Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/27409 
Year of Publication: 
2005
Series/Report no.: 
IAI Discussion Papers No. 118
Publisher: 
Georg-August-Universität Göttingen, Ibero-America Institute for Economic Research (IAI), Göttingen
Abstract: 
Following the topics discussed by Campante et al (2004), this paper contributes to the literature of the Brazilian racial discrimination by isolating the effect of intergeneration transmission of schooling and the school's quality in the race discrimination effect. Instead of modelling just one mincer-type equation like others papers, it was decided to use the Two Stage Least Square Model where the first step of modelling control the endogeneity of individual schooling instrumenting it by family background and ability tests while attending school. The paper also provide a comparative profile of urban racial discrimination in the Northeast and the Southeast recognizing the important differences across regions in Brazil both in terms of economic development and racial composition of the population. As found by Campante et al (2004), results reveal that part of the component of wage differentials ordinarily attributed to labor market discrimination is actually explained by persistent educational inequalities between races. However because they didn't control the potential bias due to the endogeneity of some variables, their discrimination effect is 15 to 19 percentage points higher than it should be. The mechanism of intergeneration transmission is correlated with financial constraints and higher education of parents because blacks have lower elasticities of education with respect to parent's education due to selection and causation. Even controlling the model using instruments, Private sector remains as the sector where race discrimination is really an issue. Moreover, the regional profile suggests that the labor market is a more important locus of the racial issue in the Southeast than in the Northeast, although the significant presence in both regions. However, we are not controlling for selection bias and consequently the results must be viewed with caution because it is not sure how precise the estimations are.
JEL: 
J15
J24
J31
J71
J78
I21
Document Type: 
Working Paper

Files in This Item:
File
Size
306.93 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.