Children's and Youth Vulnerability
-0
(DD 
°-'--(D 
Ocr,
oDQ
Cl)
CU           CD
co ,  (D O~~~~~~~~~~~~~~~~A
r-                    O~~~~~~~~~~~(
o  _~~~~~~~~~~~~N
oEQ



I



,,< J - // jSf-
^< \ >§ eG 8 \ \ r . _.>t<^.ffi>
_tVa 9 %_J \>> 9 a >-{ | [1 w
o\. r r ( ,c | ' ,= ---
, l dn li - -=-
Ar r t- {<>rk Wr<;} Ik M \
< 1 ' K n Wr
G Y rs e A a 9 k
\ x,\ t] , , , , 
tz C v ro¢> I I | l' . -
, \ \ t , _ V },
P(Xs A\>>2<n0,
- \ ) t et * C-- * v \ 
_ 1 n _ < W > 7AtE: nv vr {  \  > _  f  I 1/f y  1-----
t  ) fv  b 1 <  OL' /d         F-
A/ n r - o-
_,,<. ,, . , , t , , t ,_w -
Z V a\yreCj , rX C- t
/ l\ f at C) "
s f Cr\ l fim(\n >
no wa ,+ sr n 1va \n
>} k / 1k S u r J} %n F
CL Jrv e n 11 | - i
, \ . , \ \\ , , t V , b
r 0 lA nOc t\t / <)^' f - , >+ - \ 5
\b) r t, \ <tS 1QL         i
A // U A9 8              >r '=
CC/C\\%L d<((9                  t-
r \ \ \ } \ { \ \ \ B _ 
-'L-/ r                      rsr <s 
r  \ ( t / I  / L _l -
V )AA t ; t/<_L o/ _
.. L: _,_4^. ,_-



Cover and Graphic Design:
TDA Desenho & Arte Ltda.
Responsable director: Marcos Rebougas
Graphic Design: Marcos Rebougas e
Fabio Brumana
Copy editing: Fabio Brumana e
Tatiana Rodrigues
www.tdabrasil.com.br
The World Bank
SCN Quadra 2 Lote A
Ed. Corporate Financial Center, salas 303/304
70.712-900. Brasflia - DF
Fone: 61 329 1000
www.bancomundial.org.br
Editorial Revision
Waleska P. Magalhaes
The World Bank
Children    and     Youth     Vulnerability  -     Poverty,
Exclusion and Social Risks in Five Brazilian States
Brasilia- 2004. 1 46p - First Edition
I - Editors: Pena, Maria Valeria Junho; Rodrigues dos Santos,
Maria Madalena



Children and Youth Vulnerability
Poverty, Exclusion and Social Risks in Five
Brazilian States
Editing, Introduction and Conclusions
Maria-Val6ria Junho Pena
Maria Madalena Rodrigues dos Santos
Authors
Alinne Veiga
Barbara Brakarz
Jose Edson Gomes da Silva
Maria-Val6ria Junho Pena
Regis Bonelli
Tania Sal6m
Vania de Franco
The World Bank
Latin America and the Caribbean Region
Environmentally and Socially Sustainable Sector Unit
Brazil Country Management Unit



Director of the Brazil Country Management Unit: Vinod Thomas
Director of the Environmentally and Socially Sustainable Sector Unit: John Redwood III
Lead Economist: Joachim von Amsberg
LCSES/Brazil Sector Leader: Gabriel Azevedo
LCSES/Social Development Sector Manager: Shelton H. Davis
Task Managers: Maria Valeria Junho Pena (Lead Sociologist, LCSES); Maria Madalena
Rodrigues dos Santos (LCSHD/Brazil Sector Leader)
The findings, interpretations and conclusions in this book are attributable to the authors,
and should not be attributed in any manner to the World Bank, its affiliated organizations,
members of its Board of Executive Directors or the countries they represent.



Acknowledgements
The articles in this book were originally produced as background material for the analysis of
educational exclusion that was one part of the Brazil AAA: Next Steps for Education in Four
Selected States (2003) led by Maria Madalena Rodrigues dos Santos. Given the wide scope
of the educational issues dealt with in this study, all the materials prepared as background
analysis could not be integrated into the final version of the AAA. This book publishes these
studies in their entirety.
We owe this publication to many friends and colleagues. Gabriel Azevedo, Brazil ESSD
Sector Leader, understood and brought the youth agenda to the forefront of the World
Bank program in Brazil. In the Latin American and Caribbean Region, Vinod Thomas,
Vice-President and Director of the Brazil Country Unit, John Redwood III, Director of
the Environmentally and Socially Sustainable Development Sector Management Unit,
and Ana-Maria Arriagada, Director of the Human Development Sector Management
Unit, supported the analysis of exclusion in an education sector work. Ernesto May, Jaime
Saavedra and Shelton Davis guaranteed the technical resources needed to finalize the book
in two versions, Portuguese and English. We are grateful to Indermit Gil for his innovative,
incisive, and unexpected comments on the text. Alberto Rodriguez, Anna Sant'Anna,
Andrea Guedes, Carla Avellan, Fabiana Rezende Imperatriz, Grissel Prieto, Jorge Uquillas,
Joachim von Ambsberg, Josef Leitmann, Kathryn Johns Swartz, Lilian Santos, Maria Lucia
de Oliveira, Ricardo Silveira, Waleska Pedrosa and Zeze Weiss helped with suggestions,
patience, friendship and work. Without the generous support of the Secretariats of
Education of Maranhao, Paraiba, Pernambuco and Sao Goncalo (Rio de Janeiro) the
fieldwork could not have been conducted. We recognize that what was accomplished here
was mostly due to their professionalism and willingness to share their knowledge. Many
valuable lessons were shared by the teachers, students and families who discussed their lives,
their schools, their frustrations and their dreams with the researchers. We did our best to
merit their generosity. This book is dedicated to them.
Maria Valeria Junho Pena
Maria Madalena Rodrigues dos Santos






Presentation
This book seeks to contribute to knowledge in two areas that are key to poverty reduction
and social development: education and youth vulnerability. Brazil has vigorously sought to
expand the coverage and quality of education, and has achieved rapid results over the past
decade. Progress has been achieved on important indicators, and the quality of education is
now undoubtedly better than it was in the past.
Despite these results, this book shows that important challenges remain. Social background,
either as insufficient family income or low parental schooling, persists as a determinant of
children's lagging educational achievement. Rural, indigenous and Quilombola students
have precarious access to education, with schools often not offering the last four years of the
mandatory eight in the fundamental cycle. Dropout and repetition rates remain key areas
of concern for policymakers. With little proper schooling, many youth find themselves with
few chances of stable incorporation into the labor market, facing poverty for themselves
and for future generations.
This book also focuses on what youth think about themselves, their schools, their families
and the vulnerability provoked by their gender, their ethnic background, and sometimes
the depth of their poverty. But reading what youth think is also comforting in many ways:
for the preoccupation they demonstrate with their country and its social challenges; for the
generosity they show to those with fewer opportunities; and for the hope they hold for their
future and the future of Brazil. All efforts should be made so that these young people and
others who will come after them have better education, better occupational opportunities,
and a better chance of living a full and fruitful life. Together with the Government and
people of Brazil, the World Bank strives to contribute toward this future.
Vinod Thomas
Director of the Brazil Country Management Unit
Latin America and the Caribbean Region






Presentation
This book demonstrates the efforts that Brazil should engage in to reach a pattern of
development that is inclusive for all. As indicated, while a substantial increase in educational
enrollment has been achieved, the lack of coverage is still a concern in rural areas, as
repetition and dropout continue to be major challenges for policymakers. As increasing
numbers of children and adolescents, particularly the poorest, fail to fully benefit from the
expansion of public education and services, youth become at-risk and an increasing source
of concern.
Unschooled youth become workers without opportunities and adults without choices.
Youth need attention, care, hope and a stable source of income to survive and thrive so
that they can offer to their own children better opportunities. An inadequately educated
society impairs not only productivity, but also the quality of family and public life. The
Environmentally and Socially Sustainable Development Sector Management Unit of
the Latin America and the Caribbean Regional Office, together with the Brazil Country
Management Unit of the World Bank, is committed to making a worthy life for those
that are secularly excluded, including the children and youth who live in rural areas, in
indigenous communities, in Quilombos or in precarious conditions in urban areas.
The young students who so generously contributed to the analysis in this book deserve
committed adults and policymakers and a brighter future in Brazil. In the World Bank, we
hope we will not fail them.
John Redwood III
Director
Environmentally and Socially Sustainable Development Sector Managment Unit
Latin America and the Caribbean Region



Acronyms
CNE      Conselho Nacional de Educa,co - National Education Council
FUNAI Fundacfo Nacional do indio - National Foundation for the Indigenous
FUNDEF Fundo de Manuten,co e Desenvolvimento do Ensino Fundamental e
Valoriza,co do Magisterio - Fund for Maintenance and Development of the
Fundamental Grade and Teaching Valorization
IBGE     Instituto Brasiliero de Geografia e Estatistica - Brazilian Institute of
Geography and Statistics
MC       Ministerio da Cultura - Ministry of Culture
MEC      Minist6rio da Educacao - Ministry of Education
PETI     Programa de Erradica,ao de Trabalho Infantil - Program for Erradication of
Child Labor
PNAD     Pesquisa Nacional por Amostra de Domicflios - National Survey by
Household Sampling
SAEB     Sistema Nacional de Avalia,co da Educa,co Basica - National System of
Basic Education Evaluation



Table of contents
INTRODUCTION                                                   15
CHAPTER I
Determinants of Educational Exclusion in Five States of Brazil  25
CHAPTER II
Qualitative Aspects of Educational Exclusion in Rural Schools  63
CHAPTER III
Particularly Vulnerable Groups: Education and Schools in Indigenous and
Quilombo Communities                                           81
CHAPTER IV
Teens and Youth Perceptions on Exclusion, Inequality, and Opportunities  97
CONCLUSIONS                                                   133
BIBLIOGRAPHY                                                  141
BOXES
1: What are Quilombos?                                        17
2: Percentage of Students no Longer Attending School in 1999  34
3: Size of the Young Indigenous Population                    87
4: Sharing of Responsibility for Indigenous Peoples' Education  88
5: The Kind of School that the Indigenous Peoples Want        92
6: What Students Think About Racial Discrimination           115
7: Ethnicity and Poverty According to Students               117
8: Indigenous People and Land According to Students          119
9: Views on Indigenous People from an Urban Student with
(Recognized) Indigenous Ancestry                          120
10: What Indigenous Students Think Non-Indigenous People Think
about Them                                                121
11: Why Poor People are Poor According to Students            123
12: How Can the Poor Stop Being Poor?                         126
13: Inequality: Brazil's Paramount Problem                    128
14: What Needs to be Done in Brazil                           130
15: Brazil Today                                              131



TABLES
1: Education Indicators for Rio de Janeiro                       29
2: Education Indicators for Maranhao                             30
3: Education Indicators for Rio Grande do Norte                  30
4: Education Indicators for Paraiba                              31
5: Education Indicators for Pernambuco                           31
6: Nursery School Attendance for Children under the Age of 5     32
7: Children and Youth who are not in School but who Attended School in the Past 33
8: Dropouts for Children aged 7-17                               34
9: Age-Grade Gap Greater than Two Years                          41
10: Variables and Definitions                                     48
11: Significance of Estimated Coefficients for Children Aged 10-14 Years old 49
12: Significance of Estimated Coefficients for Children Aged 7 to 14 Years Old  51
13: Significance of Estimated Coefficients for Children Aged 15-17 Years Old 52
14: Analysis of Maximum Likelihood Estimates for 10 to 14 Years Old  53
15: Analysis of Maximum Likelihood Estimates for 15 to 17 Years Old  54
16: Analysis of Maximum Likelihood Estimates for 14 to 17 Years Old  55
17: Levels of Estimated Coefficients, State regressions           55
18: Pooled Regressions                                            57
19: Levels of Estimated Coefficients: Pooled Regressions          58
21: Explanatory Factors and Dimensions Measured                   67
22: Qualification of Teachers in Maranhao                         68
23: Qualification of Teachers in Paraiba                          69
24: Organization of Time in Maranhao                              72
25: Organization of Time in Paraiba                               73
26: Teaching Practices in Maranhao                                76
27: Teaching in Paraiba                                           77
28: Brazilian Population by Ethnicity, 1991- 2000                 85
29: Number and Proportion of Indigenous Schools that use Culturally
Appropriate Materials                                         90



FIGURES
1: Percentage of Overage Enrollment in Rio de Janeiro by Age Group  35
2: Percentage of Overage Enrollment in Maranhao by Age Group  37
3: Percentage of Overage Enrollment in Paraiba by Age Cohort  38
4: School Attendance in Pernambuco by Age Cohort              39
5: Percentage of Overage Enrollment in Rio Grande do Norte by Age Group 40
6: School Attendance, Age-Grade Gap and Dropouts in Rio de Janeiro  43
7: School Attendance in Rio de Janeiro                        43
8: School Attendance and the Age Grade Gap in Rio de Janeiro  44
9: School Attendance in Maranhao                              45
10: School Attendance in Parafba                               45
11: School Attendance in Pernambuco                            46
12: School Attendance in Rio Grande do Norte                   46
13a: Qualification of Teachers in Top and Bottom Schools According to
Teachers and Students - Maranhao                            70
13b: Qualification of Teachers in Top and Bottom Schools According
to Teachers and Students - Paraiba                          70
14a: Organization of Time According to Teachers and Students in Maranhao 73
14b: Organization of Time According to Teachers and Students in Paraiba  74
15: Divergence Between Internal and External Stakeholders in Top
and Bottom Schools in Paraiba: Teacher Organization of Time  74
16a: Teaching Practices in the Top and Bottom Rural Schools
According to Teachers and Students - Maranhao               78
16b: Teaching Practices in the Top and Bottom Rural Schools
According to Teachers and Students - Paraiba                78
17: Divergence Between Internal and External Stakeholders in Top
and Bottom Schools in Maranhao: Teaching Practices          79
18: Indigenous Students by Region                              89






INTRODUCTION
Maria Valeria Junho Pena
Maria Madalena Rodrigues dos Santos
Brazil holds over half of South America's Gross Domestic Product and population. Its
economy is approximately twice the size of Russia's and India's and comparable to China's.
Brazil is the second largest market in the world for executive jets, helicopters, cellular
phones and fax machines; fourth for refrigerators; fifth for compact discs; and ranks
third for soft drinks. Sao Paulo, its largest city, has ten times more helicopter fields than
New York City. Since 2001, Brazil became the fifth country in the world with purchasing
power parity of over US$1.0 trillion. Its 1988 Constitution recognizes all the liberal rights
of democracy: freedom of speech, equality, equity of opportunities and special protection
for the vulnerable. Nevertheless, it is generally accepted that the modernization process
that Brazil has undergone over the last century has never reached the poorest segments
of the population, transformed social relations, or significantly expanded opportunities to
groups that have traditionally been excluded from mainstream society, such as the rural
poor, landless peasants, informal workers, African descendents, and indigenous peoples. A
substantial proportion of the Brazilian population lives in poverty, at numbers and intensity
far above what would be expected given the country's per capita income.
The lack of adequate schooling plays an important role in creating, reproducing, and
expanding both poverty and inequality. Low income in Brazil is associated with low
educational achievement, with the inability of individuals to secure good jobs, and with the
inability of the economy to generate a sufficient amount of employment relative to the size
of the population. In this context, one of the main challenges for articulating policies for
youth development is to reach those who cannot access the education system fully because
they do not attend school at all or, when attending, they repeat grades several times, finally
giving up entirely and dropping out.' Because so few Brazilians have a formal education or
qualified skills, the value of education is very high for those who do have it, which adds to
income inequality. This book seeks to contribute to the debate about the Brazilian youth
' The net enrollment rate at the primary level increased from 89 percent to 96 percent between 1996-2001. In a country of
continental dimensions and as diverse as Brazil - with 42 million students - the implementation of educational policies is
a complex task. The current Government has prioritized the following elements to improve quality: implement distance
education, particularly in remote areas and poor schools; improve the quality of school materials; redefine the basic educational
curriculum; encourage foster parents' participation in school decision making processes; enhance monitoring and evaluation;
provide teacher's training; and improve infrastructure, particularly in the poorest areas. Overall, the Government expects to
make the system more cost effective and responsive to family needs by decreasing the high repetition and dropout rates.
1  15



vulnerability, the challenges of social inclusion and the risky behaviors that so frequently
attract, and then entrap young people. The questions that this book privileges are broadly,
but not exclusively, related to education. In other words, the starting point of the articles
in this book is that the absence of an adequate schooling is a nodal element of the youth
vulnerability and of the risks that youth incur - the risk of unemployment, the risk of lack
of an stable income, the risk of poverty, the risk of transmitting poverty and exclusion
to other generations, and the risk of anomic and disintegrating behaviors. Although the
articles here published tackle a diversity of issues and points of views, all are inspired by
the same question, what are the elements, internal and external to the school system, that
respond for the persistence of social exclusion, despite the unquestionable recent expansion
of the educational coverage and the progresses made in the management of the educational
system.
Conceptually, exclusion refers to situations that deny access to some types of material
or symbolic resources that are essential to a human being's survival. Poverty is the most
common reason for exclusion because of the vulnerability that it entails. However, there
are certain groups among the poor who, by virtue of some specific deprivation due to their
ethnicity, gender, childhood malnutrition (which permanently impairs their capabilities),
the violence that may be a part of their lives, the social relations they are embedded in, or
the denial of some universal social and economic rights, are even less able to access and
take advantage of the social, economic, and educational opportunities that may eventually
be available to them. The 2000 Census confirms the many progresses in the educational
indicators over the decade. But it also shows that about 15 percent of the population older
than five - 24 million of Brazilians - is still illiterate. Among indigenous peoples (652
thousand) the proportion oif illiteracy is 30.2 percent. Among the 9.8 million of blacks,
23.2 percent is are illiterate; and among the 58.7 million pardos, the proportion of illiteracy
is 21.1 percent. However, among the 82.4 million whites, illiteracy decreases to 10.9 percent.
The 2000 Census also indicates that the lower is the family income, the lower are the chances
that individuals will attend or complete any educational level. In only five states, as shown
in this book, the total of children and youth at risk because of their educational exclusion
reaches the alarming number of 2.6 million. This book analyzes the interaction among
the many socioeconomic elements that lie behind educational exclusion and explores the
specific constraints affecting the education of indigenous communities and people of
African heritage, particularly those living in Quilombos (see Box 1).
16   1



Box 1: What are Quilombos?
Quilombos are groups of descendents of fugitive Africans, brought as slaves to Brazil,
who have lived together for centuries in the same locality and are linked by community
bonds. The most notorious of all Quilombos, Palmares, dates from 1604, and consists
of several smaller Quilombos, covering a territory that now forms part of the Alagoas
and Pernambuco states. For almost 100 years, Palmares was able to successfully resist
the Portuguese and the Dutch colonizers, with the Quilombo's King, Zumbi, becoming
a revered national hero and a Brazilian symbol of endurance against oppression. In
1850, the Eusebio de Queiroz Act criminalized the traffic of africans to Brazil, although
it continued clandestinely, and in 1871, the Law of the Free Womb (Lei do Ventre Livre)
dedared that all children born of slaves were to be free. However, only in 1888 was the
institution of slavery forbidden by law, by which time there were approximately 700,000
slaves in the country, and many Quilombos continued to exist. The 1988 Constitution
guaranteed rights of Quilombos in the Article 68 of the Transitory Clauses. The Palmares
Cultural Foundation (Fundac,o Palmares), in the Ministry of Culture (Ministerio de
Cultura - MC), in conjunction with the Special Secretariat of Policies for Promotion of
Racial Equality, is responsible for the major policy issues related to Quilombos, including
identifying, recognizing and assigning these rights to them. Also, an inter-ministerial
working group was created in 1995, and is responsible for the advancement of the
Brazilian black population. The group is entrusted with formulating public policy to
foster the rights of African descendants and proposals for compliance with the prevalent
legal dispositions, including reassessing ethnic and racial stereotypes in educational
curricula. Of the about 1,000 groups identified and living in rural areas and dispersed
in states, including as Alagoas, Amazonia, Maranhao, Sergipe, Bahia, Pernambuco, Mato
Grosso, Sao Paulo, Goias and Minas Gerais, only few have had their rights officially
recognized.
This book focuses on the relationship between poverty, ethnicity, gender and educational
exclusion in four Northeast states in Brazil: Maranhao, Paraiba, Pernambuco, and Rio
Grande do Norte; and in the Southeast state of Rio de Janeiro. Notwithstanding that
the Bank has education and social development projects in other states, these ones were
chosen because of the level of their present dialogue with the Bank. In addition, the issues
discussed in this book have not yet been systematically analyzed in these states, and the
findings will also support the Bank's portfolio in the education, poverty, employment,
youth development, and social development fields. Special attention is paid to the rural
areas because, in relation to its total population, these areas present a higher proportion of
poor than urban areas, although their absolute number is smaller.
I 17



Chapter I, Determinants of Educational Exclusion in Five States of Brazil, by Regis Bonelli
and Alinne Veiga, seeks to understand why educational exclusion has been so relentless,
focusing on the factors behind the propensity to be expelled from the educational system
at the household level. Specifically, the chapter presents for the states of Maranhao, Paraiba,
Pernambuco, Rio Grande do Norte, and Rio de Janeiro: (a) dropout and repetition rates,
showing a small but persistent lack of coverage in both absolute numerical terms and as a
proportion of children in the relevant age cohorts; (b) the ratio of "cohort displacement" in
terms of the grade in which the cohort of students should be enrolled according to their age;
(c) at what level of cohort displacement the student becomes unmotivated and drops out
of school; and (d) why some children and youth are not enrolled in school. Key explanatory
variables include: (i) location (urban/rural); (ii) household structure (single headed or
jointly headed); (iii) household income; (iv) parent's education; (v) number of siblings;
(vi) ethnicity; (vii) gender; (viii) child labor; and (ix) migration. Analysis was conducted
separately for each state. For the purposes of regressions, the data was also pooled together
for the establishment of the determinants of dropping out comparatively for males and
females and for whites and non-whites. The 1999 Household Survey (Pesquisa Nacional por
Amostra de Domicilios - PNAD) was used as the data source.
Chapter II, Qualitative Aspects of Educational Exclusion in Rural Schools, by Jose Edson
Gomes da Silva and Barbara Brakarz, deals with the issue of the quality of education in
rural areas in two of the states under discussion: Maranhao and Paraiba. This chapter seeks
to present, from a school-level perspective, what factors make students fail in some of
schools and succeed in others. By focusing on the school, it complements the focus of the
previous chapter on the household, by highlighting some of the factors that enable students
to excel academically even when their economic conditions are poor. Once these factors
are identified, it is expected that, if replicated, they can improve quality, reduce repetition,
and decrease dropouts, even in poor rural schools. The analysis for this chapter is based on
fieldwork conducted in 40 rural schools whose students performed at the top and bottom of
Portuguese and Math achievement tests applied by the Ministry of Education (Ministerio
da Educacao - MEC). The Secretariats of Education in the two above-mentioned states
provided additional information on the quality of the rural schools in their respective
states. In these schools, 40 principals, 123 teachers, 294 parents, and 553 students completed
questionnaires, the aim of which was to determine what factors made some schools meet key
learning objectives as defined by MEC and what factors made others fail. The questionnaire
contained more than 50 questions, covering in detail: (i) how clear the curriculum is, how
well it is organized, how much it is followed by teachers, and how results are measured; (ii)
the teachers' practices, accountability, and responsiveness to student's needs; and (iii) time
18   1



management of the teachers, discipline they are able to convey, and their accountability to
students' efforts, with special attention to homework activities. Results are not statistically
significant given the small sample size of respondents (that is, they cannot be generalized
for Brazil). However, the results are qualitatively indicative of what enthusiastic teachers are
able to accomplish in very poor schools and under very unfavorable circumstances.
Chapter III, Particularly Vulnerable Groups: Exclusionary Schooling in Indigenous and
Quilombo Communities, by Maria Valeria Junho Pena, Barbara Brakarz and Vania de
Franco, focuses on the education of indigenous peoples and Quilombolas. Information is
based on the 1999 Census of Indigenous Schools (Censo de Escolas Indigenas) conducted
by MEC and on qualitative fieldwork in Maranhao, Pernambuco and Paraiba. Although
ethnicity is a main concern in all other chapters of the report, this Chapter specifically
focuses on schools and education of these groups, including by taking into consideration
the educational policy for this population as defined by MEC, particularly regarding a
proposed culturally diverse approach. The data source also includes a Participatory Rapid
Assessment conducted with teachers, students and parents in indigenous and Quilombolas
communities in the three states and an evaluation of indigenous education by the Secretariat
of Education of Pernambuco.
It is a truism that the youth of today will be the citizens of tomorrow and that the
processes through which their perceptions are consolidated lie at the heart of their
attitudes in the future. Chapter IV, Teens and Youth Perceptions of Exclusion,
Inequality, and Opportunities, by Tania Salem, analyzes the attitudes of youth on
the issues of exclusion, stratification and inequality in Brazil. To a large extent, this
chapter provides insights into what perceptions, values and attitudes are transmitted
by a range of different sources: families, schools, the media, and peers. The analysis
is based on responses to an open-ended questionnaire completed by 95 students (40
males and 55 females) aged 12 to 24 years old, all of whom were then in the 8th year of
the fundamental cycle2 in the states of Rio de Janeiro and Pernambuco. The main objective
of the questionnaire was to grasp how youth understand exclusion on the basis of gender
(roles and relationships), ethnicity (relations among whites, blacks and indigenous),
and social inequality and poverty. Although the survey did not aim to be statistically
representative, it encompassed students from different social backgrounds and with varied
life experiences. The questionnaire was applied to students attending private and public
schools, which, in Brazil, represents a significant difference in the quality of education
and in the socioeconomic status of students, not to mention in the physical conditions
2 An information that already indicates a high degree of repetition in the system.
19



of the schools. The private schools where the fieldwork took place were located in Recife,
the capital of Pernambuco state, and in Rio de Janeiro, the capital of Rio de Janeiro state.
A public school in Rio de Janeiro, also included in the analysis, is located in its northern
periphery, where poverty is clearly more acute than in the southern neighborhood of
the city, where the private school in the study are located. In the case of Pernambuco,
questionnaires were applied to students in a top private school of the capital, Recife, and
to public school students living in a Quilombo (located in a rural area about 80 miles from
Recife) and in an indigenous community (located in the small town of Aguas Belas, about
100 miles from Recife).
Maria Valeria Junho Pena and Maria Madalena Rodrigues dos Santos wrote Chapter V, which
presents Conclusions and Recommendations. The following are general conclusions and
recommendations of this book:
Conclusions
* There are three defining factors of educational exclusion, all of them statistically relevant: (a)
never having attended school; (b) having at least a three3 year age-grade gap; or (c) having
already dropped out of school. Results show that a large part of the 7-17 age group is at risk
of exclusion in the five states, with 46 percent in Maranhao, 39 percent in Paraiba, 40 percent
in Pernambuco, 36 percent in Rio Grande do Norte and 22 percent in Rio de Janeiro havinge
never attended school, having attended and dropped out, or probably dropping out in the
future given the level of repetition. The number of youngsters at risk of unemployment and
poverty in the five states adds up to a staggering figure of 2.6 million.
* The main determinants of educational exclusion from a statistical and household based
point of view are socioeconomic. Although in different age groups and in different states,
specific determinants may exist, household income, parent's schooling, and premature
entrance in the labor market are consistently the main determinants of educational
exclusion in all five states and in all age groups. Vulnerability to exclusion increases
among females in rural areas and among non-whites in Rio de Janeiro. In some young
age groups, inter-municipal and inter-state migration are also disruptive and contributes
to dropout and exclusion.
* Paradoxically, increasing schooling among the less poor may contribute to the exclusion
of the poor and unschooled. As youth unemployment grows, better educated youth
crowd out the less educated even from unskilled jobs. In 1992, the 16-24 age group
unemployment rate was 11 percent; in 2002 it had increased to 18 percent, resulting in
an additional 1.5 million youth unsuccessfully searching for jobs. Due to the shortage of
20   1



vacancies, poorer and less educated youth are at the end of the line for jobs, even those
less qualified. This book shows that many children and youth are candidates for unstable
and low paid jobs.
The qualitative fieldwork3 indicates the existence of strong segmentation between private
and public schools in terms of quality. Private schools tend to be more homogenous in
terms of the students' socioeconomic background, which include parents with college
education and, in the case of Rio de Janeiro, being white. Repetition is less frequent,
the school is clean and well equipped, demands on teachers' performance are high,
and the repertoire of subjects in the curriculum is broad. Parents have a stake in the
education of their children, and they demonstrate it. Public schools, on the other hand,
are heterogeneous, attended by students from different ethnic groups and socioeconomic
strata. In rural areas, particularly in Quilombos and indigenous communities, age
differences in the same classroom are noticeable. Classrooms are overcrowded, school
buildings are precarious, and safe water, latrines and electricity are very often missing.
School lunches are unpredictable and insufficient. Well-trained teachers frequently shift
to private schools where they are requested to excel, while keeping their job in the public
schools, which sometimes pay them higher and offer better benefits. Although parents
are critical of poor schools, they face income, schooling, power, and time constraints to
act as an active pressure group advocating for a better education for their children.
* The main elements contributing to explain the variation of student performance in
schools with the same poor background are related to how disciplined and well-run the
schools are. Schools where students perform better are those where it is unacceptable
for students and teachers to arrive late, classes begin on time, homework is required
and feedback provided, students are tested, and teachers follow the performance of each
student.
* Parents and students operate within a dual system to classify their priorities. The first
is based on their own private conditions, the second, on the public conditions of the
country. What is deemed as belonging to the public realm, thus demanding public
interventions, is viewed as a problem and claimed as a priority for governmental action.
Thus, corruption, or urban and rural land violence are seen as priorities. On the other
hand, unequal gender relations, domestic violence, and bad schools are viewed as a sad
reality, but not as priorities, because it is believed that nothing can be done about them
' PNAD does not allow for the discrimination between students in the private and in the public schools.
1 21



