ana MariaDrummond (WCF), Ana Lucia Dezoult (IDB),Rosemary Madden (UNFPA),DenisePaiva ,CristinaAlbuquerque (Special Secretariat for HumanRights), Rubem FonsecaFilho (Ministry o f Education), Nena Lentini (USAID), Luiz Alberto Gonplves (Special Secretariate for the Promotion of Racial Equality), Mario Volpi (UNICEF), RuthCardoso andThereza Lobo (Comunitas), Luciana Tannus andRicardo Cifuentes (Ministry of Labor and Employment), Andre Noblate, Alessandro de Leon, and the wonderful young people who represent the VozesJovens of Brazil, who generously shared their stories, work sessions, viewpoints, and proposal with the report team. Most particularly, the team i s grateful to Rodrigo Abel (General Secretariat, Presidency), whose patience, insights, and friendship were invaluable to the report team. v i I. INTRODUCTION Most Brazilian young people are well on their way to becoming productive and contributing membersof society.' Three-quarters of young Brazilians claim that they are happy with themselves and with their lives.* And they have the numbers to make substantial contributions to the present and future of their society - 19 percent of the Brazilian population is aged 15 to 24 and Brazilian young people comprise one-third of the youth population of Latin Ameri~a.~ However, some young people are not navigating their experimental youth years successfully. These people, commonly identified as at- riskyouth, are the subjectof this report. Youth at risk are young people who have factors in their lives that may lead them to behave or to experience events4 that are harmful to themselves and their communities. These include school repetition and early school leaving, inactivity (neither attending school nor working), substance use, violent behavior, early sexual initiation, and unsafe sexual practices. For the unfortunate few, these risky kinds of behavior result innegative outcomes, such as HIV/AIDS, early pregnancy, lifetime poverty, or death, affecting not just the risk-taker but also society in general. Thus, society as a whole has a responsibility to provide children with the right information to allow them to experiment carefully during the years of their youth andemerge as healthy, productive adults. This report is the first volume of a two-volume study that contributes to the emerging body of work on youth in Brazil in five specific ways. First, it attempts to make a case for making public investments in young people by summarizing the incidence of the various kinds of risky behavior that young people indulge in and by attempting to quantify, in reais, the cost of such behavior to the individual and to society. Second, the report identifies the factors that predispose young people to engage in risky behavior in an effort to highlight the roles that various actors in society play and the wide range of 1 For the purposes of this report, "youth" or "young people" are defined as those aged 15 to 24 unless otherwise noted. During the preparation of the report, there was disagreement among Brazilian policymakers and practioners about what age range comprises "youth." Ages ranging from 15 to 26, 10 to 18, and 10 to 22 were cited, depending on the sectoral focus of the respondent, the state and local legislation and institutional parameters, and the historical and cultural differences among the diverse people of Brazil. Given the lack of consensus, the UN definition (15 to 24) was used, but an empirical exercise was carried out to define the appropriate age range for Brazil, which is discussed in the next section of this report. Koller et a1(2005), Instituto Cidadania (2004). The Brazilian youth cohort will shrink over time. Population pyramids show that, while today's 15 to 19- year-old cohort i s the largest cohort in the population, the size of today's 10 to -14-year-old cohort - who are tomorrow's youth - is smaller and that of the 0 to -9-year-old cohort is even smaller. This suggests that any public policies to assist young people should be flexible enough to contract as the youth population shrinks. 4 This report differentiates between the choices made by young people and the events that they experience. These can be interpreted as extreme positions in a continuum, where some observed behavior, such as substance use, are closer to the "choice" end of the spectrum, whereas others, such as unemployment, are closer to the "events" end of the spectrum since we can assumethat unemployment i s rarely a choice. 1 policy interventions that can be adopted to address the problem. Third, the report presents comparable statistics from other countries in the world to serve as a benchmark against which to evaluate the progress o f Brazil's young people. Fourth, it asks what age range is the most appropriate for defining "youth" in the Brazilian context and the how people inthis age range transition from childhood to adulthood. Fifth, the report presents examples of programs in other countries that have been empirically proven to be effective and may be useful for the Brazilian government to consider as part of their youth investment strategy. This report (Volume I) summarizes the main research findings of the full report andpresents policy options. Fulldetails can be found inVolume II. 2 11. MAINFINDINGS Youth at Risk are a Substantial Population inBrazil Most Brazilianyoung people successfully navigate the youth years, but a smaller group is not faring as well. While there are many success stories, the status of Brazil's young people is generally worse than that of young people in comparable countries. Table 1 below presentsthe incidence of negative youth behavior inBrazil and inother countries. Brazilians have lower academic achievement than young people in comparable countries. Brazil ranked40thout of the 41 countries whose students took the PISA exam, with only Peru scoring lower (PISA, 2003). Illiteracy rates of Brazilianboys, inparticular, are double those inthe other large Latin American countries. While net enrollment rates in Brazilian secondary schools are below those only in Chile, Argentina, andthe OECD countries, gross enrollment rates in tertiary education are below all of the countries inthe table, including Mexico and Uruguay. As in most countries in Latin America, Brazilian girls perform better than boys. Low human capital accumulation among today's young people is producing a generation of Brazilians that will not be able to compete in either the region or the world. Young Brazilians combine school and work in much higher proportions than young people in most of Latin America. The lower academic achievement of Brazilian young people may be due to a higher incidence among them of combining work and school compared with other c~untries.~One in four Brazilians agedbetween 15 and 19 are both attending school and working, which i s three to five times the proportion in other countries. However, an equal share of Brazilian young people are inactive (they neither work nor attend school) or begin working intheir childhood as inthe rest of Latin America. Duryea (1998) showed that Brazilian young p"eple who work while attending school have higher rates o f repetition and eventual dropout than do young people who only attend school, after controlling for other household factors. 3 gr- II 1 I1 vlv) 00oc vl t- (uv m a %\. * v 00 2: 3 - 4 09 m(u a c 9 + c 3 3 m r 6 9 . + -- * 00 m a I r-: t-t- 30 I m a 0 m m I 2 9 t - 0 0 N - I +(u c, s 9 m 5 a h2 3 d9 0 h c cd d w k Q C d C Q O --I+ Brazilianyoung people are 3.7 times more likely than Brazilianadults to be unemployed,which is a much higher rate than in the rest of Latin America or in the OECD. While the youth unemployment rate in Brazil i s around 13 percent, which is much lower than in some other countries in the region, the degree of overall economic activity in Brazil suggests that it ought to be possible to employ more young people. Except for Uruguay, where the youth-to-adult unemployment ratio is 3.3 to 1, all countries in Latin America have rates between two and three, suggesting that Brazilianyoung people are faring much worse than they should given the state of the labor market.6 Notably, this rate is decreasing over time (Bonelli, Reis, andVeiga, 2004), but it i s still very high relative to rates inthe region as a whole. Nearly 60 percent of Brazilians aged 15 to 19 are unpaid workers or sem carteira (withouta work contract). These numbers fall to 33 percent for the 20- to 24-year-old groups. Young people aged between 10 and 14 comprise 55 percent of all unpaid worker^.