A1285
Title: Survey-to-survey imputation: From bracketed income to inequality measurement
Authors: Lorenzo Mori - University of Bologna (Italy) [presenting]
Gianni Betti - University of Siena (Italy)
Laura Neri - Universita di Siena (Italy)
Abstract: Many surveys record income in brackets, which leads to downward-biased inequality measures. A survey-to-survey imputation technique based on generalized additive models for location, scale and shape (SSIT-GAMLSS) is developed to reconstruct a continuous income variable from classed data and estimate inequality indices. A jackknife repeated replications technique is proposed to estimate the variability. The proposed method is evaluated through simulations. As an empirical application, inequality is estimated at the provincial level in Tuscany for 2021, 2022, and 2024, revealing spatial heterogeneity and temporal variation and providing timely estimates that would otherwise be unavailable.
