Class, race and access to superior education

Abstracts

This article presents evidences on the relationship between class, race and access to superior education based on data from the 2006 round of the National Household Survey, PNAD. It aims to bridge the knowledge gap on this issue, which exists in spite of the intense public debate over the adoption of policies to promote equitable access to superior education by black and by poor Brazilians. The indicators presented point out the existence of association between class, race, and access to superior education beyond the effects of the relationship between class, race and attainment of secondary education. It is concluded that if the decision - eminently political - about the groups to be benefited by such actions purports to extract its legitimacy from being evidence based, both the black youth and the poor should be contemplated.

RACE; HIGHER EDUCATION; POVERTY; ACCESS TO EDUCATION

Este artigo apresenta evidências produzidas a partir da Pesquisa Nacional por Amostra de Domicílios de 2006 sobre as relações entre classe, raça e acesso ao ensino superior. O objetivo é preencher a lacuna de conhecimento que ainda persiste sobre o tema, a despeito do intenso debate público sobre a adoção de políticas de promoção de acesso equitativo ao ensino superior para negros ou para pobres. Os indicadores apresentados apontam a existência de associações entre classe, raça e acesso ao ensino superior para além dos efeitos das relações entre classe, raça e término do ensino médio. Conclui apontando a necessidade simultânea de quotas sociais e raciais para atingir o objetivo de atacar a desigualdade de acesso à universidade.

RAÇA; ENSINO SUPERIOR; POBREZA; ACESSO À EDUCAÇÃO

Class, race and access to higher education in Brazil

Ragaela Guerreiro Osorio

Researcher from the Department of Social Studies and Policies of the Applied Economics Research Institute – Brasília E-mail: [e-mail]

Translation: Robert Dinham

ABSTRACT

This article presents evidence of the relationship between class, race and access to higher education based on data from the 2006 round of the National Household Sample Survey, PNAD. It aims to bridge the knowledge gap on this issue that still exists in spite of the intense public debate over the adoption of policies to promote equitable access to higher education for black and poor Brazilians. The indicators presented point out the existence of an association between class, race, and access to higher education that goes beyond the effects of the relationship between class, race and finalizing secondary education. It concludes by pointing out the need for both social and racial quotas for achieving the objective of fighting the inequality that exists in access to university.

RACE – HIGHER EDUCATION – POVERTY – ACCESS TO EDUCATION

Should a policy that promotes equitable access to higher education be adopted in Brazil? If the reply is “yes”, which social groups should benefit - blacks and/or the poor? Is the difference in access due to class differences or to racial discrimination?

There is intense public debate about these questions in Brazilian society and this gives rise to articles and even books, both for and against the adoption of quotas for one or other social group, legislative bills in Congress, public letters, petitions and every type of manifestation. The eventual adoption of such policies and definition of the groups that should benefit are topics, to which a significant number of Brazilians are not indifferent, particularly those who participate in the formation of “public opinion”. It is surprising to discover, however, that there are no diagnoses of the extent of the problem, just disaggregated attendance indicators. Many assumptions are made, but it is not known if there are any actual relationships between class, race and access to higher education in Brazil and what the nature of these relationships is.

The objective of this article is to fill this gap, by quantifying the inequality that exists in access to higher education, by class and race. The work focuses on the Brazilian population between 18 and 24, the age band at which people normally enter higher education. The data source used is the 2006 National Household Sample Survey (PNAD) that was conducted by the Brazilian Institute of Geography and Statistics (IBGE), covering the whole country.

The two following sections present the definitions of class and race that are used and the data that characterize and form the relationship that exists between these two dimensions. Then, the higher education access data are presented. The final section contains a summary of the facts that contribute to the debate about the adoption of mechanisms that equitably promote access to higher education.

