Educational Returns of Boarding Schools for Underserved Secondary School Children in France

Clement de Chaisemartin
Axelle Charpentier
Sourdun, France
395 academically qualified students from disadvantaged backgrounds
2009 - 2012
Target group:
Outcome of interest:
Enrollment and attendance Empowerment Student learning
Intervention type:
School choice
AEA RCT registration number:

Many high-income countries such as the United States and France have robust public school systems, yet academic success is often out of reach for students from disadvantaged backgrounds. These students may lack access to good public schools. In addition to the quality of the schools themselves, children’s home environment may also be a key factor impacting academic success. If disadvantaged children face negative influences from peers and poor study conditions at home, their academic potential could be impaired. 
Researchers conducted a randomized evaluation to test the impact of boarding schools on educational outcomes and well-being for children with disadvantaged backgrounds in France. The top third of students admitted to France’s first “boarding school of excellence” had large improvements in math scores after two years, but there were no impacts on French skills for any of the students. Students initially experienced reductions in well-being, but adjusted to their new environment by the second year.

Policy issue

Charter schools are an increasingly popular way to improve public education for underserved students in developed countries. These schools often target students from disadvantaged backgrounds, who may not have access to quality public schools in their neighborhoods. By providing more resources, higher quality teachers, and a strong academic environment, these schools are thought to improve children’s academic performance and increase their aspirations and goals, a claim supported by rigorous evaluations from the United States.

In addition to the quality of schools themselves, children’s home environment is thought to be a key factor that influences academic success. Another potential way to improve learning among deserving but unprivileged children is to enroll them in high-quality boarding schools. These schools provide a more structured, supervised, and academically-focused environment than these children would have at home. However, such boarding schools are likely to be considerably more expensive than normal public schools, and there is little evidence of their impact. As industrialized countries consider their options for delivering quality education to children who lack both quality schools in their neighborhoods and parental support at home, it is important to understand the impact of boarding schools on students’ performance and aspirations.

Context of the evaluation

Beginning in 2009, the French government began opening “Boarding Schools of Excellence,” and allowing students from disadvantaged backgrounds to apply for the limited number of slots. The Sourdun boarding school was the first of these, but was followed later by 44 other boarding schools that enrolled more than 4,000 students in 2012. Students who applied to Sourdun (aged between 11 and 16) were average students when compared to national metrics, but were generally performing better than their peers at the schools they were attending. Thirty-seven percent of the students enrolled at Sourdun came from single-parent families, and only 29 percent had legal guardians who had obtained a bachelor’s degree or higher.

These boarding schools were relatively expensive, costing an estimated €10,000 (US $13,100) per child more than a standard public school education. The French government decided to undertake a rigorous randomized evaluation to determine their impact.

Details of the intervention

In spring of 2009 and 2010, interested students were asked to fill out an application for acceptance at one of the boarding schools. In order to be eligible to attend, students had to be both academically qualified and come from a disadvantaged background. From the pool of eligible students who applied, 258 (around 65 percent) were randomly selected to attend Sourdun, while the remaining 137 served as a comparison group. Students in the treatment and comparison groups were systematically followed over two academic years in no less than 169 schools. They were given standardized cognitive tests and filled out questionnaires that included psychometric scales among others. Parents were also interviewed.

Students at the Sourdun boarding school faced very different schooling conditions than their counterparts who attended public schools. Teachers at Sourdun tended to be better educated than their public school counterparts, although they were younger and had fewer years of experience. Pupil-teacher ratios were also lower, with 8.2 teachers for every 100 students in public schools, compared to 11.2 at Sourdun. Perhaps because of this, these teachers were able to develop a more specific and tailored teaching approach—students reported that teachers were more interested in their progress, and would continue to explain lessons until everyone understood them. Additionally, because Sourdun was a boarding school and students lived and studied on the school grounds, attendees dedicated significantly more time to academic activities. On a typical school day, students spent 0.8 more hours doing homework, and watched 0.9 fewer hours of television daily. This was not reversed during weekends.

Results and policy lessons

Students who were randomly selected to attend the Sourdun Boarding School of Excellence saw significant increases in their aspirations and goals, as well as attendance and academic performance. Only 35.8 percent of Sourdun student reported having missed class in the last 15 days, compared to 51.6 percent in the comparison group. While in class, treatment group students also reported reducing the amount of time they spent not paying attention in class and had better relationships with their teachers than students who were attending public schools. Teachers were 27 percentage points more likely to express interest in the well-being of their students and 26 percentage points more likely to provide supplementary help, if necessary.

Students who attended Sourdun also had higher aspirations, and expressed greater interest in pursuing their higher education. They were nearly three times as likely to express interest in taking college-preparatory classes, and 25 percent more likely to say they wanted to attain a master’s degree. These changes were not immediate—students appeared to lose confidence in their abilities during their first year in residence at Sourdun, perhaps because they were exposed to a higher-performing peer group or were asked to do more difficult work. However, their confidence rebounded in the second year.

Mirroring the trend in students’ confidence in their work, after the first year at Sourdun there was no significant difference in French or math scores between the treatment and comparison groups, though math scores of students at Sourdun did improve after the second year. After two years, students’ average math score increased by 0.4 standard deviations, a substantial amount relative to other education interventions. There was still no effect on French scores, but this pattern of improvements in math but not language scores is common among education programs. These results were the same across girls and boys, middle- and high-school ages, and stronger and weaker students. Effects of attending Sourdun did not appear to spill over onto siblings or former classmates who were still attending public schools.

Behaghel, Luc, Clément de Chaisemartin, and Marc Gurgand. 2017. “Ready for Boarding? The Effects of a Boarding School for Disadvantaged Students.” American Economic Journal: Applied Economics 9 (1): 140–64.