Income inequality in Central America, Dominican Republic and Mexico: assessing the importance of individual and household characteristics

This study investigates the relationship between individual and household characteristics and income inequality in Central America, the Dominican Republic and Mexico from 1990 to 2002. A Theil decomposition exercise of individual and household income inequality is used to determine factors important for the level of inequality. In addition, the use of a novel semi-parametric simulation methodology from DiNardo, Fortin and Lemieux (1996) provides counterfactual income distributions of individuals and households to assess the importance of changes in their demographic, education and labour market characteristics over time. The results are very heterogeneous reflecting the differences across countries over the period. Individual characteristics can explain more of individual income inequality than household characteristics can explain for household income inequality. Regional differences, education and labour market characteristics have the greatest effects upon the level and changes of income inequality in Central America, the Dominican Republic and Mexico. Declines in the agricultural sectors of the countries and shifts to urban areas reflect the structural changes taking place in the economies, and these factors are important determinants of inequality change. In addition, the rise of the informal sector of employment with its lower benefits and job security has contributed increases in income inequality. However a significant proportion of inequality levels and trends across households and individuals could not be explained by individual or household characteristics. The explanatory power of the results for individual income inequality corresponds with other work conducted for individual income earners within and outside the countries included in this study. In contrast the inability of household characteristics to explain much, if any, household per capita income inequality is not consistent with findings from similar studies. Further work is needed to conduct an in depth investigation into the links between household income inequality and individual income inequality.