Teacher effectiveness is a concept elaborated in Economics of Education literature that describes the contribution of individual teachers to the cognitive growth of their students, as measured by standardized competence tests (Goldhaber et al. 2015). Sociological studies of educational inequalities have often overlooked the importance of this factor and its impact on school-related processes. This concept, in fact, poses challenges for sociologists for several reasons, including its alignment with an efficiency-driven educational model and the risk of reinforcing negative teaching practices (such as teaching to test), as outlined by many commentators (e.g. Grimaldi and Barzanò 2014). However, the established consensus on the critical influence of teachers in shaping student success and job trajectories (and associated disparities) calls for Sociologists to address this factor and ultimately integrate it into their empirical as well into their theoretical works (Thrupp 2001; Argentin 2018; Abbiati 2021).

In this paper we provide teacher effectiveness estimates for Italy, a country characterized by marked educational inequalities despite its high degree of centralization, particularly in the lower and upper secondary school cycles (Falzetti 2019).

To this purpose we analyze data on a nationally representative samples of Italian 8th and 10th grade students for the school years 2017-18 and 2018-19, which include competence scores in language and mathematics and socio-demographic information. We merged this data with a survey administered to their language and mathematics teachers. The choice of this school years is motivated by the inclusion of a questionnaire item that allows the identification of teachers who have followed sample students from the beginning of their educational cycle (the target of the analyses).

In line with the literature, effectiveness estimates will be retrieved by regressing students’ scores on a teacher dummy variable, students’ competences at the end of the previous school cycle (i.e. 5^th grade for models on eight graders and 8^th grade for model on 10^th graders) and a wide set of individual-level regressors (Chetty et al. 2014).

These estimates will be analyzed to assess their distribution among schools, families, and territories, and to study their predictive power in shaping students' careers in later educational stages. This study will contribute to a deeper understanding of teacher effectiveness and its implications for addressing educational disparities.

In the last few decades, there has been a notable rise in women's enrolment in higher education. This trend contributed to a decline in gender inequalities in both participation and degree attainment, eventually leading to their reversal in virtually all the OECD countries (Fenget al., 2023; Delaruelle et al., 2018; Evans et al., 2020). According to the OECD (2021) report, in 2019, 51% of women aged 25-34 attained a tertiary degree compared to only 39% of their male counterparts. On the other hand, the persistent under-representation of women in specific STEM (Science, Technology, Engineering, and Mathematics) programs continues to be a significant factor contributing to the enduring wage disparity between men and women (Barone & Assirelli, 2020; Herbaut & Barone, 2021). This paper aims to study the evolution over time (since the nineties until today) of gender differences in the participation in higher education and the choice of the field of study in Italy. We employ logistic regression models with a specific focus on examining the role of students' school performances as mediators in explaining the evolution of gender disparities in both higher education participation and STEM programs choice.

This investigation entails complex data management due to the use of different data sources needed to cover such an extensive time span in Italy. In particular, we use data from the survey on the “Educational and Professional paths of upper of secondary school graduates” conducted by the Italian National Institute of Statistics (Istat) from 1998 to 2015. Through this survey, we can encompass the cohorts of graduates from 1995 to 2011. To cover the most recent cohorts of upper secondary graduates, we will use microdata from administrative databases made available through an agreement with the Italian Ministry of University and Research (MIUR). The combined analysis of these different data sources will provide a rich overview of the evolution of gender differences and inequalities in tertiary education participation and their underlying mechanisms.

Note: The data used in this study have been processed in accordance with the RESEARCH PROTOCOL FOR THE STUDY "From high school to the jb placement: analysis of university careers and university mobility from Southern to Northern Italy" among the Ministry of University and Research, the Ministry of Education and Merit, the University of Palermo as the lead institution, and the INVALSI Institute. The reference researcher is Gianni Betti. In addition, we acknowledge financial support under the National Recovery and Resilience Plan (NRRP), Mission 4, Component 2, Investiment 1.1, Call for tender No. 104 published on 2.2.2022 by the Italian Ministry of University and Research (MUR), funded by the European Union – NextGenerationEU– Project Title Stem in Higher Education & Women INequalitieS [SHE WINS], CUP I53D23004810006, Grant Assignment Decree No. 1060 adopted on 07/17/2023 by the Italian Ministry of Ministry of University and Research (MUR).