- they are part of an unchangeable destiny of being poor. However, this should not
lead to the erroneous assumption that they are not important, but rather that they are
perceived as immutable conditions that nothing can be done about.
Recommendations
* The 15-24 age group represents one- fifth of the Brazilian population. Many are
not in school, and many others, who are enrolled, are not acquiring the adequate
skills and qualifications for the labor market and, thus, will be unable to expand
their knowledge as required by the new dynamics of globalization and advancing
technology. However, youth development has only recently become an area of
concern with initiatives in this area still being fragmented and unfocused. Effective
policies to help address the youth lack of proper qualification and chances in the
labor market should involve quality schools but also, particularly in the short run,
training programs to upgrade job skills. For those many who had dropped out with
slim chances of completing at least basic educationthe fundamental cycle, innovative
and comprehensive youth policies are needed, matching informal education to labor
market characteristics.
* The school system should define clear roles and responsibilities for all those
involved with delivering its policies and mandate. Legal, financial and management
accountabilities need to be established to boost school quality and encourage own
accountability amongto parents, students and tax-payers. Continuous, but innovative
inventive training is needed for those involved. Some experiences withof school
autonomy have shown positive results when all of these conditions are in place.
* While there is still a need to increase access to education in rural areas, in both urban
and rural areas, the quality of education, teachers and curriculum relevance are other
sources of concerns claiming for policy priority. In rural schools, education policy
should focus on solutions to address their isolation, unqualified teachers, obsolete
teaching methodologiesy, paucity of the teaching support system, and over-age
students. In urban schools, education policy should focus on integrating parents,
promoting autonomy, creating an interactive environment, and opening student's
eyes to contemporary issues such as, labor market, technology, violence, drugs and
health, including reprodutive.
* A set of specific supply and demand policies are needed for indigenous and Quilombo
schools. The main sources of concern are: location; infrastructure, teacher's training
accessibility to grades up to the fourth grade, and respect for cultural identities.
22 



These issues deserve attention from states, which should seek more appropriate
solutions. A policy development in the area of youth employment by the Ministry of
Labor and Employment should also consider special incentives for these groups and
the development of a training program based on the special advantages that their
cultural system may be able to produce and market.
1 23






(PP  -(  (7    CHAPTERI
DETERMINANTS OF EDUCATIONAL EXCLUSION
IN FIVE STATES OF BRAZIL
Regis Bonelli
Alinne Veiga
i-E0-A/ii iP



-: - :



Determinants of Educational Exclusion  L 'P
in Five States of Brazil  - ..
Objectives and Methodology
The primary goal of this chapter is to measure the rate of school dropout and explain
the process that leads youngsters to abandon the formal educational system before
completing their basic schooling. As a complement, other aspects of educational exclusion
and system failures, such as grade repetition and not starting school at all, are reviewed.
Specifically, the chapter attempts to determine: (a) the scale of the dropout problem in
absolute numbers and as a share of each age-group; (b) the size of the age-grade gap;
(c) the cohorts in which the problem is more acute; and (d) the reasons why dropouts
occur. The data source utilized is the Brazilian 1999 National Household Sample Survey
(Pesquisa Nacional por Amostra de Domicilios - PNAD) where each member of the
household is included in the survey.
The Brazilian states analyzed are Maranhao, Paraiba, Pernambuco, and Rio Grande do
Norte in the Northeast region, and Rio de Janeiro in the Southeast region. For each state,
it is determined: (i) school coverage and dropouts both in absolute numbers and as a
proportion of children in each of the relevant age cohorts; (ii) by how many years a given
child lags behind the normal school level for his/her age group; (iii) the age cohorts in
which the dropout rate is highest; and (iv) the reasons why dropouts occur. The age group
7 to 17 is used as the main reference group throughout the study.
Dropouts are defined as young people (aged 7 to 17) who have attended school at some
point, but were no longer in school when the survey was conducted. Therefore, a dropout is
a child or youth in the age group 7 to 17 years old who, in 1999, had not completed primary
(or fundamental) school with no likelihood of ever completing his or her education. For
the purpose of measurement, a slightly broader concept is used, namely, a dropout is a child
or youth (under 18 years old) who left school after completing primary school but who, in
1999, did not seem likely to ever complete secondary education.
It is commonly accepted that the process that leads someone to drop out of school is
complex and can include such factors as the low quality of schools, the loss of motivation
that comes from repeating the same grade (and is initially reflected in overage and age-
grade gaps), and the need for the child to earn income to supplement household earnings'
or to do non-paid work in the household. However, these factors will not be considered,
as they are not observable in the PNAD data. To analyze the factors that account for the
likelihood of children dropping out of school, a regression equation was used in which the
See Corseuil et al (2001).
27



Children and YouthVulnerability Primeira Edicio
Poverty, Exclusion and Social Risks in Five Brazilian States
dependent variable was binary in a Logit, Probit, or Ordered Probit model that assumes the
value of 1 if the child or youth is a dropout and 0 otherwise.2 Explanatory variables that
were included in the model were: (i) the location of the household (in a rural or an urban
area); (ii) the household's structure; (iii) per capita household income; (iv) the education
of the child's parents; (v) the number of the child's siblings; (vi) the ethnicity of the child;
(vii) the gender of the child; (viii) whether the child works; and (ix) family migration either
between states or municipalities.
State Differences in Coverage, Dropouts, and Overage
Enrollment
Education performance is different in each of the five states, with Rio de Janeiro having
higher coverage, lower dropout rates, and fewer overage students. The PNAD data also reveal
important trends that are common to all five states and probably to the whole country. They
are: (i) the increasing expansion of coverage to an extent that almost all children aged 7 to
17 are enrolled in both rural and urban areas; (ii) the persistence of considerable overage
enrollment, an indication of high repetition rates and of the low quality of teaching; and
(iii) the inability of schools to change the prevailing socioeconomic inequality because the
children of parents with low educational attainment levels are likely to do poorly in or drop
out of school themselves.
Educational Coverage
The total number of children, youth, and adults who never attended school are shown
categorized by age in Tables 1 to 5. Table 1 presents results for Rio de Janeiro, indicating
that, of the city's approximately 13.8 million inhabitants, nearly 11.41 percent (including
children under 5 years old,3 never had any kind of formal schooling when the household
survey was undertaken, a surprising amount when considering the per capita GDP of Rio
de Janeiro. However, they also reveal the important changes that have taken place in the
Brazilian educational system, as suggested by the almost universal coverage for children at
the youngest ages in Rio and elsewhere. Other results show that:
I This ad hoc procedure is justified by the lack of a structural model describing the theoretically relevant relationships. Thus, it
is entirely based on our economic intuition and conditioned by the availability of data. Note that for a student to have a non-
dropout status (value 0) he or she must either still be in school or have left school after completing a given number of years.
'The proportion of children in the 0 to 4 age bracket who have never attended school is not necessarily 100 percent, of course.
There are 841,000 children in this group in Rio de Janeiro. See Table 6.
28    1



Determinants of Educational Exclusion
in Five States of Brazil
* Only 2.9 percent of young people aged 5 to 25 have never been to school.
* The proportion of over-26 years old who have never attended school is 7.1 percent, a
figure that is low compared to the other states considered.
* Only 0.9 percent of those in the cohort age 7 to 17 had never attended school.4
Table 1 also shows that the proportion of children who had never attended school is higher
in the early and later cohorts than in the middle. This means that pre-primary education
(nursery school, kindergarten) does not yet cover all 5 to 6 years olds. This appears very
clearly in the figures shown ahead. The standard error of these estimates is typically smaller
than 5 percent, which is the standard level of confidence adopted in this study.
s!   I          6L ! 6    .0   6  ie  !      I; '
N. of people                      Never attended school
Abs. (1.000) 0|  o in State   Abs. (1.000)  Po in age bracket |   in State
State
Total         13,837           100           1,579           11.4
Age Group
5 - 25         4,878          35.3           143.3           2.9            9.1
5 - 17         3,003          21.7           116.3           3.9            7.4
7 - 17         2,568           18.6           23.9           0.9             1.5
26 or >        8,319          60.2           592.5           7.1            37.5
Source: PNAD, 1999; Standard Error ranges from .01 to .02
Education indicators for the State of Maranhao are displayed in Table 2 below. Results show
that, among the approximately 5.43 million inhabitants of Maranhao in 1999, nearly 23.4
percent (or 1.27 million people) had never had any kind of formal schooling. When the data
was disaggregated by age bracket, it showed that:
* Approximately 6 percent of children and young people aged 5 to 25 years old have never
been to school.
* The proportion of those over 26 years of age who never attended school is 28.2 percent,
substantially higher than the 6 percent for the younger cohort.
* Approximately 2 percent of those in the 7 to 17 cohort have never attended school. This
represents more than twice the percentage for the same age group in Rio de Janeiro.
This may be linked to the fact that Maranhao has the highest percentage of rural
dwellers among the states investigated (41 percent as opposed to 4 percent in Rio de
Janeiro, for instance).
4 Just for comparison, the same figure for Rio Grande do Sul, one of the leading Brazilian states in terms of the UN Human
Development Index, is nearly the same at 0.85 percent.
1 29



Children and YouthVulnerability Primeira Edi9io
>~ .       Poverty, Exclusion and Social Risks in Five Brazilian States
N. of people                       Never attended school
Abs. (1.000)       i in State   Abs. (1.000  0 in age bracket   0 in State
State
Total           5,433           100            1,269           23.4
Age Group
5 - 25          2,631           48.4           149.1            5.7            11.7
5- 17           1,872           34.5            86.5            4.6            6.8
7-17            1,620           29.8           37.9             2.3             3
26 or >         2,197           40.4           619.5           28.2            48.8
Source: PNAD, 1999; Standard Error ranges from .01 to .02
The proportions of children who never attended school are substantially higher in
Maranhao than in Rio de Janeiro for all individual cohorts. However, the standard error
of these estimates is consistently higher than the 5 percent confidence interval (in fact, it
reaches 9 to 10 percent for the older groups), which leads to a rejection of the percentage
estimates for individual cohorts.
The data for Rio Grande do Norte, presented in Table 3, reveals a higher educational
performance than Maranhao, in that 17.2 percent of the state's population never attended
school, compared to 23.4 percent in Maranhao. In the 7 to 17 cohort, the percentage is
1.17 percent, which is on the same order of magnitude as in Rio de Janeiro (0.93 percent).
The levels of confidence in the percentages of individual cohorts are similar to those in
Maranhao due to the fact that there are very few observations, zero in some cases.
N. of people                      Never attended school
Abs. (1.000) 0| in State       Abs. (1.000)  |  in age bracket 0| in State
State
Total            2.662            100           458,6           17.2
Age Group
5 - 25           1.179           44.3           38,9            3.3             8.5
5 -17             763            28.7           24,1            3.26            5.3
7 - 17            640            24.0            7,4            1,26            1,6
26 or >          1.257           47.2           217,7           17,3           47.5
Source: PNAD, 1999:; Standard Error ranges from .01 to .03
The results for the states of Paraiba and Pernambuco, shown in Tables 4 and 5, are similar
to those of Rio Grande do Norte. This leads to the conclusion that Maranhao is the
Northeastern state with the poorest results in terms of the relative number of individuals
30 1



Determinants of Educational Exclusion
in Five States of Brazil
who never attended school. This may be due to the state's low per capita income and high
rural/urban population ratio, as suggested in our equations later on in this chapter.
N. of people                      Never attended school
Abs. 1.000)    0 in State    Abs. (.000)   0O in age bracket  0 in State
State
Total           3.381           100           458,6           17.2             -
Age Group
5 - 25          1.449           42.9           52,9           3.7             8.5
5- 17            955            28.2           35,8            3.8            5.8
7- 17            821            24.3           14,6            1.8            2.4
26 or >         1.601           47.4          299,4           18.7           48.1
Source: PNAD, 1999; Standard Error ranges from .01 to .02
I~~~~~            * '                   S''L'  ;'i'ii^  1
N. of people                     Never attended school
Abs. (1.000)    % in State    Abs. (1.000)  I % in age bracket  % in State
State
Total             7.594          100      I     1,498          19.7
Age Group
5 - 25            3.312          43.6          178,7           5.4           11.94
5 - 17            2.183          28.7          125,8           5.8            8.4
7 - 17            1.867          24.6           55,7           3.00           3.7
26 or >           3.497          46.1          678,1           19.4          45.3
Source:PNAD, 1999; StandardErroris.01
However, the percentages of young people aged 7 to 17 who never attended school are
higher in Paraiba (1.78 percent) and Pernambuco (2.98 percent) than in Rio Grande
do Norte (1.16 percent), and are on the same order of magnitude as in Maranhao (2.34
percent). Thus, the current performance of Rio Grande do Norte's educational system has
been slightly better than other Northeastern states.
To complement previous findings, Table 6 presents for each state the number of children
aged 0 to 5 who have never attended school, who are presently in school, and who have
dropped out. Approximately 80 percent of the children under the age of 5 have never been
to school, while mostly all the remaining children are currently in school. Although in
absolute terms the sample is small, it is nonetheless intriguing that approximately 13,800
children in each state have already dropped out of school at such an early stage.
1 31



Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
Number of children           Never been to school                    In school                Dropouts
in Abin     Abs.         0 v   Stan)dard g. in  Abs          % g   Standarn d  0 in  AW
11.000)  age  state  (1.000)  in)  ar   error   state  1100    (n)  age     affr    state  .01
in  group                    ~~~~~~~~~~group                __group_                        _
RJ    1.064   100    7.7    841   1649    79     0.00    53.3    219   424    21     0.00    5.8    2.7     5
MA    605     100    11.1   501    608    83     0.00    39.5    103   125    17     0.00    5.0    0.8     1
RG    270     100    10.2   205    416    76     0.00    44.7    63    128   23      0.00    6.9    2.5     5
PB    331     100    9.8    270    536    82     0.00    43.4    57    113    17     0.00    5.0    3.5     7
PE    784     100    10.3   641   1938    82     0.00    42.8    139   499    18     0.00    5.6    4.3    22
Note: t Including nursery
Dropout
The scale of the dropout problem is much larger than that of people who did not attend
school, and represents almost two thirds of the excluded for both reasons in the 5-25 age
group. In this age group, more than a third of the population is excluded from education for
one of these reasons. Table 7 presents data on the number of young people aged 5 to 25, 5 to
17, and 7 to 17 in all five states who have left school at some point in their education.5
5The totals in the first line for each state have little meaning; they include, for instance, people who graduated in any year
before 1999.
32     1



Determinants of Educational Exclusion  |
in Five States of Brazil
S    Totals           Absok00    Sampfe (n) % in age bracket Standard error  % in state
_        _ _  _ 0)deviation                   _
Rio de Janeiro          8,486       16,423        61,35         0,00
5to25                   1,414        2,780        29,01         0,01         16,67
5 to 17                  163         329          5,43          0,01          1,92
7 to 17                  158         320          6,16          0,01          1,86
Maranhao                2,122        2,576        39,06         0,01
5 to 25                  626         757          23,70         0,02         29,39
50to 17                  133         162          7,13          0,02          6,29
7to 17                   133         162          8,24          0,02         6,29
Rio Grande do Norte     1,283        2,605        48,21         0,01
5 to 25                  330         670         28,00          0,02        25,72
5 to 17                   61         123          7,94          0,02         4,72
7 to 17                  58          118          9,08          0,03         4,53
Paraiba                 1,618        3,202        47,73         0,01     _
5 to 25                  366         727          25,28         0,02         22,70
5to 17                    65         129          6,81          0,02         4,03
7 to 17                  63          126          7,73          0,02         3,94
Pemambuco               3,632       12,675        47,86         0,00
5 to 25                  927         3,061        28,00         0,01         25,51
5 to 17                  200         603          9,19          0,01          5,52
7 to 17                  190         560          10,20         0,01          5,24
As demonstrated in Box 1, an estimate of 603,000 school-age children and youth, that is,
those between the ages of 7 to 17, are no longer in school but have attended in the past. In
the state of Rio de Janeiro, there are 158,000 youngsters in this situation, the equivalent
to 6.12 percent of those in the age cohort. In the Northeastern states, the proportions are
all higher. The highest dropout rates are found in Pernambuco, both in percentage and
absolute terms, where 190,000 school-age youngsters, or 10.2 percent of the age group, have
dropped out of school.
33



Children andYouthVulnerability Primeira Edig5o
Poverty, Exclusion and Social Risks in Five Brazilian States
Box 1: Percentage of Students no Longer Attending School in 1999
Pernambuco                         10,2 (190,000)
Rio G. do Norte                   9,08    (58,000)
Maranhao                          8,24  (133,000)
Paraiba                           7,73    (64,000)
Rio de Janeiro                    6,12  (158,000)
Total Number                         603,000
Table 8 presents two measures of dropouts: (i) children and young people aged 7 to 17
who dropped out of school before completing eight years of formal education and (ii)
those who dropped out after completing fundamental, or primary, education. A crucial
difference exists between the Southeastern and the Northeastern states in terms of these two
categories. In the Northeastern states, nearly all children and youth who left school did so
before completing fundamental school, while in Rio de Janeiro, only about 10 percent did
so after completing fundamental school.6
Not completed             Completed            % not completed/
State            Primary School             Primary                 Total 0
(1,000)              School (1,000)
Rio de Janeiro            143.7                   14.4                    90.9
Maranhao                  131.0                    2.5                    98.2
Rio G. do Norte           56.7                     1.5                    97.5
Paraiba                   63.0                     0.5                    99.2
Pernambuco                185.6                    4.7                    97.5
Overage Enrollment
One of the possible reasons why children leave school permanently is the increasing age-
grade gap, defined as the difference between the grade in which the student is presently
enrolled and the one appropriate for their age. When a large age-grade gap occurs, the
student often becomes discouraged and loses motivation to learn. For the purposes of this
report, overage enrollment is defined as any age-grade lag greater than two years.
° When examining this and subsequent tables, especially Tables 9 to 13, the reader should be aware of the possibility of
sampling and, especially, reporting errors.
34



Determinants of Educational Exclusion
in Five States of Brazil
The problem of overage enrollment is very severe in all five states. Altogether, a total of 2.8
million students are affected. For the school-age population (7 to 17 years old), age-grade
gaps of more than two years represent between 25 and 30 percent of total enrollment in
the Northeastern states and 15 percent of total enrollment in Rio de Janeiro. Although it is
widely accepted that repetitions of the same grade often lead a student to drop out, it is not
clear after how many repetitions the dropout occurs.
The following sets of tables and graphs present, for each state, the percentages of
overage students by age cohort and by the number of years by which the students'
present grade differs from the appropriate grade for their age. The following main
results were identified:
Figure 1: Percentage of Overage Enrollment in Rio de Janeiro by Age Group
25
20
. 5            . . .
%   10       "._
..................
5
0
11         13        15         17  Age    19        21         23         25
Age-Grade Gap:  -- Gap =3      -   Gap =5           Gap =7        -  Gap =9
Cohort   3 years gap 4 years gap 5 years gap 6 years gap 7 years gap 8 years gap 9 years gap
11        10.19
1 2       10.88      5.25
13        8.19       5.08       2.77
14        14.63      4.77       3.57       1.38
15        9.86       9.55       4.00       3.14       0.83
16        10.79      8.09       4.58       1.61        1.20      0.74
17        11.60      7.18       3.75       2.81        1.47       1.37      0.64
18        7.25       6.28       4.03       2.49       2.52       0.79       0.42
19        7.34       2.59       3.96       1.57        1.64      0.69       0.85
20        6.85       4.39       2.33        1.94       1.23       0.39       0.78
21        3.30       6.27       4.47        1.50       1.27       0.17       0.74
22        3.11       3.61       3.58       3.40        0.86       1.00       1.25
23        1.71        1.52      2.21       2.91        1.20       0.87       0.94
24        2.13       2.28        1.37       1.25       1.33       1.71       0.99
25        1.04        1.67      0.78       0.19        1.15       0.96       1.81
35



Children and YouthVulnerability Primeira Edi,io
Poverty, Exclusion and Social Risks in Five Brazilian States
* In the state of Rio de Janeiro, Table 9 shows that 10.2 percent of 11-years-old students are
three years older than the appropriate level of education for their age. That is, a student
who should be in the 4th grade is instead enrolled in the 1st grade. The three-years age-
grade gap for students aged 11 to 17 is about 10 percent of the relevant age bracket with
exception of the 14 year olds, which is a little higher. However, the percentage decreases
thereafter as the age of the students increases.7 As expected, the highest percentages are
for those with an age-grade gap of three-years, with declining proportions for the higher
age-grade gaps, notwithstanding the fact that the gap is very high for young people aged
12 to 18. Even so, it is intriguing to find that in a better-off state such as Rio de Janeiro,
1 to 2 percent of the 22 to 25 year olds are still in school with an age-grade gap of nine
years.'
* In Maranhao, nearly 18 percent of the 12 year olds are four years behind the grade
expected for their age (as represented by the solid dark line), and approximately 20
percent of the 11 year olds had repeated a class at least once.
7The total age-grade gap for a given cohort is simply the sum of all overage enrollment for that age. These totals are
shown in Tablel4.
This may be due to adult education.
36 1



Determinants of Educational Exclusion
in Five States of Brazil
Figure 2: Percentage of Overage Enrollment In Maranhio by Age Group
25
20  ..
% 1 5
10
0
11        13         15         17  Age   19        21         23         25
Age-Grade Gape:  -... Gap =3     -     Gap =5    -- Gap=7           -  Gap =9
Cohort   3 years gap 4 years gap 5 years gap 6 years gap 7 years gap 8 years gap 9 years gap
11        19.89
12        16.84     17.90
13        13.69     13.68      13.15
14        18.68     15.94      10.44       6.59
15        11.05     14.92       8.29      13.26      8.29
16        9.63       8.56      13.37       9.09       5.88      2.14
17        8.64       4.94       6.17       6.79      4.94       6.17       1.85
18        2.82       8.45       9.86       2.82       2.82      4.22       2.82
19        3.82       2.29       3.82       1.53       0.76      3.82       2.29
20        3.01       3.01       3.01       3.76       2.25      2.26       0.00
21        0.00       3.00       4.00       1.00       4.00      2.00       2.00
22        0.00       0.91       2.73       2.73       2.73       0.91      0.00
23        0.00       1.90       2.86       3.81       0.95       0.00      1.91
24        0.00       1.02       0.00       0.00       2.04       1.02      2.04
25        0.98       0.98       2.94       0.00       1.96       0.98      0.98
In Paraiba, as demonstrated in Figure 3 by the dotted line, 23.4 percent of the 1 year olds
are three years behind the level expected for their age, the highest such percentage in all
of the five states. Also, a high percentage of the 12 year olds, nearly 17 percent, present an
age-grade gap of 4 years.
37



Children and YouthVulnerability Primeira Edi,io
Poverty, Exclusion and Social Risks in Five Brazilian States
Figure 3: Percentage of Overage Enrollment in Paraiba by Age Cohort
25
20
15
%   10
S            ....                                    .
0
ll         13         15        17   Age              21         23         25
Age-Grade Gap:  .-.. Gap =3     -     Gap =5    -- Gap =7          -   Gap =9
Cohort   3 years gap 4 years gap 5 years gap 6 years gap 7 years gap 8 years gap 9 years gap
11        23.38
12        15.79      16.45
13        15.92      14.65       8.92
14        13.51      10.81      10.13       4.05
15        10.91      12.12      10.91       5.45       9.70
16        9.93       6.62       11.26       4.64       8.61       3.31
17        5.15       6.62        5.88      11.77       7.35       2.94       1.47
18        4.03       2.01        6.04       6.71       3.36       2.68       3.36
19        2.84       4.96        4.26       2.13       4.96       4.26       5.67
20        4.00        5.60       4.80       3.20       5.60       4.00       1.60
21         1.77       3.54       5.31       3.54       2.65       0.88       1.77
22        3.23        4.03       1.61       4.84       2.42       3.23       0.81
23         1.80       0.00       1.80       2.70       1.80       0.00       1.80
24        0.00        0.00       0.00       0.00       0.95       1.90       2.86
25        0.88        1.75       0.88       0.88       1.75       1.75       0.00
* In Pernambuco, a high percentage of the 11 year olds (18.4 percent), have an age-gender
gap of three years. In addition, 2.26 percent of the 25 year olds are nine years behind the
grade expected for their age, which is the highest proportion in all states.
38    I



Determinants of Educational Exclusion
in Five States of Brazil
Figure 4: School Attendance in Pernambuco by Age Cohort
25
20
1 5
% 10             \"                 "
5                                       <
0
11         13        15         17  Age    19        21         23         25
Age-Grade Gap:  - - Gap =3      -   Gap =5           Gap =7        -  Gap =9
Cohort   3 years gap 4 years gap 5 years gap 6 years gap 7 years gap 8 years gap 9 years gap
11        18.40
12        14.84      13.02
13        14.43      9.80       9,94
14        15.57      9.48       6.83       4.81
15        11.66      12.08      10.91      6.66       2.95
16        8.60       11.60      10.91      4.72       2.62       3.63
17        9.64       7.83       6.63       6.52       3.55       2.32       1.25
18        6.37       7.69       4.52       4.19       3.92       4.04       1.84
19        5.50       6.75       4.22       2.79       3.94       2.67       0.28
20        5.32       4.26       3.87       3.67       2.89        1.52      3.46
21        1.99       3.95       3.69       2.21       2.69       2.63        1.21
22        0.58       1.76       3.46       3.61       2.96        1.26       1.70
23        0.82       0.88       1.84        1.55      2.76       0.96        1.63
24        1.11       1.18       0.29       0.59       2.46        1.70      1.41
25        1.04       0.75       1.13       0.15       0.89        1.43      2.26
In Rio Grande do Norte, almost 20 percent of 12 year olds are three years behind the
level expected for their age, and almost three percent of the 17 year olds are nine years
behind in their education.
39



Children and YouthVulnerability Primeira Ediaio
Poverty, Exclusion and Social Risks in Five Brazilian States
Figure 5: Percentage of Overage Enrollment in Rio Grande do Norte by Age Group
25
20
1 5
%   10
5
0
11         13         15        17   Age   19         21         23         25
Age-Grade Gap:  .-.- Gap =3     -     Gap=5     -- Gap=7            -  Gap =9
Cohort   3 years gap 4 years gap 5 years gap 6 years gap 7 years gap 8 years gap 9 years gap
1 1       17.86
12        19.51      8.13
13        15.91      16.67      7.57
14        15.39      15.38       7.69       5.98
15        15.93      7.08       6.19        9.73       0.89
16        8.48       6.78       13.56       4.24       5.08       0.85
17        9.42       9.42       6.52        7.97       1.45       2.17       2.90
18        6.96       5.22       3.48        6.09       3.48       0.87       0.00
1 9       8.76       5.84       7.30        2.19       3.65       1.46       2.19
20        4.31        5.17       5.18       3.45       0.86       1.72       2.59
21        2.15        0.00       3.23       1.08       3.22       1.08       2.15
22        0.00        1.02       3.06       4.08       5.11       3.06       1.02
23        0.98       0.98        2.94       3.92       3.92       2.94       0.98
24        0.00        1.08       2.15       1.07       4.30       1.07       1.08
25        1.11        1.11       1.11       1.11       2.22       3.33       1.11
Table 9 presents a summary of overage enrollment in each state, displaying totals for age-
grade gaps of three to nine years. All information is significant at 1 or 2 percent. As expected,
the proportion of students in the Northeastern states who are enrolled in a lower grade than
the appropriate for their age is much higher than in Rio de Janeiro. This widespread overage
enrollment has a strong negative impact on the self-esteem of these children, often leading
them to drop out of school with little education. It also has a negative effect on the costs of
the system, for it incurs more expenses when a student repeats a grade.
40 1



