Print version ISSN 0100-1574
Cad. Pesqui. vol.42 no.146 São Paulo May/Aug. 2012
The school, the metropolis, and the vulnerable neighborhood
Maurício ÉrnicaI; Antônio Augusto Gomes BatistaII
CENPEC: Center for Studies in Education, Culture and Community Action - Centro
de Estudos e Pesquisas em Educação, Cultura e Ação Comunitária. firstname.lastname@example.org
IIHead of Coordination for Development of Research at CENPEC: Center for Studies in Education, Culture and Community Action - Centro de Estudos e Pesquisas em Educação, Cultura e Ação Comunitária; Lecturer on the Post-Graduate Program in Education at the Federal University of Minas Gerais (UFMG), and researcher of the National Council for Scientific and Technological Development - CNPq. email@example.com
This paper presents the results of a study aimed at exploring the neighborhood effect hypothesis, grasping if and how inequalities in the level of social vulnerability of school's neighborhoods impact both the education offered by school and student's performance. To do this, quantitative and qualitative research procedures were employed to collect data in the outskirts of the city of São Paulo. The results of the analysis showed that the higher the level of social vulnerability in the school's neighborhood, more limited was the quality of the educational opportunities created by the school. Results also demonstrated that the negative effect of the vulnerable territory on the school happens through five articulated mechanisms: the school's isolation within the territory; the low rates of preschool enrollments; the segregation of its intellectual population in the teaching establishments located therein; the disadvantageous position occupied by schools localized in neighborhoods with high social vulnerability in the hidden educational quasi-market; and the school's difficulties, given its unfavorable position, to offer the conditions needed to secure the institutional model of operation that guides school organization.
Keywords: educational inequities; social disadvantages; metropolis' schools
Despite their economic and cultural wealth, deeply unequal metropolises provide, in the form of spatial segregation, an objectification and reification of the social and economic inequalities that characterize the organization of their space. They give to some and deny to others the benefits of location - proximity to or distance from assets such as educational equipment, means of transportation, sanitation, hospitals, libraries; greater or lesser proximity to one's place of work; the prestige or stigma of the mere mention of the neighborhood where one lives; the chance - or lack of a chance - to select one's circle of relationships and increase one's social capital (BOURDIEU, 1997; KOWARICK, 2009).
The expressions "effect of place" (BOURDIEU, 1997), "effect of segregation" or "effect of territory" (MAURIN, 2004), and "effect of neighborhood" (MALOUTAS, 2011) designate the impact of one's place of residence and the social characteristics of its population on "the living conditions and social mobility of the inhabitants" (MALOUTAS, 2011, p. 288). They also more specifically designate the impact of territory on the school destinies of individuals (MAURIN, 2004).
In recent years there have been more studies addressing the consequences that the characteristics of a territory have for educational opportunities. In many of these studies, the assumption that the territory is important in understanding the issue of education offers a new possibility of understanding the production and the reproduction of school inequalities associated with social stratification and with conditions of educational provision (BEN AYED, POUPEAU, 2009; RIBEIRO, KOSLINSKI, 2009a).
However, the existence of this effect is the subject of debate. One of its dimensions is methodological in nature. While ethnographic studies observed connections between the characteristics of territories and the conditions of schooling, studies that prioritize quantitative data state that it is difficult to separate the influence of territorial characteristics on schooling from the weight given to other drivers of social and school destinies, such as family drivers (MALOUTAS, 2011; MAURIN, 2004). This difficulty has led some researchers to seek quantitative evidence of the existence of the effect of territory and to seek to identify the mechanisms that produce it (BEN AYED, BROCCOLICHI, 2008; RIBEIRO, 2010).
Synthesizing this debate, Ribeiro and Koslinski (2009a) point out that the difficulty in quantitatively apprehending the effect of territory may stem from the mismatch between ways of characterizing the territory on the one hand and the nature of the mechanisms or social processes thought to produce this effect on the other. There is a problem first of all in the scale of the analysis, since macro-territorial classifications prevent the understanding of social processes that can be appreciated on the micro and meso-sociological levels. Secondly, there is thought to be a problem of demarcation of the phenomenon itself, since the quantitative descriptions of the territory prioritize the social and economic status of its population, and its effect is produced primarily by two dimensions: the first one is sociocultural, in other words, the types of sociability and social models in force in the territory (TORRES et al., 2005; ALVES, 2008; RIBEIRO, KOSLINSKI, 2009b; SANT'ANA, 2009), and the second is political and institutional, in other words, has to do with the quantity and quality of public services in these territories (TORRES et al., 2008).
If the first dimension of the controversy as to the existence of an effect of territory is methodological, the second is contextual. Arguments in favor of the hypothesis that the territory is important have presumed that its effect has a different incidence according to the profile of the society and the State under analysis (MALOUTAS, 2011). Thus, the effect of territory will be more significant in circumstances marked by major socio-economic and educational inequality, by intense sociocultural segregation and/or by contexts where a Welfare State capable of making rights universal has not been consolidated.
Since the metropolises of Spanish America and Brazil are strongly segregated and rights have not been consolidated as universal, this hypothesis has been invoked by some researchers as an important element for understanding its educational problems (TORRES et al., 2005; TORRES et al., 2008; ALVES et al., 2008; KAZTMAN, RIBEIRO, 2008; RIBEIRO, KOSLINSKI, 2009a; SANT'ANA, 2009). Despite this, empirical evidence cannot always be found to authorize the establishment of explanatory relations between the social and economic conditions of territorial units and the performance of students (SOARES et al., 2008).
The general objective of the study whose main findings are presented herein was therefore to explore the hypothesis of the effect of territory so as to contribute to an understanding of the educational problems of the large metropolises. Although studies of the effect of territory in the field of education tend to dwell on performance and the educational opportunities of the individuals living in that territory, this study aimed to focus its action on the opportunities provided by the school in the territory (BEN AYED, BROCCOLICHI, 2008). The specific objectives of the study were to learn whether and how inequalities in levels of social vulnerability within the region being studied impacted the school in that region, the education provided at that school, and, through it, the performance of the students.
The purpose of this article is to present the conclusions of the investigation; to this end it is organized in four parts. Part One presents the methodological design; Part Two presents the main results concerning evidence obtained of the effect of territory; Part Three is a discussion of the mechanisms or processes that produce the effect of territory; and Part Four seeks to make explicit how the main processes of data interpretation will dialog with studies -particularly Brazilian studies - of educational inequalities in metropolises, and above all studies of the territorial expression of social inequalities.
The study was carried out in the São Miguel Paulista sub-prefecture, located in the easternmost part of the municipality of São Paulo. This sub-prefecture, with its approximately 400,000 inhabitants, generally has life quality indicators that are lower than those in the center of the city.1
Methodological choices sought to deal with difficulties in characterizing the effect of territory by means of two main strategies: one having to do with the way in which the study is segmented and the territorial classification; and the other having to do with the nature of the data. To define categories so as to spatially segment and characterize differences expressed in the urban territory, studies of the city usually consider districts or equivalent spaces as the smallest units for territorial analysis, without examining whether such units have been created on the basis of public administration needs. This option may be valid in certain circumstances, but is limited. The first constraint is that it does not allow an approach to differences that come into existence within the districts, thus sometimes reinforcing generic visions of peripheral regions. A second constraint is that these categories take very different sets of circumstances to be equivalent: for example (still in the case of the city of São Paulo) the Grajaú district, with over 400,000 inhabitants, and the Barra Funda district, with approximately 10,000 inhabitants. Finally, a third constraint appears when the boundaries of the districts are seen not to be adequate to identify the districts in the same way as they identify themselves on the basis of the social relations through which the population transforms and gives meaning to the urban space; for example, Avenida Paulista, experienced and signified as a benchmark territory for the city, actually straddles four districts and three sub-prefectures.
