Introduced in 2002 as an alternative to juvenile detention when other forms of open placement have failed, Closed educational centres (CEF) aim to create a temporary break from minors’ usual environment by involving them in intensive educational and professional activities and providing enhanced, personalized supervision. This makes them the most resource-intensive non-custodial placement of the French juvenile justice system. Building on administrative records from the Directorate for Youth Judicial Protection between 2012 and 2022, I evaluate their effect on post-placement trajectories by instrumenting actual placement with the local availability of a CEF in the prescriber’s department within pairs of minors observationally identical. I find that CEF placement raises the twelve-month probability of post-spell detention by 19 percentage points, and of a subsequent CEF placement by 22 percentage points. The effect seems to be driven by minors whose alternative would have been a community supervision order, is null on substitution from another residential placement, and is mildly protective on substitution from detention. It is also concentrated on stays cut short before the four-month threshold separating the structured and reinsertion phases of the programme. The closed centre seems to be perform closer to its political design when it displaces incarceration, and produces the worst observable trajectories when it intensifies an ongoing community-supervision. [Forthcoming: Analysis on peer effects. Depending on the possibility of merging administrative data: analysis on return to education and training, employment and health outcomes.]

This paper investigates the impact of interventions aimed at overcoming barriers that prevent refugees from accessing mental healthcare, which is widely recognised as a key determinant of their integration in host countries. Building on the French Ministry of Interior’s AGIR programme, which supports refugees in securing employment and housing, we randomly assign participants to four groups, each varying in how information and support regarding mental healthcare access is delivered. We investigate whether discussions with peer-helpers reduce stigma and informational barriers compared to standard translated materials, whether a multilingual helpline that assists with appointment booking helps overcome institutional barriers, and whether access to mental healthcare improves employment and housing outcomes within the AGIR programme. We find that facilitating access to care through a helpline increases engagement with mental health care between 3.7 and 7.3 percentage points. We find no detectable effect of the peer-helper intervention alone.

This paper studies the effects of the OEPRE programme (Opening Schools to Parents to Promote their Children’s Achievement), a large French national initiative jointly implemented by the Ministry of Education and the Ministry of the Interior. The programme targets allophone immigrant parents with children enrolled in the French school system, offering free in-school training that includes French language courses, sessions on the functioning and expectations of the education system, and sessions on the principles and values of the French Republic. Building on a survey targeting first-generation immigrant parents from 150 French primary schools, we study the effects of this programme on three domains: (i) parental involvement in their children’s schooling; (ii) socio-economic integration of parents, including civic engagement, institutional trust, and sense of belonging; and (iii) children’s learning and academic outcomes [forthcoming]. Our survey covers approximately 720 allophone parents across treatment and control schools. We find that access to the OEPRE programme significantly increases parental participation in school activities. Effects on French language acquisition and broader integration outcomes are more limited in intention-to-treat specifications, though estimates on actual programme participants point to larger effects consistent with partial compliance.

Developed by a French startup specialising in digital educational applications for children with specific needs, the AlloSchool app aims to facilitate the integration of newly arrived allophone students by providing personalised educational content, allowing them to progress at their own pace, while taking their mother tongue into account. The experiment planned to randomly allocate access to the application as well as training sessions on the application to 200 specialist teachers of French as a second language, representing around 2,000 pupils. By matching results from surveys conducted among immigrant parents and teachers with children’s cognitive test results and administrative data, we aim to assess the impact of this intervention on (i) the development, learning, and social integration of children, (ii) parental involvement in education, as well as (iii) teaching practices in a multicultural context.

This paper examines whether academic self-beliefs help explain social gaps in educational and career expectations among 15-year-olds across OECD countries. We document large SES-based gaps in expectations in almost all countries, even among equally performing students in the same schools and tracks. Lower-SES students also report systematically lower academic self-beliefs than their high-SES counterparts, reflecting lower perceived chances of success. While these beliefs are associated with expectations, they account for only a small part of the SES gap in expectations. Our findings suggest that equalising academic self-beliefs is unlikely to eliminate disparities in expectations, pointing instead to deeper structural or cultural drivers of educational inequality.