Do Institutions Reduce Gender Discrimination? Evidence from Labor Market Participation Rate in Some Selected African Countries

Abstract

The study investigates the role of social-political institutions on gender discrimination and female employment in five selected African countries, Ghana, Cameroon, Botswana, Kenya and Egypt. The methodology used is the quantile regression analysis, which is based on the premise that estimating the conditional sample median will tend to the distributional median allowing us to obtain consistent estimates. Quantile regressions have some obvious advantages over the ordinary least squares estimation technique they include the fact that they are more robust against outliers in the response measurements, it also allows for the measurement of central tendency and statistical dispersion to obtain a comprehensive analysis and finally the recent quantile regression wrapper (qreg) developed by Machando and Silva (2013) allows for obtaining heteroscedastic errors robust estimates. The results of the study show that institutions matter in improving labor market participation rate for men and women in the countries in the sample. The results of the interactive variable also show that institutions actually do not improve the effect of economic policy effectiveness in promoting labor market participation. School enrollment had a higher significant effect on labor market participation rate for women than for men showing that unskilled men are likely to get hired than unskilled women therefore schooling was probably reducing unemployment more among women than in men.

Keywords: Gender discrimination, Institutions, Labor Market Participation, Quantile regression