Thus, by opting to study inequalities within a region of the periphery - the sub-prefecture of São Miguel Paulista - which is usually thought of as a homogeneous whole, the data are organized on the basis of territorial units that intend to approach the scale of the daily experiences of local agents as much as possible. The Census-taking sector was therefore considered to be the smallest territorial unit for classifying the data. The first advantage of this is that the Census sectors are always comparable in scale, since they are defined on the basis of clusters of approximately 300 households. The second advantage is that census sectors may be grouped into larger units based on field data, giving information on how the population delimits and defines the territories in which it lives.
The second strategy has to do with the nature of the data gathered. Data was linked both quantitatively and qualitatively.
The quantitative data has to do with the population of the territories highlighted, with the equipment that seeks to assure social rights and with the characteristics of the schools and their students, and schools and students were the major units of analysis considered. The data encompasses the set of the population and the territory of the sub-prefecture, including its 61 public schools and students studying there. The data enables a descriptive panel to be created showing the correlations between differences in levels of social vulnerability of the territory and educational differences. Although the variables that have been used are presented throughout the description and analysis of the data, it should be pointed out that the description of the territory around the schools was carried out on the basis of the average of the São Paulo Social Vulnerablility Index - Índice Paulista de Vulnerabilidade Social (VS) for the census sectors touched by a 1 km radius around each school.2
The quantitative data was produced by means of an ethnographic field study. This study was carried out during the 2009 school year in five primary schools of the state and municipal networks, chosen so as to represent different types of combination of three factors retained by the quantitative analysis: greater or lesser heterogeneity or homogeneity in the composition of its student body according to family cultural resources; greater or lesser social vulnerability for the area surrounding the school; a better or worse performance by the school in the IDEB (Compulsory Education Development Index or Índice de Desenvolvimento da Educação Básica).
The main data collection procedures involved observation and interview; procedures were carried out during the daily life of the school, and prioritized teachers of the final grades of the first and second segments of basic education as well as members of the school's technical and administrative teams. When the results were organized and the first analyses produced, a debate was held involving the school directors and coordinators in order to validate the conclusions of the study; this procedure proved to be highly productive.3
EVIDENCE OF THE EFFECT OF TERRITORY
The data reveal the existence of a correlation between the variation in levels of social vulnerability of the territory where the school is located and the educational opportunities provided to students: the higher the levels of social vulnerability of the area around the school, the more limited tend to be the quality of the educational opportunities offered by them.
There are two pieces of evidence that show the correlation between levels of social vulnerability of the area around the school and the quality of educational provision. One was obtained from the schools' performance in the IDEB; the other was obtained from the students' performance in the Prova Brasil.
VULNERABILITY OF THE AREA SURROUNDING THE SCHOOL AND THE SCHOOL'S IDEB
The IDEB results for the São Miguel Paulista schools vary according to the levels of social vulnerability of the territories where the schools are located. The more vulnerable the territories where the schools are located, the lower tends to be their scores in the IDEB,4 as Figure 1 shows.
The graph shows the distribution of schools in São Miguel Paulista by their IDEB performance and their location in more or less vulnerable territories. The horizontal axis expresses the level of social vulnerability of the schools, measured by IPVS within a 1 km radius around them. The greater the IPVS, the more vulnerable the territory. The vertical axis represents the standard distance (afastamento padronizado - AP5 - in its Portuguese abbreviation) from the average IDEB score for the sub-prefecture of schools in São Miguel Paulista. This value is standardized in standard deviation units, considering school performance in the early and final years of primary school. AP = 0 means that the school has an identical performance to the average for the sub-prefecture. When a school's IDEB is below the sub-prefecture average, it receives a negative score. When above, positive.6
Graph 1 not only shows that the level of social vulnerability of the territory is an important variable for understanding differences between schools, but it also allows three groups of schools to be identified according to the level of social vulnerability of their surrounding areas. They are shown by the ellipses and may be presented in the following way: Schools located in low social vulnerability territories, whose IPVS is lower than 3.0, nearly always have IDEB scores above the local average, and a significant number among this group have results between two and three standard deviations from the local average. Schools located in territories with medium social mobility, whose IPVS is greater than 3.0 and lower than 3.6, make up a group with greatly dispersed results; however, most of the schools in this group have IDEB scores above the local average. Schools located in territories with a higher social vulnerability are those whose IPVS is greater than 3.6 and virtually always have IDEB scores lower than the local average.
VULNERABILITY OF THE AREA SURROUNDING THE SCHOOL AND STUDENTS' PERFORMANCES
The second proof of the effect of territory is given by students' performance. Children with the same cultural resources perform differently in accordance with the level of social vulnerability of the place where the school in which they study is located. When children with few family cultural resources7 study in schools located in more vulnerable areas, the set of these children tends to have a worse performance compared with students from the same group who study in schools located in less vulnerable areas. In turn, students with greater cultural resources have lower scores when they study in schools located in territories with high social vulnerability.
Graphs 2 and 3 below show the distributions of fourth-grade students by their level of proficiency in the Portuguese language test of the 2007 Prova Brasil,8 relating this variable to the level of social vulnerability of the surrounding area of the school where they study.
There is, however, one difference between the two graphs. Whereas Graph 2 represents only students with scarce family cultural resources, Graph 3 represents only students with greater family cultural resources. The dark gray bar, in both graphs, always shows the performance of students who study in schools whose surrounding area has a high degree of social vulnerability (greater than 3.6). The light gray bar represents students who study in schools with medium vulnerability (from 3 to 3.6). The medium gray bar represents students whose schools are in territories with low vulnerability (less than 3). The dotted bar is for comparison; it represents the results of all the students in that sub-prefecture who took the Prova Brasil, regardless of the location of the school where they study or of their sociocultural level.9
Graph 2 shows that when students with scarce family cultural resources study in the more centrally-located schools of the sub-prefecture with less vulnerable surrounding areas, 38% of them have a performance below basic, virtually approaching what is found for the total set of the population (33%). However, when one observes students with the same profile studying in schools with a more vulnerable surrounding area, this percentage leaps to 50%. At the opposite end of the axis of proficiencies, only 10% of students have a below-basic performance when they study in schools with a very vulnerable surrounding area, a percentage that rises to 24% when one observes students with the same profile studying in schools whose surrounding area is less vulnerable, virtually identical to the set of students in this grade (25%).