^ Comparable statistics for young people in other countries were not available. Brazil has the third highesthomicide rate in Latin America after Colombia andElSalvador. More than 100out of every 100,000young men agedbetween 15 and 29 are murdered in Brazil each year. This statistic i s twice as high in El Salvador and Colombia, but it is significantly lower in the rest of the region. Although Brazil ranks third interms of violence, this i s worrisome since Brazilian violence is more random than in Colombia, where the high homicide rates are attributable to a civil conflict, and inEl Salvador, where the high rates are mainly due to a serious youth gangproblem. The female homicide rate inBrazilis low at only 6.7 per 100,000 young women, but this i s still more than double the rate in most LatinAmerican countries. On the positive side, Brazilianyoung peoplehave a muchlower incidence of risky sexual behavior, reflecting the Brazilian government's and society's aggressive initiatives to promote safe sexual practices. Although young Brazilians initiate their sexual activity at the same (young) age as their LAC counterparts, they have higher contraception use and The youth-to-adult unemployment ratio is used rather than the level of youth unemployment in order to control for the tightness of the labor market. The adult unemployment rate serves as a benchmark - the expected level of unemployment given economic conditions - so the ratio shows how young people fare relative to the benchmark. For example, although the youth unemployment rate in Colombia is three times higher than the rate in Brazil, overall unemployment rates in Colombia exceed those in Brazil, so everyone's unemployment rate i s high in Colombia, not just youth rates. However, the ratio is higher in Brazil (3.7) than Colombia (2.3), which suggests that Brazilian young people are faringparticularly poorly given the economic situation. 'The international evidence suggests that the negative consequences for children of early child labor include greater unemployment later inlife, lower wages (when controlling for lower education levels), less job promotion, and a long list of other negative outcomes. Bradley and Anh (2006) provide a summary o f the literature. 5 lower adolescent fertility rates. Only 36 out of every 1,000 Brazilian girls aged between 15 and 19 give birth, fewer than inany other country inLatin America (Table 1). Compared with adults, young people indulge less in some kinds of risky behavior and more in others. Young people are doing better than the older generation in terms of academic achievement, particularly in terms of literacy rates - 95.8 percent of Brazilians aged 15 to 24 can readcompared with 84.5 percent of Brazilians aged 25 and older. This i s partly explained by the fact that the younger age group has completed more years of schooling - an average of 8.4 years (8.1 for boys and 8.8 for girls) compared with 7.3 years for adults. However, it may also be explained by very high repetition rates, whereas the older generation would have simply left school. Young people are also less likely to smoke tobacco than are adults. Despite the excessively high youth to adult unemployment ratios, once young people enter the labor market, they behave similarly to adults. About 35 percent of both young people and adults are employed in the formal sector (corn carteira), although self- employment i s much lower among the younger group. Also, wages are very similar; young Brazilians earn 97.75 reds for every 100 reais earned by an adult. The fact that this differential inwages is so small can be attributed to adults' greater work experience beingcounter-balancedby young people's higher educational attainment. Notably, youth unemployment tends to be shorter than adult unemployment (Maloney and Bosch, 2005), but the youth rates are higher simply because the flow of young people into unemployment i s higher than the flow of adults, leading to a clustering of young people inthe unemployment sector. Today's young people also behave in more risky ways than adults in four areas - sexual activity, violence, illegal drug use, and unemployment. Only one-third of sexually active 15- to 19-year-oldwomen use contraception compared with two-thirds of sexually active women aged 20 to 24. As discussed above, 102 out of every 100,000 young men are murdered, and40 out of every 100,000 young people are murderedcompared with 3 out of every 100,000 adults. While alcohol consumption i s similar between young Brazilians and adults, young people are twice as likely as adults (10 percent compared with 5 percent) to consume illegal drugs. And, as already discussed, youth unemployment rates far exceedthose of adults. Importantly, though, incomparing young people with adults, it is the 13 to 18 age group whose behavior is least like that of adults as 19 to 24 year olds behave very similarly to adults. 6 Brazilian Young People are a Heterogeneous Group that Requires Heterogeneous Policies Not all Brazilian young people are engaging in risky behavior. At-risk youth are concentrated in certain demographic, socioeconomic, and geographic groups in Brazil. Tables 2 and3 summarize the results.* Table 2: Young People versus Adults and Heterogeneityamong the Young iiFpq% 25+ 15-24 All Male Female White Afro-Br Poor Rich Illiteracy, % a 15.5 4.2 5.3 3.1 0.3 Mean 7.3 8.4 8.1 8.8 11.4 tl educational attainment, l l I 1 years a Formal sector 33 35 1 35 1 35 50 -39.3 I 68.4 I --- 100,000) - Suicide (per 0.7 3.4 --- --- 100,000) - Note: --- Data not available. Sources: PNAD 2001, author's calculations, DHS Statistics, Carlini et al, 2002, Waiselfisz, 2004 a * Table 2 presents frequency estimates. Table 3 presents the results of odds-ratio estimates, usingdata that were collected for this study. The sample is comprised of 3,000 school-going young people aged 14 to 24 inthe poor neighborhoods of Port0Alegre, Recife, and SHo Paulo. 7 Poverty Interms of demographic and socioeconomic differences, the most important variable is poverty. Young people from the poorest families consistently have worse indicators than those from the wealthiest families. They have illiteracy rates that are three times the national average and formal sector employment rates that are one-eighth of the national average, while those from the wealthiest families have illiteracy rates that are 93 percent below the national average and formal sector employment rates that are one-third higher than the average. Almost 90 percent of unemployed young people are from households with per capita incomes equal to less than two minimumwages (Bonelli, Reis, andVeiga, 2004). Poverty is a key explanatory factor for the more general risky kinds of behavior presented in Table 3 - school repetition, early entry into the labor force, risky sexual behavior, substanceuse, and violence. Table 3: How Demographic and Socioeconomic Factors Affect the Propensity of Young People from Marginalized Neighborhoods to Engage inRisky Behavior I School I Early Entry to I RiskySexual I Substance I Violence/ 1 Repetition Labor Force Behavior Use Suicide Poverty XX xx xx xx xx Race xx xx Gender XX xx xx xx xx Disability xx XX (suicide) Region XX xx xx xx xx Race Non-white young people generally have worse indicators than white young people. In terms of both educational attainment and performance, white young Brazilians are more successfulthan non-whites (Table 2). When the sample is limited to the poor (Table 3), race does not emerge as an important explanatory factor for grade repetition. Similar results were found in other studies (Pena et al, 2004). Notably, employment rates are very similar among races, although poor non-white young people enter the labor force at a younger age than whites do. The one area where the statistics are higher for non-white than for white young people is violence, with 68 out of every 100,000 young Afro- Brazilians beingmurdered comparedwith 39 out of every 100,000young whites. Gender Neither gender uniformly excels over the other. Women have a lower incidence of risky behavior inthe areas of educational attainment and performance, substance use, death by 8 homicide, early entry to the labor force, and ~ u i c i d e .In other instances - risky sexual ~ behavior, unemployment, andunpaid labor - women fare worse (Tables 2 and 3). Disability Young people with disabilities from poor neighborhoods have many similarities with other young people in poor neighborhoods (Koller et al, 2005). School repetition rates, early entry to the labor force, and substance use do not differ by disability status. However, risky sexual behavior - as proxied by the failure of sexually active young people to use contraception - and suicide attempts are higher among the young disabled population. Region Young people's behavior differs among the different regions of the country. Table 3 showed that the young people from each of the three regions sampled had markedly different patterns of behavior. Expanding the sample to all of Brazil and using a wider set of variables, we were able to construct the Brazilian Youth Welfare Index (for our methodology andthe variables that we used, see Annex l)." Figure 1 shows how young people in the various states fare relative to the national average. The young people in Pernambuco and Alagoas are the worst off, faring 6 and 5.3 percent worse than all young Brazilians across the full range of indicators. Both states have very low ratings in all youth behavior, performance, and interaction with local institutions, while most other states excel in some areas and are weaker in others. Young people in Santa Catarina and the Federal District fare 6.08 and 5.06 percent better than the national average. This higher than average performance can be attributed particularly to highschool performance, school advancement, pre-school enrollment rates, and formal sector employment rates in Santa Catarina and to employment opportunities, secondary education attendance, school performance, and low substance abuse in the Federal District. These differences by state suggest the need for greater investments to be made in young people in certain states and the need for state-specific youth investment strategies as opposed to a single set of priorities applied nationally. While the suicide rate is higher among men, the number of suicide attempts is higher among women. This trendis observed aroundthe world. lo The methodologyis basedon a ChildWell-BeingIndex developedby Duke University in2004 that has beenadaptedto the Braziliancontext. See Annex 1for details. 