Conference Agenda

This study examines peer effects in educational outcomes by focusing on how high- and low-skilled peers impact students' performance in different educational systems. Peer influence in schools has long been acknowledged in sociological research, yet most studies analyze the average ability among peers rather than the effects of exposure to extremes—highly skilled or low-skilled peers. Using administrative data from Italy and Norway, we explore how these variations in peer ability influence students' mathematics and language skills in secondary education. Italy's early tracking system contrasts with Norway's late tracking and stable peer environment, providing an institutional comparison. Our findings reveal a contrasting effect model, where students exposed to highly skilled peers experience a decrease in self-confidence and academic performance, aligning with the "Big Fish in Little Pond Effect" (BFLPE). Conversely, exposure to low-skilled peers has a modestly positive effect. These patterns are observed in both countries and no sizeable gender divide is present. Differences between the two countries are minor, emphasizing that peer mechanisms have a consistent effect across diverse educational contexts. This research contributes to peer-effect literature by offering a nuanced look at peer composition and gender.

Although women have surpassed men in educational attainment, they are still underrepresented in STEM fields. Understanding the factors behind these gendered patterns is crucial for tackling labor market inequalities. Recent research has started to inquire the influence of family dynamics, particularly the role of sibling, on educational pathways. This study contributes to this discourse by examining the impact of sibling gender composition on the likelihood of enrolling in science or technical schools in Italy. Findings reveal that, in mixed-gender sibling groups, educational paths often align with traditional gender roles: women are less likely to pursue technical fields, while men have more opportunities to do so. The study also examines the role of an older sibling in educational choices of his/her younger counterpart. It is found that having an older sibling of the same gender enrolled in a technical or scientific school increases the probability of following a similar path by 20 percentage points for boys and 28 for girls, a pattern consistent across educational tracks. A relatively larger spillover effect for girls suggests that girls are especially in need of extra support and encouragement, as well as female role models, to opt for STEM studies.

Creating an inclusive school environment, in which students with disabilities are integrated as much as possible with students without disabilities, has become a commitment on the policy agenda of many countries around the world. Despite this trend toward inclusive education, however, research on the effects of being exposed to peers with disabilities in school remains limited. In this paper, we provide one of the first evidences of the noncognitive effects of exposure to peers with disabilities by examining child disability spillovers on a range of internalizing behavioral outcomes (depression symptoms, positive outlook, self-esteem) and social inclusion outcomes (peer conflict, social acceptance) among students enrolled in middle and high school (grades 8-12) in the United States. Using data from the National Longitudinal Study of Adolescent Health (Add Health), we infer causality by exploiting within-school and between-grade variation in school and grade composition. Our results show that exposure to a higher share of students with physical or learning disabilities is reflected in students without disabilities' increased internalizing behavioral problems and lower social inclusion. From a policy perspective, our findings suggest the need for policymakers to improve inclusion policies so that all children can benefit from a healthier and more inclusive school environment.

Despite higher educational attainment, LGB individuals often face earnings penalties. Significant disadvantages are also found in partnership status, which has been shown to moderate LGB people’s earning outcomes. This study combines literature on the LGB educational premium and earnings penalties to better understand the relationship between sexual identity, education, partnership, and earnings. Using data from the UK Household Longitudinal Study (UKHLS) and matching techniques with stratified samples, the study validates previous findings on partnerships and earnings differences while also contributing to the literature by investigating how partnership moderates the relationship between education and earnings. Results show that the moderating effect of partnership varies by educational level. While LGB men face overall significant earnings disadvantages compared to heterosexual men and are less likely to form partnerships regardless of educational level, gay men face disadvantages in earnings only if they are low-educated and single. Similarly, having a partner is particularly beneficial for lesbian women with low education, as they earn more than their low-educated heterosexual counterparts. These findings reveal the complex interplay between sexual orientation, partnerships and socio-economic outcomes, and highlight the role of both partnership status and education in shaping these dynamics.

In the educational inequalities field, the focus often falls on the dynamics of social reproduction (Bourdieu, Passeron, 1977) and primary effects (Boudon, 1973; Jackson, 2018). Typically, researchers work on the impact of economic and cultural capitals on educational outcomes (Broer et al. 2019; OECD, 2023). Here attention is given to all cases that deviate from social reproduction, who perform better on tests than predicted by their status, i.e. Deviant Cases (Bonanni, Moreschini, 2024). We use the PISA 2022 database for the analysis and these cases are operationalized as the residuals of the regression between socioeconomic and cultural status and literacy performance, using the PISA 2022 database.

The aim of this study is to examine the effect of certain school-level variables in order to understand how the structure of the attended school and its specific characteristics influence the ability to deviate from expected performance. This will be done for seven countries with different educational systems (Italy, Spain, Germany, France, Finland, Sweden, United Kingdom) through a binomial logistic multilevel analysis. What is expected to be found are differences within countries in the impact of schools conditioned also by two sets of variables: an individual-level set and a school-level set.

The aim of the study is to assess the presence of differences in the school-to-university transition related to migration background among students who graduated from Italian high schools. By linking data from two different Italian administrative sources, we compare native and non-native students based on their university enrolment choice, conditioning on their high school performance in mathematics. Moreover, we account for potential differences related to the macroregional location and type of high school curriculum attended. Preliminary results show significant differences between the two groups of students in terms of probability of university enrolment based on previous educational attainment.

The negative impact of parental separation on children’s education is well-established. However, the underlying mechanisms remain unclear. This study examines how specific parental behaviours may mediate this relationship, focusing on parental expectations, parental home, and school involvement. Using the National Longitudinal Study of Adolescent to Adult Health (Add Health), we employ causal mediation analysis to investigate whether the impact of parental separation on college completion is mediated by parental expectations towards their children's education, parental school, and home involvement. Results show that parental expectations and home involvement mediate the relationship between family structure and children’s education, which is not the case for parental school involvement. These findings underscore the potential of policy interventions targeting single-parent households to mitigate the adverse impact of parental separation on children’s educational outcomes.

Over the past two decades, socio-economic research has highlighted and empirically demonstrated the importance of investing in early childhood development for long-term life outcomes. In response, governments and foundations have increasingly allocated resources to preschool initiatives, promoting formal childcare, encouraging access to libraries and playrooms, and facilitating parent-child activities aimed at enhancing parenting skills. While the short-term effects of many of these policies are frequently assessed, evaluating their long-term impact remains challenging due to high costs and logistical complexities. How can meaningful insights into long-term effects be derived when only short-term outcomes are available? Adopting the “surrogacy” framework proposed by Athey et al. (2024) and using data from the Millennium Cohort Study (UK), we predict long-term outcomes assuming only short-term outcomes are available, and compare them with the actual ones. This analysis introduces a valuable tool for policymakers and program evaluators, helping to identify key outcome variables for impact assessment and the prediction of long-term effects.