Events and Meetings of Italian Statistical Society, Statistics and Demography: the Legacy of Corrado Gini

Font Size: 
Bayesian Beta regression for the estimation of the Gini concentration coefficient in small domains
carlo Trivisano

Last modified: 2015-09-05

Abstract


The Gini coefficient is a popular concentration measure often used in the analysis of economic inequality. Estimates of this index for small regions may beuseful to properly represent inequalities within local communities. The literature on small area estimation for the Gini coefficient is limited. In this paper, we propose a method based on area-level models, thereby not assuming the availability of Census data at the micro-level. We propose a modified design-based estimator for the coefficient with reduced small sample bias and we consider a hierarchical Beta mixed regression model to combine survey data and auxiliary information. The methodology is illustrated by means of an example based on Italian data from the EU-SILCsurvey.

Full Text: PDF