Youth Map of the City of Santo André, Southeastern Brazil: an Instrument to Read Social Inequalities

extremo do acesso (" quase central ") e outra mais próxima à privação (" quase periférica "). As variáveis discriminantes estavam relacionadas ao trabalho, à CONCLUSÕES: As diferenças entre os grupos sociais expressaram desigualdades importantes entre os jovens que vivem, estudam e/ou trabalham na cidade, o que contribuirá para o planejamento de políticas sociais públicas dirigidas a esses grupos.

Youth is a theme that has been increasingly discussed in modern society, notably due to the complexity of problems young adults have been facing.As this is an object of interest in several fields of knowledge, it mobilizes research development and public policies.
Growth in population age distribution in the 15-to-24 -year age group in the 1990s, especially in Brazilian metropolitan areas, 3 justifies the social interest for this group as a strategic development agent.On the other hand, the so-called "job crisis" threatens the incorporation of young adults into the job market: 6 whereas those from economically stable social groups continue their studies and remain with their family of origin, the others have to submit to bad jobs and low salaries.a The perspective of collective health views youth from the social class and generation categories, submitting the reading of modern reality of youths to their class condition.Thus, there are several youth realities, and although having the generation condition in common, they end up experiencing it in different ways, according to their families' inclusion in production and consumption. 12itical epidemiology, one of the main collective health instruments, has the inequalities of social reproduction as its subject.Social class is a category of analysis capable of revealing the weight of inequalities in health and making what is abstract, underlying the population category, concrete.This category, in its turn, is conceived as a homogeneous group of individuals, without referring to their possibilities of action. 9suming that mapping inequalities constitutes a useful instrument to plan public policies aimed at young adults, this study is also founded on the Santos' concept of space, 10 "Space is understood as the environment, the material place of possibility of events (…) It is considered as something dynamic and unitary, where materiality and human action are included.Space would be the indissoluble set of systems of natural or manmade objects, the latter resulting from action systems, whether these are determined or not.In each period, new objects and actions are added to the previous ones, changing the whole, both in form and substance." The space category enables the recognition of the urban heterogeneity, once it incorporates social dynamics and structure -where and how people live; what determines the production and flow of goods and services., does not portray the social dynamics in their entirety, with the historical changes they have undergone and "which do not enable one to predict the social contents connected to housing in such places". 14us, it is necessary to incorporate variables that reflect different perspectives of social reproduction to better understand social inequalities.
Thematic maps are an instrument to show results that enable reality to be represented in a fast and broad way, providing a wider perspective of the city's social differences. 2In this way, gathering information and spatial references is a powerful tool to analyze and understand data, aiming to plan public policy actions.These maps enable the analysis of characteristics and differences present in each territorial space, beyond the simple geographical division. 8cial policies based on territorial differences are found to be important to change the strong negativity of poor areas, due to their lack of urban infrastructure services, housing, sanitation, education and public safety, among other things, which end up becoming decisive factors of reproduction of poverty. 15e present study aimed to analyze social inequalities of young adults living in an urban area by mapping socio-demographic and economic data.

METHODS
According to the Instituto Brasileiro de Geografia e Estatística (IBGE -Brazilian Institute of Geography and Statistics), the city of Santo André, Southeastern Brazil, had an estimated population of 671,696 inhabitants in 2008 and a Gross Domestic Product (GDP) of 17,066 reais per capita, practically the same as the average GDP of the state of São Paulo.Human Development Index is considered high (0.835), ranking in 24 th place in this state.b In the Atlas da Exclusão Social (Atlas of Social Exclusion), this city ranked in eighth place among the 100 cities with the highest number of rich families in Brazil and in third place among the 45 cities of the state of São Paulo. 7itially, a group of 57 demographic and socioeconomic variables, associated with young adults aged between 15 and 24 years, were selected, based on the 2000 Census.The "social reproduction" category was considered as the theoretical construction that gathers patterns of inclusion in production (working Data were pre-selected and submitted to factorial analysis, using the SPSS 2004 software.First, variables with little variability among the SDAs and, thus, little discriminating, were excluded.On the other hand, if the variable was considered qualitatively important to assess potential strengthening or weakening of the youth population, it was initially considered in the multivariate analysis, despite the standard deviation value. Next, factorial analysis 5 sought to reduce the group of variables to two, three, four (or more) factors without decreasing the SDAs' explanatory capacity of differentiation.This technique creates a measure (commonality), which enables one to analyze the proportion of variance of each variable, explained by common factors, with the variable being excluded when this measure is equal to or higher than 0.6.
The group of variables described in Table 1 was thus obtained, where two factors were identified.These factors were extracted by the analysis of main components, which explain almost 78% of the total variance of original variables.The first factor (life conditions) explains 59.6% of variance, whereas the second factor (working conditions) explains 18.0%.
Only the "septic tank" variable showed commonality below 0.6, although it remained in the final model due to its important weakening potential in certain social groups of the population studied, who live in watershed areas.
For this result of factorial analysis, a KMO=0.866value was obtained.The Kaiser-Meyer-Olkin (KMO) test is a statistical tool that indicates the proportion of data variance that can be considered common to all variables, i.e. that can be attributed to a common factor.In this way, the closer to 1 (unit), the better the result and the more suitable the group of data for the application of factorial analysis.This was also proved by Bartlett's sphericity test, which rejected the hypothesis of noncorrelation among data at a 5% significance level.Thus, the application of factorial analysis was found to be adequate for these data.
Based on these factors, cluster analysis was used, 1 following the k-means method.It was possible to identify four social groups based on the SDA clusters identified by variables.

