Research projects

The main contribution of this project is to develop new models to measure inequality from a multidimensional perspective by jointly considering economic dimensions and non-monetary outcomes, such as health and education. Due to data limitations, especially in non-economic variables, we rely on parametric copulas to estimate the multidimensional distribution of well-being using grouped data.

The project involves four core objectives:

(1) Proposal, estimation and testing of parametric models to represent the distributions of each well-being dimension. We have developed different estimation strategies for a number of parametric functional forms for income, education and length of life considering the available information in each case. In particular, the most suitable models seem to be the generalized beta distribution of the second kind for income, the generalized gamma distribution for education and a mixture of the Fisk and the Weibull distributions for the variable length of life.

(2) Estimation of copulas from grouped data on the dependence structure. We have analysed multivariate copulas which are able to provide a reliable representation of the dependence
structure between income, education and health using national data on the dependence structure. We found that the gaussian copula is a convenient model thanks to the property of additive decomposability of the covariance. We have also explored the Fretchet limits, which do not require defining a parametric structure to study the dependence between the
three dimensions.
(3) Objectives (3) and (4) are still in progress, as they are not supposed to be completed until 2024.