a carregar...


Grupo de Estudos Monetários e Financeiros

Estudos do GEMF, N.º 02 de 2001


Human Capital Investment Through Education and Economic Growth:

A Panel Data Analysis Based on a Group of Latin American Countries

Maria Adelaide Silva Duarte
GEMF/Faculdade de Economia, Universidade de Coimbra

Marta Cristina Nunes Simões
GEMF/Faculdade de Economia, Universidade de Coimbra

Our main goal is to ascertain whether investment in human capital through education (without differentiation, by gender, with gender gaps) can explain the steady state growth productivity levels and a potential convergence process in a sample of Latin American developing countries (Costa Rica, Dominican Republic, El Salvador, Guatemala, Haiti, Honduras, Jamaica, Mexico, Nicaragua, Panama, Trinidad&Tobago, Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Guyana, Paraguay, Peru, Uruguay and Venezuela). These countries will also be integrated in a larger sample with developed countries from Europe (Austria, Belgium, Denmark, Finland, France, the former Federal Republic of Germany, Greece, Ireland, Italy, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland and the United Kingdom).
We consider a neoclassical growth model with human capital as did Mankiw, Romer and Weil (1992) and later Knowles, Lorgelly and Owen (1998) who include both female and male education and gender gaps in education levels as explanatory variables. The consideration of human capital by gender and the quantification of its influence appear of major importance in the developing countries if the education of women in these countries produces positive social benefits. In these circumstances the educational policies by gender should be supported because, ceteris paribus, an increase in the levels of female schooling will produce positive total effects in the form of higher productivity levels.
The empirical analysis is based on panel data and the following estimation procedures are used: ordinary least squares (OLS), non linear least squares (NLLS), ordinary least squares with dummy variables (LSDV), ordinary least squares with first differences (OLSD) and non linear least squares with dummy variables (NLLSDV). We first present the results for the productivity equations and then for the convergence equations, considering our two samples and the different estimation procedures. The use of different samples and estimation techniques is intended to make our analysis more robust.

Download PDF

(272 KB)