9 Figure 1:BrazilianYouthWelfare Index, by State Baseline: nationalaverage 4 2 I O -2 -4 ~ ~~ Source: Dalboscoet al(2005). The age at which young people engage in various kinds o f behavior differs greatly by activity (Table 4)." Analysis of the Brazilian data suggests that the period defined as "youth" begins at 14 years old and stretches to 29 years old. The transition from school to work begins at the age of 13 for boys and 14 for girls, while changing from being a dependent to the head of a household begins a few years later, as does the transition from childhood to motherhood (there were no available data on the transition to fatherhood). The age at which certain transitions take place varies greatly by gender, poverty level, and race (Table 4). Young people from less wealthy families begin the transition out of school, out of childhood, and out of household dependency at a younger age than do young people from more affluent families. Also, they become employed at a much later age, although their transition to motherhood and household headship ends earlier for young people from poorer families than for those from richer families. Similar 11 We used a methodology developed by the OECD to identify the transition periods of Brazilian young people. The school-to-work transition age is defined as beginning at the youngest age at which 75 percent of the individuals are not in school and ending at the youngest age at which 50 percent of the population is working and not attending school. Given the invisibility of the household work contributed by married women (particularly), we separately estimated the transition ages for women who are not married. The child-to-parent transition was only identified for women, due to a lack of data on men's fatherhood status. It begins at the youngest age when 1percent of the population has given birth and ends when 40 percentof the population has given birth (half the total fertility rate of adult Brazilian women). Finally, the dependent-to-householdhead transition begins at the youngest age at which at least 1 percent of the age group is identified as a householdhead or spouse and ends at the earliest age at which half of the age group identifies themselves as having this status. 10 differences exist across races, and there is even greater variation between males and females. This suggests that it is necessary to adjust the age range for studying young people according to the transition patterns of the socioeconomic, demographic, and economic status of the group under consideration. It also suggests the need for age- appropriate youth interventions since a wide range of young people with very different experiences and levels of cognitive development are considered to be in their "youth" (see Box 1for adiscussion of cognitive development). Table4: Defining"Youth" inTerms of TransitionsfromChildhoodto Adulthood School-to-work Dependent-to-head Child-to-mother Begin End Begin End Begin End All 14.5 21.5 15.5 24.5 --- --- Male 13.5 20.5 17.5 25.5 --- --- Female* 14.5 28.5 14.5 23.5 15.5 20.5 Decile 1 11.5 29.5 14.5 20.5 14.5 18.5 Decile 10 17.5 23.5 19.5 28.5 17.5 27.5 and 24.5 years old and the child-to-mothertransition is betweenages 15.5 and 27.5. Source: Volume I1 Box 1: Adolescent BrainDevelopment Until recently, it was thought that the human brain was nearly fully formed by the time a child was five or six years old. However, very recent research suggests that significant changes in the brain's structure happen during adolescence. The frontal lobe of the brain, which controls planning, working memory, and organization and which modulates mood, undergoes most of its changes during adolescence. The changes start before puberty (age 11in girls and 12 in boys) when neural connections are significantly enlarged. This growth spurt i s followed by an extended period of pruningthat may last until the age of 25. This pruningrepresents a maturation process where neural connections that have been exercised are cemented and those that have not are discarded. The fact that young people have not yet formed these mature neural connections in the frontal lobe implies that they are at a disadvantage when it comes to controlling their impulses and makingsoundjudgments. Source: Spinks (2003) FindingWays to ReduceRiskyBehavioris a CoreComponentof Youth Development Why do young people engage in negative behavior? Much research has focused on the institutional, poverty, and economic forces that lie behind how young people behave, primarily in terms of the labor market and school attendance. The analysis in this report starts from a different hypothesis and tests the importance of the factors that serve as the models for, build the preferences of, and create the incentives for young people to make decisions. 11 Young people who engage in negative behavior are those who have reached their adolescent years after experiencing a range of negative events intheir lives.12 The young people who engage in risky kinds of behavior tend to be characterized by or have experienced certain factors that may explain their choices (Table 5). While we cannot claim any causality between these factors and the negative behaviors, the strong statistical correlation i s sugge~tive.'~These factors are: Low self-esteem A lack of spirituality Physical, sexual, or psychological abuse by other members o f the household Abuse by someone intheir communities Feelinglike an "outsider" at school Not trusting local institutions A consciousness of the limitations imposedbyhisher gender Poverty. These results have several policy implications. First, similar factors affect many kinds o f behavior so policies that target one factor may reduce several kinds o f behavior. This may obviate the need for separate policies for each kind of behavior. Second, programs to eliminate risky behavior should aim to intervene early in a person's life. Waiting until the young person i s fully formed will make it more difficult to alter his or her behavior. Finally, an array of actors are responsible for youth development - families, communities, local institutions, macro socioeconomic conditions, and social norms, as well as the young person himor herself - so support programs are needed to enhance the way inwhich each actor contributes positively to youth development. These findings are similar to those for other countries. In the United States (Blum et al, 2002) and in the Caribbean (World Bank, 2003), the individual, the home, and community factors also emerged as the most important influences on a young person's decision-making. In Brazil, there i s a special emphasis on spirituality, which i s less prominent inthe other studies. l2We used two methodologies. First, we calculated odds ratios for each behavior, and we estimated hypothesized "causal factors" and t-statistics. This methodology demonstrates which causal factors are correlated with certain kinds of behavior, but it does not give information about the extent to which the causal factors explain the variation in the observed behavior across the sample. To achieve the latter goal, we estimated a step-wise regression by regressing (separately) each behavior in question on a set of potential causal factors. Using various sets of factors, the methodology enabled us to find the set that best explains the variance in the dependent variable. The data used for this exercise were collected in poor neighborhoods in Port0Alegre, Recife, and S5o Paulo for use in this study. Thehe data are rarely collected inany country inthe world (see Koller et al, 2005 for details). l3To establish causality, panel data (which trace individuals and their behavior over time) are necessary. 12 Table 5: Key Factors Correlated with Risky Behavior among BrazilianYouth - Self Home/ Local Macro- community institution environment .- Yz h I 3 Y h h z* .I Y L1 s L1 a2 .I ma Grade repetition xx xx X X xx Entry to labor force X Early sexual initiation xx X xx xx X X xx Risky sexual practices xx X 0 Alcohol use X xx X xx X xx X X Tobaccouse X X X xx xx X Ulegaldruguse xx xx xx Violence xx xx xx X xx X X Suicide attempt xx xx xx -- X Notes: XX = key factor (significant at the 1% leve in a stem se regressi n estim . - lrand the odds-ratio is significant at the 1%level); X = significant factor (the odds-ratio is significant at the 1% level, but the variable was not significant inthe step-wiseregression). Source: Koller et al, 2005 Brazil's Current Investment Strategy does not Focuson At-risk Youth There i s a gap between the current investment strategy for human development and the optimal strategy in terms of at-risk youth. Figure 2 shows current and optimal social expenditures in Brazil by periods within a 1ife~ycle.l~The downward sloping line i s the optimal social expenditure scheme, demonstrating that investing generously early in life will mean that there i s less of a need to invest later in life. The other line in Figure 2 i s the current investment strategy in Brazil. The largest investments consist of pensions, in other words, of investments in the over 61 age group. Expenditures on young people constitute about 6 percent of total social spending, more than investments in several other l4The data for federal spending are from Palocci et a1(2005). The municipal and state spending data were collected by the World Bank. Social spending was categorizedby age (0-5), (6-17), (18-24), (25-60), and (61+), roughly corresponding to early childhood, school-age, youth, adults, and old age. Approximately 31 percent of all social spending was identified as family benefits or general spending that was not disaggregated by age group. The largest sub-set within the latter category was health expenditures. Since there is no age breakdownfor healthspending, these dataare omitted from the figure. 