Graph 3 shows what happens to the group of students with greater cultural resources in the sub-prefecture's public school system. The percentage of students with this sociocultural profile and with a below-basic performance is striking, when they study in schools whose surrounding area is highly socially vulnerable: 41%. This percentage virtually attains that found when 50% of students with scarce cultural resources study in these schools. However, when these students study in schools with a less vulnerable surrounding area, the percentage whose performance is below basic falls to 19%, far less than what is found in the set of the population in the sub-prefecture (33%) and the value found in the group of students with greater cultural resources (38%). At the opposite end of the proficiency table, one finds that only 19% of those students studying in schools with a highly vulnerable surrounding area have an adequate or advanced performance, and that this rises to 39% when examining that part of the group studying in schools with less vulnerable surrounding areas, a value above what is found for the set of the population (25%).
Although this data reinforces the importance of family cultural resources in students' performance, variations in the performance of students who have the same cultural resources prove the existence of inequalities produced by the school system that can be explained on the basis of the characteristics of a school's surrounding area. These inequalities produced by the school system may exacerbate or mitigate those produced by differences between family cultural resources.
In the following item, we argue that the correlation between the variation in levels of social vulnerability of schools' surrounding areas and variations in the quality of education provided is due to a set of mechanisms or processes determining that high social vulnerability territories tend to accumulate disadvantages that are constraints to the provision of quality education. Likewise, these mechanisms or processes lead those schools that are more centrally-located or whose surrounding areas are less vulnerable, to accumulate relative advantages, which allows them to work better, and, by extension, obtain better results.
PRODUCING THE EFFECT OF TERRITORY
Five mechanisms or processes by means of which the vulnerable territories tend to restrict educational opportunities provided by schools located in them, were apprehended.
THE UNEQUAL DISTRIBUTION OF SOCIAL EQUIPMENT
The first mechanism has to do with the distribution of social equipment in the territory: both social equipment and equipment in the public schools in the territory. On the basis of geo-coding, one can see that in territories which are highly socially vulnerable, apart from there being a scarcity of private services, there is also low coverage in terms of public facilities to guarantee social rights. The data in Table 1 show the low coverage of social equipment10 in territories with high IPVS.
It can be observed that there is a mismatch between the concentration of social equipment and the concentration of population in situations of social vulnerability. In the extreme situation (IPVS 5 and 6), 22% of the population is concentrated in areas of greater vulnerability, but only 7% of the equipment. If one goes on to include the population living in territories above level four of social vulnerability, one has 53% of the population and 31% of the equipment.
Given that in territories scoring levels five and six in vulnerability there are virtually only primary schools there (particularly the early stage of this level of education), Basic Health Units and so-called Telecentros (small-business training centers), one can conclude that schools are the major type of state-public reference equipment for families living in this territory, and they tend to be isolated within it. The field work showed that the problems inherent to the social vulnerability of families and of the territory are apparent in these schools, requiring them to take up stances without, however, being able to face such challenges, which eventually blocks their ability to carry out activities specific to schools.
The most telling data has to do with how far the school is insulated from other social equipment in emergency situations, such as when the population falls victim to floods in the region and the schools are used as emergency facilities. Other episodes are those where conflicts outside the school end up having repercussions in terms of violence within the school, such as in the case of revenge or the settling of scores between individuals or rival gangs. However, teachers also recount episodes of difficulty in dealing with students who live in borderline situations produced by violence, precarious health conditions, or barely habitable homes. In all such cases, the school must address these needs without being able to count on the support of a well-structured, accessible network of public services.
THE UNEQUAL DISTRIBUTION OF ENROLLMENTS IN CHILD EDUCATION
The second mechanism or process is a development from the first and has to do with the distribution of the provision of child education in the territory. The data show that vulnerable territories had a low provision of enrollments in child education in 2007, which tends to make the possibility of access for children who study in schools in these areas a crucial factor for success at school.11 Table 2 shows data on this, taking into consideration only enrollments in preschool.
The gross preschool enrollment rate in São Miguel Paulista is low in the year under observation, but follows the pattern for the municipality.12 However, what the data shows is that the distribution of enrollments in the territory does not follow the pattern of distribution of the population as a result of the levels of social vulnerability of these territories. In Table 2 one may first of all observe that enrollment is almost equal to the number of children living in the area of lowest vulnerability, but is only some 10% of the population living in the most vulnerable territories. Secondly, one can also observe that while 26.4% of the 4-6 year-old population (6,587 people) inhabits territories with the highest social vulnerability, only 5.1% of enrollments (656 enrollments) are located in these areas. The numbers improve if one includes level 4 social vulnerability, but even so come to only slightly over 6,400 enrollments for a total population of nearly 15,000 children, in which case enrollments total 42% of the number of children from four to six.
It was not possible to accurately calculate the percentage of students in the first grade of schools located in the high-vulnerability territories with preschool facilities; however, the data available enables one to infer that it must be a small group, which is not enough to transfer to the collective setting of the classroom and the school itself the repertoire of knowledge and school culture codes whose transmission is one of the purposes of preschool. Familiarization with this repertoire, therefore, tends to be greater when one shifts to schools located in areas of less social vulnerability, where there is a greater provision of places.
In summary, these first two mechanisms show how the presence of social equipment in the territory (and child education in particular) helps deepen the inequalities of school that are also produced by differences in families' cultural resources. Those territories with higher social vulnerability concentrate families with fewer cultural resources, and are therefore more distant from literate culture and the school universe. Thus, the gap between families' cultural resources and the school culture is exacerbated by the isolation of the school in the territory, and by the deficit in provision of child education for families in highly vulnerable situations.
THE STRONG HOMOGENEITY OF THE SOCIO-CULTURAL MAKE-UP OF THE STUDENT BODY
The third mechanism or process that produces the effect of territory has specifically to do with the intertwining of the urban social issue with the educational issue, or, to be precise, the internalization of the social issue by schools. This mechanism is revealed by an analysis of the sociocultural make-up of the student body given variations in levels of social vulnerability in the areas surrounding these schools. As a rule, schools in vulnerable settings tend to present a strongly homogeneous student body with regard to lower levels of family cultural resources and to the place of residence in the school's vulnerable neighborhood, as can be seen in Graph 4. This data provides evidence that the isolation of this population is internalized by schools located in highly socially vulnerable territories, in which this isolation takes on the form of isolation of the school.
On the vertical axis of this graph, the schools are distributed according to the sociocultural profile of their student body. When the Heterogeneity Index (HI) 13 is equal to zero, we have a school whose student distribution according to their family cultural resources (K-Cult) is identical to the total distribution of students in the São Miguel Paulista public schools network. When the HI is positive, the school tends to bring together students with greater cultural resources, 0.10 being the point which separates this group of schools. When the HI is negative, the school tends to bring together students with fewer cultural resources, -0.10 being the point which separates this group of schools. On the horizontal axis of this graph, the schools are distributed according to the level of vulnerability of their surrounding areas. Here we adopt the average IPVS of sectors encompassed in a 1km radius around the school as in Graph 1.
We find three typical situations: i. schools located in regions with less social vulnerability tend to bring together students with greater family cultural resources; ii. schools located in territories of medium social vulnerability tend to have a heterogeneous student body, catering to students with a range of family social cultural profiles; iii. schools located in highly socially vulnerable territories tend to have a student body made up of children and young people with scarce family cultural resources who reside in the surrounding area.