RESULTS
The group that concentrated adolescents with the highest access to wealth was classified as central (C) and those with the lowest access to wealth as peripheral (P).A total of two intermediate groups were identified, one closer to the highest access to wealth (almost central -AC) and another close to the lowest access to wealth (almost peripheral -AP).However, darker colors, representing more peripheral spaces with lower access to wealth, also appear among more central spaces, showing that the center-periphery relation is not so symmetrical, as observed in the case of SDAs 13 and 16 (Figure ).The lowest number of SDAs was in spaces C (4) and AC (10), while the highest number of SDAs was found in the more distant spaces, AP (19) and P (10).Group C concentrated the lowest number of young adults (Table 2).
The variables that distinguished the SDAs are shown in Table 3.As regards adolescents' place of birth, the variable that discriminated the four groups was "to be born out of the state of São Paulo".Results showed that, among young residents in group P, the highest percentage of adolescents born in other states, excluding São Paulo, was found, the majority of which came from Northeastern Brazil.
A total of five variables were selected to discriminate the SDAs in terms of access to education.Adolescents who were in primary school totaled 6% in group C and 29.4% in group P.Those who should be completing or have completed primary school at 15 years of age showed delayed schooling, with a lower number present in group C. Regardless of age, the highest percentage of adolescents who were in secondary school belonged to group AP.The poorer they were, the higher the frequencies of their having completed primary school exclusively.Thus, 54.9% of adolescents who had completed primary school exclusively belonged to group P.Among those who did not have higher education, despite their being in the adequate age group, 25.1% were in group P.
The work variable that discriminated the four groups was "not having paid work".The results showed that "not having paid work" decreased as one moved towards the city's poorer spaces, except for the group situated in the most peripheral area.This group showed an even higher index of unpaid work, compared to more central groups, once many adolescents did not work because they did not have a job.Whereas inclusion in the job market can be postponed until the end of professional qualification among richer adolescents, non-inclusion in the job market is determined by unemployment among poorer ones.
One of the youth life condition variables that discriminated groups was the position of an adolescent as responsible for the household.Among those in this position, the majority came from group P. The adolescent's responsibility to provide for the household and certainly the pressure to be included in the job market were found to increase towards peripheral spaces.This fact should cause life conditions for these families to become worse, once lower per capita income indices are concentrated in this area.Another discriminating variable was the presence of a spouse or partner.The highest percentage of young adults living with a spouse or partner was in group P, as well as the fact of their having up to two children.
A total of three variables associated with housing conditions helped to define groups: number of residents per room higher than three, condition of lot defined as property that has been handed over and household sewage disposed of by a septic tank.Of all households in these conditions, the majority were in group P.
Among the variables associated with owning assets, not to have a car was the one that statistically contributed to discriminate groups.The majority of adolescents without a car belonged to group P.
The remaining variables also showed differences among groups, although these differences will be discussed in another study.

