Measuring poverty using both income and wealth: An empirical comparison of multidimensional approaches using data for the US and Spain

This paper presents a comparative analysis of the approaches to poverty based on income and wealth that have been proposed in the literature. Two types of approaches are considered: those that look at income and wealth separately when defining the poverty frontier, and those in which these two dimensions are integrated into a single index of welfare. We illustrate the implications of these approaches on the structure of poverty using data for two industrialized countries-for example, the United States and Spain. We find that the incidence of poverty in these two countries varies significantly depending on the poverty definition adopted. Despite this variation, our results suggest that the poverty problem is robust to changes in the way poverty is measured. Regarding the identification of the poor, there is a high level of misclassification between the poverty indices: for most of the pairwise comparisons, the proportion of households that are misclassified is above 50 percent. Interestingly, the rate of misclassification in the United States is significantly lower than in Spain. We argue that the higher correlation between income and wealth in the United States contributes to explaining the greater overlap between poverty indices in this country.