CLASS

There is no consensus in sociological literature about the concept of class, or of how such a concept can be applied in operational research. On the contrary, there are various concurrent approaches (Crompton et al., 2000; Grusky, 2001; Wright, 2005). Despite this, in the public debate about access to higher education, categories like “rich”, “poor” and “middle class” are frequently employed, even though precise definitions on their boundaries are lacking. The hierarchy of these categories, however, is obvious and normally understood in terms of income. In other words, what differentiates the poor, the middle class and the rich is the average income of each group, which indicates their capacity for consumption and their level of well-being.

In this article these three “classes” will be defined according to household income per capita, in line with the criterion of the appropriation of total income into thirds, as suggested by Osorio (2003). This criterion consists in ranking the population, from the poorest individual to the richest, and then accumulating the income of the individuals so ordered. The distribution point at which accumulated income is equal to the first third of total income is the boundary between the poorest, the lower class, and the middle class. The boundary between the middle class and the richest – the upper class – occurs at the point at which accumulated income is equal the second third of total income. So, each of the classes – lower, middle and upper – appropriates one third of the total income and what differentiates them is their size and, consequently, the average income. It should be noted, however, that it is the income that is being divided into three parts of equal size, and not the population.

According to the 2006 PNAD the total income of Brazilian household groups was around R$ 89.4 billion a month. Although it is known that the PNAD underestimates the total amount of income when compared to other sources, such as the Family Budget Survey and the National Accounts System, Barros, Cury and Ulyssea (2007) proved that this does not affect studies that focus on distribution – which is the case with the division into social classes used here. By applying the criterion of thirds to the data from the PNAD to define the three social classes, each class would have a monthly income of approximately R$ 29.8 billion. The distribution of the total population and of young people from 18 to 24 and the income characteristics of the classes thus obtained are shown in Table 1.

Table 1 shows that the 5.5% richest people in the population received a third of the total income; that the 74.6 % poorest people took one third and that the 19.9% which fell between the richest and the poorest appropriated the other third. These percentages express the size of the three classes in 2006. As an example, a family of four people who had household income of up to R$ 2,024 was considered lower class; in the R$ 2,024 to R$ 6,000 band, middle class; and over R$ 6,000 upper class. The average household income per capita of the lower class was 27% of the income of the middle class and just 7% of the income of the upper class; that of the middle class was 28% of the income of the upper class.

Young people between 18 and 24 were slightly over-represented in the lower and middle classes and under-represented in the upper class – in terms of their distribution vis-à-vis the whole population. On the other hand, young people between 18 and 24 who had finished their secondary school education are over-represented in the middle and upper classes and under-represented in the lower class; just 43.8% of the young people from 18 to 24 had finished secondary school. In the last column, it can be seen that the percentages of young people between 18 and 24 who have completed secondary school were very similar in the middle and upper classes, at more than 70%. In the lower class this percentage was 34.2. Consequently, the chance of finding a young person from the lower class who had finished secondary school was smaller than finding this same person in the other two classes.

In short, although there is no strong association between being a young person between 18 and 24 and belonging to a particular class, there is a strong association between class and completing secondary school.

RACE

The definition of race and of racial groups is as controversial as that of class. To apply the concept of race from the PNAD 2006 in operational research, it is necessary to employ the requisite “color or race”, according to which, the people interviewed are classified into one of six categories: white, colored, black, yellow, Indian and not-known. Around 99% of the population place themselves into one of the first three categories. This classification system has been used almost unaltered by the IBGE since the census of 1940. Much studied and heavily criticized, it has already proved to be suitable for studying racial inequality in Brazil, despite the complexity of the phenomenon (Petruccelli, 2000; Osorio, 2003; Rocha, Rosemberg, 2007).

Oliveira, Porcaro and Araujo (1985) mention a bibliographic survey covering hundreds of studies that have various views on Brazilian racial issues published before the 1970s, in which it can be seen that academics used to indiscriminately use the word “black” to refer to a group of the population comprising both blacks and mulattos. The word “mulatto” was frequently used as a substitute for the word “colored” when presenting statistics.