The schools in this third group are strongly homogeneous. They are a microcosm of the highly socially vulnerable territory, concentrating its problems within themselves, and these problems are made intensely manifest to an extent that is sometimes difficult to control. In schools with a more heterogeneous student profile, the directors we interviewed state that such problems exist, but in a more watered-down form and are easier to manage. This dilutions occurs either because the problems happen less frequently and do not become routine or daily events, or because they affect a smaller number of students and do not become problems "of" the student body.
The field data also suggests that the very discourse of the teachers expresses differences in the ways in which social problems manifest themselves within schools located in more vulnerable settings compared with the other schools. In general, in those schools bringing together students with scarce cultural resources, vulnerability and its consequences are the attributes of a collectivity ("the student body", "the clientèle", "our students"), which is used sometimes as an explanation for the problems experienced by the school, whether disciplinary or academic. In turn, in those schools that do not bring together students with high social vulnerability, vulnerability and its consequences are held to be the attributes of individuals who stand out from the student body, which is sometimes used to justify either the importance of personal effort as a way of overcoming difficulties, or as an explanation for problems experienced by the individual in question.
To sum up, therefore, when they are isolated, when they are the main reference social facility in the territory, and when they bring together students with scarce sociocultural resources that tend not to be the fruit of preschool education, schools located in vulnerable settings end up internalizing the social dynamics of the territory, and eventually become a continuum indistinguishable from it. One of the consequences of this is that they are unable to overcome these patterns of the surrounding area in order to create a school environment that ensures teaching standards and learning for the students.
These schools thus reinforce the social and cultural isolation of the population in their neighborhood: they reproduce within themselves patterns of urban and sociocultural segregation and also not only restrict possibilities for interaction with other social groups but also hamper opportunities for cultural development that are possible in the urban territory. For this reason they tend to reproduce within themselves the urban and sociocultural territorial segregation of the population that they serve, as well as the problems resulting from the segregation, as a negative peer effect restricting educational possibilities.
This is also revealed in the statements taken from many local mothers who fear that many of the dangers of "the world outside/the street" will be reproduced in the school, above all the patterns of coexistence marked by violence or by "the law of the strongest", as well as drug dealing and drug abuse. These mothers demand that schools be a sort of rupture offering protection and care against these risks that they feel they are vulnerable to. From this point of view, schools are seen as extensions of the street and as spaces where "bad people hang out", leading them to be avoided by the families themselves for exposing their children to risks they wish to avoid (SETUBAL, 2009; LOMONACO, GARRAFA, 2009).14
DISADVANTAGES OF SCHOOLS IN VULNERABLE SETTINGS IN THE QUASI-MARKET
The fourth mechanism or process that produces the effect of territory has to do with the relations of interdependence among the school units themselves, whether in the micro-territories or inside the administrative facilities. Generally speaking, one finds that collaboration between schools is rarefied and there is a strong relation of competition among the schools for scarce resources in the territory, especially on the part of professionals and students whose characteristics are enhanced by the schools, making a hidden school quasi-market (VAN ZANTEN, 2005; COSTA, KOSLINSKI, 2009).
During our fieldwork we discovered the existence of this quasi-market both in regard to educational services and in regard to educational work. The former is defined by competition among families for enrollments in schools that better meet their educational expectations and by competition among schools for students to better meet the schools' attitudinal and academic demands. The latter is defined by competition among teachers, pedagogical coordinators and directors for positions in schools occupying a prestigious position in the hierarchy that exists among them, and by competition among schools to attract professionals with affinity for the schools' projects.
In this scenario of a dispute over scarce resources, schools located in areas of higher social vulnerability tend to be at a disadvantage, whereas schools located in medium and low vulnerability regions tend to be at an advantage, either competing in the quasi-market for educational services, or competing in the quasi-market of educational work.
This data was obtained during field work, and for this reason the relations between the quasi-market and variations in levels of social vulnerability could not be tested in a quantitative study.15 However, there was a great deal of evidence that schools located in territories with medium and low social vulnerability, owing to their advantageous position in the relations of competition, can more easily attract professionals who are more qualified and engaged, and attract students with greater family cultural resources, and who adapt better to the way the schools operate. For this reason, it is easier for them to impose demands with regard to behavior, professional engagement, and academic performance, which pressures those who cannot adapt - students or teachers - to seek other institutions. Thus these schools "exteriorize" their problems, in all three spheres: teaching staff, administration, or student body.16 As a result of this advantageous position, it is easier for them to develop good internal management of times and spaces for learning.
In turn, schools located in highly vulnerable territories accumulate disadvantages and tend to become "decanters" for a range of problems of schools in an advantageous situation. Schools in vulnerable settings tend, first of all, as we have indicated, to cluster students with scarce family cultural resources. Given their location in peripheral and highly socially vulnerable regions, enrollment in these schools is unattractive to families with greater cultural resources or who do not live in the area immediately surrounding the school. Even among families in the immediate vicinity of the schools, there are those with higher educational expectations who seek to enroll their children in better organized and better located schools, generally in less vulnerable areas.
Secondly, these schools find it more difficult to attract and maintain a stable staff of qualified and engaged professionals.17 They tend to face greater staff turnover, have unfilled vacancies, and a larger number of teachers not admitted through public competitive examinations. Every end of year, several teachers, coordinators and directors ask to be transferred to other schools. In these schools the problem of staff absenteeism in the second stage of primary school is more acute. These schools, living under tension and facing a routine that leads to dispersal and desegregation, the high turnover of teachers, the unfilled positions, and the numerous absences of teachers end up undermining the institutional conditions needed for the school to work.
The fact that these schools receive the "problems" rejected by schools in an advantageous position in the dispute for resources largely means that they help other schools work. The well positioned schools are able to "exteriorize" their problems and assure better conditions for their operation, knowing that the school system will guarantee those students who transfer away from them the right to enroll, and find positions for teachers who do not adapt to them. They benefit from and depend on these other schools where the problems decant and concentrate (YAIR, 1996). Reproduction of this interdependence reinforces the isolation of the schools in vulnerable settings, and they end up the victims of the avoidance strategies of parents, students and professionals who feel able to choose another establishment.
SCHOOLS FROM VULNERABLE SETTINGS TO BE FILLED AND THE INSTITUTIONAL MODEL OF THE SCHOOL
The fifth mechanism or process producing the effect of territory has to do with the didactic-pedagogical dimension and is specifically the institutional model that guides the school. Based on Dubet (2002, 2008), we herein assume that the institutions are relatively stable and long-lasting modes of organizing activities that are valid in a given social space and tend to impose themselves on individuals. In this sociologist's view, institutions possess an institutional program that acts intentionally to transform its participants in a specific direction.
In the case of schools, carrying out this institutional program demands compliance with some prerequisites by students and by teachers. The dispute engaged in by schools over students and teachers in the quasi-market results from this need to marshal professionals and audiences who come closer to the conditions for the fulfillment of their program. Therefore, the different positions of schools in the competition for professionals, students and resources are expressed in possibilities of greater or lesser administrative and pedagogical organization. The better situated schools within this competition scenario more easily assure students and teaching staff who meet these prerequisites, which in turn assures them of better operation. In those schools that decant and concentrate the problems of other schools, these prerequisites are not met and the institutional model is driven to the verge of non-feasibility.
