| Resumo: | A base de dados utilizada é a Pesquisa Nacional por Amostra de Domicílios (PNAD — IBGE), dos anos de 1982, 1996 e 2014, considerando os indivíduos entre 28 e 41 anos, que compreendem três coortes. A variável dependente é escolaridade (relativa e absoluta) e a principal variável independente é origem social. Também são incluídas variáveis controle, como raça, idade, estrutura familiar, sexo, região e localização (rural/urbana). Os resultados desta pesquisa sugerem que, em um cenário de progressivo aumento da escolaridade da população brasileira em anos de estudo, há uma tendência de redução da relação entre origem social e escolaridade alcançada quando a educação é medida de forma absoluta. Porém, quando a educação é medida em termos relativos, a expansão educacional ocorrida nas últimas décadas no Brasil não foi capaz de diminuir os efeitos da origem social sobre a escolaridade alcançada. Tendo em conta a posição relativa numa fila imaginária onde os mais educados estão à frente, não é possível perceber qualquer redução significativa dos efeitos de classe de origem sobre a escolaridade alcançada. Isso revela a aproximação com o diagnóstico de persistência da desigualdade para o caso brasileiro. Com essas conclusões, também foi possível confirmar as hipóteses de pesquisa.
Following a worldwide trend of the 20th century, which established the continuous development of education systems, Brazil experienced a significant expansion of the educational system in recent decades, fostered by different initiatives that sought greater inclusion of people in school and university. Considering the outline of this new scenario for Brazilian education, we seek to contribute to the analysis of the influence of social origin on educational outcomes, with the general objective of investigating whether the increasing schooling of the Brazilian population was capable of weakening the relationship between social origin and education achieved. To this end, in the light of parameters from international productions on the subject of educational stratification, which present results applicable to the Brazilian reality, relative measures of schooling are used. This means that education is measured by its positional character, in which the value assigned to the schooling achieved directly reflects the distribution of education credentials across society. The empirical analysis is based on the comparison of the results of statistical models that use absolute measures and relative measures of schooling, to measure the effects of social origin. The models are used using a quantitative approach, applying the least squares method (OLS — Ordinary Least Squares), ordinal logistic regression and multinomial model (multinomial logit).The database used is the Continuous National Household Sample Survey (PNAD — IBGE), for the years 1892, 1996 and 2014, considering individuals between 28 and 41 years old, comprising three cohorts. The dependent variable is education (relative and absolute) and the main independent variable is social origin. Control variables such as race, age, family structure, gender, region and location (rural/urban) are also included. The results of this research suggest that, in a scenario of progressive increase in the educational level of the Brazilian population in terms of years of study, there is a tendency to reduce the relationship between social origin and education achieved when education is measured absolutely. However, when education is measured in relative terms, the educational expansion that has occurred in recent decades in Brazil has not been able to reduce the effects of social origin on the level of education achieved. Considering the relative position in an imaginary queue where the most educated are ahead, it is not possible to perceive any significant reduction in the effects of class of origin on the level of education achieved. This reveals the approximation with the diagnosis of persistence of inequality for the Brazilian case. With these conclusions, it was also possible to confirm the research hypotheses. |