Souza (1971) seems to have been the first to aggregate blacks and coloreds and call the resulting group “black”, alleging the socio-economic homogeneity of the two groups and the statistical need for doing so, due to the reduced size of the group of blacks in sample surveys. The socio-economic homogeneity of blacks and coloreds was one of the two central hypotheses put forward and confirmed by the PhD thesis of Nelson do Valle Silva (1978). The public policies, which are discussed or adopted in Brazil and that aim at improving the situation of the black population, generally point out that the latter comprises people who declare themselves to be black or colored in the data gathered by the IBGE.

For purposes of this article two groups were defined: one formed by people declared to be white or yellow, and another formed by people who are declared to be black, colored or Indian. The first group is called “white” and the second “black”. People of unknown color, who are an extremely small portion of the sample, were not considered. Table 2 gives the distribution of the total population and of young people from 18 to 24 years old from both racial groups by class, the total racial composition and the composition of each class.

In Table 2 the composition statistics reveal the percentage of blacks and whites in a particular population in each class and in total. In the total population the racial composition was half and half in 2006, but in the middle class and in the upper class the percentage of blacks was very much smaller than 50%, and in the lower class was bigger. There was a considerable difference in the percentage of young people from each group who had finished secondary school. While just 54.7% of young whites from 18 to 24 had completed secondary school, among young blacks the percentage was even smaller, at 33.9%.

Due to the fact that the black population is slightly younger, its percentage in the 18 to 24 year age band in the lower class was a little bigger than the percentage of whites, but in the other two classes the percentages of young whites and blacks were very similar. Despite this, the percentage of whites in the population of 18 to 24 year olds who have finished secondary school was greater than in all classes. Around 12.3% of the whites from the lower class were between 18 and 24 and of these 42.6% had completed secondary school. But among young blacks from the lower class only 28.7% had finished secondary school. The difference between these percentages reduces as one moves up the social pyramid, but does not disappear. Therefore, the chance of finding a young black who has finished secondary school was smaller than of finding a young white with a secondary school education in all classes.

Both the total black population and the black population between 18 and 24 were similarly distributed across classes, with a concentration in the lower and middle classes. The white population was also concentrated in these two classes, but much less so in the lower class. Returning to Table 1, it can be seen that around 5.5% of the total population was from the upper class. In contrast, 9.1% of the whites were from the upper class and just 1.9% of the blacks. However, the distribution of the young, whether white or black, who have finished secondary school, was less concentrated in the lower class, with larger percentages – relative to young people from the same age band – in the middle and upper classes.

In short: there is a strong association between class and race, but there is not a strong association between race and being young in the 18 to 24 year age band. Over and above the association between class and having completed secondary school (Table 1) there is an association between race and having finished secondary school. Therefore, it can be said that there is an association between class, race and having finished secondary school education.

CLASS, RACE AND ACCESS TO HIGHER EDUCATION

To enter higher education it is necessary to have completed secondary school education. In Brazil in 2006, 18.5% of the population had finished secondary school (regardless of age and excluding the population that has already finished higher education). Young people between 18 and 24 represented 30.6% of the population with a secondary school education. Of the population who had completed secondary school education 15.6% had gone on to study on a higher education level course. More than half of the people attending a higher education course (55.1%) were between 18 and 24, but 71.9% of young people between 18 and 24 who had a secondary school education did not go on to do a higher education course.

It is reasonable to suppose that most people enter higher education when they are in this age band, considering that only 24.8% of the population attending a higher education course was 30 or more. Because of this, from this point on, all the data presented refer to the population between 18 and 24 who have finished their secondary school education.

