The vertical and intertemporal structure of educational inequalities in Italy

This paper aims to analyse the joint impact that individual and ascriptive variables (gender, family background, migratory background) and contextual variables (socio-economic and cultural context, school composition and tracking, territorial differentiation) exert on the academic performance of Italian students. The aim is to investigate the mechanisms operating at both macro and micro levels and their interplay (Coleman, 1990), which shape individual choices and how the combination of these produces collective effects (micro-macro). We investigate these goals using OECD PISA data for the years 2018 and 2022. While many studies analyse the impact of factors such as students’ family background and tracking on performance (Panichella and Triventi, 2019; Bernardi and Triventi, 2020; Giancola and Salmieri, 2022), showing the multilevel and territorial structure of these inequalities from a diachronic perspective is a relatively new topic in the Italian context. Educational inequalities at the upper secondary level arise from various factors, one of which is the territorial and school level. In our research, the territorial level, specifically geographical macro-areas, represents the highest level of aggregation for inequality factors. Individual schools represent the intermediate level. Several studies have showed the role played by the social composition of individual schools in different territories (Benadusi et al, 2010; Argentin et al, 2017). The social background of students in Italian schools plays a determining role in the level of segregation observed (Giancola and Salmieri, 2020). In Italy, schools and students' families are intertwined with diverse territorial contexts. These contexts are characterized by different socio-economic and cultural combinations, which are shaped by a range of factors, including institutions, politics, and the local environment. If this relates to the "vertical" dimension of the stratification of inequalities, then we must also consider the temporal dimension, which includes the effects of the Covid-19 pandemic such as school closures, social isolation, and the subsequent phenomena of learning loss. The study investigates these objectives using the data collected by OECD in PISA 2018 and 2022 (OECD-PISA, 2019; 2023) and refers to performance differences in Reading among Italian upper secondary students. In the first part, the study analyses the configuration of educational inequalities within and among the five macro-territorial areas considered in the PISA survey. In this section, a multilevel regression sets (Bottoni, G., 2022) will be used to construct an interpretative model of the vertical structuring (macro-area, school, individual) of inequalities. Following this phase, we analyse the data from 2018 to 2022 to investigate whether and how school performance has changed over time in the five territorial areas. We use a pseudo-counterfactual approach and apply the Difference-In-Difference technique (Morgan, S., L. and Winship, C., 2015).