13 periods of the lifecycle. However, if we eliminate university expenditures - which are limited to a small, elite portion of the population - expenditures on young people fall to nearly zero percent of total social expenditures (the evenly dotted line). Notably, the two periods of the lifecycle in which investment is the lowest in Brazil are early childhood andthe youth years -two crucial periods for preventing risky behavior amongthe young. And the expenditures on the childhood years are largely on education, which has an important pedagogicalimpact but tends to have less effect on humandevelopment. Figure2: Current and OptimalSocial Investment Strategy over the Lifecycle Investment, 71 as a share of ........ -.. all social spending 15 ..................................... ......... ............. 6 ......... ................ 3 ----------- ............ .... ...... 0-5 6-17 18-24 25-60 61+ Age Current investment strategy ..................Optimal investment strategy -____. investment strategy,excludinguniversity expenditures Current The Cost of Not InvestingEarly is Billions of Reais Per Youth Cohort This misaligned investment strategy has serious implications for the economic, social, cultural, andpolitical growth of Brazil. Negative outcomes from risky behavior by young people have significant costs not only for the individual but also for society as a whole. Conservative estimates of cash outlays to deal with those negative kinds of outh behavior are as high as $1billion reais for each age cohort of young people today." For example, very conservative estimates of the amount of money spent by the government on arresting, prosecuting, and imprisoning young criminals is equivalent to $478 million reais per age cohort (see the last row and first column inTable 6). The cost to society from lost productivity i s estimated to be even higher (see the second column inTable 6). For example, Brazil's economy i s losing $755 million reais over the next 40 years because of early school-leaving among the current youth cohort. Such high "lost benefits," as measured by what could have been if the young person had not engaged inthe negative behavior or ended up in an unfortunate situation, demonstrate the l5The estimates are qualified as being "conservative" since they cannot measure many of the costs, whether becauseof missing data or unquantifiable effects. See AssunGZo and Carvalho (2005) for a discussion of how the costs were measured for this exercise. 14 true losses to Brazil's society of not investing in at-risk young people. These estimates are conservative, since they do not include those losses that cannot be captured by numbers, such as the lost opportunities in terms of cultural development, greater efficiency in the home, growth in democratic traditions, and voluntary contributions to society (see Box 2). Table 6: Costs of Risky Behavior by Brazilian Young People, 2002 (millions of reais) Public Costs Private Costs Expenses Lost Benefit Lost Benefits (Direct Costs) (Economic Costs) R$O R$755 I R$164,518-R$297,056 outh unemployment R$120-R$224 R$7-R$13 R$641- R$1229 R$297 R$140- $R554 I R$812-R$3190 --- --- R$10,648 --- --- R$762 --- --- R$30 R$478 R$33 R$5060 The biggest losers are the young people themselves because they blight their futures by engaging in negative behavior today. One cohort of early school leavers will earn R$300 billion less over their lifetimes than they would have done if they had only finished the next level of education. Together these foregone benefits due to insufficient investment in youth development are equal to billions of re& of lost benefits for a single cohort of today's young people (see the third column inTable 6). Box 2: The Unquantifiable Costsof Early School Leaving While the numbers presented inTable 6 quantify the costs of negative youth behavior, they are an underestimate of the true costs. Taking, for example, the cost of early school leaving, various studies have shown (when controlling for household income) that lower educational attainment i s correlated with: P Poorer healthof oneself andone's family P Greater domestic violence P Poorer psychologicalindicators P Greaterjuvenile delinquency of one's children P Greater incidenceof risky sexual behavior P More substanceabuse 9 Deeper feelings of social exclusion 9 Lessvolunteerism P Less voting andcivic participation 9 Lessculturalexpression. Ifwe couldquantify these factors andaddthemto the values inTable 6, the value of staying inschool would clearly increase greatly. 15 Preventing young people from indulgingin these risky kinds of behavior would improve their health status, increase their earnings potential, and give them a greater chance of enjoying life. It would also eliminate many social costs, thus freeing up public resources to be spent on other initiatives and increasing economic growth as young people would have more human capital and thus greater productive capacity. This suggests that an analytically based investment strategy for young people should be a key part of Brazil's overall strategy for economic andsocial development. 16 111. CONCLUSIONSAND POLICY CONSIDERATIONS Our analysis suggests that the key elements of a youth investment strategy in Brazil should be a well thought out mix of prevention and remedial programs targeted to young people and their support networks. A wide array of potential interventions i s available to prevent young people from indulging innegative behavior, and this chapter discusses our four key conclusions about how best to designand implement an effective and successful youth investment strategy inBrazil andthe policy considerations arisingfrom them: 0 Base the youth investment strategy on empirical evidence 0 Get the environment right andreducerisk 0 Provide remedial"second chance" programs 0 Assign and coordinate institutional responsibilities according to comparative advantages. Base the Youth Investment Strategy on EmpiricalEvidence The report demonstrated that the negative outcomes that result when young people indulge in risky behavior have significant costs both to the individual and to society. Preventing these kinds of behavior would improve the health status of young people and increase their earnings potential and their potential for enjoying life. At the level of society as a whole, it would also eliminate many social costs, thus freeing up public resources to be spent on other initiatives and increasing growth as young people would have more human capital and thus greater productive capacity. Therefore, investing public resources in youth development i s a necessary part of any country's strategy for investing ineconomic and social development. However, the current investment strategy for humandevelopment inBrazil differs from the optimal investment strategy. This suggests two important policy considerations-knowing whatprograms are efective anddesigning a strategy around them. Any youth strategy should be underpinnedby an investment strategy for youth development, similar to the Finance Ministry's investment strategies for all areas of public policy. However, designing such a strategy will not be easy as there is very little informationinBrazil about which youth programs are effective or about the cost-benefit ratios of those programs.16 While the government could learn l6Most program impact evaluations are from the U S and Western Europe. These evaluations confirm some hypotheses such as supervised leisure time is an important protective factor and internship programs increase the employability of disadvantaged youth. They also yielded evidence that programs for which we had no expectations were effective such as the Perry Pre-school program, Big Brothersmig Sisters, and Boys and Girls Clubs. Finally, they showed that some ideas are not effective, such as gun buy-back pr&ams (Callahan et al, 1994), DARE (Lynam, 1999), and boot camps for violent young people (Satcher, 2001). 17 lessons from the evaluations of other countries' programs and apply them to Brazil (Box 3)' this i s not a perfect solution since there are some different circumstances inBrazil that may make these programs work differently than in their countries of origin.17 Therefore, it will be vital to buildup a stock of knowledge about which interventions are effective in the Brazilian context by implementing pilot programs and carefully evaluating them. With each set of evaluations, more information will be available to inform the development of effective youth investment strategies at all levels of government. Box 3: Sources of Informationfor Evaluated Programs Throughout Europe and the United States, many programs have been successfully implemented and evaluated. While these programs may not be applicable to Brazil, they give some indications of which programs may be most useful as well as ideas for new programs. These are the sources for the evaluations of these programs. General Knowles and Behrman(2005) http://siteresources.worldbank.orp;/HEALTHNUTRITIONANDPOPULATION/Resources/28 1627- 1095698140167/KnowlesEconInvestYouth.