Regarding the students, the institutional model presupposes, for example, that they will join the school in the first year with a prior familiarity with written language and having internalized certain modes of behavior and relating to others suitable for school activities. During the first part of primary education (EF1) these students will be accompanied by reference teachers who will keep an eye on them every day of the week and personify all activities, all times, all spaces, and all routines of school life. In the second segment of primary education (EF2) this model assumes that the student will have internalized and appropriated those modes of acting, thinking and relating to the school environment that were formerly personified in the figure of the teacher. Thus, in the transition from EF1 to EF2, the student begins to relate more directly, with fewer mediations, with the overall school environment, which becomes more fragmentary, impersonal, and faster than it was in EF1. It is expected, in EF2, that the student will be capable of acting autonomously in school activities, knowing, for example, how to organize himself in his spaces and times based on impersonal markers such as the bell, telling the time on the clock, or the movements of people from one place to another. This model also assumes that in the early years students will have developed relative intellectual autonomy in terms of knowledge, which will be necessary to deal with the fragmentation of subjects and with different teachers responsible for each of them.
Teachers themselves are also affected by this fragmentation, acceleration, and impersonality in EF2. For teachers, teaching in EF2 means giving classes in a number of classrooms and to a large number of students. As if that were not enough, teachers accumulate duties and their classes are normally distributed over two and even three parts of the day. Since they do not spend much time with the classes, and since their work is organized according to the false assumption that a series of prerequisites has been met in EF1, teachers at EF2 level are often faced by major hurdles to doing their work. The demands of the school model on teachers add to these difficulties. For EF2 to work well, teachers must engage in collective planning activities, preferably thinking of primary education as a whole, they must not be absent frequently, and on any given day not many teachers may be absent.
To sum up, for the institutional model to work a relative stability in the school's mode of working is assumed, which may only be obtained if these student and teacher prerequisites are met. Since "problems" shake the stability essential to good administrative and pedagogical operations, they are "externalized" by well-organized schools, thus enabling good management of learning times and spaces. In counterpoint to this, it is because these problems concentrate and "decant" in certain schools that good management of learning times and spaces faces hurdles unlikely to be overcome.
In our concluding remarks we intend to confront, in the existing literature, explanations that we have put forward herein as to the educational inequalities found in the survey. We shall pay special attention to recent studies of the metropolitan educational issue in Brazil that seek to match inputs from urban sociology with inputs from the sociology of education.18
Ribeiro and Koslinski (2009a) point out three main modes of interpreting the effect of territory. The first emphasizes aspects of educational demand and prioritizes cultural aspects arising within poorer groups, with reference to the hypothesis of social disorganization and of cultural scarcities that result from the ghettoization of this population. This point of view argues that such areas suffer from the consolidation of antisocial behaviors, the weakening of social networks, and the fragility of community mechanisms that could otherwise strengthen behaviors of collective effectiveness in favor of the school. This mode of interpretation approximates it to theories on the culture of poverty developed in the USA in the 1960s, emphasizing behavioral deviation, scarcities and absences felt by this population, and moral aspects of the poorer population, making individuals in these conditions responsible for the situation in which they live (KOWARICK, 2009, p. 34-39). The second way of explaining the effect of territory, which the present authors prefer to the first, also emphasizes the educational demand side. In this case, the US debate on poverty is also felt, since they call upon works produced by those who sought to renew this debate in the 1980s (KOWARICK, 2009, p. 39-45), prioritizing macro-social structures responsible for producing the phenomenon of urban poverty as much as the social patterns that are established within these groups, and the relationship between such groups and the middle and higher layers of society. In this approach, the emphasis falls on the isolation produced by the ghettoization of poorer populations, which denies them access to the social standards that obtain among the middle and upper classes, from the networks of social relations that enable greater social transit and the distribution of the available goods and services in the open space, and that in extreme cases of vulnerability favor disruption of social life, and the exacerbation of a culture of individualism, criminality and anti-school attitudes. Thus, based on the assumption that current social patterns in the primary socialization group and in interactions with those closest exert a great force on the construction of the subjects, this explanation prioritizes the distancing that is consolidated between these patterns and the culture demanded by the school. A third way of interpreting the effect of territory prioritizes the education and services supply side for the population living in highly socially vulnerable territories. These interpretations are based on the assumption that these individuals are influenced by the quality of the services offered to them, and that therefore the quality of provision of schooling may have impacts on students' school performance.
In their study, Ribeiro and Koslinski (2009a) seek to articulate the effects of territory on educational supply and demand, by mobilizing the second and third forms of interpreting the effect of territory. It it may be said that this is a solution that has prevailed in studies on the issue, although its explanatory weight may fall differently upon another aspect. The emphasis on sociocultural aspects of demand that are influenced by social isolation in the territory can be seen in studies such as those of Alves et al. (2008), Ribeiro and Koslinski (2009a), Ribeiro and Kaztman (2008, cf. Introduction), Sant'Ana (2009), Torres et al. (2005). Thus, especially when the schools located in these regions bring together students with a homogeneous profile as to family cultural and economic resources and social capital, there will be an effect stemming from this relatively homogeneous collective (peer effect) which will enhance the effect that is produced by traditionally studied family characteristics, when considered individually, such as race, sex, income, level of schooling of parents.
Explanations falling in the educational supply side, taking into consideration the social differences expressed in the territory, are less widely emphasized in the literature. Studies underpinned by this explanatory school of thought show that the educational inequalities associated with differences in levels of social vulnerability within territories are also produced by the unequal distribution of state resources by public policies. Torres et al. (2008) highlight inequality in the allocation of those teachers ranking highest in competitive public examinations for admission, and top scoring throughout their careers, as well as differences in guaranteeing the duration of students' school careers.
In the study carried out in São Miguel Paulista, these two prevailing explanatory procedures were mobilized in the three first mechanisms or social processes apprehended as producers of the effect of territory. However, the studies discussed so far in this paper do not highlight other aspects explored in the investigation that stem from data obtained during fieldwork.
Van Zanten (2006) shows that the competitive relationship among schools and their position in the hierarchy of prestige have consequences for their internal workings and the way in which they open outwards. In the framework of the perception of this phenomenon, studies into the educational quasi-market in metropolises have been conducted. As shall be seen below, these studies have not found strong enough correlations between educational inequalities and variations in the level of social vulnerability among territories.
Studying the case of Rio de Janeiro, Costa and Koslinski (2009) and Koslinski and Costa (2009) chose to study the effects on the running of schools and on the learning of students produced by the competition both on the educational supply side and on the demand side, although the latter dimension received more attention.
Emphasizing school-family relations, the literature that is critical of quasi market policies shows that families' search for schools with better reputations and schools' search for students who best meet their expectations of behavior and performance, lead on the one hand to a trend for schools to cluster students with higher socio-economic levels, and on the other hand for schools with less prestige to bring together students with a lower socio-economic level (COSTA, KOSLINSKI, 2009; VAN ZANTEN, 2006; YAIR, 1996). This results in the deepening of inequalities and exacerbation of the hierarchy of prestige among schools, consolidating some schools as catering to those students with fewer cultural resources that other schools avoid.