Determinants of Educational Exclusion
in Five States of Brazil
School       Absolute             OG attending  Standard  0o In  Standard
Attendance     Numbers   Sample(n)               error    total      error   in state
Rio de Janeiro    948,522     1,855      25.18      0.01     6.86    .0.01
From 5 to 25      718,071     1,411      21.64      0.01    14.72     0.01     75.70
From 5 to 17      394,320     785        14.49      0.01    13.13     0.01     41.57
From 7 to 17      394,320     785        16.54      0.01    15.36     0.01     41.57
Maranhao          811,401     985        39.75      0.02    14.94     0.01
From 5 to 25      731,501     888        39.36      0.02    27.80     0.02     90.15
From 5 to 17      570,042     692        34.50      0.02    30.44     0.02     70.25
From 7 to 17      570,042     692        39.34      0.02    35.18     0.02     70.25
Rio G. do Norte   303,436     616        32.99      0.02    11.40     0.01       -
From 5 to 25      256,645     521        31.69      0.02    21.77     0.02     84.58
From 5 to 17      161,577     328        23.82      0.02    21.18     0.02     53.25
From 7 to 17      161,577     328        28.13      0.02    25.25     0.02     53.25
Paraiba           427,890     849        37.39      0.02    12.66     0.01       -
From 5 to 25      370,436     735        35.96      0.02    25.56     0.02     86.57
From 5 to 17      246,452     489        28.87      0.02    25.82     0.02     57.60
From 7 to 17      246,452     489        33.18      0.02    30.02     0.02     57.60
Pernambuco        873,612     2,775      35.53      0.01    11.50     0.01       -
From 5 to 25      757,446     2,362      34.37      0.01    22.87     0.01     86.70
From 5 to 17      491,359     1,449      26.49      0.01    22.51     0.01     56.24
From 7 to 17      491,359     1,449      30.33      0.01    26.32     0.01     56.24
In Rio de Janeiro, approximately 949,000 students, or 25.2 percent of all those in school,
were at least three years behind the appropriate grade. In Maranhao, 811,000 young people,
nearly 40 percent all students in school, were over age in 1999. This means that the children
of this state are being very poorly served by their education system. In Rio Grande do Norte,
the proportion was 33 percent, or 303,000 students, while in Pernambuco it was 874,000
and in Paraiba nearly 428,000. What is more worrisome is that the percentages of overage
students in the 7 to 17 years old group are also very high in the Northeastern states: 39
percent in Maranhao (570,000), 33 percent in Paraiba (246,000), 30 percent in Pernambuco
(491,000), and 28 percent in Rio Grande do Norte (162,000), again as compared to Rio de
Janeiro's 16.5 percent (394,000).
41



Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
It is generally accepted that, if a student has to repeat a grade more than once, he or she
is likely to drop out of school eventually. However, it is not clear after exactly how many
years of repeating the same grade the student is likely to leave school permanently. Tables
8 to 14 do not provide the answer to this question, as they show overage enrollment as
a continuum. However, they do help determine what proportion of youth in the 7 to 17
cohort are enrolled in the "correct" grade,9 defined as those who are up to two years behind
their age-appropriate grade. This is achieved by subtracting from the total percentage
the proportions of overage enrollment, those who have never attended school, and the
dropouts. Thus, for Rio de Janeiro, the result is 76.4 percent (100 - 16.54 - 0.93 - 6.16), 50.1
percent for Maranhao, 61.6 percent for Rio Grande do Norte, 57.3 percent for Paraiba, and
56.5 percent for Pernambuco.
The Dropout Process
This section presents a graphical analysis of the dropout process in each of the states
analyzed. It is very useful in analyzing the dynamics of the school attendance cycle, since
the interactions and effects of the three aspects of educational exclusion-age-grade gap,
dropout rates, and school attendance-are exposed.
As represented in Figure 6, school attendance in Rio de Janeiro is almost universal for
students aged 7 to 11, but begins to decrease quickly from age 11 onwards. For instance,
while 65 percent of the 18 year olds attend school, only 44 percent of the 19 year olds are
still in school, implying that, given the small age difference, approximately 21 percent of
students either graduated or left school between those ages. Figure 6 also shows the curve
for the age-grade gap of more than two years (dashed line). The increase is so steep from 9
years old onwards that, when it reaches the 15 years old group, 28 percent of them are more
than two years behind the expected level for their age.
The dotted line displays the dropout pattern for those young people who have left school
without completing primary school. The percentages are very small (indeed, negligible) up
to the 11-12 year olds, as already suggested. From then on, the curve rises steeply, following,
9 An estimate of the proportion of pupils repeating a class before the age of 16 in France is illustrative. Although the concepts
are not the same as Brazil's: "... only 42.8 percent of them have not repeated a class before the age of 16, and 53.7 percent repeat
a class once or twice'" Magnac and Thesmar, 1998 (p. 18). The average ratio of those who have not repeated a class and those
who have repeated is nearly 0.8 and increases from year 6 (of the first grade) on upwards. This suggests that repetition rates
accelerate as students get older.
42



Determinants of Educational Exclusion  f.
in Five States of Brazil  L.9
with a lag of two to three years, the overage enrollment curve. The proportion of 17 years
old who left school before completing primary education is 16.1 percent - approximately
one out of six.
Figure 6: School Attendance, Age-Grade Gap and Dropouts in Rio de Janeiro
0.9
0.8
0.7
0.6                                                      --- *-  -      . -..
0.5
% 0.4
0.3                               - ,.    -
0.2
0.1
0
5           10           15           20          25           30
Age
-Attend    - - Age-grade gap> 2ys  *.- Dropout < 11 years old -Dropout 11 or older
Figure 7 shows the percentages of all students currently attending school, those who have
never attended school, and young people who have attended school at some time in the
past with no additional constraints. These curves add up to 100 percent at all points. In
Rio de Janeiro, school coverage is almost universal for school-aged children, as shown
by the dotted line. However, starting at age 15, the proportion of students that have
dropped out of school increases dramatically until age 22 from (8 to 69 percent), and
then increases at a slower rate.
Figure 7: School Attendance in Rio de Janeiro
0.9…
0.8
0.7
0.6
0.5
% 0.4
0.3
0.2
0.1
5           10           15           20          25           30
Age
-Never Attended           -     Attend           -.-- Attended at some point (pure dropout)
43



Children and YouthVulnerability Primeira Edi,io
Poverty, Exclusion and Social Risks in Five Brazilian States
Figure 8 shows the percentages of young people who are currently attending school with
and without any age-grade gaps. Note here that, beyond the age bracket of 19 to 20 years
old, the percentages of the former group are larger than those of the latter. Note also that the
four curves add up to 100 percent for all cohorts: (a) He may have attended school in the
past but not anymore (the broadest measure); (b) She may have left school with less than
eight years of education; (c) He may have left school with less than 11 years of education;
(d) She may have left school after having completed at least eight years of school (having
completed primary school); and (f) He may have left school after having completed at least
11 years of school (having completed secondary school).
Figure 8: School Attendance and the Age Grade Gap in Rio de Janeiro
0.9
0.7                                                       -               3
0.6
% 0.5
0.4
0.3
0.2- 
0.1
5           10          15          20          25          30
Age
-     Never attended        --- Attended but dropped out
-     Attended expected year    Attended but are overage
The process of educational exclusion, with emphasis on the dropout process as it relates to
school attendance patterns, is demonstrated for the four Northeastern states in the graphs
below:
The educational process for Maranhao is represented in Figure 9. The graph illustrates the
percentage of children and youth, according to age, who attend school (dotted line), who
have never attended school (solid line), and those who attended school at some point but
have dropped out (dashed line). Once again, the three curves add up to 100 percent for
all age cohorts. In Maranhao, attendance rates are highest between the ages of 7 and 15,
corresponding to some degree to the secondary level of education. In fact, the pattern present
in all five states for children in this age bracket is a nearly universal educational coverage
accompanied by low non-attendance and dropout rates. The proportion of students who
drop out of school in Maranhao rises sharply between the ages of 15 and 27 (from 10 to 66
percent), then continues to increase, albeit at a slower rate. At age 24, 14 percent of those in
that age group have never attended school, as indicated by the solid line.
44 1



Determinants of Educational Exclusion
in Five States of Brazil  J
Figure 9: School Attendance in Maranhio
0.9
0.8
0.7
0.6
%  0.5
0.4
0.3
0.2
0.1---- 
0
5            10           15           20           25            30
Age
-   Never Attended          -   Attend           - - - Attended at some point (pure dropout)
As demonstrated in Figure 10 by the dashed line, dropout rates in Paraiba are very low
until the age of 13 (it goes up to three percent), at which point it begins to rise sharply and
steadily until the age of 30. School attendance is highest from ages 6 to 13.
Figure 10: School Attendance in Paraiba
0.9
0.8
0.7
0.6
%  0.5
0.4
0.3
0.2
0.1
0
5            10          I15           20           25            30
Age
-  Never Attended         -    Attend            - - - Attended at some point (pure dropout)
Unlike the other four states analyzed, Pernambuco's dropout rates are high even in the
earlier years, that is, for varying reasons, children dropout of school during the fundamental
cycle of education. This pattern is evident starting with the 4 year olds, whose dropout
rate is 3 percent. Then, starting at age 12, there is a steady increase in the incidence of
dropout cases. Also, it is worth noting that in terms of educational coverage, Pernambuco's
educational coverage presents the poorest performance of all states analyzed.
1 45



Children and YouthVulnerability Primeira Ediaio
Poverty, Exclusion and Social Risks in Five Brazilian States
Figure 11: School Aftendance in Pernambuco
0.9          An                  -    e                 - -
0.8                                                              -
0.7
0.6
0.5
0.4
0.3 
0.2
0.1 
0
5             10            15           20            25            30
Age
-Never Attended             -Attend                - - Attended at some point (pure dropout)
As demonstrated in Figure 12, the dropout pattern in Rio Grande do Norte is similar to
those of Rio de Janeiro and Maranhao: there is a very low incidence of dropout cases in the
early years (until age 13, it does not surpass 2 percent), then a sharp increase in dropouts
from age 13 until 22, and then the incidences continue to rise, but the increase is much more
moderate. In fact, for students between the ages of 23 and 25, the cases of dropout remain
almost constant at about 72 percent.
Figure 12: School Antendance In Rio Grande do Norte
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1                                - --
0
5             10            15           20            25            30
Age
-Never Attended             -Attend               - - - Attended at some point (pure dropout)
46



Determinants of Educational Exclusion
in Five States of Brazil
One important conclusion that can be drawn from this graphical analysis refers to the age
at which the dropouts increase noticeably. Except in Pernambuco, where dropout rates are
apparently high even in very early years, the percentages of dropouts begin to increase from
12 years old onwards. Typically, dropout rates increase from 1 percent or less to 2 percent or
more between the ages of 11 and 12 years old. Not surprisingly, this is the same age bracket
in which overage enrollment is highest in most Brazilian states.
One important conclusion to be drawn from this graphic analysis is that the dropout process
is more intense from age 12 onwards. Typically, going from age 11 to age 12 increases the
dropout probability from 1 to 2 percent.
Explaining Dropouts
When children drop out of school, this may be due to factors within the child's household
and to factors that pertain to the school system. Our descriptive analysis so far indicates that
the probability of dropping out depends on the child's age (the older they are, the higher the
dropout rate), and on which state he or she lives in (the lowest rates are in Rio de Janeiro,
the only Southeastern state in our analysis). Below, other household variables are added
into the model to provide a more comprehensive view of the constraints to education that a
child may face that are not strictly related to the quality of the educational system.
First, in order to analyze what factors are important in determining dropout rates in each
state, a separate regression was run for each state for three different cohorts: (i) 7 to 9 years
old; (ii) 10 to 14 years old; and (iii) 15 to 17 years old. Second, in order to understand what
each of the states may add to the likelihood of dropping out, all the states were pooled in
one regression, with each state being represented by a dummy variable.'° Third, separate
regressions pooled data for boys and for girls to discover whether different factors affect
their respective dropout rates. And fourth, the same exercise was carried out for whites and
non-whites to see if there are peculiar factors affecting their respective dropout rates. The
dependent variable adopted in Logit and Probit models is "dropout," which represents those
children aged from 7 to 17 who left school with less than eight years of formal education
(in other words who left school without completing primary school). Logit, Probit,
and Ordered Probit models were used in which the parameters were adjusted using the
' After deducing missing values for the explanatory variables, the sample sizes by state are: 4,826 in Rio de Janeiro; 1,914 in
Maranhao, 1,273 in Rio Grande do Norte, 1,593 in Paraiba, and 5,429 in Pernambuco.
1 47



(  )  | Children and YouthVulnerability Primeira Ediaio
Poverty, Exclusion and Social Risks in Five Brazilian States
Maximum Likelihood Method of estimation. The units of analysis were children and young
people in the relevant age brackets. Some explanatory variables expressed the characteristics
of their households. The following variables were used in the estimation with the expected
coefficient signs appearing between brackets after the respective symbols:
Expected
Name        Coefficient                        Definition
Sign
RURAL               +       Location of the household, rural assuming the value 1
Household structure, indicates whether only one (value =
SINGLE              +       1) or both parents live in the household
INCOME                      Household income per capita in logs
Parent's education, The highest number of years of
PERENEDU            -       education completed by either father or mother.
DEPRATIO11          +       Household dependency ratio: No. of dependents per
working adult in the household.
SIBLINGS            +       No. of siblings
NONWHITE            +       Indicator of race, non-white assuming the value 1 (the
value 0 for this variable includes Indigenous peoples)
GENDER              ?       Gender of the child, female assuming the value 1
WORKS59             +       Child labor. Children 7-9 years old working received the
WORKS59      +      ~~~value 1
WORK 10             +       Adolescents 10-17 years old working received the value 1
MIGEST              +       Migrant born in another state assumes the value 1
MIGMUN              +       Migrant born in another municipality but same state
assumes the value 1
Separate regressions were run for each different model, state, and age, with both weighted
and unweighted data. As unfortunately the results obtained for the 7 to 9 cohort were not
robust because the small sample yielded unstable regression coefficients, this group was
pooled with the 10 to 14 group in separate estimations. In this case, however, the work
variable could not be used (see footnote 12).
The regression coefficients that yielded statistically significant results (significantly different from
zero) are shown in the next three tables, one for each cohort. An X indicates any coefficient that is
significant at the 5 percent (or greater) level of confidence. When the level of significance is only a little
higher than 5 percent, it is indicated by the number of the estimated confidence level.
" This variable was not used in the final estimation because all of the relevant information is already conveyed in both the
Single and Siblings variables.
48 a



Determinants of Educational Exclusion  |
in Five States of Brazil  |
Determinants of Dropouts for Children aged 10 tol4
Three conclusions clearly stand out from Table 11. First, the models with unweighted
and weighted variables yield essentially the same results, in terms of the significance
of the coefficients. Second, the Logit and Probit models also yield the same results
with respect to the significance of the regression coefficients. Third, the regression
coefficients on the Location (rural/urban) and Gender variables were not significant in
any regression equation.
S. ff32kM I. m-   ..     0                 -   .~ 
Significance of estimated coefficients
Single  Income   Parenedu  Siblings  Nonwhite  Won    Migmun   Migest
10
L    P   L    P   L    P   L    P   L    P   L    P   L    P   L    P
RJ  Unweighted            X    6            X    X             X   X
Weighted              X    X            X    X   X    X   X    X
MA   Unweighted                                                X   X    X   X    X   X
Weighted                                                  X    X   X    X   X    X
RN   Unweighted            ?   ?    X                          X   X
Weighted              ?    ?   X                          X    X
PB  Unweighted             X   X    X                                   X   8
Weighted              X   X   X                            _       X    9
PE  Unweighted    X   X    6   X    X        6                 X   X    6   7
Weighted     X    X   6       X     X   X    X            X    X   7    6
Notes: (1) coefficients on the rural and gender variables were not significant in any equation; (2) L = logit; P probit
(3) Indicates significant but with the opposite sign; (4) numbers indicate level of significance if higher than 5 percent (but lower than
10 percent).
In four equations (all for the state of Rio Grande do Norte), the estimated coefficients for
the household income variable yielded significant results -but with the opposite sign from
what was expected! The positive estimated coefficients suggest that, in Rio Grande do Norte,
the rate of school dropouts for the cohort aged 10 to 14 increase with increasing income.
One possible explanation for this may be the size of the household; in small households, a
child is often needed to help with work at home.
49



Children andYouthVulnerability  Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
Other results include:
* Single-headed households have a significant effect on increasing dropouts in
Pernambuco.
* Ethnicity is significant in Rio de Janeiro, indicating that being a non-white youth has a
negative impact on the probability of a child becoming a school dropout.
* Being an inter-state migrant in Maranhao increases the probability of a child becoming
a dropout.
* Child labor, parents' education, migration, and household income are the variables that
are statistically significant in all five states.
Determinants of Dropouts for Children aged 7 to 14
Table 12 presents the corresponding results for the 7 to 14 cohort. The results are very
similar to those for the 10 to 14 years old cohort, including the unexpected sign for the
household income variable in Rio Grande do Norte. However, in this case, the location
variable (rural = 1) is significantly different from zero in the state of Rio de Janeiro. Gender
continues to be insignificant in all states and models.
Again, Rio de Janeiro is the only state where being non-white has a significant negative effect
on becoming a school dropout. This may be sociologically explained by the fact that Rio
has a smaller proportion of non-whites relative to the total population (38 percent) than,
for example, Maranhao (75 percent), making it more likely that some racial discrimination
may exist. Pernambuco is the only state in which belonging to a household headed by a
single parent increases the probability that a child will drop out, as expected.
50



Determinants of Educational Exclusion
in Five States of Brazil
_6~        S    -  -g -      .-g   6-    -    S     S~   .O~O    S 1   --    O@
Sificance of estimated coefficients
Rural   Sin le  Income  Parenedu Sibilings Nonwhite Migmun  Migest
L   P   L    P   L   P   L    P   L   P   L    P   L   P    L   P
RJ Unweighted   X   X   =    =      =9              X   X    xX
Weighted    X   X _           X               X   X   X    X
MA Unweighted                            X   X                     X   X   X   X
Weighted            _                  X                       X   X    X   X
RG  Unweighted                   ?   ?   X    X
Weighted                     ?   ?   X    X                        _
PB  Unweighted                   X   X   X    X   ?                6   9
Weighted                     X XX         X X X                6   9
PE Unweighted           X   X   X    X   X    X   X   X            X   X
Weighted            X    X   13  12  X    X   X   X            X
Notes: (1) coefficients on the gender variable were not significant in any equation; (2) L = logit; P = probit
(3) Indicates significant but with the opposite sign; (4) numbers indicate level of significance if higher than 5 percent, but lower than 10 percent
Determinants of Dropouts for Children aged 15 to 17
Table 13 presents results for the 15 to 17 cohort. There are some differences from the
previous equations. First, all variables have been significant in at least one equation or state.
However, as before, all models yield nearly the same information.
* Being a single-headed household is significant only in Maranhao and even there only
barely.
* Household income is significant in Rio de Janeiro, Paraiba, and Pernambuco and,
fortunately, does not appear to be wrongly significant in Rio Grande do Norte.
* The parent's education and work variables are significant in all models for all states
(albeit at a lower level of confidence in Paraiba and Rio Grande do Norte.) Together with
household income, they remain the most important determinants of dropout.
* The siblings variable, a traditional indicator of social capital at the household level, is
barely significant in Rio de Janeiro and is not significant overall.
* Being non-white is significant but only in Rio de Janeiro (although at a lower level of
confidence than among those aged 10 to 14).
* Gender is significant in Maranhao and Paraiba but especially in the former, indicating
that being female increases the probability of dropping out in these states.
* The location of the household is relevant in all models, now including the Ordered Probit
Model (see Johnston and Dinardo, 1997, p. 434) for the states of Rio de Janeiro and
Paraiba - and almost in Pernambuco.
51



X   {' |   Children and YouthVulnerability Primeira Edicio
Poverty, Exclusion and Social Risks in Five Brazilian States
Snificance of estimated coefficients
Rura   Single   Income    Pareneau  S,DIngs   Nonwh,le  Gender  Work10  Migmun   Mlgest
L     P OP L    P  L p OP L     p OP L P OP L P Op L p OP L P OP L p OP L P
RJ X   X 8           X XX     X XX X          5   5 X          I    X X X      I _    8 8
X   X  8         X X I    X  X X    7 8   X   9 8               X XX
MA             6  6  6 6 6 9  X   X X   ?                  X X X    X   X    X XX X
RG __X                           X X
PB          X   X X           X  X X                       7 7      X        XXX 
PE 8    7            XX     X  X   X                       - -      9 97     - 
Notes: (1) L = logit; P = probit; OP = Ordered Probit; (2) Indicates significant but with the opposite sign; (3) numbers indicate level of
significance if higher than 5 percent (but lower than 10 percent); (4) for each state, the first line is unweighted and the second line is
weighted
Logit Estimates for Weighted Data
Logit models estimated with the weighted data are used to analyze in greater detail the
estimated coefficients as shown in Tables 14, 15, and 16, in which only significant results
are reported. Pooling the data for cohorts 7 to 9 and 10 to 14 yields more robust results.
In some cases (cohorts), the small number of observations yielded unstable and/or non-
robust results. This happened for the 7 to 9 cohort in the states of Rio de Janeiro (where
only the siblings variable had a significant result), Maranhao (where no variable was found
to be significant), Rio Grande do Norte (where only the Parents' Education variable was
significant), Paraiba (where no variable was found to be significant), and Pernambuco
(where no variable was found to be significant).
Table 17 summarizes and organizes the information from the regression coefficients. Some
variables are significant in only a few states. Thus:
* Surprisingly, rural area location is significant in Rio de Janeiro for the 15 to 17 cohort, and
somewhat in Paraiba.
* Single-headed households have a bearing on increasing school dropouts only in Paraiba.
* Being a non-white increases the probability of dropping out of school in Rio de Janeiro.
* The number of siblings increases dropout rates only in Rio de Janeiro and Pernambuco.
* Gender is important only in Maranhao, and only for the 15 to 17 cohort.
* Inter-state migration is relevant only in Maranhao (for the younger cohorts) and in Paraiba
(for the 15 to 17 cohort).
52    1



Determinants of Educational Exclusion  r'?
in Five States of Brazil
Child labor, parents' education, migration, and the level of household income all have a bigger
effect on dropout rates than the remaining variables. However, concerning the first of these
variables, it is impossible to affirm that child labor causes dropouts or vice versa, rather only
that the two processes are associated in some way. Young people may drop out of school
because they need to work for their households or they may enter the labor market because
they failed their studies (probably after repeating the same grade several times).
-_____ Parameter Standard  Wald   Pr >    Standardized  Odd
5w   Variable      DF  Estimate  Error                      Estimate    Ratio
n                                          square   square
Rio de Janeiro
INCOME         1    -0.207    0.093    4.934    0.026     -0.129      0.813
SIBLING        1    0.379     0.082    21.361   0.000     0.265       1.461
NONWHITE       1    0.644     0.294    4.794    0.029     0.176       1.905
WORK 10        1    2.264     0.382    35.210   0.000     0.204       9.620
Maranhao
WORK 10        1    1.432     0.391    13.413   0.000     0.377       4.188
MIGMUN         1    1.545     0.398    15.056   0.000     0.290       4.687
MIGEST         1    1.723     0.612     7.927   0.005     0.191       5.599
Rio Grande do Norte
INCOME         1     (77)     0.261    4.135    0.042     0.311       1.700
(??)
PARENED        1    -0.319    0.089    12.920   0.000     -0.734      0.727
WORK 10        1    1.284     0.494     6.747   0.009     0.220       3.609
Paraiba
INCOME         1    -0.437    0.148     8.777   0.003     -0.278      0.646
PARENED        1    -0.120    0.061    3.913    0.048     -0.329      0.887
_   MIGMUN      1    1.008     0.516    3.815    0.051     0.187      2.739
Pernambuco
SINGLE         1    0.890     0.195    20.910   0.000     0.205       2.435
PARENED        1    -0.105    0.029    12.871   0.000     -0.243      0.900
SIBLING        1    0.095     0.041     5.468   0.019     0.105       1.100
WORK_10        1    0.886     0.210    17.794   0.000     0.193       2.425
153



Children andYouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
MiMe kAX       a1fWIM- Hog*g                               -
Parameter Standard  Wald      Pr>        Standardized   Odd
< Variable     DF  Estimate   Error    Chi-square Chi-square Estimate      Ratio
RJ
RURAL           1     0.647     0.277     5.448     0.020        0.084        1.910
INCOME       1    -0.296     0.075    15.628     0.000        -0.174       0.744
PARENED      1    -0.153     0.027    32.210     0.000        -0.357       0.858
WORK 10      1     0.801     0.207     15.050    0.000        0.163        2.228
M A__                                       _  _   __  _ _
PARENED      1    -0.137     0.038    13.353     0.000        -0.316       0.872
SIBLING      1    -0.2751    0.069     16.135    0.000        -0.315       0.759
(??)__                                   _  _ _ _  _
GENDER       1     0.662     0.260     6.474     0.011        0.183        1.938
WORK 10      1     1.192     0.288     17.088    0.000        0.329        3.294
MIGMUN       1     1.001     0.296    11.421     0.001        0.222        2.722
IPARENED    I 1 1 -0.116   [ 0.044  J   7.004  1  0.008  1    -0.287    1   0.891
PB
RURAL        1     0.951     0.292    10.591     0.001        0.255        2.588
PARENED      1    -0.118     0.037    10.203     0.001        -0.323       0.889
_  MIGMUN    1     0.720     0.351     4.209     0.040        0.150        2.054
PE
I INCOME    | 1 | -0.255   | 0.060   | 17.763  |  0.000  |    -0.155        0.775
| PARENED   I 1 | -0.122   | 0.020   | 36.024  |  0.000  |    -0.281        0.885
54



Determinants of Educational Exclusion
in Five States of Brazil
Parameter   Standard Wald     Pr>         Standardized  Odd
c  Variable    DF Estimate     Error    Chi-square Chi-square  Estimate   Ratio
_     PARENED    1     -0.150    0.059    6.555      0.011       -0.330        0.861
MIGMUN       1     1.637     0.355    21.247     0.000       0.285        5.139
MIGEST       1     1.534     0.525    8.548      0.004       0.181        4.639
RG
IINCOME      1    0.502 (??)  0.232    4.674     0.031        0.308        1.652
IPARENED   11 1    -0.343  J 0.079  1 19.019 1   0.000   1    -0.793       0.710
PB
INCOME       1    -0.417     0.139    8.972      0.003       -0.258       0.659
PARENED      1    -0.117     0.059    3.956      0.047      -0.317        0.890
SIBLING      1  -0.2818 (??)  0.128   4.819      0.028      -0.302        0.754
MIGMUN       1     0.951     0.498    3.649      0.056       0.171        2.589
PE
SINGLE       1     0.798     0.171    21.883     0.000       0.178        2.220
PARENED      1    -0.113     0.026  | 19.587     0.000      -0.264        0.893
SIBLING      1     0.116     0.037 | 9.963       0.002       0.124        1.123
MIGMUN      1      0.533     0.206    6.663  J   0.010       0.098    -   1.704
| Variables  | Ages 10 - 14               Ages 15 - 17          Ages 7 - 14
gRural       |                           10.65 RJ 0.95 PB       0.32 RJ
Single       10.89 PB                                           0.80 PB
Income        -0.21 RJ -0.44 PB           -0.30 RJ -0.25 PE    -0.42 PB
Parenedu      -0.32 RN -0. 12 PB -0. 11 PE  -0.15 RJ -0.14 MA -0.12 -0.15 MA -0.34 RN -0.12
RN -0.12 PB -0.12 PE  PB -0.11 PE
Sibling      0.38RJ 0.10PE              |                      0.15 RJ 0.12PE
Nonwhite      0.64 RJ                                           0.30 RJ
Gender                                    0.66 MA
Work_ 10      2.26 RJ 1.42 MA 1.28 RN 0.89  0.80 RJ 1.19 MA
_________PE0.0R119M
Migmun        1.54 MA 1.01 PB             1.00 MA 0.72 PB       1.63 MA 0.95 PB 0.53 PE
|Migest       1.72 MA                     0.72 PB               1.53 MA
RJ= Rio de Janeiro; MA = Maranhao; RN = Rio Grande do Norte; PB = Paraiba; PE = Pemambuco
1 55



