Determination of wage inequalities in the urban labor market of El Salvador: A segmented and decomposition analysis

The two essays that constitute this dissertation address some issues on the wage inequalities in the urban labor market of El Salvador. The first essay studies the determinants of wage differentials in the urban labor market. The second essay studies gender wage differentials and its changes over time in El Salvador.The first essay uses the labor market segmentation (LMS) approach to examine wage differentials. According to the LMS theory, wage differentials emerge and are perpetuated over time by confining some workers to a labor market that offers low wages, bad working conditions, little or no opportunity for advancement and low technology. This essay determines whether there are distinct sectors in the labor market and whether the sample of workers in each segment is a random sample of the population. A multinomial choice model is used to determine sector allocation. Then, a wage equation for each sector corrected for sector selection bias is estimated. This study as well as the second essay uses micro-level data from the National Multipurpose Household Surveys of 1989 and 1992 conducted by the Ministry of Planning in El Salvador. This study shows that the urban labor market of El Salvador is divided into three separate segments--the private, public, and the informal sectors--distinguished by different wage setting mechanisms. The results indicate that Salvadoran urban workers are not randomly allocated to the three different sectors.The second essay studies changes in the male-female wage differential over time due to changes in productivity characteristics and inequality differentials. A single index model is used to measure the effect of changes in observable characteristics, relative prices, and residual wage inequality on changes in the gender wage gap. It introduces a sample selection model into the gender wage decomposition to account for male-female offer wage differentials. The results indicate that overall the male-female wage gap converged over the period of study. Without correcting for selection bias, the gender wage differential over time is overestimated. Changes in the predicted gap or observables explained one third of the gender wage convergence. Convergence in observables had a stronger effect on the male-female wage differential over time than the increase in residual wage inequality. The male-female wage ratio would be higher than it was had women not increase their schooling and experience and had rates of return not fallen.