DISCUSSION
Among a great group of variables, the youth map of Santo André enabled the identification of the following two factors, capable of discriminating the different spaces of the city between themselves: life conditions and working conditions.By grouping these factors by similarity, based on the cluster analysis method, the existence of four different social groups, distributed throughout the spaces of the city of Santo André, was revealed.Although the difference between rich center and poor periphery is significant, the map revealed the location of poor social groups out of the so-called peripheral space and close to more central spaces.Results of the present study are similar to those from other studies.In the city of São Paulo, Southeastern Brazil, "higher levels of privation are present in more distant districts, especially in the north, east and south areas of the metropolitan region, while richer groups are highly concentrated in central São Paulo." 14 The map, constructed to reflect adolescents' social reproduction profiles, also shows great similarity with the map of poverty distribution in Santo André, designed from the "half a minimum wage per capita income" variable.7The lower number of SDAs comprised spaces C and AC, while the higher number of SDAs comprised the more distant spaces AP and P, thus confirming the national pattern of exclusion and wealth accumulation.The average monthly family income of rich young adults was about R$ 22,000.00 in 2003, a value 14 times higher than the average monthly family income in the country and about 80 times higher than the line of poverty, below which the poorest 20% individuals are. 7As regards the youth portion of the population, 31.3% can be considered poor, with a per capita family income of up to half a minimum wage.Only 8.6% come from families with a per capita income higher than two minimum wages and about 60% belong to the intermediate stratum, with a per capita income between half and two minimum wages.a This distribution follows the proportionality among classes, found in a sample from a study performed in the city of São Paulo in 2001, showing the growing proportion of young adults present in the population as poorer spaces are approached. 14e pattern of rich center and poor periphery is not sufficient to portray inequality among the city's adolescents, once there are poverty spaces close to the center, as observed in the city of São Paulo in 2003. 4hus, although the dual periphery-center relation, which marked the consolidation of the city of São Paulo as a Brazilian industrial metropolis in the 1960s and 1970s, has evolved towards a less polarized model with the rise of sub-centers, there was no rupture with consolidated inequality patterns, but rather the creation of new designs, with the spread of poverty in spaces closer to the center. 4ta on level of education from the 2005 National Household Sample Survey (PNAD) were in agreement with those on young adults of Santo André: per capita family income was an explicit divider of level of education, once this level in the poorest 20% of the population was almost half of that in the richest 20% (25.1% against 48.6%).b On the national level, it is known that many adolescents interrupt their studies in primary school and less than half of Brazilian adolescents aged between 15 and 17 years go to secondary school.c In Santo André, social inequality is even greater when it comes to the percentage of young adults in higher education courses, which is similar to that of the country, whose rates vary according to per capita income, ranging between 4.7% among those who earn from half to one minimum wage and 59.8% among those who earn five or more minimum wages.a Thus, it can be affirmed that the young adults' educational situation in the city of Santo André follows the national context and can be summarized as "the existence of 1.5 million illiterates; persistence of high student age-school grade distortion, thus compromising access to secondary school at the right age; low frequency of young adults in higher education; and restricted opportunities of access to professional education".d In its turn, level of education of heads of family is associated with per capita family income in a directly proportional relation.d In terms of working conditions, the analysis of unemployment in Brazil reveals that the highest rates are among young adults, responsible for 50% of national unemployment.A total of 26.2% of unemployed adolescents belong to low income families, whereas 11.6% belong to higher income ones.Young adults from higher income families have greater access to paid work and a formal working contract, which means that those from low income families are more excluded from social and labor law benefits.The phenomenon of inter-generational regression is observed, i.e. although having a higher level of education than their parents, these adolescents cannot obtain better working and life conditions. 7 addition, the number of hours worked varied according to the level of education: whereas 41% of those who only completed up to the fourth grade of primary school work more than eight hours a day, this excessive workload affects only 13% of those who have a higher education level.Similarly, a study in the city of Rio de Janeiro, Southeastern Brazil, showed that households with a high density of residents are situated in areas of great poverty, in a clear relation with income inequality. 13is relation is also valid for access to urban infrastructure services.If, on the one hand, Brazilian indices show an improvement in sewage system coverage in homes, with an increase in the connection to the sewage system and stability in the number of homes with a septic tank, on the other, this improvement was not sufficient to improve life conditions of all social groups homogeneously. 14e decision to use Census data to construct the youth map reveals some limitations and potential aspects.Among the main limitations, the lack of full correspondence between the "social reproduction" category, a theoretical framework of analysis, and the set of secondary data provided by the Census, selected as variables to be treated statistically, could be emphasized.Mapping has a macro-spatial character and, as a result, does not enable the assessment of social differences in the groups/spaces formed.Variables that can be potential strengthening or weakening aspects, such as debts or number of hours spent in transportation, are not usually collected in the census.Although data were from 2000, which may not reflect the city's current reality, the method of map construction enabled the identification of social inequalities.In addition, the map is an important instrument to perform new studies that consider young adults as individuals, which may enable the process of setting the limits of proportional sampling to be available to several groups.Moreover, this instrument is important to plan public social policies aimed at different groups in the spaces of social life.By considering unequal territories, social policies have the potential, in the medium and long terms, to promote social integration of youth groups segregated by poverty.

Table 3 .
Sociodemographic and economic variables discriminating areas of statistical data, according to social group.City of Santo André, Southeastern Brazil, 2000.
Source: Instituto Brasileiro de Geografia e Estatística (Brazilian Institute of Geography and Statistics).

Table 2 .
Distribution of the total population and that of young adults aged between 15 and 24 years, according to social group.City of Santo André, Southeastern Brazil, 2000.
en_05.indd 4 28/1/2010 16:08:55 Once income comes from work and determines housing conditions and possibilities of consumption, adolescents living in more peripheral areas are expected to have worse living conditions and lower access to durable goods.