Gad Yair (1996), who carried out a survey in a large city in Israel, states that competition between schools is not global, but takes place between schools ranking close to each other in the relation structure connecting them. Studying what he calls the market ecology, the author shows that schools differ from each other in the hierarchy of prestige and in their profiles. There will be competition between schools that are close in the market; there will be cooperation between distant schools, especially when they are in inverted hierarchical positions. Thus, as we learned in São Miguel Paulista, less prestigious schools abet the more prestigious ones, since they assure the right to enrollment to those students avoided by schools in advantageous positions in the quasi-market. Competition, therefore, produces interdependencies between schools and self-reproducing profiles, while they reproduce each other, a conclusion similar to what was reached when studying the schools of São Miguel Paulista.
Studies of the quasi-market can enter upon a fruitful dialog with studies mobilizing the concept of the peer effect to study educational differences. Although they do not make explicit the concept of the quasi-market, one example would be the studies (in the case of Belo Horizonte) of Soares et al. (2008) and Alves (2010). The authors mobilize the concept of the peer effect to explain differences between students' performances resulting from the socio-economic make-up of the student bodies in the schools. To explain this socio-economic make-up, the authors raise the hypothesis that families with more intense educational investment and with a higher socio-economic level tend to seek schools which have better reputations and a student body of a higher socio-economic level, thus reproducing within the public school system the patterns of inequality found in society (ALVES, 2010).
Soares et al. (2008) attempted to relate this explanation to the distribution of schools in the territory of Belo Horizonte, dividing it into 121 spatial areas deemed homogeneous; however, they did not find sufficient results to support the existence of this relation. However, as the authors state, these relations with the territory very likely went unnoticed because they worked with large territorial units instead of prioritizing the area immediately surrounding the schools.
In the study that is the subject of this article, we perceive that the structure of the hidden school quasi-market relates to variations in the level of social vulnerability of the territories. However, rather than being a closed explanatory model, we have sought to show herein that there are a wide range of interrelated processes that tend to lead to an accumulation of educational disadvantages in territories that are more highly socially vulnerable and that, by definition, possess fewer resources to meet the challenges they face.
Observing Graphs 1 and 4, one can see that the fourth school marked as a square, whose IPVS is slightly above 4, is the only one located in a high vulnerability area with an IDEB score above the local average (cf. Graph 1), and is the one notable exception. However, Graph 4 reveals that this school has a student body that accumulates more cultural resources than might be expected from schools located at this point of the IPVS axis, a difference that may be observed between the actual position of the school in the graph and the point where the straight line of the trend meets the position of its IPVS. In other words, since the school has consolidated itself as a unit that attracts and maintains students with greater cultural resources within this area of high vulnerability, it manages more easily to develop very effective forms of managing the learning spaces and times.
This is a school that has prestige in its location, which is a small patch of older urbanization that is surrounded by a large slum area. Over the course of years this school has managed to develop mechanisms to protect and defend itself against the effect of territory through its favorable position in the quasi-market. Since it concentrates those students with greater family cultural resources from the micro-region, teachers from neighboring schools state that it faces few problems. According to these teachers, the school has few problems because it has exteriorized them to their own schools.19
This case, which appears to be exceptional, actually confirms the regularity of the phenomenon we seek to demonstrate. This is where the school quasi-market intersects with variations in the levels of vulnerability in the territory, mediated by the prerequisite demanded by the school model. Thus, in seeking to bring together conditions for this model to operate, schools establish disputes and competition among themselves, and those located in more vulnerable settings are at a disadvantage in these relations, although there may exceptionally be schools in vulnerable areas that escape the effect of territory through the quasi-market, and schools in less vulnerable areas that are disadvantaged in the competition for students and teachers. However, as an overall pattern, the data in this study suggest that the existence of a quasi-market - certainly hidden and not encouraged by deliberate policies - is an important mechanism enabling the effect of territory.
It is from this point of view that we may return to our final process/mechanism that produces the effect of territory. The affirmation that the organization of educational activity presupposes students from families with educational or school capital is well-known in the sociology of education (BOURDIEU, PASSERON, 2008). We set out from this thesis to argue that the institutional model organizing the school presupposes a profile of the student and the teacher, and is not suited to the new actual situation of metropolitan public schools stemming from two factors that came together as of the 1990s: the explosion of the new urban social issue, and the enabling of universal access to primary school.
Everything indicates that this collapse has to do with an institutional model clearly not meeting the needs and possibilities of its real audience. Therefore, when the school selects pupils and teachers that have affinity with its institutional projects, and when it is not invaded by the disruptive demands and practices of poorer territories, when it "exteriorizes" its more disruptive cases, then it seems to possess suitable conditions to enable the operation of this institutional model. Therefore, the presuppositions on which the institutional model depends to reproduce itself begin to disappear when the school body is very large, when more than 1,500 students may attend, overwhelmingly with scarce cultural resources, when dynamics of poorer territories encroach upon the school, when its teachers are besieged and assailed by the violence of the surrounding area, and by violence in the facility itself, and when the school suffers from a high turnover of staff. In summary, the institutional model cannot work properly when the school becomes a focal point for decanting the problems of the system and of the territory.
Thus, if the institutional model only works at the expense of the "externalization" of an audience that does not meet the prerequisites demanded, it might then be prudent to conclude that the population has been given universal access to a school model that cannot be universalized. This model was structured under different historical and cultural conditions, to meet the needs of an audience that was very different from today's; the model led to high rates of exclusion in the recent past, to which the high rates of failure and truancy bear witness. One expression of this mismatch between the expectations of the institutional model and the characteristics of the population seeking public schooling, may be the burden of prejudice against students and lack of belief in their ability to learn, as presented in the studies by Alves et al. (2008) Costa and Koslinski (2009), Torres et al. (2008).
The result one may expect of this state of affairs is the profound reproduction of social inequalities, presented as a form of "harmless" exclusion (BOURDIEU, CHAMPAGNE, 1997; FREITAS, 2002), since the children and adolescents remain enrolled, but in schools unable to assure them either of protection or the conditions for their learning and development. Furthermore: since inequalities in the social space in metropolises find their expression in the territory (BOURDIEU, 1997) and do so by means of avoidance mechanisms actuated by those in higher positions towards those in lower positions (MAURIN, 2004), we conclude that the school system not only reproduces social segregation in the make up of its student bodies, but also distributes unequally the supply of quality education throughout the social hierarchy made objective in the space.
What we have tried to show is that beyond the major inequalities that are clearly present in the whole range of the social space of the city of São Paulo, this process is also present within a peripheral region through the articulation of four dimensions that penetrate and influence each other: policies of distribution of social equipment; schools internalizing the urban social question; the consequences of competition in the school quasi-market; and the didactic-pedagogical organization of the schools.
