1. Residential and school segregation are persistent features of large cities worldwide (Gutiérrez et al., 2020; van Ham et al., 2021). These trends raise serious concerns because segregation sort young people from different social and ethnic backgrounds into unequal neighborhoods and schools that affect their opportunities and outcomes (Galster & Sharkey, 2017). Consistent with recent calls for a more integrated examination of neighborhood and school segregation processes (Oberti, 2007; Rich & Owens, 2023), this article examines the segregation of children at the secondary school level between neighborhoods and between schools within large French metropolitan areas.

   Prior studies have shown that neighborhood and school segregation are highly correlated, not only in residence-based school allocation systems (Frankenberg, 2013; Oberti & Savina, 2019), but also in context that allows more freedom of school choice (Boterman, 2019). However, most existing studies employ a cross-sectional perspective to examine segregation at a single point in time, such that there is less knowledge concerning the evolution of the relationship between residential and school segregation. Moreover, previous studies have compared residential segregation of adults and school segregation of students, which are two groups with different residential segregation patterns (Owens, 2017). As such, there is still a prevailing lack of work that compare segregation of children across neighborhoods and schools. Another critical issue in the literature relates to the factors that shape residential and school segregation. Scholars have been generally concerned with understanding how contextual features at the metropolitan level are associated with either school segregation (Bonal et al., 2019) or neighborhood segregation (Jähnen & Helbig, 2023). Conversely, little research has examined the mechanisms that affects the relationship between the segregation of neighborhood and schools and thereby explain the correlation or mismatch in their social composition (Bischoff & Tach, 2020; Candipan, 2019; Rich et al., 2021).

   Against this background, we adopt a longitudinal approach to examine how changes in neighborhood and school segregation of children are related to each other and how the characteristics of the local educational markets shape the association between residential and school segregation. We create a novel longitudinal dataset combining data on children, neighborhoods, and schools spanning from 2007 to 2019. Our analysis starts with segregation indices showing that residential segregation is higher than school segregation, especially among socially advantaged children. Decomposition analysis further shows that school segregation between public and private schools has increased, which is correlated with declining residential segregation over the period. Next, we use panel regression models with city and time-period fixed effects to examine whether spatial inequality in public and private school options and social selection in private schools predict gaps between residential and school segregation among children. In the discussion, we connect our findings to the debate about the consequences of the expansion of school market mechanisms to organize education systems for residential and school segregation

School segregation has emerged as a core challenge in Barcelona (Spain) in the past few years. Driven by wide margins of school choice, a large share of private subsidized schools, and sustained inaction from local government bodies, the uneven distribution of students across schools became a persistent issue (Bonal et al, 2021). In response, the city’s educational authorities implemented the Shock Plan Against School Segregation (SP) in 2018. This policy aims to promote a more balanced distribution of socially disadvantaged students through targeted funding, preassigned school seats, and quotas for vulnerable students (CEB, 2019).

While SP has been evaluated in terms of its design and implementation (Bonal & González, 2023), its impact on school segregation, concentration, and social stratification between public and private subsidised schools has not yet been analysed. Our research has two objectives. First, to estimate to what extent the changes in the distribution of socially disadvantaged and migrant students among schools in the city can be attributed to the impact of the SP. Second, to examine how the impact of the SP varies among disadvantaged groups and the characteristics of the different areas of the city (i.e., catchment areas and districts).

To do this, we design a quasi-experimental study employing Interrupted Time Series regression analysis (McDowall et al., 2019) to estimate the impact of these policies on school segregation, concentration, and social stratification between public and private schools. This allows us to estimate the causality between the implementation of these policies and the evolution of school segregation in Barcelona for the past decade. While school segregation and concentration have significantly decreased during the period analysed (2011-2023), the findings show that overall, the causality between this decline and the implementation of SP is relatively limited. We observe significant variations among different groups of students—such as those with immigrant backgrounds or those considered vulnerable—and across different areas of the city. Despite these limitations, the impact of SP in reducing the stratification of socially disadvantaged and migrant students between public and private subsidised schools is particularly significant. Regarding the geographical factors mediating the impact of the SP, the results show that the level of residential segregation, the concentration of socially disadvantaged students within the area and the school supply diversity considerably affect the capacity of the SP to promote a more balanced distribution of students among schools.

The paper concludes by exploring the various factors and mechanisms behind these differential impacts, shedding light on the complexities of policy effects in diverse urban contexts.

School segregation is widely studied as a process underneath the reproduction of inequalities, also in the Italian context. Less attention was paid to teachers’ segregation. The aim of this paper is to investigate a hidden mechanism of inequalities’ reproduction in the Italian education system, namely the allocation of students to teachers. In particular, we focus on teachers with temporary contracts, a large ratio of the teaching workforce in the Italian. The widespread precariousness of teachers implies their turnover during the study cycle and educational discontinuity for students. The situation has its origin in the paradoxical coexistence of: i. the precarious status of a large proportion of teachers, employed on annual contracts or on shorter substitute contracts directly employed by schools; ii. the difficulty in filling certain types of teaching positions, especially for STEM subjects; iii. the seat transfers requested by tenured teachers (Abbiati et al., 2021). Several empirical studies have assessed how teachers' preferences are strongly influenced by non-pecuniary factors such as students' characteristics and from this it follows that teachers ask to move from the most problematic schools to easier contexts (Barbieri et al., 2011) and how a higher propensity to leave has a negative impact on students' learning (Barbieri et al., 2017). A possible mechanism of perpetuation of inequality in the functioning of the Italian school system is thus configured, with the best teachers concentrated in the most favourable contexts and the others in the most disadvantaged ones (Abbiati et al., 2017).

The phenomenon here analysed is the assignment of unstable teachers to students. Starting from the present literature, and using the available data, we test the following specific research hypotheses:

H1: unstable teachers (temporary contractually or in the school) at the working level are more likely to be assigned to socially disadvantaged students; students with lower values of the ESCS indicator of socio-economic status, with migrant background and in specific Italian territorial areas are therefore taken into account.

H2: we expect this allocation to operate both at school and class level, within schools; this issue was not adequately investigated in previous analysis, despite the relevance of breaking down the two processes.

The data used in this work come from the combination of information regarding both students and teachers in the INVALSI databases. Our dataset results from the harmonisation of the school years 2017/2018, 2018/2019, 2020/2021 and 2021/2022. The school grades considered: 2, 5 (primary school) 8 (secondary school) and 10 (high school).

Overall, our analysis confirm the presence of a not-neutral teacher-student pairing, with an increasing intensity of the unequal matching as the school grade increases. In addition, we detect that the largest part of the process takes part at class level. Overall, the precariousness of teachers interplays with students’ school segregation and seems to be part of the mechanisms underneath the reproduction of educational inequalities in the school system.