The debate on how socioeconomic status affects students’ performance and educational opportunities dates to the 1960s, a period when this issue was first brought to light by what we can consider the pioneering work, the Coleman Report (Coleman et al., 1966). Among the findings of the report, one of the most important is that a student’s level of education is correlated with both their family background and the background of other students in the school. Another line of literature has shown that, regardless of the family's socioeconomic status, a significant factor affecting students' academic performance is the average socioeconomic status of the school, highlighting that schools with a higher average socioeconomic statustend to have higher test scores (Rumberger & Palardy, 2005; OECD, 2004). Thus, in a world increasingly leaning towards equality, it is in the interest of institutions to monitor this trend between socioeconomic status and school performance to ensure equal opportunities for students regardless of their socioeconomic conditions, both family and school, aiming to improve education. Indeed, higher education can enable a better standard of living by providing broader access to various job opportunities. Individuals with a stronger cultural background can more easily access to qualified and better-paid jobs, enjoying concrete prospects for professional advancement. On a larger scale, this phenomenon contributes to reducing the income gap between different segments of the population (Abdullah et al., 2015). 

In this vein, assessing students’ achievements has become of paramount importance worldwide to highlight any disparities between schools, regions, and geographical areas. In Italy, this work is carried out by the INVALSI (National Institute for the Evaluation of the Educational System of Education and Training) through the annual administration of the so-called “INVALSI Tests,” which are periodic and systematic assessments of students’ skills and knowledge at the end of the second and fifth grades of elementary school, the third grade of middle school, and the second and fifth years of high school. The tests are prepared and provided by INVALSI and are intended by the Ministry of Education to assess students’ proficiency in Italian, Mathematics, and English. The purpose of the INVALSI tests is multiple: to verify the level of student learning, monitor school performance, and analyze if there are differences between different schools, regions, or geographical areas of the country. 

The present work aims to investigate the impact of the socioeconomic status of the schools attended by Italian students on the mathematics scores obtained in the INVALSI tests, with a particular focus on regional differences. Indeed, Italy is a country characterized by significant regional differences that manifest in many aspects, from the economic aspect to the quality of life up to education. These disparities are particularly evident when comparing the northern regions with those of the south. The study proposes a multilevel regression discontinuity based on the integration of Regression Discontinuity (RD) and Multilevel regression (Hox et al., 2017). On the one hand, the RD design represents a non-experimental strategy for analyzing causal effects, which allows us to estimate the causal effect of attending a school with a low socioeconomic status on the results of the INVALSI Mathematics tests. On the other hand, the multilevel specification provides new insights into Italian territorial heterogeneity by examining regional variations in the effect of school socioeconomic status. 

The remainder of the paper is organized as follows: Section 2 focuses on the methodology employed, specifically detailing how the multilevel approach was integrated with the regression discontinuity technique to achieve the research aims. The data used in the analysis are described in detail in Section 3, while Sections 4 and 5 present the empirical results and some conclusions, respectively.