Considering that 28.1% of young people between 18 and 24 who finished their secondary school education went on to do a higher education course, if there were no inequalities of class and/or race in access to higher education it could be expected that this proportion would be repeated in all groups that are formed by the crossing of these two dimensions. In other words, if young blacks or whites from any class were to be considered, this total attendance rate would be found. As can be seen in Table 3, this was not what happened. The attendance rates of the middle and upper classes were higher than 28.1%, and the rates of whites were always higher than those of blacks, even in the lower class.

If there were only racial, but not class inequality, the total attendance rates for blacks and whites, 34.8% and 18.2 % respectively, would be repeated in all classes. Otherwise, if there were just class inequality, blacks and whites would have attendance rates equal to the totals for their class: 13.3%, 43.1% and 75.2%. In analyzing attendance rates by class and race in Table 3 as a function of these expected values, based on the hypothesis of an absence of association between race and class - or both - and access to higher education, the association appears net of the effect of the association between class and race and having completed secondary school education, as represented in Table 2.

The total attendance rate can be split down to analyze the contribution of each class and racial group. To do so it is sufficient just to divide the number of members attending higher education, not by the population of the group, but by the whole population between 18 and 24 who have finished secondary school education. This was done in the three columns of intermediary data in Table 3. Here it can be seen, for example, that 3.1% of the young people from 18 to 24, were blacks from the lower class who were doing a higher education course. Once again, breakdown shows that whites have higher education attendance rates than blacks in all classes. However, it reveals an additional fact: despite the attendance rate being higher among young people from the upper class, 14.4% of the young people from 18 to 24 who finished secondary school and who attended higher education courses were from the middle class. Therefore, more than half of the young people attending higher education courses were from the middle class. Another interesting aspect from this outline is that it reveals that there are more young people from the lower class than the upper class attending higher education courses.

However, this does not mean that access is equitable from the point of view of class. In fact, returning to Table 1, it can be noted that the distribution by social class of young people between 18 and 24 who have finished their secondary school education is as follows: 58.8% lower class; 33.5% middle class; 7.7% upper class. Consequently, if access were equitable, the young people from the lower class should be in the majority among students in higher education. If the associations between class, race and higher education, as revealed by the attendance rates, did not exist, i.e. if the attendance rates of each group were exactly 28.1% instead of the values seen in the breakdown, we would have the counterfactual (simulated) values that are presented in the last three columns of Table 3.

Comparing the rate broken down by simulation with the breakdown observed, it can be noted that access to higher level education is unequal by class and race, even if one considers as given the preexisting associations between class and having completed secondary education (Table 1); race and having completed secondary education (Table 2); class and race and having completed secondary education (Table 2).

In Table 4 the information from Table 3 is represented in a different way to facilitate understanding. The values of the cells were divided by the reference value: the total percentage of young people between 18 and 24, 28.1 %. It can be noted that the higher education attendance rate of young white people from the upper class was 2.7 times the total rate, or that the rate for blacks from the lower class was 0.4 times the total rate.

In the case of the breakdowns, Table 4 contains the weight percentage of the contribution made by each class and racial group to the overall rate (observed or counterfactual). For example, young people from the lower class contribute 27.8% of the total rate and young whites from the upper class were 17.5% of all students in higher education. However, if there were no class and race inequality in access to higher education then young people from the lower class should represent 58.8% of all students and young whites from the upper class would be just 6.4%.

Therefore, in addition to the association between class, race and having finished secondary school (Tables 1 and 2), there are also associations between class, race and access to higher education for young people between 18 and 24. With regard to young whites from the upper class, all the groups suffer from a degree of access deficit to higher education. The lower class is the only one that has attendance rates below the totals, but the middle class also shows a considerable access deficit relative to the upper class. This deficit is always greater among blacks, even among those who come from the upper class.

CONCLUSIONS

There is currently a warm debate in Brazilian society about the adoption of affirmative action for access to public universities. The debate revolves around reserving places by establishing quotas for certain social groups. Among those who are in favor of the adoption of quotas, there are those who argue that they should be social, for people who have a low family income and/or those coming from public schools; those who argue they should be racial, for blacks and coloreds, and possibly indigenous; and there are those who argue for both types of quota.