Q          Children and YouthVulnerability Primeira Edicao
Poverty, Exclusion and Social Risks in Five Brazilian States
Life Patterns in Determining Dropouts
The fact that most coefficients for the same cohort are on a similar order of magnitude
in different states prompted an attempt to pool regressions for each cohort. This has the
advantage of increasing the robustness of the results and makes the coefficients for the 7 to 9
cohort statistically significant. Accordingly, a regression results for each cohort after pooling all
the states together. This was done by adding one new (dummy) variable for each Northeastern
state to differentiate among them. Rio de Janeiro represents the base line; it is included in the
intercept of each equation. As before, the results reported below include only the coefficients
that were significant at a 5 percent or higher confidence level (see Table 18).
Beginning with the 7 to 9 cohort, none of the state-characteristic (dummy) variables
were significantly different from zero. This indicates that there are no differences in the
probability of dropping out among the states in this cohort. This conclusion will be
reaffirmed below. The household income, parent's education, and municipal migration
were the only significant variables for this cohort. In fact, this is the only group of variables
that are significant in the three age cohorts.
In the 10 to 14 cohort, most variables are significant. Single-headed households, low
household income, parents' low levels of education, a large number of siblings, child labor,
and migration (both inter-state and inter-municipality) all affect the probability of a child
dropping out of school. Maranhao and Paraiba have a different pattern of dropouts from
Rio de Janeiro in the sense that, when the other variables are included, the probability of
dropping out of school is significantly smaller in these states than in Rio de Janeiro.
For the 15 to 17 cohort, the list of significant coefficients is similar to those of the 10 to 14
cohort ? household income, parents' education, child labor, and municipal migration plus
the location of household. Maranhao has a distinctive pattern compared with those of the
other states and, in particular, with that of Rio de Janeiro. When the other variables are
included, the probability of dropping out of school in Maranhao is significantly smaller
than in Rio de Janeiro.
Overall, when all states are pooled together, household income, parents' education, and
migration are the main variables that explain the probability of dropping out. Child labor
is significant only for the cohort of children aged 10 and over. Interestingly enough, this
variable is not significant for children under age 10.
56 



Determinants of Educational Exclusion
in Five States of Brazil
Parameter Stadard  Wald        Pr>      Standardized  Odd
i    Vui ba ble  DF Estimate  Error     Chi-square  Chi-     Estimate     Ratio
co ~~~~~~~~Isquare
Children 7 to 9 - All states together
INTERCPT     1    -2.964    0.739     16.087     0.000        .
INCOME       1    -0.212    0.103     4.219      0.040     -0.136      0.809
PARENED      1    -0.103    0.044      5.407     0.020     -0.252      0.903
MIGMUN       1    0.990     0.367     7.278      0.007      0.152      2.690
Children 10 to 14 - All states together
INTERCPT     1    -2.346    0.387     6.788      0.000        .
SINGLE       1    0.405     0.144     7.956      0.005      0.091      1.500
INCOME       1    -0.153    0.055     7.785      0.005     -0.097      0.858
PARENED      1    -0.112    0.021     29.745     0.000     -0.277      0.894
SIBLING      1    0.066     0.032     4.342      0.037      0.067      1.068
WORK 10      1    1.083     0.152     50.495     0.000      0.224      2.954
MIGMUN       1    0.693     0.164     17.808     0.000      0.122      2.000
MIGEST       1    0.670     0.231      8.436     0.004      0.082      1.954
MA           1    -0.668    0.205     10.612     0.001     -0.155      0.513
PB           1    -0.584    0.246     5.615      0.018     -0.103      0.558
Youths 15 to 17 - All states toqether
RURAL        1 | 0.335      0.102     10.773     0.001      0.082      1.398
INCOME       1    -0.248    0.038     43.549     0.000     -0.161      0.780
PARENED      1    -0.130    0.013     100.221    0.000     -0.325      0.878
WORK_10      1    0.562     0.096     34.157     0.000  -   0.144      1.753
MIGMUN       1    0.476     0.115     17.247     0.000      0.093      1.609
MA           1    -0.377    0.138      7.426     0.006      -0.085     0.686
Determinants of Dropouts According to Gender
Pooled regressions run separately for male and female dropouts and for white and non-
white dropouts are shown in Table 19. The table presents only those coefficients that are
significantly different from zero at the 5 percent level of confidence.
57



Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
Variables             Ages 7 - 9             Ages 10 - 14              Ages 15 - 17
Females                                                                Rural 0.78
Income - 0.35           Income - 0.28            Income - 0.27
Parenedu - 0.14         Parenedu - 0.06          Parenedu -0.11
Work 10 1.10             Work 100.27
Migmun 0.89              Migmun 0.52
Migest 1.00
Maranhao - 0.59
Males                                         Single 0.62            Income - 0.20
Parenedu -0.17           Parenedu -0.17
Work 10 1.05             Work 10 0.90
Migmun 1.53             Migmun 0.47              Migmun 0.32
Migest 0.45
Maranhao - 0.68          Maranhao - 0.82
Paraiba - 0.48
Nonwhites                                                              Rural 0.32
Single 0.53            Income - 0.25
Parenedu - 0.16        Parenedu -0.10            Parenedu -0.10
Work 10   1.06           Work_10 0.44
Migmun 1.33             Migmun 0.77              Migmun 0.52
Migest 0.71              Migest 0.48
Maranhao -0.70           Maranhao - 0.36
Parafba - 0.79           Paraiba - 0.39
Rural 0.43
Whites               Income - 0.40           Income - 0.27            Income - 0.21
Parenedu -0.12           Parenedu -0.17
Work_10 1.16             Work_10 0.87
Sibling 0.13           Maranhao - 0.72
These results are more robust than the previous results because they make categorical
differences stand out more clearly. Thus, for those in the very young group (aged 7 to 9),
the following results:
* There are differences between males and females. For females, the only significant variable
is (municipal) migration, while for males, the explanatory variables are household
income and parents' education. In other words, when children are young (aged 7 to 9),
the socioeconomic status of their parents seems to cause boys to drop out more than
girls, whereas municipal migration causes more girls to drop out than boys.
* There are differences between white and non-white children. Among white children,
only household income is significant (as was the case for males). For non-white children,
58    1



Determinants of Educational Exclusion  I'.
in Five States of Brazil
only parents' education and municipal migration (as was the case for females) were
significant. In other words, among young white children (aged 7 to 9), low household
income significantly increases the probability of a child dropping out of school, but the
chances of non-white children dropping out are more affected by their parents' education
and by municipal migration.
The results for the 10 to 14 cohort are more diverse, in the sense that more variables were
significant than were significant in the 7 to 9 cohort.
* The parents' education and child labor variables are significant for boys and girls and for
whites and non-whites.
* Household income and inter-state migration are significant for females, but not males.
* Being in a single-parent household, is significant for males, but not for females. In other
words, the lack of a male parent (statistically female-headed households constitute the
vast majority of these households) seems to affect boys more than girls in terms of
dropping out.
* For both boys and girls, being from Maranhao decreases the probability of
dropping out.
* There are more differences between white and non-white children in terms of their
likelihood of dropping out of school in this cohort than in the younger cohort; only
parents' education and work are common significant variables in both cohorts.
Finally, turning to the 15 tol7 cohort:
* Household income, parents' education, child labor, and municipal migration are
significant for determining dropout for both genders.
* Household location is significant for females but not for males, which means that
girls are more likely than boys to be influenced to drop out of school by living in a
rural area.
* In Maranhao and Paraiba, inter-state migration is significant for males but not for
females.
* There are several variables that are significant in influencing both white and non-white
children to drop out-rural location, household income, parents' education, child
labor, and living in Maranhao.
* Migration, both state and municipal, once again seems to be significant for non-
whites.
59



Children and YouthVulnerability Primeira Ediaio
Poverty, Exclusion and Social Risks in Five Brazilian States
Conclusions
Coverage
Educational coverage in the five states analyzed, Maranhao, Paraiba, Pernambuco, Rio
de Janeiro and Rio Grande do Norte, is quite high. Nonetheless, educational exclusion
persists even in light of significant progress on the part of the government towards the
goal of universal coverage. A disturbingly large number of school-aged children and
youth do not enjoy full access to the educational system, either because they do not
attend school at all or when attending, repeat grades several times, eventually giving up
and dropping out entirely.
Also, educational exclusion, as shown by the number of people who never attended school,
is more common in the Northeastern states than in Rio de Janeiro. This is true for all cohorts
but more noticeable in the 26 or older cohort, indicating that past education coverage was
not only much worse than it is today but also more unequal regionally, which in turn may
reflect regional differences in per capita income and level of urbanization.
The coverage of preschool education for children aged 5 and 6 is still quite modest and
approximately 80 percent of children younger than age five have never been to school. This
is particularly evident in the Northeast states where a few thousand children at this age
have already left school. Nonetheless, this picture represents a substantial improvement
over time as shown in the increased school attendance levels of the 7 to 17 age cohort, as
compared to older cohorts.
Overage Enrollment
Multiple grade repetitions lead to increasingly larger age-grade gaps, which significantly
increase the risk of dropout. Age-grade gaps of two years or more are very common in all
five states considered. For the school-age population, this problems affects about one-third
of the students and is particularly acute in Maranhao (39 percent) and Paraiba (33 percent).
The proportion of dropouts increases steeply for children aged 12 and over.
60   1



Determinants of Educational Exclusion
in Five States of Brazil  ,J
Dropout Rates
The decision to drop out of school is the most important determinant of educational
exclusion. For school-age children (7-17 years) the dropout rate is 8.2 percent in Maranhao,
7.7 percent in Paraiba, 10.2 percent in Pernambuco, 6.2 percent in Rio de Janeiro, and 9.1
percent in Rio Grande do Norte. In other words, dropouts account for the educational
exclusion of more than 600,000 youngsters in these five states. Except in Rio de Janeiro,
those who dropout do so before completing the fundamental level of education, and this
propensity increases significantly from age 12 onwards.
The econometric analysis of the reasons why students drop out of school focused on
socio-economic variables, omitting school-based factors due to the limitations of the
database used. The most salient results from this analysis show that parents' education and
household income are negatively associated with dropout while youth work and migration
are strong contributing factors.
Determinants of Dropout
In all states, when children in the 7 to 14 age group drop out of school, they do so because
their family income and their parents' education is low and because they have experienced
migration, which forces them to leave school in their place of origin and face difficulties in
re-entering school in the new location. For the 10 to 14 age group, a number of additional
factors increase the probability of dropping out of school, such as living in a single parent
household, having more siblings, and having to work. Adolescents between the ages of 15
and 17 share most of the same reasons for dropping out of school with the 10 to 14 cohort,
except that at this time in life [KJS1]rural location becomes a significant factor, especially
for young women.
Gender as a determinant of educational exclusion
Gender differences in the decision to drop out of school are found particularly among the
younger age cohort, since girls aged 7 to 9 are more prone to drop out when their family
is poor and their parents have low education levels. Boys at this age are only significantly
affected by migration. For the 10 to 14 age cohorts, boys differ from girls in that household
income is not a significant determinant, but living in a single parent household is; however
girls at this age are affected by their family's income
1 61



Children andYouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
Ethnicity as a determinant of educational exclusion
Racial differences are also apparent, especially among the younger groups in which
migration and parent's education significantly affects the behavior of non-whites, but not
that of whites who are prone to drop out particularly because of lower family income. For
older children, there are more common and different reasons for dropping out of school
between the two groups. Particularly in Rio de Janeiro, consistently, in all age cohorts, the
probability of non-whites to drop out is higher than that of whites.
In conclusion, the combination of the three factors that define educational exclusion
- never having attended school, having a larger than two year age-grade gap, and having
dropped out of school, leads to the conclusion that a large part of the 7 to 17 generation is at
risk of social exclusion and lack of stable income opportunities in the five states. In relation
to the population in this age group, the young people at risk represent about 46 percent in
Maranhao, 40 percent in Pernambuco, 39 percent in Paraiba, 36 percent in Rio Grande do
Norte and 22 percent in Rio de Janeiro. The number of youngsters at risk in the five states
adds up to 2.6 million, a staggering figure.
62   1



-                  U_
r      -AK
i i                       !;mpr    - CHAPTER II
QUALITATIVE ASPECTS OF EDUCATIONAL
EXCLUSION IN RURAL SCHOOLS
Jose Edson Gomes da Silva
Barbara Brakarz
I          ,&T






Qualitative Aspects of Educational  | L
Exclusion in Rural Schools   | 
As shown in Chapter I, parent's education, family income, child and youth labor, and
migration are the most influential factors determining school dropout rates in the five
states considered. In rural areas these factors are more common, and rural children and
youth have a higher propensity to be excluded from educational opportunities than their
urban counterparts. However, student performance varies among rural schools and, in
fact, in some rural schools, student performance is above the state average. The first part
of this chapter explores, from a school-level perspective, what factors make students excel
in some rural schools. The second provides some conclusions.
Study Design
This analysis is based on a study that ranks rural schools in Maranhao and Paraiba according
to student learning achievement test scores in the 1998 Brazilian System of Basic Education
Evaluation (Sistema Nacional de Avaliacao da Educacao Basica - SAEB). The SAEB results
are overwhelmingly disappointing in that: (a) learning performance is inferior to that
expected in private, state and municipal schools. In public schools, students only near the
expected performance in the fourth grade of the fundamental cycle; (b) although private
schools are solidly ahead of public schools, the performance of their students is also below
expectations; (c) more than 50 percent of students in public schools perform below average,
which is already lower than expected; and (d) curriculum content of the fourth grade of
the fundamental cycle is only mastered in the eighth grade; students in the fourth grade
barely understand the meaning of words they read, and students in the fourth and eighth
grades are unable to read and understand a simple newspaper article. Consequently, a vast
majority is not prepared to proceed to the secondary level of education (Araujo e Oliveira;
Schwartzman, 2002). Nonetheless, there are differences among school performance levels,
and students in some rural schools performed above their state's average. Forty rural
schools among those that ranked highest and lowest in students' test scores were selected
for detailed analysis.
While the previous chapter focused on the determinants of educational exclusion from a
household perspective, i.e. factors outside the school, the analysis in this chapter centers on
three possible explanatory factors of school excellence in terms of learning achievement:
Teacher Qualifications, Organization of Time, and Teaching Practices. A set of variables was
defined to measure different dimensions of each of these three factors, and then organized
into a questionnaire. The questionnaire was completed by 40 principals, 123 teachers,
294 parents and 553 students. Although results are not statistically significant, given the
small size of the sample, they indicate some of the factors that may account for successful
165



I   -     Children and YouthVulnerability Primeira Edi,co
I'D ;     Poverty, Exclusion and Social Risks in Five Brazilian States
experiences in schools even when a majority of its students are poor. The basic assumption
underlying the analysis is that once the success factors are identified, they can be replicated
in other schools as part of programs such as, staff development, teacher training, and
community involvement.
The variables designed to measure different dimensions of each factor of schools success
in terms of learning achievements are listed in Table 21. They were presented in the
questionnaire as statements of good practices leading to successful student performance.
The respondents were asked to rank each variable according to how much they agree
or disagreed with the statement, on a scale of 1 - strongly disagree, to 5 strongly agree.
Respondents were categorized into four groups: Internal or External Stakeholders in Top
or Bottom Schools. All schools visited were located in poor rural areas in Maranhao and
Paraiba, two of the poorest states of Brazil.
66 1



Qualitative Aspects ot Educational
Exclusion in Rural Schools
Qualification of Teachers
The school knows the official curricula
Teachers know the content of the curricula they teach
Teachers organize the contents well
Teachers know what was taught in previous grades on the subjects they teach
Teachers clearly define the learning results expected
Leaming results are measurable
Organization of Time
The teaching process is not disrupted
The time allocated for each subject, class and grade is well defined
For the most part, teachers are present during school hours
Teachers use a pre-defined teaching plan
Theoretical and practical elements of the curricula are well integrated
There are norms defining times of arrival and departure for teachers and students
Teaching Practices
The time of students is mainly spent on leaming
The rhythm of teaching is adequate for the class
Emphasis is given to Math and Portuguese
Teachers are clear
There are links between one class and the next
Student curiosity is stimulated
Student comprehension is verified
Students receive feedback for homework and tests
Criticism is constructive, and student self-esteem is stimulated
Homework is assigned regularly
Homework is regularly completed
Homework content is appropriate for the age of the students
67



X   A\  }Children and YouthVulnerability Primeira Edicao
Poverty, Exclusion and Social Risks in Five Brazilian States
Qualification of Teachers
With respect to the qualification of teachers, perceptions differ between bottom and top
schools. Also, perceptions of teacher quality differ according to stakeholder standing:
internal stakeholders, i.e. teachers and principals, have a more positive perception than
external stakeholders, i.e. students and parents.' Tables 22 and 23 show the results that
indicate teacher qualifications in the top and bottom schools according to main educational
stakeholders in Maranhao and Paraiba, respectively:
Intemal Stakeholders              External Stakeholders
Dimension.          Pnncipals           Teachers            Parents             Students
Top     Bottomn   Top     Bottom    Top     Bottom    Top     Bottom
Schools Schools Schools Schools Schools Schools Schools Schools
official curricula        4.0      4.0      4.2      3.8      1.5      1.3      1.6      1.3
Teachers know what        4.4      4.2      4.5      3.9      1.7      1.8      2.9      2.1
they teach
Contents are organized    4.6      4.0      4.4      4.2      1.6      1.4      2.2      2.0
in sequence
Teachers know what was    3.5      3.7      4.0      3.9      1.6      1.8      2.7      2.1
taught in previous year
Expected results are      4.0      3.7      4.1      4.1      1.4      1.5      2.8      2.0
cleady defined
Leaming results are       4.3      4.3      4.0      3.0      1.3      1.2      3.5      2.0
measurable
Range: 1 - Strongly Disagree to 5 - Strongly Agree
'In the Escola Vista Por Dentro, Oliviera e Araujo and Schwartzman had already shown in more detail and in a much broader
context how uncritical of their own performance teachers are. According to them, "teachers consider as normal the actions,
behaviors, practices and results which are strongly associated with school failure" (2002: 47).
68    1



Qualitative Aspects of Educational  - '
Exclusion in Rural Schools
Internal Stakeholders          External Stakeholders
Dimension.        Principals        Teachers          Parents           Siudents
Top    Bottom   Top    Bottom    Top    Bottom   Top    Botlom
Schools Schools Schools Schools Schools Schools Schools Schools
School knows            1.8     1.8     3.2     3.1     1.5     1.5     1.6     1.5
official curricula
Teachers know what      1.8     1.9     3.6     3.6     1.7     1.7     2.4     1.9
they teach
Contents are organized  2.0     1.9     3.5     3.5     1.5     1.6     2.0     2.1
in sequence
Teachers know what was  1.8     1.8     2.0     2.0     1.4     1.6     2.7     2.0
taught in previous year
Expected results are   1.9      1.9     3.1     3.2     1.4     1.4     2.2     2.0
clearly defined
Learing results are    1.9      1.8     4.0     3.0     1.3     1.3     2.7     2.6
measurable          I_           _ _   _ _ __ _ _ _ _   _    I__ _ _ _ _
Range: 1 - Strongly Disagree to 5 - Strongly Agree
These numbers indicate that stakeholders in Maranhao have a more optimistic view than
those in Paraiba about the performance of their teachers-even in schools where students
were ranked at the bottom. According to students, schools in which students fail (and therefore
have a higher probability of being excluded) are mainly characterized by having teachers who
are not familiar with the materials they teach and who plan their classes poorly.
Except for few dimensions, all answers from teachers and principals in Maranhao averaged
4.00 or higher, showing a high degree of satisfaction with their own performance and with
how the curriculum is organized. On the other hand, answers from parents and students
almost never averaged higher than 3.0. Parent perceptions are particularly negative.
The distinction between the top and the bottom schools is more clearly demonstrated in
student perceptions. Students in the top schools tend to agree more that their teachers know
the curriculum, are familiar with what was taught in previous grades, and establish clear
learning indicators and measurable learning targets, whereas the perceptions of bottom
school students are more negative.
Figures 1 3a and 1 3b trace the rankings assigned by teachers and students in top and bottom
schools to the explanatory factor "Qualification of Teachers" in Maranhao and Paraiba,
respectively. Teachers represent internal stakeholders and students represent the external
ones. The patterns indicate that in both states, there is a significant divergence between top
and bottom schools.
69



Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
Figure 13a: Qualification of Teachers in Top and Bottom Schools According to
Teachers and Students - Maranhao
5.0                                                                           - - - Teachers-
4.0             _______      - - -                             -   _          TopSchool
- --  Teachers -
3.0                                                                           Bottom School
-     Students -
2.0                                                                            TopSchool
1.0                                   .                                           Students -
Bottom School
0.0
School know s  Teachersl'ow k Cntents are  Teachers know Epected resuis  Pesuis are
officelcurriclium  whatthey teach  organized  prevbus year cbarlydefined  mBasurable
insnuctbDn
Figure 13b: Qualification of Teachers in Top and Bottom Schools According to
Teachers and Students - Paraiba
5.0                                                                           - - - Teachers-
4.0                                                                           TopSchool
- - - Teachers -
3.0     5--5-                     >vv                                         Bottom School
-   Students -
2.0                                                                            Top School
1.0                                                                           -     Students -
Bottom School
0.0                                                                          I
School know s  Teachers know  Odntents are  Teachers lnow Expected resuts  Fesults are
oficialcurriculum whatthey teach  organeed  prevbus year clearly defied  msasurabt
insfuctbn
In Maranhao, teachers in both top and bottom schools have very similar perceptions of
their qualifications, whereas teachers in top schools appear somewhat above those in
bottom schools. Moreover, students in top schools have a more positive perception on
their teachers' qualifications. Stakeholders diverge the most in their perception of the
organization of material, while for the dimension "teachers know what they teach," the gap
between students in top schools and teachers in bottom schools is small.
In Paraiba, teachers in top schools have only a slightly more positive perception of their
qualifications than teachers in bottom schools for every dimension, except for whether
"results are measurable." Teachers in both top and bottom schools rank themselves lower in
regards to knowing what was taught the previous year, and this score is as low as the ones
given by students in bottom schools.
70 1



Qualitative Aspects of Educational
Exclusion in Rural Schools
As compared to students, teachers assign a higher score for all dimensions, except
when asked if they know what was taught the previous year as noted above, where their
scores slide to a level lower than that given by students in top schools and as low as
the scores by students in bottom schools. As expected, students in top schools have a
more positive perception of the qualification of their teachers than students in bottom
schools, except for the one dimension "contents are organized," for which both have the
same low opinion.
Teachers and principals agree that teachers in schools in Maranhao are familiar with
the official curriculum. In both states however, this familiarity is more often recognized
by students in schools at the top. However, in questions concerning the curriculum
and organization of content, top schools gave a lower score, close to that of bottom
schools.
Other significant results include:
* Qualification of teachers regarding the content of what they teach. In Maranhao, all the
stakeholders except parents, agree that in the top schools, teachers are better qualified
with regard to the content of their classes than in the bottom schools. According to
students, such a difference also occurs in Paraiba.
* Qualification of teachers to organize teaching practices logically. All stakeholders in
Maranhao mention that teachers are more apt to organize teaching in logical sequences
at the top schools than at the ones at the bottom. However, in Paraiba, the differences
between both types of schools in this regard are not as evident as in Maranhao.
* Clear definition of steps and results. Students in Maranhao agree that clarity in teaching
and in learning targets are more prevalent in schools at the top than at the bottom.
Surprisingly, in Paraiba the difference between the two types of schools is insignificant.
* Measurable results. Teachers in both states, but only students in Maranhao, point to
the importance of having measurable learning results as a distinction between top and
bottom schools.
Organization of Time
Perceptions of Organization of Time as a factor in school quality indicate the existence of
significant differences between top and bottom rural schools. Tables 24 and 25 show the
perceptions of "Organization of Time" in top and bottom schools according to the main
educational stakeholders in Maranhao and Paraiba, respectively.
7 7 1



. .    Children and YouthVulnerability Primeira Edi,o
Poverty, Exclusion and Social Risks in Five Brazilian States
Figure 14b: Organization of Time According to Teachers and Students in Paraiba
5.0                                                                              - - - Teachers -
4.0      -     - -                    - - - - - - - - - - - - - - - - -          TopSchool
BoTeachers-
3.0             ---- ------- ---__                 ,,"          ---              Bottom School
-    Students -
2.0                                                                               TopSchool
1.0                                                                              -     Students -
Bottom School
0.0 
No disruption  Time for each  Teachers in  Teaching is  Theoretical and  Norms for
during classes subject clearly school most of the  according to  practical time well attendance and
defined        time     predefined plan  organized   tardiness
VWhen evaluating the dimension "teachers define teaching time clearly" illustrated in Figure
15, a consistent pattern for school types is observed: top schools have a higher perception
than bottom schools. Within each school type, another pattern is maintained: internal
stakeholders have a higher perception than external ones. In top schools, teachers evaluated
themselves very high with high scores, whereas students in the same top schools assigned
the teachers a much lower scores. But it is clear that scores in bottom schools are lower in
general than top schools for both internal and external stakeholders. Teacher attendance
tends to be less reliable in schools at the bottom than at the top. Teachers in Paraiba confirm
that unreliable teacher attendance may be a reason leading to educational failure and
eventual exclusion.
Figure 15: Divergence Between Internal and External Stakeholders in Top and
Bottom Schools in Paraiba: Teacher Organization of Time for Practice and Theory
5
4
3
2
0
Teachers              Students             Teachers              Students
UTop Schools                                Botton Schools
74



Qualitative Aspects of Educational
Exclusion in Rural Schools
Other noteworthy results include:
* Class Disruption. Students and teachers in Maranhao suggest that class disruption is an
important reason for why students fail in some schools.
* Planning of the learning process. Teachers in Paraiba and students in both states agree
that it is more common for teachers to plan their classes in schools where students rank
at the top than in those where they rank at the bottom of the achievement tests.
* Organization of theoretical and practical material. Students in both states and teachers in
Paraiba agree that the schools where student achievement is higher balance the practical
and theoretical learning content of the curriculum better.
Teaching Practices
External and internal stakeholders in Maranhao and Paraiba show differences in teaching
practices between top and bottom schools, as demonstrated in Tables 26 and 27. In general,
the pattern is maintained, where internal stakeholders have more positive evaluations of the
schools than external stakeholders.
75



f~ / >. |Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
Internal Stakeholders            Extemal Stakeholders
Dimensior .        Principals        Teachers           Parents             Students
Dimensions           Top     Bottom    Top    Bottom    Top    Bottom    Top    Bottom
Schools Schools Schools Schools Schools Schools Schools Schools
Time of students is       4.5      4.5      4.0      3.9     1.5      1.8     2.0      1.6
mainly used to learn        .
Teaching adjusted to      3.9      3.8      4.3      4.1     1.4      1.4     1.6      1.6
class
Math and Portuguese       4.1      4.1      4.4      4.3     1.7      1.8     2.1      1.8
receive priority
Teachers are clear        4.2      3.9      4.4      4.5     1.7      1.8     2.0      2.2
Classes are connected     4.2      3.6      4.4      4.2     1.7      1.6     2.0      1.4
Curiosity is stimulated   4.2      3.8      4.7      4.1     1.4      1.8     1.8      1.9
Comprehension is          err?     4.2      4.1      3.6     1.3      1.5     2.2      2.1
verified
Students receive          4.0      4.0      4.8      4.2     1.3      1.9     2.6      1.9
feedback
Criticism is positive and  4.2     4.2      4.8      4.2     1.9      1.7     2.2      2.1
self esteem is stimulated
Criticism is positive & self  4.2  4.2      4.8      4.2     1.9      1.7     2.2      2.1
esteem stimulated
Homework regularly        4.1      4.1      4.6      4.6     3.1      3.0     3.0      2.1
assigned
Homework is regularly     3.7      3.7      3.8      3.8     3.1      3.0     2.9      2.1
completed
Homework adequate to      3.4      3.4      4.4      4.4     2.4      2.4     2.6      2.6
Range: I - Strongly Disagree to 5 - Strongly Agree
76    1



Qualitative Aspects of Educational
Exclusion in Rural Schools  J
MIwMSeM                                      0M Su
Internal Stakeholders             External Stakeholders
Dimnsions         Principals        Teachers           Parents             Students
Top     Bottom    Top     Bottom    Top     Bottom    Top     Bottom
Schools  Schools  Schools  Schools  Schools  Schools  Schools  Schools
Time of students is   3.3      3.1      4.0      3.1      1.5      1.5      2.0       1.9
Teaching adjusted to  3.9      2.3      4.3      2.3      1.5       1.4     2.0       1.8
class
Math and Portuguese   4.1      3.6      4.4      3.7      1.8      1.7      2.2      1.8
receive pnonty
Teachers are clear    3.9      2.8      4.4      2.4      1.8      1.6      1.6       1.4
Classes are           3.6      2.6      4.4      3.1      1.7      1.6      1.8       1.2
connected
Curosity is           3.8      2.6      4.7      2.5      1.4      1.4      1.5       1.3
stimulated
Comprehension is      4.1      2.6      4.1      2.5      1.3      1.3      1.7       1.4
verified
Students receive      4.0      2.7      4.7      3.7      1.9      1.2      1.4      1.1
feedback
Criticism is positive
and seH esteem is     4.2      1.1      4.7      3.7      1.3      1.3      1.9      1.3
stmulated
Criticism is positive
& seHf esteem         3.6      2.6      4.6      2.6      3.1      3.6      2.2       1.4
stimulated
Homework regularly    3.3      3.1      4.0      3.1      1.5      1.5      2.0      1.9
assigned
Homework regularly    3.9      2.3      4.3      2.3      1.5      1.4      2.0       1.8
completed
Homework adequate     4.1      3.6      4.4      3.7      1.8      1.7      2.2       1.8
to age             I______________________________
Range: 1 - Strongly Disagree to 5 - Strongly Agree
Figures 16a and 16b trace the rankings assigned by teachers and students in top and bottom
schools to all dimensions of Teaching Practice for both states. In Paraiba, teachers in top
schools have a considerably more positive perception of teaching practices than teachers in
bottom schools in all dimensions. In Maranhao, perceptions are not as clear.
1 77



