Matching practices in secondary schools – France

This country profile is part of a collective effort by the network members to map matching practices across Europe. If you find it useful and want to refer to it in your own work, please refer to it as “Hiller, Victor and Olivier Tercieux (2013), Matching practices in secondary schools – France, MiP Country Profile 16.

Download full profile pdf


Relevant country background

The French education system is divided into public schools and private schools. Overall 85% of primary school students and 80% of secondary school students attend public school (this has been a rather stable proportion over the last decade).[1] Private schools are mostly made of schools that have a contract with the State, which specifies that they should respect the official curriculum (in return, teachers are paid by the State) – these are mainly catholic schools. A small proportion of private schools do not have such a contract (because they do not respect the curriculum) and rely on a strong financial participation of families.

School education is compulsory from age 6 to age 16. There are no tuition fees in public schools and they are low in publically-funded private schools. Primary school covers age 6 to 12. Secondary school covers age from 12 to 18 and is divided into junior high school (age 12 to 15) and high school. Preschool for children aged 2.5 and above is also offered and publically funded, but it is not compulsory.

After the strong increase in the number of students in secondary schools in the sixties, France chose a system of residence-based assignment of students to schools in 1963. The system geographically divides France into catchment areas. There is a single school corresponding to each catchment area and except in case of a waiver, a student has to go to the school associated with the catchment area in which s/he lives.

One of the main concerns of this system often pointed out by its detractors is that it failed to ensure social diversity in schools. There are several reasons for this. One of the most often cited reasons is that the housing market is already socially segregated and this then mechanically translates into socially segregated schools. In some cases, this social segregation has been reinforced by the mere existence of this residence-based assignment system (see Gravel, N., Michelangeli, A. Trannoy 2006 and Fack and Grenet 2010). Another reason is that parents can escape the system using some strategies like fake addresses or waivers based, for instance, on the choice to study a rare language that is only taught at a school outside their catchment area. While these strategies are a priori available to all parents, they were used significantly more by social groups having good information on the education system such as teachers (see Chausseron 2001).

These concerns partly motivated the introduction of school choice regulation in France. The residence-based assignment of students to schools was abandoned and replaced by a school choice procedure in 2007. While the goals of the education policy are decided at the national level, France is divided into 30 districts and each has some freedom in defining the exact ways to reach these goals (see map of the districts below; Paris forms a single district). To foster social diversity, it has been decided to give priority in all districts to students receiving means-tested scholarships (low socio-economic status students).

So far, this school choice procedure applies in all districts for students entering high school and in some districts for students entering junior high school. Private schools (around 20%) are not integrated in the procedure.[2] The rest of this profile describes the system applicable to public high schools (and some junior high schools).

Figure 1: School districts (académies) in Metropolitan France

france

Summary box

ORGANIZATION OF HIGHER EDUCATION Mix of public and private schools (bulk of private schools are confessional and publically funded)
STATED OBJECTIVES OF ADMISSIONS POLICY Reduce school segregation by promoting social diversity in schools, promote freedom of choice for families, transparency, simplicity and no need to strategize
WHO’S IN CHARGE OF ADMISSIONS? Districts (France is divided into 30 districts)
ADMISSIONS SYSTEM IN PLACE SINCE 2007-08 for high school level and junior high school level.
AVAILABLE CAPACITY Decided by district authorities. Depends on infrastructures and government funding.
TIMING OF ENROLMENT Applications take place in May. Assignments are sent to families in late June. The exact timing varies across districts.
INFORMATION AVAILABLE TO STUDENTS PRIOR TO ENROLMENT PERIOD Each district operates a website providing information about the schools within the district. There is no league table but some information on performance is available online. Many schools advise their students on “suitable” high schools.

RESTRICTIONS ON PREFERENCE EXPRESSION

Varies across districts. Students can only apply to schools within their districts..
MATCHING PROCEDURE The assignment uses a variant of the school-proposing deferred acceptance algorithm (with restrictions on the number of schools that can be ranked). There is a second round for students who were unassigned in the first round.
PRIORITIES AND QUOTAS There are no quotas for minorities. Priorities vary across districts. They are based on a score that depends on a combination of geographic, academic, social and strategic factors.
TIE-BREAKING Given the way priorities are defined, tie-breaking is usually not needed. In some districts, the date of birth is used as a tie-breaking rule.
Other special feature In many districts, applications benefit from extra points at schools that are ranked first.

Description of current practices

The current system has been in place since 2007 (i.e. starting for academic year 2007-08). It is mainly centralized and is run at the district level. Specifically:

1. In Mid-May, parents receive a single form for each of their children in age to register. Parents can rank up to k schools on the form where k ranges from 3 to 8 depending on the district. Some districts, such as Paris, also require a minimum number of schools to be ranked (at least three schools must be ranked in Paris). Others impose that at least one of the schools listed be a school close to home (most of the time in what used to be the catchment area). Children are otherwise eligible for admission at all schools within their district. Parents must submit the form to the school that their child currently attends and this school sends the form to the headquarters of the district.

2. First stage matching – centralized matching at the level of each district. Priorities for students at schools are determined by a system of points based on a number of criteria that are common across all schools of the district:

a)    Priorities in high schools outside of Paris and junior high schools: Students obtain a great number of points in the school of their former catchment area. This ensures that students have a top priority to this school and can always be assigned to it if they want to.[3] Some points are also awarded for medical reasons and to students benefitting from means-tested scholarships. Finally, some districts give points to a student’s application at the school that s/he ranked first.

b)    High schools in Paris: The district of Paris is divided into 4 sub-districts and students obtain 600 points at schools in their sub-district (it is important to stress here that, in contrast to catchment areas that only include one school, each sub-district has many schools). Students receiving a means-tested scholarship get 300 points at all schools where they apply. Up to an extra 600 points are given to students based on their past academic performance (grades are normalized to account for differences across schools). Finally, a student that has a sibling in a school gets 50 points in that school.

