The Gender Wage Gap in Developing Countries

Abstract

The labor literature documents a gender wage gap between the earnings of men and women in the developed countries but research on the gap in developing countries is less common and usually focuses on individual countries. Research typically identifies a part of the wage gap that is due to observable characteristics, or endowments, and an unexplained part attributable to differences in gender wage equation coefficients sometimes interpreted as gender bias. The focus of recent gender wage gap research has been on understanding how the gender gap varies along the earnings distribution. In this paper, we investigate the determinants of the gender wage gap across 12 developing countries using data from the World Bank’s STEP (Skills Toward Employment and Productivity) Measurement survey that provides information on household demographic characteristics, education and training, employment, and job skills. We apply the classical Blinder-Oaxaca decomposition and combined with quantile regression and also use decomposition methods based on the recentered influence function to obtain distributional decomposition of the wage gap by covariates and by quantiles. Our results show that there are significant gender pay differences that generally decrease with earnings and are mostly due to differences in the intercepts and a slightly offsetting difference in returns to education that are higher for women. The explained part of the gender wage gap, although much smaller, is attributed almost entirely to lower educational level of women and the motherhood penalty. The main conclusion is that improvements in women’s education and training, as well as child friendly labor policies, remain the most important factors that can help to narrow the gender gap in developing countries, which is consistent with previous work.