The previous literature in education largely acknowledges the role of family socioeconomic conditions in the intergenerational transmission of inequalities. In this framework, the achievement of a high level of education for less-advantaged students is a strategic asset in reducing disparities. Indeed, higher educated people have a greater awareness of their abilities and opportunities to build a better future, which ensures better economic rewards in the long term: higher educational levels are generally associated with higher employment rates and salaries. However, data from Italy show that achieving a tertiary education degree in Italy is still strongly associated with family background. The probability of getting a degree for individuals in the cohorts 1960-69, 1970-79, and 1980-89 coming from highly educated families is about .5 higher than the one of those coming from low-educated families (Busetta et al., 2023). Moving from this framework, this research aims to evaluate the factors affecting schools’ capability to foster fairness and inclusion in education by offering the same opportunities to reach the highest level of education to students coming from disadvantaged backgrounds (Field et al., 2007; OECD, 2012a; Agasisti & Longobardi, 2014; OECD, 2016a; Sulis et al., 2020). Based on the PISA OECD definition of resilient students, which identifies as resilient those students who achieve a high level of proficiency in the PISA test (score in the third quartile) despite facing adverse socioeconomic circumstances (ESCS index in the first quartile), we define the resilient students at the university by focusing on the regularity of students’ career at the end of the 1st year. In particular, we define resilient students as those from families with low ESCS index who achieved a number of formative credits higher than the median observed in their disciplinary field. For this aim, the database MOBYSU.IT, which contains the Italian National Student Archive (ANS) microdata related to the cohorts of students enrolled for the first time in an Italian university, has been merged with the data from the INVALSI surveys, which provide information about student families’ socioeconomic conditions. A multilevel approach that considers as Level-1 units the students and as Level-2 the school-field of study combinations has been adopted to identify the school and peers’ effect on the probability of fostering resilient behaviors at the university, taking into account the differences in the choice of the field of study and the source of heterogeneity in the data related to individual, school and territorial area’s characteristics. The combined use of multilevel modeling approaches to deal with a complex data structure (i.e., nested observations in schools, fields of study, and universities ) and latent profile regression analysis allowed us to detect the role played by schools’ characteristics in determining the capability of the institutions to boost students’ resilience and to combine quality and equity in terms of expected results of their students at the university.

A large body of researches in both economics and sociology has focused on peer effects in education highlighting as social interactions at primary and secondary levels school affect student outcomes (Coleman et al., 1966; Sacerdote, 2011; Patacchini et al., 2017), aspirations and expectations among adolescents (Raabe and Wölfer, 2019; Lorenz et al., 2020), and secondary school-related choices (Zwier et al., 2023). At university context, peer effects resulting from interactions between students are mainly investigated to explore their effect on academic performance (Winston and Zimmerman, 2004; Griffith and Rask, 2014; Vitale et al., 2016). To the best of our knowledge, studies on university choices have paid little attention to the peer environment (Porcu et al., 2022; Usala et al., 2023 and references therein).

Within this study, we examine how individuals’ educational choices in tertiary education dependent on socio-demographic background, the family context, the school environment, and social interactions among peers. Starting from these dimensions, it is important to understand the factors playing a critical role within a perspective where the quality of educational systems can occur only within a scenario of non-reproduction of social inequalities. How and which these contextual effects can interact in the individual choice’s process from high school to university is investigated by data gathered through ad hoc surveys. Data integration procedures are adopted to reconstruct a unique dataset of around 4000 students who have participated to three surveys devoted to a high schools’ sample in Campania region. Hence, logistic regression models are applied to assess the effect of peers and other contextual variables on the probability to enroll at university in a specific disciplinary field as well as on the probability to move in another region to attend a degree program.