~df Crime & Violence Satcher (2001) http://www .surgeonaeneral.gov/librarv/vouthviolence/default.htm Health Focus on YoungAdults (2001) http://www.pathfind.ordpf/pu bs/focus/pubs/eop report.pdf Maddaleno and Schuitt-Aine (2002) http://www.paho.orp/Enalish/HPP/HPF/ADOL/SRH.pdf Manlove (2004) http://www.teenpre~ancv.or~works/pdf/NotimetoWaste.pdf Labor Betcherman et al(2004) Http://siteresources.worldbank.orrt/SOCIALPROTECTION/Resources/SP-Discussion-papers~abor- Market-DP/0402.pdf Castro & Verdisco (1999) http://www 1.worldbank.ordeducation/secondarv/documents/Castro.htm Aedo. andNuiiez (2001) Http://www.cinterfor.ora.uv/public/enalish/renion/ampro/cinterfor/temas/youthdoc/aedo/index.htm SubstanceAbuse Office of Justice Programs (2000) http://www.oip.usdoi.gov/docs/psrsa.pdf l7For example, for years, evidence from Europe showed that youth training programs do not work, but recent evidence from Latin America has shown that they do under certain circumstances (Betcherman et al, 2004). The reasons for these contradictory conclusions have not been identified, but one possibility is that unemployed young people in Europe have high levels of education so their unemployment is due to tight labor markets. Conversely, unemployed young people in Latin America lack fundamental skills (due to early school-leaving) and have little knowledge about marketsoutside their immediate sphere (due to social exclusion) so intensive programs that provide these skills to Latin American young people may be more effective than equivalent programs inEurope. 18 Get the EnvironmentRightto ReduceRiskyBehavior Prevention programs focus on ensuring that young people grow up in healthy and positive environments. As discussed in this report, experiences early in life can increase or decrease the likelihood that young people will engage in risky behavior. To maximize the healthy youth population of tomorrow, it i s important to start building a positive environment today. Cost-benefit analyses of programs inother countries have shown that preventing negative behavior i s more cost-effective than taking remedial actions. Three priority areas for takingpreventive action inBrazil emerged from the report. Helping Families and Communitiesto Build Healthy YoungPeople There i s an immediate need for programs to support families and communities and to give them the tools to support young people. Early child stimulation, conflict management, parenting teenagers, and managing family relationships are challenges regularly faced by parents and caregivers, but there are some tools to help them to improve the home environment. Such information can be communicated to parents in various wa s: (i) providing parenting and conflict management classes for parents or caregivers It7 in the family community centers that have been built in over 900 municipalities in Brazil, (ii) requiring parental participation at such classes as a condition for receiving the Bolsa Familia tran~fer,'~ and (iii) the media, particularly during using the evening telenovelas, to communicate parenting skills. At the community level, several initiatives could be expanded: (i) investmentfunds social could include young people in the planning and implementation o f their projects or could provide services for tomorrow's young people such as early child development programs, (ii) supervisedleisureprogramscouldbeprovidedincommunityspacessuchas regular public squares, parks, churches, and community centers, and resources could be spent on program materials and adult supervision2' rather than on building expensive leisure '*TheU S Federal Government's Office of Juvenile Justice and Delinquency Programs (OJJDP) launched the Strengthening America's Families initiative in the mid-1990s to address youth violence and measure the impact of programs. The program evaluation results show that to be maximally effective interventions need to: start as early in life as possible; train parents and caretakers in effective positive discipline; help to improve parent-child communication; teach parents non-violent coping skills; provide high-riskfamilies with intensive and repeated family and youth interventions by professionals; encourage weekly family meetings in order to change internal family dynamics and communication patterns; and tailor interventions to the types of risks that the family faces and to the development stage of the young people and parents being targeted (Bilchik, 1998). '' l9 Attendance at regular presentations is a condition for receiving the cash transfer in the Colombia Familias en Accion program. The topic of the presentation is selected by the commuity, and the resentation itself is provided by the program. Research has shown that constructing community centers or sports fields does not affect young people's behavior. Instead, supervised youth activities where there is a structure, stimulation, and supervision by a caring adults are effective (Clark, 1988). The Project Learn in the US is based on the observation that most risky behavior tends to occur between the hours of 3 pm and 5 pm (after school). The program aims to help young people spend their afternoon leisure time on activities that have an academic focus but are also fun in nature. The 25 to 35 hours of activities a week provided by Project Learn include homework 19 spaces, and (iii) investments could be made to increase the safety of urbanneighborhoods like those being implemented in poor communities in Salvador and Ilheus, including upgrading water and sanitation infrastructure, paving streets, and enabling the implementation of social programs, such as creches, inthe community. Early Childhood Development Programs Expanding early childhood development programs is an effective way to prevent young people from indulging in risky behavior as they get older (see Box 4 for an example o f such a program). Fewer than 20 percent o f Brazilian children benefit from early childhood programs (many of which are simply private babysitting arrangements), which suggests that transfers given by the national government to the municipal governments to fundearly childhood services are not reaching young children. The FUNDEB(Fundo de Desenvolvimento e Manuten$o do Ensino Bbsico e Valoriza@o do Magisterio) i s likely to facilitate the expansion o f such services to young children since some resources will be earmarked for pre-school services. The taxes earmarked for municipal educational expenditures could be directed only to pre-school and basic education, while taxes for state educational expenditures could be used to fund basic and secondary schools. Early childhood development can also be fostered at home. Parents often do not know what techniques to use to stimulate their children's development, so it is important to include the teaching of these skills to parents in these early childhood development programs. Get Children in School and Keep Them There Schools are a vital protective factor. Simply by keeping young people in school, the incidence of negative behavior and unfavorable outcomes will decrease. The government's current effort through the FUNDEB to give municipalities and states incentives to increase secondary school enrollment i s an important advance for youth development. In addition to the pedagogical benefits of providing education, a safe school environment with capable teachers who instill discipline in their students also reduces many other social ills. The recent discussions to increase public funding to secondary schools i s timely given the current bulge in the youth population and the country's highrepetitionrates. help, discussions with adults, time for leisure reading, and games that require young people to use their cognitive skills (such as Scrabble and Monopoly). According to a 30-month impact evaluation (Schinke, 2000), program participants increased their overall grade point averages by 11 percent on average in mathematics, history, science, spelling, and reading and decreased the number of days on which they were absent from school by 66 percent. 