In defending social quotas it is frequently argued that there is a coincidence between having a low income, studying in a state school and being black. By applying to the poorest, these quotas would mainly benefit blacks and those coming from state schools. In this approach, it is believed that racial quotas would benefit a black middle class in detriment to the poorest whites, a problem that would not occur with social quotas. So, social quotas would resolve both class inequality as well as racial inequality in access to higher education, without creating “dangerous divisions” that would racially segregate the nation, thereby instilling hatred between groups.

Only part of this argument can be analyzed from the facts of the case. It is impossible to know if the introduction of racial quotas will create “dangerous divisions”; this is an exercise in predicting the future. Likewise, it is possible to carry out an opposite exercise to predicting the future, in which the non-introduction of racial quotas would create “dangerous divisions” due to Brazilian society's inability to integrate half of its population and provide it with equal opportunities. There is no way of deciding in favor of one or other scenario.

Study demand in state schools also causes considerable problems when it comes to empirically analyzing its pertinence. There are no data sources in Brazil that allow for an estimate to be made of the proportion of possible university candidates who always, or for a time, studied in state public schools. Although there certainly exists an association between studying in a state school and having a low income, establishing study in a public school as a criterion for claiming benefit of the social quota would be problematic. It would be necessary to decide when and for how long a person who had had a mixed path in the education system would need to have studied in state schools. There would be a risk of doing a grave injustice to those families which, despite their low income, decided to sacrifice their present consumption to invest in the future of their children by paying for private schools for them in the hope of providing them with a better quality education. There would also be the problem of those who had studied in private schools on scholarships. There is, however, no way of estimating the incidence of each one of these situations.

Social quotas based on family income would undoubtedly provide many of the benefits alleged by those who defend them. Most of the underprivileged would obviously be people who have attended state schools at some time, or have always studied in them. Scholarship-holders who have studied for free in private schools, and the children of poor families, which have made a sacrifice to pay for private schools for them, would also benefit. Poor whites would not be overlooked and a good part of the problems that are rightly or wrongly attributed to racial quotas would be eliminated; and given the fact that blacks are the majority among the poor their children would be favored without benefiting the black middle class.

However, in the light of the facts here presented, the adoption of social quotas may be considered insufficient from the point of view of racial inequality, because a requirement for entering higher education is the conclusion of secondary school education. The probability of a young, low income person concluding secondary school is remote, and it is even more remote if the young person is black. Consequently, the adoption of social quotas would benefit young whites more. Furthermore, it has been shown that both blacks from the middle and upper classes have access deficits relative to whites from the same classes. The supposition underlying the allegation that young blacks from these classes are unlikely to benefit because they compete for access to higher education under equal conditions with young whites cannot be sustained. Young blacks from the middle class, although not so prejudiced as those from the lower class, both whites and blacks, have a considerable access deficit when compared with their white peers. In fact, all young people have access deficits when compared with young whites from the upper class, of whom more than three quarters are in a higher education course.

Given the coincidence between race, class and attending a state school, purely racial quotas would also suffer from many of the effects claimed for social quotas, with the advantage that they are more efficient from the point of view of racial inequality. Given the fact that blacks are over-represented among the poorest, many of those who would benefit would come from the lower class and/or be from state schools.

However, the adoption of purely racial quotas would be insufficient from the point of view of social inequality – and there is obviously a public call for attacking this problem. In fact, there is a reasonably large contingent of young whites from the lower class which has finished secondary school. These young people have a considerable access deficit to higher education, which is bigger than that of young blacks from the middle class. Purely racial quotas, in addition to overlooking them, would prejudice them, because they would have to compete for a smaller number of places with whites from the middle and upper classes, who are better prepared to sit the higher education entrance exams; and it is for this reason, better preparation, that purely racial quotas would be of greater benefit to blacks from the middle and upper classes than those from the lower class.