ALVES, F.; FRANCO JR., F. C. J.; RIBEIRO, L. C. Q. Segregação residencial e desigualdade escolar no Rio de Janeiro. In: RIBEIRO, L. C. Q.; KAZTMAN, R. (Org.). A Cidade contra a escola? Segregação urbana e desigualdades educacionais em grandes cidades da América Latina. Rio de Janeiro: Letra Capital; Montevidéu: Ippes, 2008. p. 91-118. [ Links ]
ALVES, M. T. G. Dimensões do efeito das escolas: explorando interações entre famílias e estabelecimentos de ensino. Estudos em Avaliação Educacional, São Paulo, v. 21, n. 46, p. 271-296, 2010. [ Links ]
BEN AYED, C.; BROCCOLICHI, S. Quels liens entre réussite scolaire, inégalités sociales et lieu de scolarisation? Revue Ville École Intégration Diversité, n. 155, p. 66-71, déc. 2008. [ Links ]
BEN AYED, C.; POUPEAU, F. École ségrégative, école reproductive. Actes de la Recherche em Sciences Sociales, Paris, n. 180, p. 4-10, déc. 2009. [ Links ]
BOURDIEU, P. Efeitos de lugar. In: ______. (Coord.). A Miséria do mundo. Petrópolis: Vozes, 1997. p. 159-166. [ Links ]
______. Os Três estados do capital cultural. In: ______. Pierre Bourdieu: escritos de educação. Petrópolis: Vozes, 2008. p. 71-79. [ Links ]
BOURDIEU, P.; CHAMPAGNE, P. Os Excluídos do interior. In: BOURDIEU, P. (Coord.). A Miséria do mundo. Petrópolis: Vozes, 1997. p. 481-586. [ Links ]
BOURDIEU, P.; PASSERON, J-C. A Reprodução: elementos para uma teoria do sistema de ensino. Petrópolis: Vozes, 2008. [ Links ]
CAMPOS, M. M. et al. A Contribuição da educação infantil de qualidade e seus impactos no início do ensino fundamental. Educação e Pesquisa, São Paulo, v. 37, n. 1, p. 15-33, abr. 2011. [ Links ]
COSTA, M.; KOSLINSKI, M. C. Escolha, estratégia e competição por escolas públicas: pensando a ecologia do quase mercado escolar. In: ENCONTRO ANUAL DA ANPOCS, 33., 2009, Caxambu. Anais... São Paulo: Anpocs. v. 1. p. 1-26. [ Links ]
DUBET, F. Le Déclin de l'institution. Paris: Sueil, 2002. [ Links ]
______. Faits d'école. Paris: EHESS, 2008. [ Links ]
FREITAS, L. C. A Internalização da exclusão. Educação e Sociedade, São Paulo, v. 23, n. 80, p. 301-327, set. 2002. [ Links ]
FUNDAÇÃO IBGE. Censo demográfico 2000. 2000. Disponível em: <http://www.ibge.gov.br/home/estatistica/populacao/censo2000/default.shtm.> Acesso em set. 2011. [ Links ]
KOSLINSKI, M. C.; COSTA, M. Competing for public schools in Rio de Janeiro: reflexions on a hidden quas-market. In: PENSER LES MARCHES SCOLAIRES, 2009, Genève. Genève: Rapp, Université de Genève, 2009. p. 1-22. [ Links ]
KOWARICK, L. Viver em risco: sobre a vulnerabilidade socioeconômica e civil. São Paulo: Ed. 34, 2009. [ Links ]
LOMÔNACO, B. P.; GARRAFA, T. C. A Complexidade da relação escola-família em territórios vulneráveis. Cadernos Cenpec, n. 6, p. 27-38, jan./jun. 2006. [ Links ]
MALOUTAS, T. Efeitos de vizinhança e desempenhos escolares. In: VAN ZANTEN, Agnès (Org.). Dicionário de educação. Petrópolis: Vozes, 2011. p. 288-291. [ Links ]
MAURIN, É. Le Ghetto français: enquête sur le séparatisme social. Paris: Seuil, 2004. [ Links ]
MOORE, R. Capital. In: GRENFELL, M. (Ed.). Pierre Bourdieu: key concepts. Stocksfield, UK: Aucumen, 2008. [ Links ]
PADILHA, F. et al. As Regularidades e exceções no desempenho no Ideb nos municípios. Estudos em Avaliação Educacional, São Paulo, v. 23, n. 51, p. 58-81, jan./abr. 2012. [ Links ]
RIBEIRO, L. C. Q. Desigualdades de oportunidades e segregação residencial: a metropolização da questão social no Brasil. Caderno CRH, Salvador, v. 23, n. 59, p. 221-233, maio/ago. 2010. [ Links ]
RIBEIRO, L. C. Q.; KAZTMAN, R. (Org.). A Cidade contra a escola? Segregação urbana e desigualdades educacionais em grandes cidades da América Latina. Rio de Janeiro: Letra Capital; Montevidéu: Ippes, 2008. [ Links ]
RIBEIRO, L. C. Q.; KOSLINSKI, M. C. A Cidade contra a escola? O caso do município do Rio de Janeiro. Revista Contemporânea de Educação, v. 4, n. 8, p. 221-233, ago./dez. 2009a. [ Links ]
______. Efeito metrópole e acesso às oportunidades educacionais. EURE, Santiago, v. 35, n. 106, p. 101-129, dic. 2009b. [ Links ]
SANT'ANA, M. J. G. O Papel do território na configuração das desigualdades educacionais: efeito escola e efeito vizinhança. In: CARNEIRO, S. M. S.; SANT'ANA, M. J.G.. (Org.). Cidade: olhares e trajetórias. Rio de Janeiro: Garamond, 2009. p. 167-192. [ Links ]
______. São Paulo em perspectiva, São Paulo, v. 20, n. 1, jan./mar. 2006. [ Links ]
SETUBAL, M. A. Os Desafios de uma educação de qualidade em comunidades de alta vulnerabilidade. Cadernos Cenpec, n. 6, p. 7-19, jan./jun., 2006. [ Links ]
SOARES, J. F.; RIGOTTI, J. I. R.; ANDRADE, L. T. As Desigualdades socioespaciais e o efeito das escolas públicas de Belo Horizonte. In: RIBEIRO, L. C. Q.; KAZTMAN, R. (Org.). A Cidade contra a escola? Segregação urbana e desigualdades educacionais em grandes cidades da América Latina. Rio de Janeiro: Letra Capital; Montevidéu: Ippes, 2008. p. 119-144. [ Links ]
TORRES, H. G.; BICHIR, R. M.; GOMES, S.; CARPIM, T. R. P. Educação na periferia de São Paulo: ou como pensar as desigualdades educacionais. In: RIBEIRO, L. C. Q.; KAZTMAN, R. (Org.). A Cidade contra a escola? Segregação urbana e desigualdades educacionais em grandes cidades da América Latina. Rio de Janeiro: Letra Capital; Montevidéu: Ippes, 2008. p. 59-90. [ Links ]
TORRES, H. G.; FERREIRA, M. P.; GOMES, S. Educação e segregação social: explorando as relações de vizinhança. In: MARQUES, E.; TORRES, H.G. (Org.). São Paulo: segregação, pobreza e desigualdade. São Paulo: Senac, 2005. p. 123-141. [ Links ]
VAN ZANTEN, A. Efeitos da concorrência sobre a atividade dos estabelecimentos escolares. Cadernos de Pesquisa, São Paulo, v. 35, n. 126, p. 565-593, set./dez. 2005. [ Links ]
______. Compétion et fonctionnement des établissements scolaires: les enseignements d'une enquête européenne. Revue française de pédagogie, n. 156, p. 9-17, jui./sep. 2006. [ Links ]
YAIR, G.. School organization and market ecology: a realist sociological look at the infrastructure of school choice. Britsh Journal of Sociology of Education, v. 17, n. 4, p. 453-471, 1996. [ Links ]
Received: DECEMBER 2011 TRANSLATED BY David Coles
Approved for publication: JANUARY 2012
1 Population data is based on Fundação Seade estimates for 2007. To consult indicators for social inequality in the city of São Paulo, see the Índice Paulista de Vulnerabildiade Social - IPVS (http://www.seade.gov.br/projetos/ipvs/) and the Observatório Cidadão da Rede Nossa São Paulo (http://www.nossasaopaulo.org.br/observatorio/index.php).