20 Box 4: The Perry Preschool Program In 1962, 123 high-risk African-American children aged 3 and 4 years old were selected to participate in a high-quality, active learning preschool program. The children were randomly divided into a program group, who attended the preschool program, and a non-program group, who did not attend any preschool program. The program focused on the children and their families. It featured daily classes of 20 to 25 children and weekly home visits by staff to discuss the children's development with their parents. Topics discussed at these meetings included health, education, and family support and the children were actively involved in planning, executing, and reviewing their own learningactivities. The children were then tracked through to the age of 40, which made it possible to measurethe long-termeffects of this intensive family and child-based program: 0 More of those inthe pre-school education group graduatedfrom high school (65 versus 45 percent) 0 More of those inthe pre-school education group were employed at the age of 27 (69 versus 56 percent), a pattern which remained consistent at the age of 40 (76 versus 62 percent) 0 More of those inthe pre-school education group earnedmore than US$20,000 annually at the age of 40 (60 versus 40 percent) 0 The group who received the pre-school education had fewer arrests: 36 versus 55 percent were arrestedfive times or more 33 versus 48 percent were arrestedfor violent crimes 14versus 34 percent were arrestedfor drug crimes 36 versus 58 percent were arrestedfor property crimes. This evaluation found that there was a return to society of more than $17 per tax dollar invested inthe program. Source: Schweinhart (2004) In addition, policymakers should consider: (i) employing mental health experts (to deal with behavioral issues) and guidance counselors (for school-to-work preparation) in schools; (ii)reducing physical and psychological violence and integrating conflict management techniques into school management; (iii) upholding the quality of the curriculum, teaching staff, and materials, particularly in math and the sciences; (iv) expanding and tailoring the Bolsa Familia program to provide the correct incentives for secondary school completion, particularly taking into consideration the age-grade distortion; and (v) using school buildings for out-of-school activities, primarily supervisedleisure programs along the lines of theAbrindu Espaps program. Provide Remedial Programs Ideally, the appropriate investments would be made early in life, but some young people will inevitably fall through the cracks. It will never be possible to eliminate all risks in children's lives. Furthermore, the current generation has already been shaped by its environment and already engages in risky behavior and may have experienced negative outcomes from such behavior. This group should not be forgotten in the youth development strategy. Although remedial programs are costly, they are the only way to 21 reach young people who are currently at risk. Based on the report's findings, we recommend two specific types of remedialprograms. Targetedand Individual Treatmentfor At-risk Youthand their Families Effective remedial programs focus on human development rather than on treating the behavior. For example, putting young people in prison will prevent them from being violent in their communities, but it does not change the set of core beliefs and experiences that caused the young person to be sent to prison in the first place. Instead, effective remedial programs focus on changing the value system that lies behind the young person's choice to indulge inthe risky behavior. An effective way to re-program young people who are already engaged in negative behavior i s to offer intensive counseling and productive (employment-type) activities outside their own communities, which may be full of negative influences. This is not a recommendation for institutionalizing young people but for offering carefully planned productive activities in a setting that does not contain the same negative influences as prevail in the young person's own community. For example, some countries have instituted long-term volunteer programs2' in which young people are actively engagedin organized volunteer work and live in group houses with a team of social workers who specialize in youth development. These programs use the volunteer experience to teach job and life skills, and the whole program includes intensive social services to address behavioral issues. This differs from Brazil's Servip Civil Volunturio (SCV), which focuses on education and training with the volunteer work being a smaller component focusing on building civic responsibility.22 Models of volunteer programs that might work in Brazil are a Conservation Corps focused on preserving the Amazon or the Pantanal or a program in which young people help to improve the quality of life in large cities by, for example, removing graffiti and cleaning parks while learning how to make good decisions andto avoid risky kinds of behavior. Facilitating the School-to-work Transitionfor At-risk Youth. Some young people find it particularly difficult to move from beingfull-time students to full-time employees. This is one of the reasonsbehindthe high numbers of young people who work at the same time as attending school and of those who neither work nor attend school. The difficulty that they find in making this transition i s not surprising since young people have no experience in navigating the labor market. At-risk youths in particular tendto lack the social and technical skills to find and keepjobs. 21Often, a small stipend is provided, but it i s not a salary, since the programs use community work as a means of providing training rather than as employment. 22The structure of the models also differs in four important ways. Inthe long-term volunteer programs, the young people work and live in communities that are not their own; they live with, are supervised by, and are counseled by a team of social workers; the duration of the community work ranges from six months to two years depending on the time the young person needs to complete the program; and the volunteer opportunities are selected and designed by the program and the community. 22 The most effective policies for addressing this problem would be to expand labor intermediation services and support services for young people and to provide more job training and encourage companies to offer more internships. This is an expensive proposition but could be achieved by adapting existing programs. For example, a new component of the Sistema Nucionul de Emprego (SINE), Brazil's national employment system, could be developed that is aimed specifically at young job seekers. This component could not only provide young people with information about job vacancies but also offer classes on social skills needed in the labor market, advice on job search techniques, andfollow-up services. The technical skills of at-risk youth need to be upgraded. While a wide range of technical schools exists and the President recently announced further expansion o f this network, these programs (Sistema-S, CIEE, and technical schools) are largely inaccessible to the most at-risk youth since they are not designed to serve this population. Again, it shouldbe possible to develop either special components to these programs or to contract with the private sector to provide intensive support to at-risk youth who lack the educational'andbehavioralprerequisites to participate inthe labor market.23 Apprenticeships can help young people to make the transition from the education sphere to the work sphere. The apprenticeship needs to provide on-the-job training for the worker so that the program has a positive rate of return for the beneficiaries. Further, only those firms that can benefit from hiringthe apprentices should participate as forcing all firms to participate will lead to shirking and negative rates o f return for apprentices placedinfirms that do not need them. Alternatively, the government couldrestrict public financing o f (public or private) training firms to those that have formed partnerships with firms that will provide apprenticeships for their students upon completion o f the training course. Finally, at-risk youth need a more general support system if they are to become good workers. Social service agencies that specialize in the factors that may prevent a person from being a good employee (such as a lack o f childcare, illhealth, the absence of transportation, and domestic violence) need to collaborate to assist young people to overcome these obstacles. Box 5 outlines a successful model that has been implemented across Latin America for at-risk youth, thus demonstrating that this is not an impossible task. 23Specific recommendations are not given since the SINE and Sistema-S deserve serious analysis before any reforms are recommended. 