Children and YouthVulnerability Primeira Edicao
Poverty, Exclusion and Social Risks in Five Brazilian States
Figure 1 6a: Teaching Practices in the Top and Bottom Rural Schools
According to Teachers and Students - Maranhao
5.0                                          - --   _ -      --        - - - - -  - -
4.0                                                                            -
3.0
2.0
1.0
0.0                *               I                                                      I         X
Timne mainly Teahring  Malh  Teaer Cassesar  Cursity  Compehen. Students Self-steemr  HW     HW      HW
usWd for  adjusted  & Port.  are dear connected stinulated  verHled  cive  simulated  regulerly  regulady  arJequate
eaming   to dasm  pority                                      eedback           assigned  completed  to age
- - - Teadcs - Top School  ---Teahes - Bottom School  -    Students - Top S&hod    Students - BooM School
Figure 16b: Teaching Practices in the Top and Bottom Rural Schools
According to Teachers and Students - Parafba
5.0…-- --- -                                     ---   ---   --   --  
4.0
3.0  1-- ;,        ...               -- ;". :
2.0
1.0
0.0                X                                       I                                   I
iTme mainly Teaching  Math  Teachers Clesm are CunosIty  Copwehe. Students Self-esteem  HW   HW      HW
used for  adjusted  & Port  are dear connected stinutated  vernted  rceive sitnulated  regulady  regularly  adequat
larming  to dass  priot                                      feedback           assigned  oot,leed  to age
- - - Teachers - Top School  - - - Teachers - Bottom Sdool  -   Students - Top School  -   Studanb - Botorm Schoo
Figure 17 compares divergence between students and teachers in Maranhao in both types
of schools. Again, teachers represent internal stakeholders and students represent external
ones. Teaching processes in Maranhao are clear cases of divergence of perceptions between
internal and external stakeholders, for both types of school. Teachers have a significantly
higher perception than students across all dimensions. In Paraiba clearly, teachers in top
schools have a considerably higher perception of teaching practices than teachers in bottom
schools in all dimensions.
78     1



Qualitative Aspects of Educational
Exclusion in Rural Schools
Figure 17: Divergence Between Internal and Extemal Stakeholders in Top and
5                   Bottom Schools in Maranhao: Teaching Practices
4
3
2
0
Teachers           Students           Teachers           Students
* Top Schools                           Top Schools
Other relevant results include:
* Learning Priority. Teachers in Paraiba in particular, but only students in Maranhao, agree
that students perform better in schools where priority time is for learning activities.
* Adjustment of the rhythm of teaching to the student's needs. Teachers and principals
in Paraiba point out that adjusting the rhythm of teaching to the needs of the students
occurs more often in top schools than in schools where students perform poorly.
However, in Maranhao, this adjustment is not seen as an important feature of the top
schools.
* Clarity in teaching. For students in Paraiba (and slightly in Maranhao), clarity in teaching
tends to be more frequently recognized as a characteristic of schools at the top than at the
bottom. This helps explain inadequate performance by students and their educational
exclusion.
* Links between classes. According to all stakeholders in both states, classes are not
connected to each other in schools at the bottom; while classes are viewed as better
connected in schools where students perform better.
* Curiosity stimulation. In Paraiba, this dimension seems to distinguish between the
two types of schools, although parents do not deem it as an important discriminating
element. In Maranhao, only principals and teachers agree that when there is stimulation
of student's curiosity, students tend to succeed.
* Regular assignment of homework. According to students, the schools where homework
is not regularly required are the ones they fail more probably.
79



Children and YouthVulnerability Primeira Edicao
Poverty, Exclusion and Social Risks in Five Brazilian States
Conclusions
Several findings of this study highlight the exclusionary characteristics of rural schools in
two of the poorest states in Brazil. Overall, a small but consistent divergence of perceptions
between top and bottom schools and, morever, between internal and external stakeholders
within the schools, was observed for all three factors of education quality.
A primary characteristic of the study is uncritical and self-congratulatory teachers and
principals. A second is unhappy and voiceless parents. In general, the views of teachers and
principals contrast markedly with those of parents and students who tend to have a dim
view of the schools. Moreover, educators seem unable to distinguish between the schools
where students are performing at the top and those where they are at the bottom. This fact
could be related to their lower qualification and training.
In Paraiba, although educators see their schools in a more positive light than the parents
and students, they tend to be much more critical than their colleagues in Maranhao. More
importantly, educators in rural Paraiba tend to indicate more clearly the distinctions
between top and bottom performing schools, particularly with respect to teaching practices
and organization of time.
Echoing their poverty, rural parents whose children attend public schools generally see the
conditions of these schools as very deficient in both Maranhao and Paraiba. Independently
of whether the school performance ranks at the top or at the bottom in terms of student
success, the average response from parents with respect to all aspects analyzed does not
frequently exceed level 3 on the 1 to 5 scale. This finding suggests that parents may not be
fully aware of the way in which school practices influence student performance.
Students are slightly more positive than their parents and certainly more discriminating in
their views. Students seem to be able to perceive small, but consistent differences in school
practices between top- and bottom-performing schools. This feedback may provide helpful
insights on the characteristics of the schools in which students fail and thus are more likely
to eventually repeat grades and drop out. Finally, it can be deduced from students' answers
that regular homework assignments, evaluation and feedback on homework, effective
planning of school time, and discipline are important factors for success, and can make a
difference even when the environment is unfavorable and the students are poor.
80   I



I             , ,a1 X i1is _, I,   '  1
'
r               5
I    ~~A-    ,
t \\-/ o- U - -_x CHAPTER III
PARTICULARLY VULNERABLE GROUPS: EDUCATION
AND SCHOOLS IN INDIGENOUS AND QUILOMBO
COMMUNITIES
Maria Valeria Junho Pena
Barbara Brakarz
Vania Medrado de Franco
EIi,.'..,






Particularly Vulnerable Groups: Education And Schools
In Indigenous And Quilombo Communities
How Colorful is Brazil?
Brazil is one of the most diverse of all Latin American societies, with descendents of several
ethnic groups blending together, including original Amerindian inhabitants, Portuguese and
Dutch colonizers, French invaders, European and Asian immigrants, and Africans brought as
slaves since the beginning of the colonization period. Contrary to popular belief, miscegenation
was not always harmonious. It has produced inequitable and unequal results in terms of
access to resources and opportunities and control over assets. Some types of discrimination
are pervasive and reflected in the educational system. This chapter summarizes some traits
of the persistence of educational exclusion among two of the most vulnerable groups in the
Northeast region and the country: indigenous peoples and Quilombolas.
Almost all great Brazilian scholars, such as Darcy Ribeiro, Caio Padro Jr., Celso Furtado,
Sergio Buarque de Hollanda, and Gilberto Freyre have stressed the triple characteristics
of the Brazilian population: Amerindian, European and African with the Portuguese
conquering markedly different from the British colonization in North America. While the
latter tried to replicate the social format of its original Empire (bringing colonizing families
and recognizing the local autonomy of the new settlers), the former tried to subjugate
local inhabitants by a centralized Crown, bringing adventurers and sometimes criminals
(degredados) to inhabitant the newly discovery territory.
In the Brazilian model, the integration of the local indigenous population into the
Portuguese Empire was deliberately pursued through a series of instruments that included:
land donations, religious missions and cathequese, forced learning of the Portuguese
language, and the imposition of Portuguese laws. This was a model that, while radically
exploitative, was also incorporative into the dominant Portuguese pattern. Since the
Portuguese traveled to Brazil without wives and without women, miscegenation was
inevitable'. As Pena (2002) has demonstrated through the study of the Brazilian human
genome, 60 percent of Brazilians who consider themselves as whites have indeed at least one
Amerindian or African female ancestor.
' In 1552, in a letter addressed to King Joao, Priest Manule da N6brega asked the Crown that white women be sent to Brazil
so men could marry and live afar from the sin they now live with" ("casem e vivam (... ) apartados dos pecados em que agora
vivem" ). After 1755, the Crown formally, through Alvara de Lei by Pombal Marquis, responded with incentives for inter-
marriages between local indigenous women and Portuguese men. Clearly, the idea was to promote the spatial occupation of
the colony. Blacks were excluded from the incentives.
83



A .9 | Children and YouthVulnerability Primeira Edicao
Poverty, Exclusion and Social Risks in Five Brazilian States
Thus, to a large extent, due to miscegenation, the colonization pattern was binary: from one
side, it was discriminatory, exclusive and hierarchical; from other, it was integrative, cordial
and mobile. Nonetheless, the power structure, based economically on large donations
of land (feitorias), followed the discriminatory mandate of the Portuguese Blood Purity
Statute (Estatuto de Pureza do Sangue) with access to positions within the state and the
church monopolized by those possessing "sangue limpo" (clean blood).2
In many ways, this duplicity of cordiality and exclusion still marks racial relations in
Brazil. First, Brazilians consider themselves a result of triple ancestry. Almost half of all
Brazilians do not consider themselves either blacks or whites, confirming the intense
identification with the miscegenation (mesti,agem) process. Dozens of skin identities were
built throughout the centuries, the most conspicuous being mamelucos, the result of the
miscegenation of Portuguese and indigenous; cafusos, the result of the miscegenation of
Africans and indigenous; and, as the result of the miscegenation of Africans and Europeans,
mulato, moreno, escuro, claro, roxinho, among others. Pardo, the denomination used by
Censuses and household surveys, encompasses more than one hundred different skin color
identifications, which are not black, white, indigenous or Asian descendent, but mixed.
Second, given such a degree of miscegenation, color could not be, and is not considered a
family characteristic, but rather an individual one. Members of the same family can have
different colors, depending on how their skin is seen with regard to other's skin colors in a
given interactive environment, including the family. Third, perception of skin color varies
regionally. Someone may have one color in a region, another color in other. And fourth,
upward mobility is whitening; downward mobility is blackening. Nonetheless, over and over,
Censuses and household surveys have indicated that income, education, employment and
access to services are more inadequate among blacks, pardos and indigenous than among
whites. Asian descendents have the highest standard of living of all the color ranges.
The 2000 Brazilian Census reports a population of 170 million inhabitants, in contrast with
146 million in 1991. In absolute numbers in 2000, Brazil had 20 million more whites, 3
million more blacks, 400,000 more indigenous, and 2.5 million more pardos.
The proportion of whites increased during the same period from 51.6 to 53.4 percent;
also the proportion of the black population increased from 5 to 6 percent; and finally,
the proportion of indigenous followed the same trend, increasing from 0.2 to 0.4 percent.
Only the proportion of pardos decreased from 42.4 to 38.9 percent. Table 28 shows that
2 Boxer, apud Souza Ramos, 2002.
84 1



Particularly Vulnerable Groups: Education And Schools
In Indigenous And Quilombo Communities
the absolute number of blacks grew by almost 40 percent between Censuses, whereas the
absolute number of pardos grew only 4 percent. It also shows that the size of the indigenous
population in 2000 is more than double what it was in 1991.
1991                   2000           Absolute   Relative     00
Color/Race    Total    _             Total       oto     Diference  Diference  Variation
Total      1 49,925,350   100     170,143,862    100     20,218,512    13.5       -
White       77,308,354    51.6    90,831,475    53.4     13,523,121    17.5      3.5
Black       7,490,497      5      10,423,567     6.1     2,933,070    39.2       22.6
Yellow       644,016      0.4       868,732      0.5      224,716      34.9      18.9
Pardos     63,635,915     42.4    66,150,797    38.9     2,514,883      4        -8.4
indigenous   300,361      0.2      702,886       0.4      402,525      134      106.2
Ignored      546,207      0.4      1,166,405     0.7      620,198       -         -
Source: IBGE
The increase in the proportion of blacks and the decrease in the proportion of pardos cannot
be attributed to the fertility rate being higher among the former than among the latter.
Indeed, it was a change in the ethnic self-identification that occurred among those older
than nine years of age that is responsible for the recent increase in the black population3. As
part of this renaissance of black identity in Brazil, remaining groups of slave descendents,
the Quilombos, began reentering Brazilian history.
Similar to blacks, there has also been a surge related to indigenous peoples identity,
particularly in the Northeast. Until very recently, indigenous peoples living in that region
were considered as sertanejos, the Brazilian word for rural poor in that area. Slowly, a process
of recognition of their new identity began taking place. At the beginning, this recognition
came, generically, as Northeast indigenous-in other words, several groups who were not
culturally, historically, or institutionally identifiable, but geographically. Hence, the term
"indio misturado" (mixed indigenous) a stigma that still constitutes an obstacle for their
cultural identity and legal identification.
The fact is, that, in the 1950s, ten indigenous ethnicities were recognized in the Northeast
and fifty years later, the list had increased to 23. As keenly observed by Pacheco de Oliveira
(1999) this increase consists an absolute contradiction: the recent appearance of peoples
whose definition is to be "originario" (since the origin) or first nations. It was only in
3According to Petrucelli (2002) even considering mortality, there was a substantial increase in the black identification
during the 1990s.
85



Children and YouthVulnerability Primeira Edicao
Poverty, Exclusion and Social Risks in Five Brazilian States
the context of the demand for land and assistance from the National Foundation for the
Indigenous (Funda,co Nacional do Indio - FUNAI) that Northeast indigenous groups began
to be recognized as such, with the 1975 Annual Meeting of the Brazilian Anthropological
Association playing a key role in this process.
In sum, ethnicity has never been a racial matter in Brazil. It has always consisted of a
social process in which identities are constructed and/or destroyed. It is in this context
that it plays a role in promoting educational exclusion. The more a group is identified
with its original excluded status, the more the exclusion persists. This chapter draws
from the following sources: (a) rapid appraisals in indigenous and Quilombo areas in
Maranhao, Pernambuco and Paraiba; (b) interviews with the local population and leaders
in Pernambuco; (c) qualitative questionnaires answered by 8th grade indigenous and
Quilombo students themselves; (d) the 1999 Census of Indigenous Schools, undertaken by
the Ministry of Education (MEC) in all indigenous schools in Brazil; and (e) a 2002 report
of the Secretariat of Education of Pernambuco (Secretaria de Educacao) about each of the
indigenous schools in that state.
Indigenous Education in Brazil
Until the 1980s, indigenous education in Brazil promoted linguistic and cultural integration
as pursued since colonial times to subsume indigenous languages and culture into a holistic
Brazilianism-in other words, into an European anthropocentrism. This stance changed
with the 1988 Citizenship Constitution, which gave indigenous peoples the right to a free
education that would be inter-cultural and, wherever necessary, bilingual. The Constitution
also stated that indigenous culture is a Brazilian heritage belonging to all, part of the
national identity, which the state has a duty to protect and preserve.
The 1988 Constitution provides the legal support for the current Directives and Basis
for Education Law (Lei de Diretrizes e Bases da Educa,ao) that guarantees to indigenous
peoples, in its articles 78 and 79, the type of education defined in Article 210 of the 1988
Constitution. The Resolu,co 3/99 (Resolution 3/99) of the National Education Council
(Conselho Nacional de Educa,co - CNE) complements the legislation by creating the
category "indigenous schools," establishing norms and procedures for their functioning,
and defining administrative responsibilities for their management and funding. Indigenous
schools are understood as those "schools located in territories inhabited by indigenous
peoples, which provide exclusive attendance to these communities, where teaching is
86 I



Particularly Vulnerable Groups: Education And Schools  |   L
In Indigenous And Quilombo Communities  Jj
provided in the mother language of these communities, with an adequate schooling
organization.
Starting in 1991, MEC took over the responsibility for indigenous education, previously
assigned to FUNAI, which had inherited this function from the National Service for the
Protection of the Indigenous (Servico Nacional de Protec,o ao Indio). The transfer of
educational responsibilities to the MEC did not occur without conflict. Until today many
indigenous peoples oppose the downsizing of FUNAI's role, which, despite, criticisms, is
still seen by the indigenous peoples as the only institution within the government that
understands and can provide protection to these groups.
In principle, indigenous schools, as all other
fundamental level schools (the eight grades of the  Box 3: Size of the Young
previous primary and gymnasium cycles) are run  Indigenous Population
by municipal or state governments with funding
from federal transfers that are proportional to the  Age groups  Total (thousand)
number of students enrolled. Additional financing   0-4            77.6
is provided by the Fund for the Maintenance and     5-9            72.2
Development of Education and Value of Teachers     10-14           67.4
(Fundo de Manutencao e Desenvolvimento do          15-14           43.8
Ensino e Valorizacao do Magisterio - FUNDEF) also  18-19           29.5
on a student per capita basis. State governments have  Source: 2000 Census
the mandate to supervise the implementation of the
inter-cultural guidelines and to train indigenous teachers. These activities are financed by
federal funds via transfers from the MEC, which also pays for teaching materials, including
textbooks and teaching guides that have been developed through a partnership between
universities, NGOs, and indigenous organizations. In practice, FUNAI still runs some
schools while others are state-run.
87



|      Children and YouthVulnerability Primeira Edigao
Poverty, Exclusion and Social Risks in Five Brazilian States
Box 4: Sharing of Responsibility for Indigenous Peoples' Education
The Federal Government should (a) plan and monitor; (b) transfer funds to municipalities
for operating indigenous schools, according to number of students enrolled; (c)
transfer funds to state governments for teachers training and development of culturally
appropriate teaching materials and guidelines; (d) through special programs, support
construction and rehabilitation, distance education, and equipment such as computers,
TVs and videos; and (e) promote parent's participation and empowerment.
State Governments, within their geographical boundaries, should: (a) supervise
indigenous education and schools; (b) define a program of multi-cultural education; (c)
produce teaching materials; and (d) promote teacher's training.
Municipal Governments should: (a) create and maintain indigenous schools at the
fundamental level; (b) provide these schools with adequate teachers; and (c) look for
additional support from special educational programs of the federal government.
Civil Society Organizations, NGOs and Universities should collaborate with state
governments on the design of specific teaching materials and guidelines.
Indigenous Organizations should: (a) participate with civil society organizations,
universities and state governments on the design of specific teaching materials and
guidelines; and (b) participate in internal decisionmaking processes at the school level
through programs to enhance community and parental involvement financed by MEC.
Results from the 1999 Census of Indigenous Schools
There are 1,392 indigenous schools in Brazil with the majority located in the North
region, where the majority of the indigenous peoples live. Fifty-five percent of these
schools are municipal schools, with the states being responsible for the remaining 45
percent. Most of the time, indigenous schools are treated as standard rural schools that
operate in a small room and are not necessarily located in indigenous communities.
88   1



Particularly Vulnerable Groups: Education And Schools
In Indigenous And Quilombo Communities  3
Figure 18: Indigenous Students by Region
18%
North
3%                                        Northeast
* Southeast
_|F                          *~~~~50/  Soulth
2200                                 U Midwest
* About 75,000 students are enrolled in these indigenous schools, with 43 percent in the
first year of the first grade of the fundamental cycle, 23 percent in the second year, 15
percent in the third year, 9.4 percent in the fourth year, and the remainder in the fifth to
the eighth years of the first grade. Also, about 15 percent of students are in pre-school or
literacy classes and a small proportion are in the subsequent grades.
* About 4,000 teachers work in indigenous schools.
* The vast majority of teachers are indigenous (75 percent).
* The vast majority of teachers are male (73 percent).
* Teachers are generally not well qualified-28 percent have not completed the first eight
years of primary education and only 4.5 percent have completed secondary education.
Only a very small proportion (1.5 percent) of teachers have a college education. Table
30 and other figures describe the major findings of the census with regard to indigenous
teachers.
* In almost half of the indigenous schools (46 percent), indigenous culture is not part
of the curricula. Only a third of the indigenous schools use specifically designed and
culturally appropriate didactic materials, but in general there is no connection between
what is taught in school and the indigenous traditions of the students.
* Education outcomes in indigenous schools are generally poor, with a decreasing number
of students along the educational trajectory.
Table 29 displays the findings of the Census on intercultural education in indigenous
schools in the Northeast and Southeast of Brazil, the pertinent regions in this study.
1 89



Children and YouthVulnerability Primeira Edigao
Poverty, Exclusion and Social Risks in Five Brazilian States
I able 29: Number and Proportion of Indigenous Schools that use Culturally
Appropriate Materials
Schools                                    Enrollment
Regions and        Schools that use adequate materials  In schools that use adequate materials
States
_ __________    Total       Total                   Total       Total        0
Brasil             1,392        424         30.5       93,037     28,999       31.2
Northeast           283          10         3.5        20,141      1,499        7.4
Maranhao            138          1          0.7        6,148        30          0.5
Ceara                24          4          16.7       1,829        181         9.9
Paraiba              29          0          0.0        3,396         0          0.0
Pernambuco           46          1          2.2        4,036        192         4.8
Alagoas              10          1          10.0       1,374        575        41.8
Sergipe              1           0          0.0         128          0          0.0
Bahia                35          3          8.6        3,230        521        16.1
Sotheast             24          7          29.2       2,755       1,620       58.8
Minas Gerais         5           5         100.0       1,575       1,575       100.0
Espirito Santo       10          1          10.0        552         21          3.8
Rio de Janeiro       2           0          0.0         27           0          0.0
Sao Paulo            7           1          14.3        601         24          4.0
Southeast            89          46         51.7       6,336       2,544       40.2
Parana               24          15         62.5       1,819        833        45.8
Santa Catarina       25          21         84.0       1,085        631        58.2
R. G. Do Sul         40          10         25.0       3,432       1,080       31.5
Indigenous Education in Pernambuco, Maranhao and Paraiba
In the state of Pernambuco, there are nine distinct indigenous groups making up a
population of approximately 20,000 individuals. There are 96 indigenous schools with 288
teachers, attended by about 6,336 students. The highest proportion of these students is
enrolled in the first four years of the fundamental cycle (1,056 in the first year, 721 in the
second year; 477 in the third year; 310 in the fourth year, with a drop to 140 in the fifth
year). In almost half of these schools, indigenous culture is not part of the curricula, only a
third use specially designed and culturally appropriate materials and generally there is no
connection between what is taught in school and the cultural traditions of the students. In
the majority of cases, schools offer only the first four grades of the fundamental cycle, within
a multi-grade format taught by teachers who have not been trained in this pedagogy.
90 1



Particularly Vulnerable Groups: Education And Schools
In Indigenous And Quilombo Communities
The decreasing numbers of students throughout the eight grades of the fundamental cycle
indicate high dropout rates. Qualitative analysis with students and field visits to indigenous
schools indicate that there are two main reasons why indigenous students drop of school after
the fourth grade, not completing the fundamental cycle. Both reasons are related to the lack
of easy, comfortable and speedy access that a public school should provide4. First, schools that
offer higher grades are not located in indigenous villages. Because of cultural reasons and fear
of prejudice, parents are reluctant to send their children away to school; and second, because
usually indigenous villages are located far away from the municipal center, students would
need to walk great distances to reach them even if they wished to attend. As a result, a vicious
circle reproduces indigenous poverty by virtue of their ethnicity, through a process in which
poor children, whose access to school is denied because of an inadequate location outside the
boundaries of their communities, become poor adults with limited opportunities.
During the first four grades, successive repetition seems the main reason for dropping
out from schools. As results from questionnaires applied in indigenous schools indicate,
repetition is higher among boys than girls. In contrast to students in Quilombos, child and
youth labor do not seem to be a reason for dropout among the indigenous children or youth
since few work. In general, those that do work are girls helping their mothers at home. Lack
of interest and apathy associated with the poor quality of the schools seem to be the major
causes of the repetition/dropout process.
About 20,000 indigenous peoples live in Maranhao. They belong to eight different ethnicities
that are not only culturally, but also linguistically diverse: Tenetehara/ Guajajara, Krikati,
Gaviao/ Pykobje, Apaniekra/ Canela, Ramkokamekra, Urubu/ Ka'apor, Awa/ Guaja, and
the descendents of the Timbiras. After the states of Amazonas and Mato Grosso, Maranhao
has the most numerous indigenous schools, a total of 157 in 16 municipalities, with 297
teachers-most are indigenous people recruited from within their own communities.
The inadequacy of indigenous education in the state of Maranhao is not very different from
Pernambuco-with almost no students enrolled in the secondary level and very few in the
last four years of the fundamental cycle. Indeed, the vast majority of students attend the
first four years of the first grades of the fundamental cycle (3,625 in the first year; 801 in
the second year; 351 in the third year; and 125 in the fourth year). Only very few (22) are in
the 5th year and none in any higher levels of education. Only in 2001, Maranhao began the
process of establishing the four final years of the primary cycle.
On the relationship of spatial location and fairness see Bigman and Diechmann (2000).
1 91



Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
The reasons for students not to proceed to the final four years of the first grade are not very
different from Pernambuco, including: (i) lack of accessible schools in the locality offering
the final four years of the primary education; (ii) by the fourth year, repetition has already
taken its toll and students are overage and lack the incentive to go further; and (iii) by this
time, they are considered adults and they are requested to perform adult roles. Also, the girls
sometimes get pregnant and will leave school.
In the case of Maranhao, as in the other two states, indigenous schools are run by the state.
However, as this responsibility is not as recent as in the case of Pernambuco, indigenous schools
are eligible for state-run programs, such as school lunches, school maintenance, and the training
and payment of indigenous teachers, whereas in Pernambuco, sometimes, they experience a kind
of institutional "limbo." Indigenous peoples also benefit from a program that trains community
members to participate in the school's administration. In addition, the state, in partnership with
indigenous teachers, produces teaching materials addressed to the specific cultural needs of the
groups. The production of these materials, however, is slow and inadequate, particularly given
the many different groups and languages involved. Nonetheless, some progress has been made.
In 1999, six Indigenous geography books were published and in 2000, three books for early
education were produced, two in Tenetechara and one in Timbira.
In Paraiba, although the schools attended by indigenous students are run by the state, in
contrast to Maranhao, the state has no policy regarding indigenous education and has
no criteria for defining what kind of school qualifies as "indigenous." Therefore, the state
does not supervise indigenous education, as required by MEC, nor does it enforce the
constitutional mandate to ensure cultural diversity. As in the two other states, dropout is
evident in the enrollment numbers of the first five years of the fundamental cycle: 994; 724;
508; 238 and zero respectively, according to the 1999 Census of Indigenous Schools.
Box 5: The Kind of School that the Indigenous Peoples Want
* Quality and diversity
* Adequate planning that meets the needs of students
* Respect for indigenous culture
* Modern teaching methodologies
* Curriculum and calendar that respects indigenous organizations, culture, and
economic needs
* Enthusiastic teachers
* School autonomy over financial resources
* Participation of parents
Sowrce Site veI, Marc2 0l2
92   1



Particularly Vulnerable Groups: Education And Schools  C * 
In Indigenous And Quilombo Communities  ;J
During a fieldwork visit for this study in Bahia da Traicao, a municipality with majority
of indigenous population, it was noted that none of the teachers working with indigenous
students had any special training. Culturally appropriate materials, as mandated by the
Constitution and by MEC are not available. The fieldwork revealed that the schools in Bahia
da Traicao are characterized by:
* No connection between what is taught in school and indigenous traditions
* No playful activities as part of the teaching process
* Excessive emphasis on making students memorize information rather than making
them think
* Absence of bilingual education
* Absence of state government responsibility for the policies and legal regulations
* Insufficient supply of books
* Misrepresentations of indigenous organizations and traditions
* Students have poor academic performance
Education in Quilombos
In the state of Pernambuco, Quilombos consist of a population of approximately 33,300
people5, distributed among 39 communities, but only five of them have a school. This does
not imply that children are not enrolled in some school, but points out the lack of attention
to providing them with easily accessible educational services.
Field visits and qualitative questionnaires filled by Quilombo students in Pernambuco
found a particularly worrisome situation among the young people. Respondents to the
study questionnaire were youngsters between the ages of 12 and 24, but all of them were
enrolled in the 8th grade. Clearly, the age-grade gap among Quilombo students is wider
than in the student population as a whole, reaching as much as 10 years or more in the
case of the 24 year-old students. As seen in Chapter 1, since repetition is the initial step of a
process that ends with complete dropout from the school system, the chance that children
and youth in this group have of being excluded is higher than in the general population.
' There is no existing data on the exact Quilombo population. However, data shows there are 8,318 families in Quilombo
communities, and it is estimated that each family comprises on average of four members.
1 93



Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
Three factors explain the observed age distortion when analyzing questionnaires completed
by 8th grade students. First, children living in Quilombos enter school later than the
officially prescribed age of seven years old. Second, half of the enrolled number of students
repeat grades, and up to three repetitions of the same grade is not uncommon. Third, some
of the students tend to dropout temporarily from school. This behavior is seen among
girls and boys, but in the case of boys it is more frequent and tends to extend over a longer
period of time. Some boys stay out of school for as long as seven years. In the case of girls,
the reasons they give for dropping out of school center around school-based factors such as
"no place available in the school" or "the school is too far from home." Boys systematically
justify their temporary absences from school in terms of work saying: "I had to survive" or
"I had to help my family."
Quilombo schools and students present some, but not all of the same characteristics of the
indigenous schools:
* Schools located in Quilombo villages are few abd only offer the first four years of the
fundamental cycle. To enroll in the final grades, students need to walk sometimes several
miles to the nearby municipal center. Most of them do not attend the final grades.
* Schools are impoverished . The physical conditions of the schools are appalling-they
lack electricity, latrines and running water, and often school meals. Because many of
these schools are unofficial, they do not receive funds from FUNDEF.
* Schools are multi-grade and teachers do not have training in this methodology.
* In contrast to indigenous children/youth, almost all Quilombo students, both girls and
boys, work some way.
* Constant repetition is associated with labor, distance and lack of quality of teaching.
Dropout is associated with the same factors, plus repetition.
* Enrollment is associated to the social norm that states one should be in school within
a certain age range and to incentives, such as the Program for the Eradication of Child
Labor (Programa de Erradicac,o do Trabalho Infantil - PETI) and school lunches.
* Quilombo students are similar to indigenous students in the sense that they both
feel that they suffer discrimination by "whites." Perhaps because of this similarity of
experience, indigenous and Quilombo students also disagreed about who suffers more
discrimination. In some cases, they even minimized the discrimination against the other
group as a way of underscoring their own discrimination. For example, "The indigenous
have suffered from discrimination but not as much as the blacks" (Quilombo youth);
"Blacks have already been able to win a place in society, much more than the indigenous"
(indigenous youth). For example, there are already black judges and lawyers. If there are
any indigenous in these professions, they are very few" (indigenous youth).
94    



Particularly Vulnerable Groups: Education And Schools
In Indigenous And Quilombo Communities
Conclusions
For the indigenous and Quilombo populations, educational exclusion seems to be
based less on the socioeconomic characteristics of the individuals, although they are
very poor, but rather on the characteristics of the group. Everything takes place as if
the proximity to their original status in the Brazilian society is still undermining the
achievement of their citizenship rights today. The problem is not that schools and
teachers are inadequate, but rather, that they are barely there, which for the most
part is the result of the institutional incapacity of the educational system to deal with
groups that are in a constant reaffirmation of a different identity, ancestral rights
and unbreakable pride. Principally, exclusion derives from the incapacity to listen to
different voices and, for those who do have a voice, to speak on behalf of the voiceless.
As land titling and the safety of their frontiers are the central axes of their lives, and as
these are, at the same time, daily sources of conflict, in the end, locating an indigenous
school or a school where quilombolas children and youth attend far away from where
they live, actually seems to generate more conflicts for those who stubbornly insist in
staying in their original land.
The main characteristics of educational exclusion among indigenous and Quilombo
students are frequent repetitions and high dropout. they also rests in the characteristics
of the schools, when they exist, and when the children or youth actually attending it.
Other characteristics of indigenous and Quilombo schools are as follows:
• Lack of emphasis on multi-culturalism, characterized by:
* Lack of connection between what is taught in schools and indigenous and Quilombo
traditions and history.
* Absence of sufficient bilingual teachers.
* School materials are culturally inappropriate and few indigenous schools use
specifically designed and culturally appropriate materials. Only four out of 39
Quilombos have schools in Pernambuco.
. Poor learning methodology:
* Too much emphasis on making students memorize information rather than
stimulating individual thinking.
* Absence of playful activities as part of the teaching process.
* Irregularities between schools, the municipal, state and federal government, with none of
them taking full responsibility for the follow up of policies and regulations in indigenous
education.
95



Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
. Insufficient supply of textbooks, poor infrastructure, lack of electricity, clean water, and
bathrooms.
. Indigenous
* One salient phenomenon is that indigenous students tend not to advance to the fifth
grade of the fundamental cycle or beyond. The reasons they do not proceed are:
(i) indigenous schools do not offer the secondary cycle of education in indigenous
villages, thus forcing students to leave their communities in order to complete their
education, (ii) by the fourth year, repetition has already taken its toll and students who
are overage lack incentives to proceed, and at this time they are considered as adults
and are requested to perform adult roles. In addition, indigenous students manifested
feeling prejudice from non-indigenous students with whom they would attend school
after the four first years.
* According to the qualitative fieldwork, the age-grade gap distortions are also very
acute among indigenous students. It is observed that they do not progress after the
first four years of fundamental education. Sixty-two percent of indigenous students
have from 1 to 4 grade failures.
* As indicated in the Census, of the indigenous students in Pernambuco, 43 percent are
in the first year of the fundamental level, 23 percent in the second year, 15 percent in
the third, 9.4 percent in the forth, and the remainder in the fifth grade and above.
* Quilombos. In the case of the Quilombo students who do complete the final four years
of fundamental education, racial prejudice is not felt as strongly, but those interviewed
mention the distant location of schools from their communities as one of the main
constraints to advance to fifth grade.
* Overage is particularly high among Quilombo students. As shown by the fieldwork,
only 6 of the 16 students in the 8th grade were 15 years of age or under. Age distortions
among the Quilombo students can be explained, first by their entrance in school later
than the officially prescribed age of seven; second, by frequent repetition; and third, by
temporary dropouts from school-a behavior more frequent and for longer periods
among boys.
96 1



