This thesis employs advanced econometric methods to understand the determinants of race inequalities in labour markets and in higher education in Brazil. It then investigates whether race diversity can be used as a policy to reduce existing inequalities in pay and college campuses. It uses data from the National Household Sample Survey PNAD of 2005 and the National Examination of Higher Education Courses (ENC/Provão) of 2003. The main methodological contributions of this thesis are: 1) extending wage models to include several variables which can explain more than 40 percent of the total variation in wages; 2) computing a proxy for parental education (this has not been possible to estimate using PNAD since 1996); 3) correcting wage equations for selection bias using a robust instrument (most studies ignore the sample selection problem by using employed males only); 4) implementing a new algorithm that combines Heckman Two-Step, complex sample weights and constrained least squares (this increases robustness of the detailed decomposition of the discrimination term). This is done in a generalized wage decomposition setting where the level of discrimination is invariant to the choice of the reference wage group. Results show an existing pay-gap and a significant level of discrimination against nonwhites even after corrections are made. Selection bias appears to underestimate the discrimination term considerably. This study also develops a theoretical framework for the study of the impact of diversity on labour productivity and on discrimination simultaneously. Results support policies which seek greater diversity in order to reduce the existing inequalities in labour markets and on higher education campuses. However, the outcome of policies aimed at increasing diversity on campuses can significantly differ depending on the existing level of diversity and the subject majors attended by students.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:542651 |
| Date | January 2011 |
| Creators | Garcia Ozemela, Liana M. |
| Publisher | University of Aberdeen |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://digitool.abdn.ac.uk:80/webclient/DeliveryManager?pid=167792 |
Page generated in 0.9439 seconds