Abstract

Authors: Annamaria Bianchi, Antonella D'Agostino, Chiara Gigliarano, Laura Neri

Italy has made some progress in recovering from the economic crises induced by COVID-19, the Ukraine war, and inflation, but it is still grappling with significant economic and social challenges. The multiple crises have led to a significant reduction in wellbeing, drawing in individuals who were previously not poor and trapping those who might have escaped poverty if not for the crises. Local factors interacting with these shock events can create varied impacts across different sub-regions. It is essential to identify which territories and population groups have experienced the most significant changes in inequality and determine if certain domains exhibit greater resilience. Therefore, it is necessary to investigate these impacts using data collected at a sub-regional level, employing a multidimensional perspective. Thanks to the availability of a newly gathered database from the PRIN 2022 PNRR project, "MYPEOPLE: Measuring Inequality, Poverty, and Living Conditions for Planning Local Strategies," whose data have been collected through probability-based sample surveys in three different Italian regions—Lombardy, Tuscany, and Campania—this work aims to illuminate the varying degrees of wellbeing among different demographics and geographical areas of the aforementioned regions. Traditionally, assessments of wellbeing and inequality have been confined to national or broad regional perspectives, often based on disparate data sources that lack a comprehensive micro-level view. This paper seeks to address this gap by providing a more detailed analysis of wellbeing and horizontal inequality. This step is crucial for proposing intervention strategies that account for the territorial heterogeneity and the specific needs of different groups of individuals and families. For these reasons, it becomes very important to concentrate the analysis on specific domains, territorial levels, or peculiar groups of the population, commonly labeled as “small areas.” To produce accurate small area-level estimates, even with very limited sample sizes, we adopt small area estimation (SAE) methods. The idea behind SAE is to leverage auxiliary variables to obtain indirect estimators that may exhibit a lower mean squared error than direct estimators.