w, /EX~
A - <^\t Ji'm       CHAPTER IV
TEENS AND YOUTH PERCEPTIONS ON
SfXr*^~~~~
~~XCUIO,INQALIY     N   POTNTE
~~~~~ ii ~ ~ ~ Tii SI-






Student perceptions on exclusion, inequality,
and opportunities
Objectives and Methodology
It is a truism that the young of today will be the citizens of tomorrow and that the processes
through which their perceptions are consolidated are at the heart of how they will act in
the future. Analyzing how students think about Brazil involves, to a large extent, analyzing
what a network of institutions-their families, their schools, the media, and their peers-is
teaching and preaching. A qualitative analysis was carried out of responses to a relatively
unstructured questionnaire filled out by 95 students (40 boys and 55 girls) aged from 12 to
24 years, all of whom attended the 8th year of the first (primary) grade in the states of Rio
de Janeiro and Pernambuco. The objective of the questionnaire was to gather information
about the way students see and evaluate the main issues related to exclusion on the basis of
gender (roles and relationships), ethnicity (blacks and indigenous), and social inequality
and poverty. All of these are groups suffer from acute disparities and discrimination in
Brazil, as described in Chapter I. The answers to the questionnaire also shed light on the
country's paramount problems and what solutions Brazil's future citizens envisage.
Although the survey did not aim to be statistically representative, it encompassed youth
from different social backgrounds and with varied life experiences. The questionnaire was
applied to students attending private and public schools, which, in Brazil, represents a
significant gap in both the quality of education and in the socioeconomic status of students,
not to mention in the physical conditions of the schools.'
The private schools where the fieldwork took place were located in Recife, the capital of
Pernambuco state, and in Rio de Janeiro,2 the capital of the Rio de Janeiro state.3 The public
school in Rio is located in a northern suburban area where poverty is clearly more acute
than in the southern neighborhoods of the city. In the case of Pernambuco, in order to
take into account possible variations in students' perceptions due to their racial and ethnic
' One of the private schools in Rio, for example, charges around R$ 2.000,00 per month, which is the equivalent of II Brazilian
minimum wages. The disparities in the quality of the infrastructure in public and private establishments can be deduced from
some of the students' complaints about their schools. While those attending public schools very often refer to their filthiness,
those attending private schools lament that there were no air conditioning units in some of the classrooms.
2 One of the private schools in Rio is a traditional establishment for middle-class students run by nuns. Lately, however, this
school has been granting scholarships to attract a less affluent clientele. Since it has a more diverse clientele than most private
schools, it will be referred to as "mixed private" whenever necessary.
3 Due to the similarities of responses from students attending these private establishments, their regional location is not
specified in our analysis except where necessary.
99



Children and YouthVulnerability Primeira Edi,ao
Poverty, Exclusion and Social Risks in Five Brazilian States
backgrounds, the questionnaire was also applied to teenagers living in a Quilombo (situated
in a rural area about 150 miles from Recife) and in an indigenous community (located in a
small town, Aguas Belas, about 180 miles from Recife). All of the indigenous students and
the students living in the Quilombo attended public schools.4 In this analysis, these students
will be referred to as Quilombo or indigenous students, while the designation "public
school" is reserved for students attending the public education establishment in Rio. Of the
total of 95 respondents, 43 attended private schools (24 girls and 19 boys); 20 came from the
public school in Rio (10 boys and 10 girls); 16 lived in the Quilombo (8 boys and 8 girls),
and 16 were indigenous (13 girls and 3 boys).
The following points should be considered given that they have an impact on the analysis
that follows:
* First, students' gender and the type of school that they attended (or the social class from
which they came) affected the way they expressed themselves. Besides writing better and
making fewer spelling mistakes, girls are, in general, more fluent than boys. They expanded
more in their responses, suggesting a tendency to be more reflective on the matters posed
by the questionnaire. Among the girls, the indigenous were the most succinct in their
answers but notably less so than the indigenous boys. Girls, moreover, seemed to be more
conscientious about responding to the questionnaire; most of the non-answers or the
terse "I don't knows" (which were particularly noticeable in responses concerning the
indigenous and the best/worst assets of their own genders) came from boys.
* Second, students from private schools were able to convey their ideas relatively clearly,
especially when compared to those from public, Quilombo, and indigenous schools. It
was difficult to decipher the writing of students from the latter schools, both in terms
of the calligraphy and content, and they made frequent spelling mistakes. At least in
part, the differences in how private and public school students express themselves can
be attributed to-and, at the same time, attest to-the disparities in the quality of the
education provided to rich and poor pupils.'
When the research was carried out, classes in the public schools attended by the Quilombola and indigenous students
were suspended because of a teachers' strike for better salaries. Therefore, the teenagers were contacted and answered the
questionnaire in the communities where they lived.
I Although extreme, the quotation below is illustrative of the difficulties to express themselves clearly. A boy from a public
school in Rio answers the question about why are there rich and poor people in Brazil: "Dependendo de quais pessoas, se elas
forem honestas ou nao. Porque se elas forem honestas dignas de ser um cidadao umas tem sortes e outras nao. Agora se e
por causa da corruptura eu considero essas pessoas falsas no caso dos politicos" (Depending on which people are honest and
which are not. Because if they are honest and worthy of being a citizen, then one is lucky and becomes rich, the other one is
not. Now, if it is because of corruption, I consider those people unreliable in the case of politicians")."
100 I



Student perceptions on exclusion, inequality,  \   U
and opportunities
Third, among Quilombola youth, girls gave more careful answers than boys, and their
handwriting was very clear. However, both made many spelling mistakes-even more than
the students from the public school in Rio and as many as indigenous students.6 Yet, both
boys and girls from the Quilombo seemed to take the task of filling in the questionnaire very
seriously. There were no questions that had been left blank, and, when the students did not give
a direct answer, they justified this carefully and some even apologized for it.7 Moreover, they
were sometimes very assertive in their responses, which reflects the fact that they had more
political awareness and social commitment than students from other kinds of schools. For
instance, although they disliked teachers' constant strikes, many of them understood why they
happened: "I don't like (in my school) the strikes. But the government is to blame: teachers
strike because they don't receive a fair salary, and in this case the students suffer the most. Not
because of the strike but because of the suspension of dasses" (girl). Nevertheless, when asked
what they did not like about their schools, some of them were able to generalize (something
that not even middle- and upper-middle-dass students did), stating, for example, that "not
only my school but all public schools in Brazil must improve the quality of instruction, to
teach more efficiently" (boy). In sum, beyond their spelling mistakes and the illegibility of
their handwriting, some of the Quilombola students presented surprisingly well-articulated
arguments.
Characteristics of the Respondents
The appropriate age for students attending the 8th year of the first grade in Brazil is 13 to
15. Almost all of the respondents from the three private schools and from the public school
in Rio fell into this age category; the only two exceptions were 15-year-old girls from one of
the private schools who had each repeated a grade once. Given that repetition and dropout
rates are high, particularly among the poor, the respondents from the public school in Rio
should be considered as exceptional.
In the case of indigenous and Quilombola students, they clearly demonstrated some of the
main reasons why poor students tend to do badly in school. Quilombola students enrolled
in the 8th grade who answered the questionnaire about their views were 12 to 24 years old,
6 For example, "nessefario" instead of nascenfa; "viulencia" instead of violencia; "misera" instead of miseria; "disenpengo"
instead of desemprego; "sovao" insteado of s6 vao; "emgearia" instead of engenharia; "baguza" instead of bagunfa, etc.
'As expressed by one of the boys: "I cannot speak about the indigenous because I don't know them very well. I only hear about
them on TV and that's why I don't want to speak of something I don't know. That's my answer. I am sorry."
I 101



Children and YouthVulnerability Primeira Edi,ao
Poverty, Exclusion and Social Risks in Five Brazilian States
and, on average, the girls were younger than boys. Only 6 of the 16 Quilombola students
were 15 years old or under, and among these, four were girls and two were boys. From those
who were 19 years old or more, four were girls and four were boys. Three factors explain
this age distribution among the Quilombola students. First, some seem to have entered
school later in life. Second, 50 percent have had school failures (up to three repetitions).
Third, some of them mentioned dropping out of school temporarily for up to seven years.
The role played by gender in this respect is noticeable; although both boys and girls suffered
from interruptions in their schooling, boys experienced them more often and for longer.
Furthermore, the reasons given for these temporary dropouts were clearly differentiated
by gender. While girls talked about "a lack of school places" and "school was too far from
home," boys inevitably justified the interruptions by stating that they had to work: "I had to
survive" and "I had to help my family."
Indigenous students were aged from 13 to 19 years old, and their relative youth compared
with Quilombolas seems to be due to the fact that neither the boys nor the girls had ever
been engaged in the labor market. The comings and goings in and out of school that is
characteristic of students who work does not prevail here, and there were no eventual
dropouts among these indigenous students. On the other hand, indigenous students, in
comparison with all the other groups who gave their views on exclusion, had the highest
rates of grade repetition, with 62 percent of them having from one to four grade failures.
The majority of respondents in the two states lived in households with a "standard
nuclear configuration" (two adults and their children),' although this format is more
predominant among those attending the private schools (84 percent) and less so among
,he indigenous (43.7 percent). Sixty-five percent of students attending the public school
in Rio and 62.5 percent of those from the Quilombo lived in this kind of household. There
was no significant difference between students from the private and public schools in Rio
in terms of living in single-parent homes, with 9.3 percent and 10 percent respectively
living in female-headed households. This contrasts clearly with the home situations of
indigenous and Quilombola teenagers, with 37.5 percent of each group living in female-
headed households. The "extended family" format (which consists of at least one parent,
usually the mother, and children as well as other relatives) is most prevalent among
students from the public school in Rio and among indigenous students, 25 percent and
8 Included in this category are cases in which the students lived with one of their parents and with a stepmother and a
stepfather. There are six such cases among students from the private schools and three from among those attending the public
school in Rio.
102 1



Student perceptions on exclusion, inequality,  -
and opportunities
18.5 percent respectively. In 10 percent of these public school households and in 12.5
percent of the indigenous households, the father was absent). In the three private schools,
only one of the students lived in an extended family (2.3 percent), while no Quilombola
student fell into this category.
Parent's Education and Occupation
Despite the fact that there was no direct question on family income in the questionnaire,
it is possible to infer students' social class through their answers to questions about their
parents' level of education and occupations. The significant majority of both mothers
and fathers whose children were attending the private schools had a college degree, and
students often used the expression "se formou" (graduated) to indicate this. A few parents
(around 16 percent) had a Masters degree, while others (also around 16 percent) had
only completed the secondary level. In households containing nuclear families, around
70 percent of both parents were involved in a professional activity. In the single-headed
families (both those with and those without other relatives), all but one woman worked
outside the home, totaling around 12 percent. The sum of these cases-reaching more
than 80 percent-points to the remarkable percentage of middle-class adult women
(mothers or stepmothers) who are engaged in the labor market. The mothers worked
as teachers, lawyers, business owners, biologists, technicians in medical laboratories,
secretaries, physicians, computer analysts, and dentists. Meanwhile, the fathers worked as
engineers, physicians, journalists, lawyers, computer analysts, public employees (mainly in
Recife), and executives, while one father was a director of a public enterprise and another
was a football trainer. It is noticeable that there is no acute gender distinction either
in terms of formal instruction or in terms of occupations among upper, middle, and
upper-middle-class parents. Only very few women were involved in what is considered
"traditional feminine" professions and, with some exceptions, both genders were engaged
in careers that require college degrees and/or are generally regarded as prestigious.9 There
is no available information on whether mothers earned less money than the fathers, but
9 Once more here, some students from the "mixed private school" in Rio pointed to exceptions, uor example, that their fathers
and/or mothers, when working outside home, were engaged in semi-professional activities (such as being a vendor in a
store). This reconfirms the relative social diversity of the students attending this particular private school. Nevertheless, there
is no doubt that most of them are middle class (for example, their mothers had jobs such as accountants or administrative
consultants ).
103



t  $)   Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
it is interesting that teenagers from the middle- and upper-middle classes are very aware
of gender discrimination in the job market (see below).
Answers from students attending the public school in Rio reveal sharp differences from
students' in private schools in terms of their parents' education levels and occupations.10
The majority of the parents of students in Rio's public schools had never completed first
grade; a few had attended second grade but only two had completed it. The expression "se
formou" that middle- and upper-middle class students use to denote that their parents have
a college degree was sometimes used by teenagers from Rio's public school to indicate that
their parents had completed first grade, as if adjusting their expectations to reality.
Sixty percent of the mothers of these students did not work outside home, even those
who had a higher degree of formal instruction than their husbands. The great majority
of mothers who worked outside the home were the heads of their households. To put it
in reverse, in households where both parents were present, only a minority of women
(25 percent) worked outside home. The greater "domesticity' of mothers from the lower
classes (their role as homemakers) is a clear contrast with the middle and upper-middle
class mothers with children in private schools. This characteristic, as will be discussed
below, seems to have an important impact on how girls attending the public school see
gender issues. The fathers of students from Rio's public school work as plumbers, bakers,
electricians, clerks, or hold informal occupations in the streets. The few mothers who did
work were cooks, seamstresses, and receptionists. Thus, both the fathers and the mothers
were basically engaged in semi-professional activities; that is, in jobs that do not require
a higher educational degree and/or which have a lower rank on the occupational prestige
scale. It is also noteworthy that at this level, women's and men's occupations are clearly
more differentiated compared to the occupations of middle-class men and women.
The fact that the poor have only limited access to educational opportunities is even more
perceptible in the Quilombo. The number of mothers and/or fathers who were illiterate is
significant (around eight cases). Most of the remaining parents had attended elementary
school, but only very few had completed it. Almost all of the fathers worked as masons
(only one was a rural worker, farming together with the rest of the family), which may be
evidence of their lack of employment alternatives. The great majority of mothers did not
° Some students from the public school (as well as some from the "mixed private" school) did not know the level of education
attained by their mothers and/or fathers. In two cases, the teenagers explained their ignorance by saying that they did
not know their fathers, but there were a lot of "don't knows"- particularly about fathers' levels of education - among the
Quilombola and indigenous students.
104 1



Student perceptions on exclusion, inequality,  |
and opportunities  a  J
work outside home, including some who headed their households. The few women who
worked outside home were involved in informal and inconstant work whenever available.
Some worked in their own fields or vegetable gardens (roca) for subsistence; however, the
students did not consider this kind of occupation as "outside' work" but rather as part of
housework. Quilombola mothers, thus, were more domesticated than other mothers, which
seems to have consequences in the way in which their daughters perceived gender issues.
It can be said that poverty (in terms of both means and resources) is more acute in the
Quilombo than in urban areas.
Indigenous students' parents had a wider range of educational achievement than the
Quilombola parents, from four cases of illiteracy to cases of mothers and/or fathers
who had attended the second grade (six according to the students, had even completed
it). Nevertheless, most of these parents (around 50 percent) had attended but had not
completed the first grade. Their involvement with work inside and outside the home is
particularly interesting. In 10 of the 16 indigenous households, no adult worked outside
the home, and in the two households that contained a nuclear family, both parents had
jobs. Only three indigenous fathers received salaries, and of these, one worked as a driver
at FUNAI, one worked as a wall painter, and the third worked in the fields. Only six
women were engaged in a professional activity, and even those who were the heads of
their households did not necessarily work outside the home. When they did, one took with
her sewing machine to others' homes, two worked at FUNAI, and three were teachers in
the indigenous school located in the area."' Thus, the mothers of the indigenous students
were also notably domesticated. Despite the fact that the majority of fathers do not work,
"domesticity" is not a pertinent way in which to categorize their situation since, as noted
later on, indigenous girls very often complained that men-fathers and brothers-do not
do any domestic work. This information is also crucial to understanding indigenous girls'
perceptions about gender.
In sum, the analysis of parents' levels of education and their occupations reveals a clear
overlap between public school attendance and being in a lower-income group, on the
one hand, and private school attendance and being in a middle-and upper-middle
group, on the other.
" One of the indigenous girls complained that there were very few alternatives for women; they are either teachers or health
agents, always at FUNAI.
105



Children and YouthVulnerability Primeira EdiSao
Poverty, Exclusion and Social Risks in Five Brazilian States
Child and Youth Labor
In spite of their socioeconomic and cultural differences, there is a striking similarity
between students in the schools in Rio de Janeiro and indigenous students; with very few
exceptions, they do not work. Students from the public school in Rio were well aware that it
is common for poor young people like themselves to work and that this labor often results
in those students dropping out of school, and some of them indicated that they were aware
of being in a privileged position. "I like going to school, because I have the right [sic] to
study, and many people don't" (girl from the public school). Their recognition of the effort
and costs implied for their parents is translated into a sense of obligation and a desire to
reciprocate. When answering questions about their "dreams" in the future, many say they
wanted a "good job" to be able to help their parents "in their needs," "so that they can stop
working," and/or to "buy them a house." 12
The privileged situation of indigenous students and of students from the public school
in Rio in this regard is particularly noticeable when contrasted with those living in the
Quilombo, of whom 6 out of the 16 students (five males and one female) were engaged in
outside work, especially those aged 16 and over. However, even more important than age,
gender seems to be a key determinant in this matter; while females stayed at home and were
responsible for home chores (see below), young men were charged with doing the "outside"
work. Indeed, the only Quilombola young woman who worked outside the home was a
monitor of the official Program for the Elimination of Child Labor (PETI), while the five
boys worked in slaughterhouses and informal jobs. One of them said he was "unemployed."
Two of them worked in the fields with the rest of their family in subsistence farming, but,
as already noted, students considered this to be part of domestic work. Those who received
a salary earned, at most, one minimum wage. While the few urban students who worked
said that they spent their wages on "personal expenses," the Quilombola students stated that
they "buy food" and "help the family."
Apart from boys and girls from private schools and indigenous boys, all of the other students
said that they had to do domestic chores. Nevertheless, there were differences among these
students in terms of the kind of chores they had to do. Teenagers who attended the private
schools mentioned, regardless of their gender, being responsible mainly for "making their
beds (and, more rarely, tidying their bedrooms) and washing the dishes." In contrast,
12 This same sense of obligation is expressed by some of the Quilombola students: "I want to help my family in their unhappy
life" (boy).
106



Student perceptions on exclusion, inequality,
and opportunities
those from the public school not only had more varied chores to do but also, especially
in the case of boys, heavier ones. The boys mentioned "cleaning bathrooms," "waxing the
floor," "washing the rugs," and "weeding the backyard," while the girls alluded to "cooking,"
"washing clothes," and "helping to clean up the house," as well as being in charge of their
younger siblings. So it is noticeable that the domestic chores of students from the public
school are much more differentiated by gender than those of private school students. Taken
with the evidence above on the occupations of the students' parents, this leads to a more
general conclusion-that the distinction between the feminine and the masculine domains
tends to be sharper among the lower class than among the middle class.
Looking at the domestic chores of Quilombola and indigenous students reinforces this conclusion.
Girls were less specific than boys about their home chores, very frequently stating that they were
involved "in all" household tasks. It is possible to take this answer as evidence that they are more
involved in housework than middle- and upper-middle-class girls. Boys, by contrast, were much
less involved with domestic chores and their chores were clearly differentiated from the ones
assigned to girls. The Quilombola boys unambiguously worked "outside"- either working in
the fields (which, as noted, is considered by students to be part of housework) or taking care
of the animals. Considering that more Quilombola boys than girls were in the labor market, it
can be said that a distinction is made between feminine/the inner space and masculine/outer
space both in the domestic realm and outside it. The three indigenous boys who answered the
questionnaire either did not help at home or had a very few chores such as "taking the garbage
out." The marked differentiation between the feminine and masculine domains is, therefore,
very perceptible among the indigenous students.
Despite the fact that they had domestic chores to do, both students from the private schools
and from the public school in Rio referred to their homes as places where "one has nothing
to do" and, for this reason, they sometimes saw them as "boring." This is one of the reasons
why the majority of students preferred going to school than being at home. Socializing
with friends and "guaranteeing a good future" (stressed particularly by students from the
lower classes) were also invoked to justify their preference for school. Quilombola students
unanimously chose school as the better place to be. None of them referred to their homes
as places of leisure and, even more important, they explained their preference by stating
that in school one "learn things." Although this assertion is not peculiar to the Quilombola
students, it sounds stronger and less rhetorical coming from them compared with the
answers from middle- and upper-middle-class students: "My future is there, in spite of
everything" (boy, 19 years old); "Study is very important and I would have liked to complete
the second grade, but as I have said (referring to interruptions) I didn't have time" (boy, 20
years old); "In school, you become aware of the importance of learning to read" (girl). The
1 107



) X    Children and YouthVulnerability Primeira Edi,o
Poverty, Exclusion and Social Risks in Five Brazilian States
fact that illiteracy, and its consequences for prospects in the labor market, is a very familiar
experience to the Quilombo dwellers seems to illustrate, at least in part, the value imputed
to formal instruction by students from the Quilombo. Middle-class students, in contrast,
seem to take education for granted as an indisputable privilege.
Ethnicity
The last general characteristic of the respondents that we want to examine is "ethnicity"-
or, in more typical Brazilian terms-"skin color." The questionnaire posed this issue in an
open way, requiring students themselves to clarify its meaning. The answers reveal Brazilian
idiosyncrasies regarding the matter of race compared with the U.S., for instance. In American
culture, a dichotomous logic governs the perception of race, dividing the population into
two mutually exclusive groups: black and whites. In contrast, in Brazil, a logic of gradation
prevails. Between the two poles of black and white, there are other intermediate categories
such as "morenos," "pardos," and light.'3 The teenagers' answers indicated that they prefer
imprecision. One of the female youth from the public school in Rio even described herself
as "close to white." Only five students (5 percent of the sample) considered themselves to
be "blacks," and two of them were Quilombo dwellers; 42 percent designated themselves
as "whites" and the remaining 53 percent classified themselves as in-between, mostly as
"morenos" with the inevitable gradation from "dark" to "light" moreno. There is no perfect
correlation between skin color and type of school (or social class). For example, 31 percent
of the students from the private school in Recife, which is undeniably an upper middle-
class establishment, defined themselves as "morenos." On the other hand, a clear correlation
between these variables is perceptible when the self-definitions of the students from one of
the private schools in Rio are contrasted with the self-definitions offered by the students
from the public school in the same state. All but one of the private school students classified
themselves as "white" (94 percent). In contrast, the majority of the students from the public
school (60 percent) defined themselves as "morenos" or "mulatos;" 15 percent said they
were "black" and only 25 percent considered themselves as "white."
13 One of the respondents in Recife refused to classify himself in terms of color, alleging that "color does not exist." This is, by
the way, a frequent way of expressing opposition to color discrimination or "racism:'
108 I



Student perceptions on exclusion, inequality,   | _
and opportunities
'What makes "ethnicity" even more complicated is the fact that probably the meaning of
"moreno" is not necessarily the same to all students; or to put it in more general terms, skin
color does not necessarily have the same weight in the construction of their identities. As
noted below, race seems to be most important to Quilombola students in terms of defining
their identities, regardless their self-ascribed skin color.
Perceptions About Gender and Inequity
Gender is an important dimension of social classification. This classification, however,
is not neutral (such as having blue or dark eyes) and students point out stringent social
discrimination against women, whereas men are still viewed as holding the responsibility
as breadwinners.
What Female Youth Think About Gender and Inequality
According to urban students, the most commonly mentioned "asset" that women have,
regardless of whether they go to private or public schools, was their ability "to give life to
a new being" ("a gerac,o de um novo ser"). Conversely, not having this "asset" was seen
as the main disadvantage of being a man, since men are "unable to experience such an
emotion."14 However, nature was not always perceived as siding with women. Almost all of
the female students cited menstruation and cramps (and some others also add the pains of
childbirth) as one of the greatest hardships endured by their gender and, conversely, one of
the advantages of being a man: "They don't menstruate,' "Men only attend childbirth and
don't feel any pain and don't have to have a huge scar caused by a cesarean."
Some females also mentioned benefiting from certain social privileges such as "Not having
to enroll in the Army," "Having preference in lines thanks to the saying: ladies, first," and
"Being spared from hard or heavy work." It is not always clear whether these prerogatives
rest ultimately on natural differences between genders or are merely a product of cultural
forces. Nevertheless, a slight difference between poor and non-poor students emerged
on the issue. While the girls from public schools tended to emphasize differences based
'4 Girls very often referred to this feminine "gift" stressing its physical aspect; that is, they stressed in particular the physical
experience of pregnancy or of "having a small person inside your belly."
1 109



Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
on nature (they are more at ease in referring to men as the "strong sex"), the ones from
private schools were more reluctant to do so. The quotation that follows, from one of the
latter women, expresses on one hand, an ambivalence between not endorsing a stereotype
and, on the other, recognizing the advantages of being considered the fragile sex: "Women
have an image of being a weak and fragile sex, even if they are not. That's why we women
are excluded from certain things in which, really, it is not agreeable to participate as, for
example, hard jobs, those that require a lot of physical strength."
For natural or social reasons or both, the fact is that females are aware of the burdens
carried by men due to their representation as the strong sex, literally or as the breadwinner.
This is, according to many of the male students, the most difficult thing about being a
man: "Sometimes men are overburdened in occupations that require physical strength or
whatsoever," "Society overcharges men with the responsibility of maintaining a family,"
and "People think that (men) are machines, and put too much pressure on them." These
attitudes prevailed even though the great majority of middle-class mothers not only worked,
but also did so in professions similar to men. Despite most middle- and upper-middle-class
mothers working outside their homes, the idea nevertheless persists among the girls that
men are ultimately more responsible than women for the family's livelihood.'5
Other than hardships stemming from their biological condition, young women evoked
social constraints as an additional disadvantage of being a woman. Girls perceived men
as having more freedom than women. But, here again, some differences are noticeable
among urban girls depending on their social class. Those from private schools repeatedly
emphasized man's greater physical freedom and stated that men do not have to pay as
much attention to their appearance as women do; they "Can urinate almost everywhere,"
"Can wear larger and more comfortable clothes," "Don't have to care about appearance
as much as we do," and "Don't have to wax their legs." Girls from the public school, in
contrast, referred to men's freedom more frequently: "They can go anywhere," "They are
less controlled than us," and "They don't have to explain where they go to their wives." In
addition, girls from the public school affirmed that gender differences are more marked
in the lower class than in the upper classes. They also often asserted that, because they are
1 The answers from the girls attending the private school in Recife suggested a clearer separation between men's and women's
domains, although, as we have seen, all of their mothers worked outside home. For instance, some girls said that one of the
bad things about being a woman is that "men don't help in taking care of children and the home." It is well known that the
Northeast region tends to be more culturally conservative than the Southern region, and this is also manifest in the relations
between genders.
110 I



Student perceptions on exclusion, inequality,
and opportunities
stronger," men "can defend themselves" and/or "have less chance of being assaulted and
raped." These statements indicate that the threat of violence is a very familiar experience
to them. Yet, the worst social disadvantage suffered by women according to these girls is
discrimination either from men (some of whom still consider women to be "inferiors"
and/or "mistreat them") or in the labor market (difficulty in finding a job or wage
discrimination). Although some girls stressed that a few positive changes have occurred
in recent years, they insisted that discrimination persists. For instance: "Although they are
perfect idiots, it's easier for them to find a job" (girl from Recife).'6 "Even nowadays, there
are women who suffer prejudice, for example, in that they earn a lower salary in spite of
working the same hours as men" (girl from a private school). "Women suffer prejudice for
being women because there are people who still think in medieval terms, believing that
women should pipe down in front of men, and because women are still not allowed to
occupy certain jobs" (girl from a private school).
Here, again, there is a very intriguing socioeconomic differentiation among these
females. Although some young females from the public school in Rio de Janeiro explicitly
mentioned women's discrimination in the job market, their major complaint was a
sort of "discrimination" [sic] that exists before they go out into the labor market? their
confinement at home. Two young females pointed out the fact that "men can stay away from
home the whole day" while others stated that: "No woman likes to be a homemaker. They
like to work outside home, as men do" and "(The worst thing about being a woman) is the
discrimination; for instance, many people think that women were born to take care of the
home."
It is interesting to note that Quilombola and indigenous females were particularly
radicalized by their dissatisfaction with the relationship between genders. Many of
them stated explicitly that there is no advantage in being a woman and/or that there is
no disadvantage in being a man ("everything is good for men"). Some mentioned their
discrimination in the labor market, but the justification for their perception of gender
differences went far beyond that. First, they unanimously believed that men enjoy "greater
(social) freedom." "When he wants to go somewhere, he goes and no one can stop him." "A
woman is more locked up at home and can never go where she wants without her parents'
" Some girls in Recife pictured men as "jerks" (babacas) and "gross." Correspondingly, one of the bad aspects of being a
woman was to have to "bear men" or "bear men's stupidity':
111



5"      Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
permission." "Mothers don't pick on them and they do whatever they want." "Girls are too
governed by our brothers just because we are women."17 Second, these women stressed more
than women from other backgrounds that women are mistreated by men and experience
"humiliation," "abandonment," and sexual abuse and domestic violence. "Women suffer
violence and are obliged to do things that they don't want to."'8 "When (the couple) has
discussions, men beat women" (indigenous girl); "Men suffer less (referring to betrayals of
love)"and "Men "are stronger and thus can defend themselves from bandits and robberies."
Third, they pictured men as "womanizers," "violent," and "egocentric." They only think of
themselves," "Gross," "Aggressive" and "Don't give love to women."
Indigenous females added a specific and very recurrent complaint that men are "lazy." They
"do nothing besides eating and sleeping," they "wake up whenever they want and are too
lazy to go to school." They do "not help women at home, either with the children or with
housework." Fourth, although some Quilombola young women recognize that they are
spared from the hard work that is expected of men (including being the breadwinner),
neither they nor the indigenous girls idealized or had romantic fancies about their gender.
On the contrary, some claimed a link between a woman's role and slavery. "(The worst thing
about being a woman) is having to marry and take care of the house, children, and husband.
I don't want this for me. I want to marry but I don't want to be a slave" (Quilombola female).
This aversion to women's confinement to the home is widespread; only one Quilombola girl
identified taking care of home as an advantage of her gender. Indigenous girls in particular
emphasized that "women work too much in the home without men's help." In sum, their
discontent is the result of the combination of women having to work too hard in the home
and men's laziness. Finally, Quilombola and indigenous females hardly ever mentioned
women having any "natural" assets; only two of them were enthusiastic about "having
children." Nor did they stress any female disadvantages based on "nature," but they tended
to stress instead the social dangers that women face.
'' Some of the Quilombola and indigenous girls answered the question on women by thinking of themselves as daughters.
However, the indigenous girls were the only ones who referred to their brothers as controlling them. The high rate of
single-parent homes among this community may explain, at least in part, this peculiarity if brothers take the place of
fathers in these girls' lives.
1" This statement seems to be an allusion to sexual abuse since this same Quilombola girl said that the worst aspect of men
was that "they are violent".
112 l



Student perceptions on exclusion, inequality,
and opportunities
What Young Males Think About Gender and Inequality
Males, particularly those from the mixed private school, the public school in Rio de Janeiro,
and the indigenous schools, were more laconic than girls in their answers regarding gender
and were sometimes even ironic. Some answered with a terse: "I don't know." Others, when
facing the questions on the female gender, gave the same response: "I don't know because I
have never been a woman." Otherwise, many of their responses on feminine and masculine
genders coincided with the answers given by girls. They agreed that "giving life to a new
being" and "menstruation, childbirth, and PMS" were respectively the best and the worst
aspects of being a woman.
The burden of being the main if not the sole family breadwinner was stressed by many
young males from the private, public, and Quilombo schools as the major responsibility
implied by their gender. "Women don't have to sweat so much everyday." Quilombola boys,
in particular, added the fact of being poor to their definition of being a male, which may
be an indication of their difficulty in separating the two conditions. For example, some
students said that the worst aspect of being a man is to feel responsible for "Being poor,
unemployed, and with no education " of having to "See his children go hungry" and "When
he reaches adolescence, he has to get a job."
All of the males acknowledged that men have an advantage in the labor market and in
other spheres as well. As noticed by the Quilombola boys: "(Women) have won the right to
be citizens, but there is still a lot to be done. They are still very discriminated against" or,
stressing the double discrimination suffered by Quilombola women, "They suffer from a lot
of moral and racial prejudice." Further, the boys all recognized that they are "more free" than
girls, and they exemplified this masculine asset in terms of: "not having to wear a bra," and
"we can pee standing up" and in terms of being less controlled ? "We can go anywhere."
The concept of freedom is particularly noticeable in the writings of males who attend the
public school, as is the perception of themselves as the "strong sex." An interesting contrast
is offered by the statement made by an upper-middle-class boy who wrote that the major
burden of being a man is that "people say that (men) are strong, shouldn't cry and should
have no feelings. In many situations, they say one thing, but think otherwise due to this
social imposition." To put it in more general terms, while males living in poverty tended to
explain the differences between men and women by referring to natural factors (which are
hard to change), the boys from the middle- and upper- middle classes seemed to be more
inclined to invoke cultural constraints.
113



Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
In contrast, males attending the Rio de Janeiro public school and the indigenous school
very rarely expressed awareness of discrimination against women. Only one male from the
Rio public school mentioned discrimination against women in the job market ("for us, it's
easier to find jobs"). Another youth from the public school agreed with the assessment of
the girls in noticing that one of the worst things about being a woman is "to be a housewife
because I think that women want the same freedom to work outside the home as men do."
One of the indigenous males said that one of the bad things for a woman is that "she has to
do everything at home." Besides these exceptions, there were no other signs in the answers
of the boys from the public school and the indigenous school that they had any perception,
much less any criticism, of gender inequity. From this point of view, it seems that females
from the same social class or ethnic group as these boys have good reasons to complain.
However, these remarks should be taken carefully since what the boys do not say may be as
important as what they do say.
Perceptions of Ethnic Discrimination and Inequality
The analysis that follows is based on youth views on color and discrimination answering two
open questions. Do you think that black and white people have the same opportunities in
Brazil? Second, they were asked to write what do you know and think about the indigenous
in Brazil. In general, students agree that there is racial discrimination in Brazil, that blacks do
not have the same opportunities as whites, and that the white society oppresses indigenous
peoples, including by taking their land. These assertions, although sometimes expressed
naively, reflect an increasing concern about racial inequalities in Brazilian society"9.
Color, Discrimination, and Poverty
Three out of the 20 males that attended schools in Rio did not answer this last question on
indigenous ("nothing to declare"), while another one simply stated that "it must be cool."
Aside from the indigenous students, only one female from the mixed private school in Rio
de Janeiro admitted to being descended from indigenous ancestors.
" See Silva (2001). The author indicates increasing awareness about racial discrimination with recent surveys showing that although
the majority of the surveyed affirm not having prejudice, they also affirm the existence of discrimination in the country.
114 1



Student perceptions on exclusion, inequality,
and opportunities  zJjj
Only 5 percent of the respondents described themselves as black, while a significant number
-(53 percent) identified themselves as belonging to the intermediate categories ("morenos,"
mulattos, and "pardos"). It is difficult to tell whether these young people perceive themselves
as being closer to black than white or vice versa. Some of them did refer to black people as
"they," but the majority of the students, regardless of their self-ascribed color, responded
to the question on discrimination without the use of "we"/"they." Quilombola students,
not surprisingly, were the exceptions to this. Although their self-ascribed color did not
differentiate them from other students (only two of them saw themselves as black, while
the majority saw themselves as "morenos"), being descended from slaves undoubtedly is an
important part of their identity. Thus, in answering the question about what makes them
proud to be Brazilians, 25 percent of the Quilombola students (all of them "morenos")
responded "my color, for sure." In other words, much more than color, the fact that these
teenagers were part of a Quilombo is crucial to their identity in terms of their "ethnicity."
However, it is noteworthy that there were no significant differences in the contents of
students' answers regarding racial discrimination when self-definition in terms of skin
color was taken into account.20
Box 6: What Students Think About Racial Discrimination
'I think that prejudice is very wrong because no matter the color of the person, everyone
is equal" (white boy from private school).
"They are people" (eles sao gentes) (morena young female from public school).
"[Black people should have the same opportunities] because everybody is flesh and
bones and we are all children from the same father who is God" (Quilombola female).
All teenagers expressed the view that black and white Brazilians do not have the same
opportunities. One exception, curiously, was a "moreno" from the public school in Rio who
wrote that "we all have the same opportunities, we have only to fight for what we want."
All of the others maintained that "racism" is widespread throughout the country. Even
those few who spontaneously maintained that there have been some improvements in race
relations added that "discrimination persists." Many commented on the unfairness of this
situation and explicitly emphasized that they were against racial discrimination-usually
saying that color doesn't make (or shouldn't make) any difference. Their assessment of the
20 Students classified their own color in an open question in the questionnaire.
1 115



Children and YouthVulnerability Primeira Edi,co
'  . ,^t  Poverty, Exclusion and Social Risks in Five Brazilian States
erroneousness of this discrimination is based on the idea, expressed mainly by students from
private schools, that differences between blacks and whites are "a creation of the whites"
(white boy) rather than a rule of nature. Since it is a social construct, it can and should be
changed. There was a consensual agreement that whites are the agents of discrimination.
"Almost all white people don't like blacks" and "many white people think that blacks are
worthless" (Quilombola girls). "Many white people are bosses and they see black people
as slaves; they want to give them orders because of their color" (black girl from the public
school). Most of those who justified their antagonism to racial prejudice did so by stressing
the "equality of all human beings."
Students immediately identified being black with poverty. One middle-class white boy
stated bluntly that "racism engenders poverty." Others depicted blacks as stuck in the same
vicious circle as the poor (see below); that is, since black people are poor, they have few
educational opportunities and thus, are unable to get good jobs so they remain poor.2' Most
students, on the other hand, explicitly emphasized that race adds to poverty, which is a
further complication of race relations in Brazil.
Discrimination is especially noticeable in the job market, but racism goes far beyond the
workplace. The first two quotations below reminds, as does the last one above, that youth
not only hear about discrimination against blacks, but they also witness it. "There is a lot
of racial prejudice in this country. Just look around: how many blacks are there in this
school?" (white boy from the private school); "if a black person tries to go to a very high-
class restaurant in a neighborhood where there are only rich white people, most likely he
won't be allowed in. White people have more privileges than black"; (black boy from a
private school). "there is a lot of color prejudice in Brazil, both regarding job admissions
and membership admission in clubs" (white boy from a private school). Finally, "the whites
have priority in everything, and the blacks always come second place" (Quilombola female
student).
21 Yet, in continuing his reasoning, this same boy slipped into racism: "That is why black people don't even try to find a job or
to organize their lives in an honest way." This is, by the way, the only student who showed this kind of bias.
116 1



Student perceptions on exclusion, inequality,
and opportunities
Box 7: Ethnicity and Poverty According to Students
The majority of black people are poor, and this makes it difficult to them to take
advantage of opportunities" (white boy from private school).
"Brazil is a very prejudiced country, and many people can't find jobs because of their
color. Also, very often, they are mistaken for thieves, criminals, and murderers" (white
girl from private school).
"Discrimination in Brazil is high. Let's imagine that a black and a white person are both
trying to get a job in an office. It is obvious that the white person will win" ("pardo"
female from public school).
"We have, as an example (of the discrimination against blacks), the stores, supermarkets,
and pharmacies in Garanhuns: the majority of the people who work there have light-
colored skin and are very good-looking" (Quilombola female).
Some students showed some sense of connection to history when, in speaking about blacks
in Brazil, they referred to the slave system that operated in Brazil in the past. Slavery can be
seen either as a cause of present racial discrimination or as evidence of the persistence of
the unfair conditions endured by black people throughout our history. For instance, "black
people are more discriminated against because they used to be slaves" (Quilombola girl).
"The idea of the black as an inferior being comes from slavery and still persists in the minds
of many" (girl from a private school). "Due to our slavery heritage, prejudice still exists"
(white boy from a private school).
Indigenous Peoples and Discrimination
As the idea that all people are born equal prevailed among the students, they deplored