23 Box 5: The LatinAmericaJ6venes Models Throughout Latin America, governments have adopted an experimental program for at-risk youth, which i s producing remarkable results. The program i s based on the following assumptions: (i) young people from poor neighborhoods do not have the skills to enter the labor market and do not know how to look for or secure training or employment; (ii) the opportunity costs of job preparation i s high for the young person and hisher family (particularly young women); and (iii) it is impossible to predict the skills that the market demands. As a result, a model was developed on the following general principles: Job training is provided by-NGOs, which-are only accepted into the program if they have arranged internships for their students once their students graduate from the classroom instruction portion of the program and if the NGOs pass the accreditation criteria for traininginthe skills that the internship demands. The NGO provides services in the communities where the young people live in order to encourageparticipation ina familiar environment andprovide a personalized service. The program includes an intensive "life skills" component, which requires students to develop a life plan (education and employment), to study conflict management, and to learnjob search techniques. Social services are a key component of the program. The programenforces discipline in terms of dress, tardiness, absenteeism, and behavior andretains the rightto expel students who do not comply. A stipend for transportation, lunch, andchildcare (for women with children) is provided throughout the program. The employers do not pay the full cost of employing the apprentice and do not have an obligation to hire the apprentice. Impact evaluations of the program have shown that it i s successful in getting employment for at- risk young people and that they stay employed longer than those who do not participate in the rogram. :ource: ~ o p et ai (2002) o Assign and CoordinateInstitutional Responsibilitiesaccordingto Comparative Advantages So who should implement these programs? As the report has shown, common factors can lead young people to indulgein multiplekinds of negative behavior but the incidence of these kinds of behavior differs greatly by social, economic, geographic, and demographic factors. This suggests that there i s a need to abandon institution-specific youth strategies and to adopt a holistic, multi-dimensional strategy that puts young people, not institutions, at the center of the policy. In practice, each institution or individual would contribute towards the overall goal of eliminatingthe negative behavior in question. The heterogeneity of the factors that influence young people's behavior points to the need to tailor programs and policies according to these geographic, demographic, and socioeconomic differences, thus requiring local institutions to define and support to the very specific needsof the young people intheirjurisdiction. The report has also shown that people and institutions at all levels of society affect the development of young people and are affected in turn by the actions of those young 24 people. Young people are most influenced by those closest to them - their families, their communities, andlocal institutions -but even state andnational governments andgeneral social norms play a role. Inturn, the negative behavior of young people imposes costs on everyone, not just on the young people themselves. This suggests that actors at all levels of society have a role to play inand a responsibility toward youth development. Based on the comparative advantage of different levels of governmental and non- governmental actors, clear roles can be identified (Table 7). The various actors inBrazil are well positioned to collaborate in providing youth development services with little overlap. For example: The Secretaria da Juventude has recently been created as a central guiding body strategically positioned within the Secretaria Gerd to facilitate collaboration across ministries and develop a national youth strategy. By focusing on developing national priorities and guidelines to enable actions at the local level andon providing technical and financial support, it can advise on how to allocate public and private budgets for maximumimpact. Other ministries are also doing this with their own youth strategies. Young people themselves are mobilizing both at the community and national level. The recently organized Vozes Jovens (see Box 6) i s a group of leaders from youth NGOs that is giving young people a voice on the national stage and that has developed a national youth policy. The government should ensure that this group i s involved in the preparation of the national strategy and in the implementation o f the strategy at lower levels of government. At the local level, young people continue to contribute via their own programs and actions Box 6: Youth Policy by andfor YoungPeople The youth community of Brazil has recently become organized and, in an inclusive manner, has defied a youth strategy to present to the government as their input to the national debate on youth policy. The group was originally formed in2003 to respondto the World Bank's invitation for young people to provide inputs to the Country Assistance Strategy that was under development. In 2004, representatives from more than 100 youth NGOs met - in a process named Vozes Jovens - to begin defining a youth strategy to present to the national government. The findings of the Vozes Jovens Iconference were published later that year. This process culminatedinMarch 2005 when the youth organizations andtheir partners inthe government and inthe internationalcommunity heldthe VozesJovens IIConferencewith the objective of refining the strategy created at the previous conference. More than 200 young people from across Brazil who representeda cross-section of the country in terms of race, gender, income status, ethnicity, disability status, and age spent three days together and emerged with a strategy that was outlined inthe Relatorio Final - VozesJovens Rita Janaina Quadros. The strategy recommends actions in the areas of education; labor markets; citizenship and social participation; human rights, affirmative action, gender, andrace; sports, culture, leisure, technology, andcommunications; the environment; and health. The next step will be for the young people's representativesto discuss the recommendationspresentedinthe strategywith the government. 25 State and local governments have developed youth strategies and are channeling resources to local civil society organizations and private sector firms to implement these programs. Better coordination between the state and local levels in defining target groups, priorities, and the division of labor at each level of government would make the delivery of services more efficient (see Box 7 for an example). Box 7: Coordination betweenCity and State Government inRio de Janeiro The state and municipalgovernments of Rio de Janeiro both have very active but very different youth development programs. Comparing these programs highlights not only the different possible models for youth development but also the difficulties involved in ensuring coordination across levels of government. Institutional Arrangement. The state government has a State Secretary for Childhood and Youth (SEIJ) with a mandateto coordinate all programs targeted toward children andyoung people. All existingprograms were put under the supervision of the SED. Incontrast, the municipality of Rio de Janeiro established a coordinating committee in 2002 (the Mucrofin@o Jovern) that i s responsible for coordinating and integrating all of the municipal programs that focus on young people. Target Populations and Strategies. The state's mission i s to "change the status quo of Rio's young people and create new perspectives for human development for these individuals, their families, and communities." The target group is 10to 22 year olds of mixed race, low education, living in poor neighborhoods, with family problems, and unable to meet their basic needs. The programs have a particular emphasis on at-risk youth and their families. Priority i s given to communities where the risks are most recurrent according to studies done by the Secretary of Justice and UNESCO. The programs include supporting sexually exploited young people, nutrition programs, programs for disabled young people, and family support programs and use methods including art, sports, and community work. The municipalities, on the other hand, focus on excluded youth, defined as those who have no access to cultural activities, social activities, and public services outside of their communities and who are discriminated against in the labor market. Most programs focus on education and labor market integration. Partnerships. Both the state and municipal governments have strong partnerships with civil society. NGOs were involved in conceptualizing and implementingthe SEWSprogram and some of the municipalities' programs. Because of the differences in their target groups and methodologies and the small number of program beneficiaries, there is no overlap between programs. However, there i s also little complementarity between the programs, primarily due to a lack of coordination between the two levels of government. This highlights the need for a concerted effort for different levels of government to do a better job of working together to support young people and those on whom they depend. Source: Prates andNogueira (2004) 26 2h 4 .. 