The statistical evidence of the relationship between class, race and access to higher education that has been presented here leaves no doubt; the decision cannot be about adopting either purely social or purely racial quotas. If the objective is to attack unequal opportunities in access to university it is necessary to have both social and racial quotas, simultaneously.

Two additional considerations are necessary. The first is that the debate about quotas is centered on access to state universities and the evidence presented here was about access to higher education, without distinction as to the system – state or private. This is due to questions of a technical order: the size of the public system would demand additional statistics about the reliability of the indicators for sampling purposes. In any event, there is no reason to consider that class and racial inequality in access to state universities is any less, or very different from what has been characterized here.

A second consideration has to do with both social and racial quotas. Quotas are somewhat rudimentary mechanisms for promoting equal access opportunities to higher education. But there is no doubt that, even though they are rudimentary, from the point of view of equal opportunities it is better to have quotas than not to have them. There are better and more sophisticated alternatives than quotas. But these alternatives also demand the definition of the groups that will benefit. The current debate about quotas is precisely about which groups should be favored. At the right moment, when the decision about the groups has been taken, more refined ways of promoting equitable access to higher education can be thought of.

The decision about the groups that should benefit, however, is political; it cannot come from groups of the enlightened. In as far as is possible it must involve the whole of society.

BIBLIOGRAPHIC REFERENCES

• BARROS, R. P.; CURY, S.; ULYSSEA, G. A. Desigualdade de renda no Brasil encontra-se subestimada? Uma análise comparativa usando PNAD, POF e Contas Nacionais. In: BARROS, R. P.; FOGUEL, M. N.; ULYSSEA, G. (eds.) Desigualdade de renda no Brasil: uma análise da queda recente, 1. Brasília: Ipea, 2007. p.237-273.
• CROMPTON, R. et al. (ed.) Renewing class analysis Oxford: Blackwell, 2000.
• FUNDAÇÃO IBGE. Pesquisa nacional por amostras de domicílios: microdados. Pesquisa básica e aspectos complementares de educação, afazeres domésticos e trabalho infantil. Rio de Janeiro, 2007. CD-ROM.
• GRUSKY, D. B. (ed.) Social stratification: class, race and gender in sociological perspective. 2.ed. Westview: Boulder, 2001.
• OLIVEIRA, L. E. G.; PORCARO, R. M.; ARAUJO, T. C. N. O Lugar do negro na força de trabalho Rio de Janeiro: IBGE, 1985.
• OSORIO, R. G. Mobilidade social sob a perspectiva da distribuição de renda Brasília, 2003. Dissert.(mestr.) Departamento de Sociologia, Universidade de Brasília.
• ______. O Sistema classificatório de cor ou raça do IBGE. Brasília: Ipea, 2003. (Textos para Discussão, 996)
• ROCHA, E. J.; ROSEMBERG, F. Autodeclaração de cor e/ou raça entre escolares paulistanos(as). Cadernos de Pesquisa, São Paulo, v. 37, n. 132, p. 759-799, set./dez. 2007.
• PETRUCCELLI, J. L. A Cor denominada Rio de Janeiro: IBGE, 2000. (Texto para Discussão, 3)
• SILVA, N. V. Black-white income differentials: Brazil, 1960. Ann Arbor, 1978. Thesis (Ph D.) Department of Sociology, University of Michigan.
• SOUZA, A. Raça e política no Brasil urbano. Revista de Administração de Empresas, Rio de Janeiro, v. 11, n. 4, p.61-70, dez. 1971.
• WRIGHT, E. O. (ed.) Approaches to class analysis Cambridge: Cambridge University Press, 2005.

Publication Dates

• Publication in this collection
  27 May 2010
• Date of issue
  Dec 2009

History

• Received
  Aug 2008
• Accepted
  Dec 2008