prejudice against the indigenous, whom they believe are discriminated against. However,
students hardly ever thought of themselves as having an indigenous heritage. Indigenous
peoples are thought of as "different" and as the other.
In general, all non-indigenous students tended to think alike about indigenous people,
making comments ranging from a neutral "I have nothing against them" (black girl
from the public school) to more emphatic words of sympathy. The imaginary qualities
1 117



Children and YouthVulnerability Primeira Edi,ao
C   /     Poverty, Exclusion and Social Risks in Five Brazilian States
with which the students endowed indigenous people include "pacific:' "intelligent,"
"extroverted," and "brave," and they also "admire" them for their "culture," "harmony
with nature," and for "preserving and "taking care of our forests." Females, in particular,
very often felt "sorry" for them.
Many teenagers and youth in discussing the indigenous focused on the "terrible and
unfair conditions" in which they live. They lamented their increasing impoverishment,
deplored the fact that "they are being extinguished:' and regretted that nobody cares
for them. "They don't live as normal people [sic] because nowadays the indigenous
settlements are very precarious. Many don't even have schools and they live as illiterates
without working. They eat only what they plant or fish. But since there are many of them,
they often experience hunger and they suffer because they are aware that, despite this
horrible situation, nobody cares for them."
The quotations above show that for some students, indigenous peoples are almost
synonymous to poor to the point of confounding the two categories. From this point of
view, some students envisaged the same solutions for the problems of the indigenous as they
had suggested for the poor: more schools, more jobs, and better housing. Others, instead,
felt that the problems of the indigenous stem mainly from land disputes. As a result, these
students blamed "whites" (including miners and the "landless movement") and, once more,
the government, saying that while the whites "attack" and "steal from" the indigenous, the
government and politicians "don't do anything about it."
Almost all students pictured the indigenous as being culturally different from "us." In
other words, the distinctiveness of indigenous culture is precisely what defines them,
and, according to some, this is the reason why they are discriminated against. "They are
discriminated against because of their culture" (student from the public school). Teenagers
reckoned that the indigenous have this separate cultural identity because "they learn it from
their parents since they don't have schools." "They believe that rivers, the sun, and nature
are gods." "They are not as developed as we are: they don't have shopping centers, produce
only what is necessary to survive, and survive on selling their crafts".
118 I



Student perceptions on exclusion, inequality,  I
and opportunities  .  $-
Box 8: Indigenous People and Land According to Students
"The indigenous are fighting not to have their lands invaded by the landless movement,
by the gold seekers, and even by white men who go there to cut trees to make things like
pencils...." (male youth from public school).
"It's an absurd invasion of their territories by white man. Each of us has our own space
and should not trespass on the space of another. We should leave them alone so we don't
harm them" (female youth from public school).
"White men are taking their land, forcing them to go to the cities to find somewhere to
live" (boy from private school).
However, there is also a pervasive recognition that the indigenous are "losing their culture."
Some students saw this as having a negative impact on cultural diversity in Brazil, while
others saw this as a positive development precisely because the indigenous will be able to get
rid of the factors that make them poor. In other words, some of the students valued cultural
diversity and its preservation and, therefore, explicitly regretted the fact that the indigenous
are becoming integrated into Brazilian society. Some even suggested that contact with whites
is pernicious to the indigenous since, as a result, they are exposed to new diseases; thus, the
best thing to do is to "leave them alone." Others admire "their resistance to technology"
(girl from a private school). "The indigenous are losing their customs and traditions. This
makes me sad because I think that one of the most valuable things a group of people has is
their culture. If they lose it, they stop making a difference as a particular ethnic group and
become the same as everybody else. I think that the indigenous should go naked or wear
their native clothing" (girl from a private school).
On the other hand, there are those youth who praised the indigenous for "adapting" to
"civilization." These students did not see this as a problem but in fact as part of the solution
to indigenous poverty. "The government forgot that the indigenous are also human beings as
we are. It is government's obligation to help them to move out of their isolated communities
and bring them closer to the rest of the world. And the government should help them to
find housing and jobs." (male from the public school). "People insist on thinking of the
indigenous as characters from tales who only wear loincloths. But if they paid more
attention to the indigenous, they would realize that they dress like civilized people do and
want the same rights that we have. Therefore, these people should try to help the indigenous
to acquire those rights as some people already do" (boy from a private school).
1 119



Children and YouthVulnerability Primeira Edicao
Poverty, Exclusion and Social Risks in Five Brazilian States
Box 9: Views on Indigenous People from an Urban Student with (Recognized)
Indigenous Ancestry
"They were landowners and land was taken from them. They were superstitious [sic],
had their own religions and beliefs, and their culture has been destroyed. We owe so
much to the indigenous because they gave us the land, and in return we humiliate them,
burn them, and throw them off their land. My great-grandparents were indigenous
and Portuguese and, in spite of that, they got married. I am very proud to be their
descendant because they really are people who deserve a country like Brazil."
Indigenous Peoples' Rights. Students used three kinds of arguments to justify indigenous
people's "right to rights." First, despite being "different," the indigenous are equal to "us"
because they are "human beings" (or, as stressed by some Quilombo dwellers, "because
they are the children of God"). As a result, according to students, they should be given,
not only respect but also equal rights and opportunities. "There is no reason to think that
they are not equal to us just because they have a different culture from ours and belong to a
different ethnic group" (girl from a private school.) Second, they were the first inhabitants
of Brazil. This "right of those who arrived first" was one of the reasons invoked by teenagers
to claim that the indigenous deserve the same opportunities, merit being helped, and/or
are entitled to land. "They were the first inhabitants of Brazil and now they don't have even
one-tenth of the territory. So they have the right to expect to be helped by others" (girl
from a private school). And third, "our culture owes a lot to their culture " as one student
wrote, underscoring a sense of obligation. "They have contributed a lot to the culture of our
country. Many words of our vocabulary, for instance, come from the Tupi language. They
are a marvelous race that should not disappear" (boy from a private school). "Many of the
things we wear, eat, drink and practice come from them like the chimarrao, football, and
the peteca. And we should follow their example since they don't harm nature" (youth from
the public school).
120 1



Student perceptions on exclusion, inequality,
and opportunities
Perception of Indigenous Students About Indigenous People
From the answers given by indigenous students in their questionnaires, they did not have
very different views about indigenous people compared to those of the non-indigenous
students. However, there were one or two exceptions. First, indigenous students are much
more forthcoming on the discrimination that they suffer because of their "color," "race," and
"beliefs." And, in speaking of this subject, they often mention their own school-labeled by
some of them as a "school for whites"22-as a place where they experience discrimination.
Box 10: What Indigenous Students Think Non-Indigenous People Think about Them
"My classmates - not all but some - keep saying that the indigenous are lazy and boring
and this and that" (young female from privat school).
"They (white students) never want us to participate in collective tasks at school because
they say that the indigenous don't know how to speak and when they speak, they don't
speak correctly" (young female from privat school).
"They say that the indigenous are ignorant and should not study in the same school as
whites. But no prejudice will make me deny my origins" (young indigenous female).
Second, they repeatedly declared themselves to be "proud" of being indigenous. "The best
thing that has existed in the country: they are brave, and I'm very proud of being indigenous."
And third, they are the only ones who re-wrote the question in the questionnaire (What
you know about Indigenous People in Brazil) to refer instead to the "Brazilian indigenous."
They sometimes were very clear about claiming their double identity: "If we had more
opportunities, and if our talents were properly recognized, we would be seen as worthy
people, but we wouldn't stop being what we are: indigenous" (female youth).
2 The indigenous school in Aguas Belas only offers the first four years of the primary cycle, but for the remaining four,
students need to go to schools outside the indigenous neighborhood together with other students of the municipality. A visit
that was made to that area confirmed the discrimination mentioned by indigenous students. Staff in the local Secretaria de
EduLcacao mentioned that they are lazy, dependent on the government, and are unwilling to work. As has been seen before, this
is a view shared by indigenous girls themselves at least with regard to males. It is true that among the indigenous in the region
there is a sense that the government owes them their survival in return for their land that "has been stolen."
1 121



Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
Differences and Similarities in the Views About Indigenous and Blacks
There is an important difference in the way in which students referred to racial and ethnic
minorities-compared to blacks, the indigenous are clearly more "distant others."23 First,
indigenous were systematically referred as "they" in contraposition to "we." Second, quite a
few of the students state that they "know very little" about indigenous people or that they
never "had the opportunity to meet one." The distance between the non-indigenous students
and indigenous is noticeable in the remark of one of the boys from private school who wrote
that, if he had the chance to meet an indigenous boy, he "wouldn't be afraid of him." Third,
while the students never refer to a particular and distinctive black culture, they implied that
the indigenous were culturally different from "us." Even more important, while blacks are
seen as victims of social and economic discrimination, the indigenous are deemed to be
suffering from social exclusion. Three girls from private schools explained their ignorance
about indigenous people by stating, "in our society they are very much excluded, which
prevents us from knowing more about them." Two boys from private schools said explicitly
that the indigenous are "marginalized" or "are excluded" from our society, and another boy
suggested that "the indigenous have not won a place in our country"-meaning, maybe,
that they are not yet part of the nation. A boy who attends the public school, after writing
about the discrimination from which the indigenous suffer, concluded, "they are not very
present in Brazilian life." In sum, the students portrayed the indigenous very differently
from the blacks as being apart or as a non-part of the nation.24 Some indigenous teenagers
seem to agree: "we don't have a definite place in Brazil" (female youth).
231 am not including the replies of the indigenous students in these observations.
24 It is also interesting to note that they never speak of "Brazilian Indigenous " but rather of "Indigenous in Brazil." However,
the question that we asked contained the same bias. Only one boy from the private school in Recife spoke of "Brazilian
Indians" - besides some Indians students themselves, as noted above.
122 1



Student perceptions on exclusion, inequality,
and opportunities
Box 11: Why Poor People are Poor According to Students
"The majority of the poor don't have the opportunity to study because the money that
should go to education is spent elsewhere" (boy from private school.)
"Income is unevenly distributed, the job's supply is low and the few (jobs) available are
taken by those who have a higher educational level" (boy from private school.)
"We all know that there are poor and rich people in all countries, but in Brazil's case, as
well as in others, the degree of poverty is higher due to a general lack of education. The
government is responsible for this." (Quilombola female.)
"Brazil is a country with few industries and too many social inequalities" (Quilombola male.)
Finally, Quilombola youth are closer to indigenous students in the sense that they both suffer
discrimination by "whites "(as we ) indigenous are very discriminated against by the whites"
(Quilombola female). Maybe precisely because of this similarity of experience, they also
disagreed about who suffers more. In some cases, they even minimized the discrimination
against the other group as a way of underscoring their own discrimination. "The indigenous
have suffered from discrimination but not as much as the blacks" (Quilombola boy);
"Blacks have already been able to win a place in society, much more than the indigenous.
For example, there are already black judges and lawyers. If there are any indigenous in these
professions, they are very few" (indigenous female). Moreover, there are some subtle mutual
criticisms. One Quilombola girl, in an exception to the way in which most students referred
to the indigenous, wrote that, "What I know about the indigenous is that they do whatever
they want and don't think about anybody else. I think that they should think first about
God." And conversely, an indigenous boy wrote, "I think that some black people don't want
to have the same opportunities as the whites."
Perceptions of Social Inequality
Perceptions of poverty were approached by asking two questions. First, why are there poor
and rich people in Brazil? And second, how can a poor person rise out of poverty? When
answering the first question, the students raised five issues, regardless of what kind of school
they attended: first, unequal income distribution, which is one of the trademarks of the
country; second, unemployment; third, fortune (or lack thereof); fourth, corruption or
plain robbery; and fifth, the egotism of the wealthy.
1 123



Children and YouthVulnerability Primeira Edicao
Poverty, Exclusion and Social Risks in Five Brazilian States
Inequality as the reason for poverty
Most of the teenagers from the public school emphasized the "unevenly distribution of
"income" (or "capital" "payments" or "money"). Those attending private schools, as well
as Quilombola and indigenous students, tended to be more specific. They thought that
social disparities were caused by differences in opportunities or "privileges"-particularly
in education and, as a result, in the labor market. "Some people have more teaching and
education than others." "Some people have more privileges while others have none." Some
students even stressed a sort of vicious circle, saying that the poor are poor because they
have few opportunities in life and they have few opportunities because they are poor.
Poverty and unemployment are regarded as the same, particularly by Quilombola and
indigenous students. "There are poor and rich because some people have a steady job,
while others don't" (Quilombola male). At the same time, since illiteracy is a very familiar
experience to them, these teenagers were particularly sensitive to the role of a lack of
education in promoting unemployment and, thus, poverty. Recognizing that education
means becoming literate, one student wrote, "nowadays for almost all jobs the person must
know how to write and read" (indigenous male).
Youth also very often underscored the fact that social inequalities, far from being due to
merit, are caused by some kind of unjust destiny. "Life in Brazil is unfair."25 This means that,
regardless of their social origins, the students did not blame the poor for being poor. None
of them suggested that the poor are lazy or are responsible for their fate. On the contrary,
the students diagnosed poverty and inequality as social problems "created by men" [sic],
and they depicted those who were experiencing material deprivation mainly as victims of
unfair conditions. As a result, pessimism pervades students' answers on this subject, which
is perceptible in their assertions about a deepening socioeconomic gap. "The rich are getting
richer and the poor are getting poorer." Two indigenous females were extremely fatalistic in
answering the question about why are there both poor and rich people, stating that "God
wanted it to be so" or "It's God's will."
25 The only answer that ascribed social disparities to merit came, curiously, from a girl attending the public school. "Half of the
population knows how to administer their money, and the other half doesn't."
124 1



Student perceptions on exclusion, inequality,
and opportunities
Together with (and as important as) differences in opportunities and/or the uneven
distribution of income, almost all students invoked "robbery," "exploitation," "the bad
government that rules the country," and particularly "corruption" as explanations for social
inequality.26 For example, "There are people who steal from others, they exploit people
and don't give the poor any opportunities to make money and also because of corruption.
Finally, some students-particularly those from private schools and those who live in the
Quilombo-attributed social inequities to the selfish egoism of the wealthy, who are seen
as lacking social commitment. "The wealthy don't show any interest in giving up something
they have to help the poor." "The wealthy don't understand that with a minor contribution
they could help a lot of people" (Quilombola young woman). "The rich do not share the
money with the poor: that's why they have nowhere to live." (Quilombola young woman).
"This is a question that we don't need to speak too much about and it can be answered with
three words. the Government, the powerful, and their relationship" (Quilombola young
man). "The rich want to be richer and don't want to give any opportunities to the poor.
They think that because they are rich and have studied in better schools, they are more
capable" (Quilombola female). "The rich are egoists and do not offer jobs to the poor"
(Quilombola boy). At the same time, some of the most deprived remember that they are
dependent on the wealthy. "The poor must be humble because they depend on the rich (for
jobs)" (Quilombola male).
How Can Poverty Be Overcome
In answering the second question, how can a poor person stop being poor, youth emphasized
two factors. First, they answered that individual effort and will are key. They saw "studying
hard" and "running after" a good job as means by which the less privileged could improve
their situation and even reduce the distance that separates them from the wealthy. Second,
but no less important, youth answered good fortune, a factor that is out of the control of
individuals.27
21 Six girls and boys from the private schools in both Rio and Recife mentioned "capitalism" as an explanation for the existence
of rich and poor in Brazil, and elsewhere. "All capitalistic societies have large social disparities. To say that there are economic
classes [sic] means that people do not have the same conditions and opportunities in life."
Quilombola students mentioned "God" - his help or faith in - even more than luck.
1 125



Children and YouthVulnerability Primeira Edicao
Poverty, Exclusion and Social Risks in Five Brazilian States
Box 12: How Can the Poor Stop Being Poor?
"The poor should work hard to reach stability in life, although this is almost impossible
to achieve nowadays"
"When the poor own their own business and do not have too much expenses because
with our earning we can only pay the expenses"
"This is impossible in Brazil"
"Leave Brazil"
These two considerations are not mutually exclusive, and students sometimes referred to
both. However, those who exclusively mentioned luck or who gave it the highest priority
clearly believed that personal effort is not powerful enough to overcome social forces and/or
poverty. In other words, they perceive Brazilian society as being too rigid to allow upward
economic mobility. The quotations below exemplify this pessimistic opinion. One girl
pointed out that not even the lottery is accessible to everybody: "It is very complicated (for
a poor person to become rich) because sometimes people are poor because they don't have
enough education to guarantee themselves a good job in the future. So the way out is to win
the lottery. But some of them aren't even able to afford to participate in the game and
never win" (girl from private school). Another girl wrote, "In most cases, a poor person
becomes rich due to luck. Because nowadays social discrepancies are so strong that is very
difficult for a person to overcome poverty exclusively through work" (girl from a private
school). Finally, from a Quilombola young female, "Only through a miracle of God (can
a poor person become rich)."
The public school, indigenous, and Quilombola students were more likely than those from
private schools to stress the importance of individual's efforts rather than luck on surpassing
poverty. They emphasized "working with honesty and with dignity" and "with dedication
and determination" as a way the poor could better themselves. However, this emphasis was
not frequent or strong enough to conclude that these students believed that it was possible
for poor people to prevail over poverty by sheer strength of will. Nor are they actually more
optimistic than the students from the private school. On the contrary, their awareness of
Brazil's resistant socioeconomic structure is clear from their remarks about how little the
poor can achieve as well as in the insurmountable difficulties that they face.
126 l



Student perceptions on exclusion, inequality,
and opportunities  .
Perceptions of Brazil
Overall, students are gloomy about Brazil, Brazilian society and Brazilian politics. Also in
general, they convey a certain despair about its future.
Brazil's Major Problems: Inequality
In their responses to the first question, youth reiterated that they perceive social
inequalities as the country's major problem. Also, "racism" or "discrimination"
were frequently mentioned, particularly by white, indigenous, and Quilombola
students28 but, paradoxically, less so by non-white urban students. Some students,
particularly those from the Northeast (indigenous, Quilombo, and private schools), also
cited drugs and, to a lesser extent, prostitution (particularly mentioned by indigenous
and Quilombola students) as national problems. They also mentioned inequality and
discrimination as reasons why they are less than proud of being Brazilians.
However, youth wrote more expansively about social disparities between rich and poor:
Many of them, particularly those from private schools, referred to "social inequalities"
directly, while others pointed out some of its consequences or manifestations such as the
fact that the poor have little access to health care, and suffer misery, illiteracy, and famine.
Quilombola and indigenous students, for reasons already suggested, put a lot of emphasis on
the lack of education as a major impediment to getting a job. Some of the students' writings
conveyed the impression that, even more than poverty, Brazil's paramount problem is the
gap between the poor and the rich, particularly given the rigid stratification.
28 It should be noted that "discrimination" was often used imprecisely. Students used it to refer either to particular groups
(women, blacks, or the indigenous) or, more generally, to the deprived - "the discrimination against the Brazilian people
[povo]" (Quilombola girl).
127



\ .>  Children and YouthVulnerability Primeira Ediao
Poverty, Exclusion and Social Risks in Five Brazilian States
Box 13: Inequality: Brazil's Paramount Problem
"The major problem is the great difference between social classes. Very few have a lot,
and many have very little" (boy from private school).
"It seems that we are going back to the Middle Ages when those who were born poor died
poor and those who were born rich, died rich" (boy from private school).
Brazil's Major Problems: Violence and Unemployment
The students, particularly the Quilombola and indigenous students and students attending
the public school in Rio, also ranked "violence" and "unemployment" as important problems
in Brazil.29 This seems to indicate that teenagers from the lower class are more exposed to
these problems than richer students, even though they included kidnapping (which is more
likely to affect the rich than the poor) in their definition of violence along with robbery and
murder. One of the young students from the public school in Rio was even more specific,
stating that what made him least proud of his country was "the death rate in the slums" as
an indicator of violent crimes.
Students from the private school also often mentioned "violence" as one of Brazil's most
severe problems, and they tended to regard it as a symptom of social inequity. "Poor people
with no education cause violence," "The poor steal because they are dying of hunger," and
"Hunger generates violence." In naming violence as a major problem, middle- and upper-
middle-class students seemed to be aware of the negative consequences of poverty not only
for the poor themselves but also for society as a whole.
Brazil's Major Problems: Corruption and Politicians.
Together with (and as important as) these social and economic problems, students also very
frequently referred to the government's incompetence and/or mismanagement as one of the
country's major problems. "Politicians' disrespect and lack of consideration for the poor"
and "the inactivity of justice that does nothing to protect or defend the citizen" were also
among the issues that made them less than proud of their country. However, most of all, and
almost unanimously, students underlined a moral problem ? corruption in government and
29 Quilombola students even mentioned violence inside their own schools.
128 1



Student perceptions on exclusion, inequality,
and opportunities
politics - or to put it in a Quilombola male's words, politicians"'lack of ethics." All students,
regardless of their gender, social class, or region mentioned "corruption" in their answers
to the questionnaire.30 Yet the students referred to the "government" ("the racketeering
and mess of our government") not only as the root of the problems but also as the answer
to those problems. They considered corruption to be such a serious and uncontrollable
problem that one girl from the mixed private school maintained that "even if the whole
government is replaced, corruption will persist." Corruption is even often blamed for the
social inequality among classes since it diverts public funds away from education and/or
from the creation of jobs.
These considerations lead to who is to blame for the country's miseries. As already
mentioned, although some students invoked the selfish egoism of the wealthy, they all
regarded politicians and government as the main villains. They blamed the government
for everything-from inequality between the rich and the poor to the problems of the
indigenous-either because they take no action or because their actions are misleading
(there are rich and poor "because of the President and of the Government" and "because
politics helps the rich more than the poor " (Quilombola boys).
How to Overcome Brazil's Major Problems
Not surprisingly, therefore, youth answers to "how do you think the major problems in
Brazil can be solved' regardless of gender, social class, racial or ethnic background, stressed
the need for changes in government and politicians, mainly by electing more moral
representatives.
God, morality, and punishment are the main recommendations the students had for
solving the paramount problem in the country-social inequality. However, teenagers also
identified some other feasible actions to overcome this problem. However, while students
from the private schools and also some Quilombola teenagers suggested that public funds
should concentrate on education, those attending the public school, the indigenous
students, and some Quilombola students emphasized the need for government to create
jobs. Indeed, a significant number of students manifested a clear belief in education as
the most efficient corrective for poverty and inequality and not just in a spirit of radical
3 Compared with the other students, the indigenous made fewer references to the government - either as the villain or
the savior.
129



N         Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
egalitarianism. According to one of the girls, "It is utopian to believe that the problem of
social inequality can be completely solved." Some proposed that the government should
provide "educational opportunities to all." The poor would then be able to take advantage
of these opportunities to acquire qualifications that, when used in the labor market, would
help them to climb the social ladder and, thus, lessen social disparities in Brazil: "The
government should increase educational opportunities because in doing so people would
be able to attend a good university and increase their chances of getting a good job (girl
from private school).
Box 14: What Needs to be Done in Brazil
"More dignity in politics"
"Increasing punishment for corrupt politicians"
"Putting all of them in jail"
"Having honest and good people in the government"
"Electing better politicians and government than the ones who are harming the country"
"Honest politicians could change the country"
"Praying to God"
The public, indigenous, and Quilombola students in contrast felt that the government
should prioritize the creation of jobs as a solution for social inequity. Two of them, a male
and a female, went even further in suggesting the adoption of a labor-intensive industrial
policy: "Instead of using machines, we should put them aside and give more jobs to people."
The main conclusion to be drawn from their comments is not that these students do not
value education but rather that the problem of unemployment is very close and familiar to
them and/or their social class.
130 1



Student perceptions on exclusion, inequality,  '
and opportunities
Who Can do What?
Despite the fact that there were no specific questions on who should be responsible for
taking action to solve our problems, students' answers gave some indications on the subject.
They seem to feel that what could, and should, be done cannot be left to the poor themselves
since they are mainly seen, as already pointed out, as victims of major social forces. A few
students said "All of us." A few others also proposed that the wealthy should make some
sacrifice, either through redistribution or philanthropy: "We should distribute the leftover
goods" (boy from a private school.)
From the perspective of the vast majority of students, regardless of gender and social
class, civil society either "has nothing to do" with solving our problems or, at most, should
carefully elect the best representatives possible (thus stressing the importance of the vote).
According to a boy and a girl, both of whom attended the public school, "We can solve the
problems in Brazil by voting in someone who is worthy" and "All we can do is choose our
representatives well in the election."
Box 15: Brazil Today
"The crisis is huge: Brazil is going from bad to worse. And it seems that it's going to
get even worse. I don't see any solution to the country's problems" (male from public
school).
"I was born in Brazil, but since my father is Italian I have opted for Italy as my nation.
See, if Brazil were a decent country, I would have opted for it ... I don't feel proud of
Brazil in any way" (male from private school).
31 Blaming the state for the persistence of poverty and inequality is a recurrent finding in Reis' research on Brazilian
elites. The author shows the feeling of alienation that elite members have towards the state, even when they are state
players (Reis, 2001).
131



'> H; sk Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
In turn, it is indisputable that students regard the "Government" and "Politicians" (and
sometimes, the President) as being responsible for solving our problems.3" These people
are seen, therefore, both as the major villains and the major rescuers of our country. For
instance: "We can help but we cannot solve (our problems). The people who have to solve
them are the politicians, who are becoming worse every day" (boy from a private school)
or "We cannot solve them, but I think that the government could do something about (our
problems)" (Quilombola girl); and, "We cannot resolve our problems, but the politicians
can" (girl from the public school.)
The Source of Pride for Brazilians
Given youth awareness of the severity of our problems and inequities, it is not surprising
that they draw their pride in being Brazilian from sources that are not linked to collective
achievements but rather are the result of nature in the sense that they are "endowments."
The only exception is football (particularly mentioned by boys), but many added that even
the Brazilian team and/or its politics is nowadays a motive for "shame." Thus, the reasons
cited most frequently (and in fact monotonously) for having pride in Brazil were, its natural
beauty and specific character traits attributed to Brazilians, particularly the persistence of
their "joy of living" despite the difficulties that they have to endure. It is noteworthy that
this good humor - interpreted as a sort of resilience - binds together rich and poor as if
concealing the social differences that separate them. "No matter whether it is a rainy or a
sunny day, the rich and the poor are always smiling" (girl from a private school).
Students, in sum, seem to be implicated in a perverse vicious circle. They are, on the one
hand, aware of the multiplicity and magnitude of the problems faced by the country, but,
on the other hand, they put the solutions in the hands of those who are, according to them,
the least reliable of all-the government and politicians. Therefore, the pervasive lack of
hope, trust, or reliance in their writings is not surprising. "Nothing" is the answer some boys
gave in answer to the question about what gives them pride in Brazil, and many others said
explicitly that they saw "no solution" to the country's problems.
132



(1 _. -  -o            -V     Jn   Pena
.iX  a  t--  ;>d      1 \
E a
Maria Madalena Rodriguez dos Santos
i  'I l;Xr (fIP
NA)






Conclusions
This study indicated that a high degree of educational exclusion persists in Brazil despite
the fact that educational coverage has undoubtedly expanded. A disturbingly large number
of school-aged children and youth in the five states analyzed still do not enjoy full access to
the educational system, either because they do not attend school at all or when attending,
they repeat grades several times, eventually giving up and dropping out entirely. The levels
of dropouts are related to high repetition rates, but also to the lack of incentives to continue
attending school, as pointed out in Chapters 1 and 4. In light of this discrepancy between
increasing coverage and persistent exclusion, one of the main challenges for educational
and youth development policy makers policymakers in Brazil is to reach those children and
youth who do not have full access to the educational system and provide them with the
benefits of education and with future employment opportunities.
General Conclusions
The following general conclusions can be drawn from this study:
* The combination of the three factors that define education exclusion-never having
attended school, having at least a three year age-grade gap, or having already dropped out
of school-produce a large number of the 7-17 age group at risk of youth exclusion in
the five states: 46 percent in Maranhao, 40 percent in Pernambuco, 39 percent in Paraiba,
36 percent in Rio Grande do Norte and 22 percent in Rio de Janeiro. The number of
youth at risk in the five states adds up to 2.6 million-a staggering figure.
* The main determinants of educational exclusion, from a statistical point of view,
are socioeconomic. Although in different age groups and in different states, specific
determinants may exist, household income, parent's schooling, and premature entrance
in the labor market are consistently the main determinants of educational exclusion in all
five states and in all age groups.
* Vulnerability to exclusion increases among females in rural areas and among non-whites
in Rio de Janeiro. In some young age groups, inter-municipal and inter-state migration
are also disruptive and contribute to exclusion.
* The qualitative fieldwork indicates the existence of strong segmentation between private
and public schools. On one hand, private schools tend to be more homogenous in terms of
the students' socioeconomic background, which includes parents with college education
and, in the case of Rio de Janeiro, being white. In these schools, classrooms are clean,
repetition is less frequent, the demands on teachers' performance are high, the repertoire
of subjects in the curriculum is broad, and parents have a stake in the education of their
children and they demonstrate it. On the other hand, public schools are heterogeneous,
1 135



(; ,)>    Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
attended by students from different ethnic groups, races and socioeconomic strata. In
rural areas, particularly in the Quilombo and indigenous communities, age differences
in the same classroom are very noticeable, classrooms are overcrowded, school buildings
are precarious, water, latrines and electricity are very often missing, and school lunches
are unpredictable and insufficient.
* Well-trained teachers in public schools frequently work shifts in private schools as well,
where a higher performance is demanded from them, while maintaining their jobs in the
public schools, which sometimes pay higher wages and offer better benefits. This seems
to be one more subsidy from the educational system to those who need it less.
* Although poor parents are critical of the schools attended by their children, they face
income, schooling, power and time constraints to act as an active pressure group to
advocate for a better education for their children.
* The main elements contributing to high student performance in poor schools are related
to how disciplined and well-run the schools are: schools where students perform better
are those where students and teachers are not allowed to arrive late, classes begin on time,
students are given homework and tests, feedback is provided, and teachers follow each
student's performance.
* Both parents and students operate within a dual system to classify their priorities. The first
is based on their own private conditions, the second, on the public conditions of the country.
What is deemed as belonging to the public realm, thus demanding public interventions,
is viewed as a problem and seen as a priority for governmental action. Thus, parents
and students consider the problems of corruption and urban and rural land violence as
priorities. In contrast, poverty, unequal gender relations, domestic violence and bad schools,
are considered a sad reality not subject to action, since they are understood as part of an
unchangeable destiny. This does not imply, however, that they are not important, but that
they are viewed as immutable conditions that nothing can be done about.
Specific Conclusions
The following specific conclusions were reached through an analysis of the relationship
between poverty, ethnicity, and educational exclusion in the five states:
Exclusion in Absolute Numbers
* Those who are excluded are poor in their majority. However, it is not the poverty of
their households that accounts for their exclusion but rather the lack of institutional
136 1



Conclusions
mechanisms that are able to create for them the same educational opportunities and the
same educational equipments that are in place for the non-poor.
Educational coverage is quite high in all states, representing a substantial improvement
over time as shown when comparisons are established with older age cohorts.
However, the size of the population that has never been to school is larger in
the Northeast region than in Rio de Janeiro. This is true for all age cohorts but
particularly for those older than 26.
* Attendance in pre-school (nursery, kindergarten) is far from satisfactory. Four-fifths
of children under age five have never been to school. It is curious that a few thousand
children in each state have already been enrolled and dropped out from school at such
early stage. Although those children will most likely return to school, this illustrates how
early the fragmentation of the school experience begins, which particularly characterizes
poor students in the Northeast.
* Part of this fragmentation is due to the high levels of repetition, which eventually lead to
definitive dropout. About one-fourth of all students in Rio de Janeiro were at least three
years behind the grade expected for their age with higher percentages in the Northeastern
states for example, one out of every two students in Maranhao have an age-grade gap of
at least three years.
* In the Northeastern states, nearly all children and youth who left school did so before
completing fundamental education, in comparison with only 10 percent in Rio de
Janeiro. The proportion of dropouts increases steeply from 12 years old and on. Typically,
going from 11 to 12 years old increases the dropout rates from 1 percent or less to 2
percent or more. Perhaps not surprisingly, this cohort has the highest overage students in
most states.
* There are differences in the reasons for dropouts between males and females, which also
vary according to age cohorts. For instance, for the 7 - 9 cohort, being a male migrant
increases the probability to dropout, but for females, the variables are income and
parent's education. For the 10 - 14 cohort, the only difference between genders is being
an interstate migrant, which increases the dropout probability for females, but not for
males. For the 15 - 17 cohort, location is a variable for girls in rural areas, who are in a
more disadvantaged situation than boys.
* There are differences in the reasons for dropout between whites and non-whites
and they also vary according to age cohorts. However, many variables are equally
important for both. For the non-white 7 - 9 cohort, the probability to dropout
is partly dependent on the migration status and parent's education. For whites,
the only difference is household income. For the 10 - 14 and 15 - 17 years old
cohorts, there appears to be a higher degree of differentiation between ethnic
backgrounds. For the 15 - 17 cohort, it also matters if the youth lives in the rural area.
1 137



Children and YouthVulnerability Primeira Edi,co
Poverty, Exclusion and Social Risks in Five Brazilian States
Lessons from Rural Schools
* Parents in rural areas, although they see themselves as voiceless, are neither happy with
the status of their children's education, nor with the way the schools are run.
* Fieldwork conducted with parents, teachers, principals and students in rural schools
demonstrate that: (a) teachers and principals are extremely positive about their own
performance, in contrast to critical parents and students who are unhappy with
the education system, their schools and their teachers; and (b) small, but consistent
differences exist between schools with high student performance (top schools) and the
poor student performance (bottom schools). The most significant differences are: (i)
how knowledgeable teachers show to be with regard to what they teach; (ii) the school
calendar is not usually disrupted and classes are not interrupted; (iii) teachers plan their
classes in advance; (iv) Math and Portuguese share the bulk of the teaching attention; (v)
classes are linked and teachers refer to contents from previous years and other disciplines;
(vi) teachers correct exercises and provide students with feedback; (vii); students are
asked to do homework and receive feedback; and (viii) students receive constructive
criticism and constant positive reinforcement.
Indigenous and Quilombo Exclusion
* The case of indigenous and Quilombo students included in this study clearly shows some
of the main traits that affect the school progress of those who have been excluded for
centuries. High overage rates among students can be explained by at least three factors:
(a) some seem to have entered school later than the expected age; (b) there is a high
incidence of grade repetitions; and (c) temporary dropouts, clearly differentiated by
gender, are particularly high among boys and in some cases last for very long periods.
All these factors are related to their deep poverty but also to the exclusion they face from
public institutions.
Recommendations
* The 15-24 age group represents one-fifth of the Brazilian population. Many are not in
school, and many others, who are enrolled, are not acquiring the appropriate skills for
the labor market and will be unable to expand their knowledge as required by the new
dynamics of globalization and advancing technology. However, education for youth has
only recently become an area of concern. Effective policies to help address youth lacking
138 1



Conclusions
adequate schooling should include quality schools but also, particularly in the short run,
training programs to upgrade job skills.
* The school system should define clear roles and responsibilities for all those involved in
delivering its policies and mandate. Legal and financial and management accountabilities
need to be established to boost school quality and encourage accountability among
parents, students and taxpayers. Continuous but innovative training is needed for those
involved. Some experiences with school autonomy have shown positive results when all
of these conditions are in place.
* While there is still a need to increase access to education in rural areas, in both urban
and rural areas, the quality of education, teachers and curriculum relevance are other
sources of concerns that must be considered policy priorities. In rural schools, education
policy should focus on solutions to address their isolation, unqualified teachers,
obsolete teaching methodologies, paucity of teaching support systems, and over-age
students. In urban schools, education policy should focus on integrating parents,
promoting autonomy, creating an interactive environment, and opening student's eyes
to contemporary issues of labor market, violence, drugs, and technology.
* A set of specific supply and demand policies are needed for schools in indigenous and
Quilombo areas. The main sources of concern are: location, infrastructure, teacher's
training, accessibility up to grade four, and respect for cultural identities. These issues
deserve attention from states, which can seek more appropriate solutions with support
from the Ministry of Education and community involvement.
139






R-             C ;{ 
J A  N ( Ž )   ,-  t  ,'  -( S  
'n  0_%R  j  C  ai
C .L         I !, ,_X )fp 
s~~ C)    fiS'i,  l






Bibliography
Araujo e Oliveira, Joao; Schwartzman (2002) A Escola Vista por Dentro. Alfa Educativa
Editora. Belo Horizonte.
Bigman, Davud; Uwe Deichmann. 2000. "Spatial Indicator of Access and Fairness for the
Location of Public Facilities." In David Bigman and Hippolyte Fofack, Geographical
Targeting for Poverty Alleviation. The World Bank (Washington D.C.)
Corseuil, Carlos Henrique; Santos, Daniel D., and Foguel, Miguel N. (2001) "Decisoes
criticas em idades criticas: a escolha de jovens entre estudo e trabalho no Brasil e em
outros paises da America Latina" Economia Aplicada, v. 5, n° 4 (Outubro - Dezembro).
Departamento de Ecocomia, USP/FIPE. Sao Paulo, SP, p. 819 - 860.
Costa Ribeiro, Sergio (1993) "A educacao e a insercao do Brasil na modernidade', Cadernos
de Pesquisa n° 84, p. 63 - 82, Fevereiro, Sao Paulo, SP.
De Leon, Fernanda Leite Lopez and Menezes Filho, Naercio Aquino (2001) "Uma analise
descritiva da repetencia, avanco e evasao escolar e de seus determinantes para o Brasil no
periodo de 1984 a 1997", Texto para Discussao n° 20/2001, Universidade de Sao Paulo,
Instituto de Pesquisas Econ6micas, Sao Paulo, SP 18 p. (processed)
Fletcher, Philip R. (1997) "As dimensoes transversal e longitudinal do modelo profluxo",
MEC - Departamento de Estatistica Educacional, Rio de Janeiro, RJ, 32 p. (processed)
Fletcher, Philip R. (w/d) "A demographic perspective on duration and participation in
Brazilian education" 68 p. (processed).
Iwakami Beltrao, Kaiz6 (2003) Alfabetizac,o por ra,a e sexo no brasil: um modelo linear
generalizado para o periodo 1940-2000. Texto Para Discussao N. 1 IBGE/ENCE
Johnston, Jack and DiNardo, John (1997) (4th edition) Econometric Methods. McGraw-
Hill International Editions, Economic series. New York, London, etc.
Klein, Ruben (1995) "Producao e utilizac,o de indicadores educacionais" Laborat6rio
Nacional de Computacao Cientifica (LNCC/CNPq), Preliminar, Fevereiro, 35 p. +
pictures (processed).
Magnacv, Thierry and Thesmar, David (1998) "Identifying dynamic discrete choice models:
na application to school-leaving in France" Preliminary version. CREST, Paris, Franca 49
p. (processed)
Paes de Barros, Ricardo and Pinto de Mendonca, Rosane Silva (1996a) "O impacto da
gestao sober o desempenho educacional" Banco Interamericano de Desenvolvimento,
Abril, 48 p. + pictures and tables (processed).
Paes de Barros, Ricardo and Pinto de Mendonca, Rosane Silva (1996b) "O fluxo educacional
no Brasil", IPEA - Instituto de Pesquisa Econ6mica Aplicada, Junho 11 p. (processed).
Paes de Barros, Ricardo and Pinto de Mendonca, Rosane Silva (1996c) "Trabalho infantil
e evasao escolar", IPEA - Instituto de Pesquisa Econ6mica Aplicada, Junho 9 p.
(processed).
143



Children and YouthVulnerability Primeira Edi,ao
-o o   ~ Poverty, Exclusion and Social Risks in Five Brazilian States
Paes de Barros, Ricardo and Pinto de Mendonya, Rosane Silva (1996d) "O impacto do
ambiente comunitario sobre o desempenho educacional" Relat6rio Preliminar. IPEA -
Instituto de Pesquisa Econ6mica Aplicada, Junho 6 p. + pictures and tables (processed).
Parra, F.C, Amado, R.C. Lambertucci, J.R., Rocha, J., Antunes, C.M., Pena, S.D.J. (2003)
Color and genomic ancestry in Brazilians. Proceedings of the National Academy of
Sciences of the U.S.A.
Pena, Sergio.D.J. (org.) (2002) Homo brasilis: Aspectos Geneticos, Linguisticos, Hist6ricos e
Socioantropol6gicos da Forma,ao do Povo Brasileiro. Ribeirao Preto, FUNPEC.
Ramos, J.S. (2002) "O Brasil sob o paradigma social: sociologia hist6rica de uma
representa,co." In: Homo brasilis: Aspectos Geneticos, Linguisticos, Hist6ricos e
Socioantropol6gicos da Forma,co do Povo Brasileiro (S.D.J.Pena,ed.) Ribeirao Preto,
FUNPEC-RP.
Reis, Elisa P. (2001) "Perceptions of Poverty among Brazilian Elites" Paper presented to the
Rio Workshop on Inequality. Rio de Janeiro, July 3-6, 2001.
Schwartzman, S. (1999). "Fora de Foco: diversidade e identidades etnicas no Brasil."
Estudos CEBRAP 55, Nov. Pg 83-96. Sao Paulo.
Torres, Fabio Jose Sanchez and Mendez, Jairo Augusto Nnfiez (1995) "?Por que los nihos
pobres no van a la escuela? (Determinantes de la Asistencia Escolar em Colombia)" Dep.
Nacional de Planeaci6n, Republica de Colombia, Noviembre, s/p (processed).
Iwakami Beltrao, Kaiz6 (2003) Alfabetiza,co por ra,a e sexo no brasil: um modelo linear
generalizado para o periodo 1940-2000. Paper presented in the Latin American and
Caribbean Meeting Todos contamos. Banco Mundial, Lima.
Velez, Carlos Eduardo; Soares, Sergei and Medeiros, Marcelo (2001) "Reducing school
inequality in Brazil: demographic opportunities and inter-cohort differentials"
Texto para Discussao n° 850, IPEA - Instituto de Pesquisa Econ6mica Aplicada,
December 17 p.
144 1