2 In the State and in the City of São Paulo, a 2 km radius is considered for enrollment at a school, measured by the postal code (CEP) that each family declares. However, given the proximity between schools in the region, we opted for a smaller radius allowing greater differentiation between the schools and their surrounding areas.
3 With the fieldwork data it was possible to analyze the real situation of each school on the basis of information about the workings of the others. This comparative analysis procedure enabled the apprehension of certain general processes or mechanisms that operate among the schools and that ensure that the territory has consequences on the quality of education provided that are difficult to be identified quantitatively. As will be seen, these are processes concerning interdependencies that the schools establish among themselves and concerning the institutional model that governs the schools, and both have implications for the type of sociability and the school management conditions in each unit.
4 Data from the 2007 Prova Brasil and IDEB were used, since the micro data for the 2009 edition had not been released by November 2011.
5 Two reasons determined this form of standardization for IDEB for the schools. Firstly, it enabled identification of internal variations from the local average, which is similar to the average for the city (municipality) of São Paulo. Secondly, because in 2007 there were 61 schools in the region, not all of them offering every segment of primary education. Thus, if we were to work in isolation with data for EF1 (1st to 4th years) and EF2 (5th to 8th year), we would have two non-coinciding samples of schools, each one with up to 40 units. To calculate the AP for each unit, on a case-by-case basis, we took into consideration either the average AP of two segments, or the AP of the only segment provided by the school.
6 In 2007, the IDEB for the city of São Paulo was 4.5 in EF1 and 3.8 in EF2. Schools in the sub-prefecture had IDEB scores of 4.17 in EF1 and 3.72 in EF2. According to the classification of the Observatório Cidadão da Rede Nossa São Paulo, this data places the São Miguel sub-prefecture among the worst ranking results in the city, particularly in the first years of primary education. (http://www.nossasaopaulo.org.br/observatorio/index.php).
7 We chose to construct an indicator that would express cultural capital, which speaks more directly to the resources demanded by the school system. We therefore chose a set of variables from the socio-economic questionnaire of the Prova Brasil that would give more information about two dimensions of cultural capital (BOURDIEU, 2008; MOORE, 2008): institutionalized cultural capital and objectified cultural capital. In the tests carried out, this indicator was more sensitive to capturing educational variations than the traditional NSE, which was particularly important when working with a universe of only 61 schools. We preferred to use the expression "cultural resources" because here we are addressing a social group that largely does not have the cultural capital that is profitable in the school world and in the social space.
8 For each grade that was assessed, the SARESP (São Paulo State School Performance Evaluation System-Sistema de Avaliação do Rendimento Escolar do Estado de São Paulo) translates the results of the scale of proficiencies into the following levels: below basic, basic, adequate and advanced. SARESP and SAEB (Compulsory Education Evaluation System-Sistema de Avaliação da Educação Básica) of which IDEB is part, use the same methodology, Item Response Theory (IRT) and the same scale ( http://saresp.fde.sp.gov.br).
9 As has been shown, the students' results were categorized in accordance with the learning expectations of SARESP. However, in the two graphs, only the extreme expectations or categories have been retained ("below basic" and "adequate and advanced"), because these are the categories that best show the differences we wish to highlight. The "basic" category comes between them, data for which is difficult to interpret since it may both indicate improvement over the lower level and stagnation or deterioration over the higher levels. Thus the differences between the sums of percentages of each color bar and 100% represent the percentage of the "basic" category.
10 Social equipment means a state-public, private or third sector institution whose goal is to help make concrete a social right. In the second place, we took into consideration types of equipment linked to civil rights, such as public safety. We geocoded the following types of equipment: formal education, culture, sports and leisure, health, social help and public safety equipment.
11 On the issue of child education and its impact on schooling, see Campos et al. (2011).
12 Cf. data for the Observatório Cidadão da Rede Nossa São Paulo: http://www.nossasaopaulo.org.br/observatorio/index.php.
13 This indicator was compared to the Index of Dissimilarity put forward by Reynolds Farley of the University of Michigan's Center for Population Studies (http://enceladus.isr.umich.edu/race/calculate.html). Similarity between the indices was of the order of 95%.
14 These analyses are reinforced by data gathered in 2011 during a different stage of this survey from a group of families living in a highly socially vulnerable micro-territory in São Miguel Paulista. As one of the mothers residing in this region, internalizing the social stigmas attributed to its population, put it: "I had a very poor opinion of school Z [because of violence and lack of discipline], you know? But I came to realize that the problem was not the school. The people who go to that school were no good. Not the school itself. You have to go to school and you have to want to study." This mother has high educational expectations, and a certain sense of distinction vis-à-vis the neighborhood and among her educational strategies avoids the neighborhood.
15 A study on public competitive exams for the transfer of teachers at the 61 schools of the sub-prefecture is under way using data obtained from the official Government Daily Gazette and the School Census (Diário Oficial and Censo Escolar, respectively).
16 On the student body and family side, even among the most vulnerable, practices of avoidance of schools with poor reputations or schools that bring together students from highly socially vulnerable territories, such as attempting to enroll in schools with better reputations and those that better protect their children, were common; on the school side, (and an ongoing study is broadening this data, obtained through interviews with Education Secretaries in different sub-prefectures in the city of São Paulo) we found "harmless" student selection practices (for example, a preference for enrolling students transferred from private schools and the refusal of enrollment to children and young people according to their neighborhood of residence, or if they are lagging behind, have low grades, a history of bad behavior or due to other socio-educational measures); we also found "harmless" practices of the expulsion of students showing undesirable behavior.
17 It must be pointed out that there are engaged and qualified professionals at these schools, who are ethically and politically committed to the school and the families of pupils. However, what stands out here is that these professionals are not in the majority and it is very difficult for them to find peers with whom they can work as a team.
18 This article does not seek to provide a deep bibliographical balance of studies on the issue of education in metropolises. Brazilian studies are nourished above all by debates produced in English-language and French-language research. Some of these international studies will be cited only insofar as they are important for clarifying aspects we debate more directly.
19 This is from a dialog carried out between the teachers of different schools in a meeting to validate the results.
Received: DECEMBER 2011
TRANSLATED BY David Coles