There is a centralized platform called Affelnet that is used in all districts for students entering high school and in some districts for students entering junior high school. The actual matching mechanism used in all these districts is equivalent to the school-proposing deferred acceptance mechanism except for the fact that (i) because of the existing extra points associated with first-ranked schools, priorities are in part determined by the rank of the school in the rank-order list, and (ii) the size of the rank order list is limited, the details are provided in Hiller and Tercieux (2013). Students are informed in late June of their assignment and have about five days to confirm this choice. The exact timing varies across districts.

3. Second stage matching: For students who did not get an assignment in the first stage of the matching procedure, they can reformulate their wishes in early July. However, their choices are restricted to schools in which available seats remain (they have lost their priorities on the seats that have been filled during the first stage). Then, the assignment procedure is run again for those students.

Performance

There have been a couple of official evaluations of the current enrolment policy (see for instance, Fack and Grenet (2012) and Oberti, Préteceille and Rivière (2012)) which are reports for public administrations. These studies broadly underline the lack of available data to properly assess the impact of the reform. Fack and Grenet (2012) cautiously suggest that the 2007 reform has not had much impact so far on the trends that were observed prior to the reforms. They argue that the reform has yielded a slight social desegregation of public high schools. On the other hand, they also observe a simultaneous increase in social inequalities between the public and the private sector (the private sector attracts relatively more students from favored social groups). Overall, these effects tend to compensate. With respect to admissions in high school, there is no significant change compared to what was observed before the reform. However, they observe that very unpopular schools are now avoided: relatively good students from very disadvantaged areas can now escape those schools. In Fack and Grenet (2012), a special focus is put on the Parisian case and they provide some evidence suggesting that social diversity has increased in public schools.

One of the reasons that could explain the weak impact of the new school choice program is the important weight put on the geographic criterion in most districts, associated with the limited capacities of schools. Indeed, the fact that students have a top priority in the school of their former catchment area, together with the use of a matching procedure that favors the school side, is likely to yield situations where all seats in a school are filled by students of the area in the first stage of the procedure.[4] Overall, there is limited room for students from outside the former catchment area and the final allocation does not differ much from the earlier residence-based assignment system. Even in the few cases where the priority based on location is weaker (like in Paris), demand for popular schools outside of former catchment areas has been relatively low. This is true even for students on scholarships who hence do have a high priority. Consequently, students from disadvantaged families have not so far benefitted much from the reform even though the reform had a clear objective of social desegregation that the “scholarship criterion” was supposed to fulfill.

In general, assessments of the matching procedure by districts are based on figures such as the proportion of students who are assigned to a school at the end of the first stage of the matching procedure and, among those students, the proportion who are matched to their very first choices. Those figures are quite different between vocational schools and non-vocational schools but do not vary much from one district to the other. As an illustration, in the Paris district, for non-vocational (resp. vocational) schools, about 85% (resp. 70%) of students get assigned to a school at the end of the first stage of the procedure while – among those — 75% (resp. 78%) get matched with the school that they ranked first.

Of course, these figures are computed using reported preferences, which need not coincide with true preferences (recall that under the mechanism used, there is no reason to report preferences truthfully). So those assessments should be taken with much care.

Recent policy change

The current system was preceded until 2007 by a strict residence-based assignment system called « carte scolaire ».

Perceived issues

See the Section “Performance”.

Existing data

The official website of each district provides some data on high school enrollment at the district level (here is the website for the district of Paris: http://www.ac-paris.fr). There is no publically available data on the enrollment procedure.

Legal texts

Legal texts describing the general principles of the French’s enrollment policy are available at: http://goo.gl/yVdAqm

References

Chausseron, C. [2001], “Le choix d’un établissement scolaire au début du secondaire”, Note d’information n°01-42, ministère de l’Education Nationale –ftp://trf.education.gouv.fr/pub/edutel/dpd/ni0142.pdf

 

Gravel, N., Michelangeli, A. and A.Trannoy [2006], « Measuring the social value of local public goods: an empirical analysis within Paris metropolitan area», Applied Economics, 38, p. 1945-1961

 

Fack, G. and J. Grenet [2010], “When do better schools raise housing price? Evidence from Paris public and private schools”, Journal of Public Economics, vol. 94, p.59-77.

 

Fack, G. and J. Grenet [2012], Rapport d’évaluation de l’assouplissement de la carte scolaire, document du CEPREMAP, http://goo.gl/1Kri9I

 

Hiller, V. and O. Tercieux [2013], Choix d’écoles en France : une évaluation de la procédure Affelnet, Revue Economique, forthcoming, www.pse.ens.fr/IMG/pdf/Affelnet.pdf

 

Oberti, M., E. Préteceille and C. Rivière [2012], Les effets de l’assouplissement de la carte scolaire dans la banlieue Parisienne, Rapport de la recherche réalisée pour la HALDE – Défenseur des Droits et la DEPP – Ministère de l’Education nationale – http://goo.gl/rzjPKb


[2] For these schools the admission procedure is decentralized. Parents apply to the school(s) of their choice. Each school’s principal is free to accept or not this application.

[3] There are no such strict catchment areas for vocational high schools. For this kind of schools, the geographic criterion plays a smaller role than in non-vocational schools.

[4] This is the case when, for each school, the number of seats of that school is equal to the number of students in its catchment area.