Speaker
Silvia Scarpa, Università di Modena e Reggio Emilia
Abstract:
Economic inequality is a heterogeneous and continuously expanding phenomenon nowadays. The provision of an exhaustive statistical methodology for the measurement of economic inequalities is a necessary premise to help policymakers reducing them. This work proposes a complete framework for the estimation of the quantile ratio index (QRI), an indicator based solely on quantiles and that considers the entire distribution of the variable of interest. A direct estimator is defined to this end, which uses only sample observations in the context of finite population inference. The problem of its measurement in small areas - where survey data may be insufficient to ensure the direct estimator reliability - is addressed by introducing a small area estimation model at the area-level. The proposed methodology is applied on Italy’s income and wealth data to measure the inequality in small areas defined on a socio geographical basis. The performance of the QRI – in terms of bias, mean squared error, and sensitivity to outliers – is compared through simulations based on EU-SILC data with that of other popular quantile-based inequality measures, such as the quintile share ratio, P90/P10, the Palma index, and the Gini coefficient, used as a benchmark due to its popularity.