0 . vl rr a Y k a, 3 0 a M 9 C R C 0 -1a, 0 U 9 . s 3 3 0 . B B b NGOs are already very active in implementingprograms and giving feedback to the government at all levels. The government can maximize this creative and energetic activity by providing more incentives for NGOs to get involvedin youth development, increasing financial and technical support for this work, and improvingprogram designthrough the development of monitoring and evaluation (M&E) systems. Of course, the successful implementation of this model will require a group of people at all levels of society who are well acquainted with youth issues and effective policies for addressing them. This is not a trivial prerequisite, given the complexity of the youth years. However, the recent organization of the youth issue, elevation to more visible levels in government, and the rich dialogue around the issue is a start to the education process. The coordination challenge i s more difficult, but there have been some successful experiments in Brazil. For example, in some municipalities, the Bolsa Familia program has municipal andNGO services that are tailored to the specific needs of the clients (Box 8). At the national level, the Inter-institutional Committee meets regularly to coordinate youth work across ministries and to discover the state-specific needs of young people. Another example is how the Ministries of Labor and Justice collaborate on the SCV. Box 8: PuttingYoung Peopleat the Center of Policy Cross-agency collaboration to improve service delivery i s an evolving initiative in the Bolsa Familia program that may benefit young people. Specifically, the Bolsa Familia teams in some municipalities are not only ensuringthat families comply with the program's conditions but meet regularly to discuss the difficulties that certain families have in meeting those conditions. The assumption underlying this strategy i s that external factors may be responsible for preventing people from fully complying with the program, so these families may need additional support. Thus, the client andhisherneeds arethe primary focus of the programteam's efforts. Many sectors, levels of government, and non-governmental actors have a role to play even in what may seem to be sectoral-specific policies. For example, further expanding secondary education will require the Ministry of Education to provide the schools and pedagogical guidelinesto be further developed by localeducation departments. However, it will also require the Ministry of Work andEmployment to provide guidelines to local governments for hiring guidance counselors and school-to-work services, the Ministry of Social Development andEradication of Hungerto provide resources to the Bolsa Familia to encourage school attendance and to local institutions to support the process, and local NGOs and young people to provide supervised after-school leisure programs. While coordination is achallenge, the multi-agency approach is key for helping at-risk youth. 28 REFERENCES Aedo C. and S. Nunbz (2001) The impact of Training Policies in Latin America and the Caribbean: The case of "ProgramaJoven". CinterforLLO. Assun$io, J. andL.Carvalho (2005) "Brazilian Youth at Risk: Estimating the Costs of Not Preventing RiskyBehavior." Mimeo Betcherman, B., K.Olivas, andA. Dar (2004) "Impacts of Active Labor Market Programs: New Evidence from Evaluations with Particular Attentionto Developing and Transition Countries." The World Bank:Washington, DC Bilchik, S. (1998) Juvenile Justice Bulletin. Department of Justice: Washington, DC Blum,R., C.A. McNeely, andP.M. Rinehart (2002) Irgxoving the Odds: The Untapped Power of Schools to Lmprovethe Health of Teens. Center for Adolescent Health and Development, University of Minnesota: Minneapolis, MN Bonelli, Regis, Jose Guilherme Reis, andAlinne Veiga (2004) "Characteristics and Determinants of Youth Unemployment inBrazil" LCSES/World Bank, Mimeo Bradley,Steve andAnh Ngoc Nguyen(2006) "The School-to-Work Transition" inG. Johnes andJ. Johnes (eds.) The International Handbook on the Economics of Education. Edward Elgar Publishing. Callahan, C., F.Rivera, andT. Koepsell(l994). "Money for Guns: Evaluation of the SeattleGunBuy-Back Program". Public Health Report 109:472-477 Carlini, E. (2002) "Levantamento sobre o Us0 de Drogas Psicotropicas no Brasil- 2001." CEBRIDAJNIFESP: SZo Paulo. Castro C. andA. Verdisco (1999) Training Unemployed Youth inLatin America: Same Old SadStory? Inter-American Development Bank: Washington, DC. Clark, R.M. (1988) CriticalFactorsinWhy DisadvantagedChildren Succeed or Failin School. Academy for Educational Development: New York Dalbosco Dell' Aglio, Debora, Vicente Cassepp Borges, and Joana Severo Leon, (2005). "Indice de Bem-EstarInfanto-Juvenil no Brasil." Mimeo. DukeUniversity (2004) "The Foundationfor ChildDevelopment Index of Child Well- Baing (CWI), 1975-2002" Mimeo 29 Duryea, Suzanne (1998) "Children's Advancement through School inBrazil: the Role of Transitory Shocksto HouseholdIncome" IADB Office of the Chief Economist Working Paper #376 (IADB: Washington, DC) Instituto Cidadania (2004) Proiecto Juventude: Documento de ConclusZo Brasilia, Brazil Knowles, J.C., andJ. R.Behrman(2005) "Assessing the economic benefits of investing inyouth indevelopingcountries." WorldBank: Washington, DC Koller, Silvia, J'amis Ribeiro, Elder Cerqueira-Santoa,Normanda Araujo de Morais, and MaycolnLeoniTeodoro (2005) "Juventude Brasileira: Comportamentos de Risco, Fatoresde Risco e de Proteggo." Mimeo. Lynam, R. (1999) "Project DARE: NoEffects at 10-YearFollow-Up". Joumal of Consulting and Clinical Psychology 67(4):590-593 Maddaleno, M.andJ. Schuitt-Aine (2002) Sexual Health andDevelopment of Adolescents and Youth inthe Americas: Program andPolicv Implications. Pan- AmericanHealth Organization: Washington, DC. Manlove, J, K.Franzetta, K.McKinney, A Romano, E.Tery-Humen (2004) NoTime to Waste: Programs to ReduceTeen Pregnancyamong MiddleSchool-Aged Youth. National Campaign to Prevent Teen Pregnancy: Washington, DC. Maloney W. andM.Bosch (2005). "Labor Market Dynamics inDevelopingCountries: ComparativeAnalysis using Continuous Time Markov Processes." World Bank: Washington, DC Unpublished Nopo, Hugo, Miguel Robles, andJaime Saavedra (2002) "Una Medici6ndel Impact0 de Capacitacih Laboral Juvenil Projoven," Documento de Trabajo GRADE. Office of Justice Programs (2000) Promising Strategies to Reduce SubstanceAbuse. US Departmentof Justice: Washington, DC. Palocci Filho, Antonio, Marco de Barros Lisboa, Octavio RibeiroDamoso, Roberto Pires Massenberg, andRozaneBezerrade Siquiera (2005) OrqamentoSocial do Govemo Federal 2001-2004Ministerio da Fazenda, Brazil, April Pena, Maria-Valeria, Barbara Brakarz, andVania Medrado de Franc0 (2004) "Particularly Vulnerable Groups: Education andSchools inIndigenous and Quilombo Communities" inMaria-Valeria Pena and Maria Madalenados Santos (eds.) Children's and Youth Vulnerability: Poverty. Exclusion and Social RiskinFive Brazilian States World Bank: Brasllia PISA (2003) "Literacy Skills for the World of Tomorrow: Further Results from PISA 2000". OECD andUNESCO Institute for Statistics: Paris 30 Prates, Ceres and MariaBeatrizNogueira(2004) "Juventude e PoliticasNblicas: Um Olhar sobre as A@es Executadas no Estadoe no Municipio do Rio de Janeiro". Satcher, D. (2001) "Youth Violence: A Reportof the SurgeonGeneral." Departmentof HealthandHumanServices:Washington,DC Schinke, S. (2000) Thirty-monthImpactAssessment of ProjectLearn. Columbia UniversitySchoolof SocialWork New York Schweinhart,Lawrence(2004). The HighlScopePerryPreschoolStudy throughAge 40: Summarv, Conclusions, andFrequentlyAsked Questions. High/ScopePress: Ypsilanti, Michigan Spinks, Sarah (2003). "Adolescent BrainsareWorks inProgress." PBS Frontline. Waiselfisz, J. (2004) Mapa du violencia IV: osjovens do Brasil: juventude, violencia e cidadaniu. UNESCO:Brasilia WorldBank (2005) Youth inNumbers:LatinAmerica andthe Caribbean.World Bank: Washington, DC World Bank (2003) CaribbeanYouth DevelopmentWorld Bank:Washington, DC 31 ANNEXI:THEYOUTHWELFAREINDEX The index presentedinthis section is an adaptation of the Child and Youth Index (CYI) created by Duke University in the United States. The original index i s based on the concept that young people are multi-dimensionalpeople whose development i s not easily measuredby a small set of indicators. Inthe United States, the CY1has been calculated annually since 1975, which has yielded a series of data on youth development over 30 years. The Brazilian Youth Welfare Index (BYWI) was created for this report as a way to illustrate the conditions experienced by young Brazilians. It includes indicators that measure socioeconomic conditions, health, behavior, school performance, and participation in civil society. We deliberately used data that were widely available to make it easier to continue producing the index infuture years. The data were drawn from a variety of sources (see Dalbosco et al, 2005 for the sources), and all the data could be disaggregatedto the state level to allow for state-level analysis. The specific variables included inthe calculation of the BYWIare: Percentageof 10to17 year olds and 18 to 24 year olds with AIDS Pregnancyrate of 15 to 19-year-oldgirls Suicide rate of 15 to 24 year olds Homiciderate of 15 to 24 year olds Morbidityrate due to external causes of 15 to 19 year olds Percentagesof 12to17 year olds who use alcohol, marijuana, and cocaine Rates of grade advancement in primary education (5th to 8th grade)and in secondaryeducation Literacy rate of 15 to 24 year olds Scores in Portuguese and mathematics on the 8" grade Ensino Fundamental test and on the 1lth Ensino Mkdio test grade Average years of study of 14 year olds Ratesof school attendance of 7 to14 year olds andof 15 to 17 year olds Unemployment rate of 15 to 24 year olds Percentageof 10to 17 year olds who work Percentageof 10to 17 year olds who neither work nor study Percentageof 16to 24 year olds who work inthe formal sector Percentageof 16to 17 year olds who vote. Additional variables are included in the Brazilian Child and Yout,, Welfare Index (BCYWI) and inthe BrazilianFullChild and Youth Welfare Index (BFCYWI), including general factors that affect the development of today's young people, such as family poverty or the share of physicians in each state. Each variable is standardized, using as the "mean" the national average for that indicator, and the standardized variables are average by state to create the index. 32