evidoria

← Back to browse

Good practice Imported

Parcoursup's Local Algorithms — France's Constitutional Council Orders Post-Hoc Publication of University Selection Criteria

France · Paris · See the France profile · See the Paris profile

Top 62% 53/100 · Ask Evidence Copilot about this practice

After a 2020 QPC by the UNEF student union, France's Constitutional Council ruled that universities must publish, once admissions end, the criteria used to rank Parcoursup applications, including whether algorithms were used. A 2023 Senate question says many still refuse.

Parcoursup's Local Algorithms — France's Constitutional Council Orders Post-Hoc Publication of University Selection Criteria

Details

Promoter
Conseil constitutionnel; Ministère de l'Enseignement supérieur et de la Recherche
Period
2018–present (ruling 3 April 2020)
Keywords
higher education admissions, algorithmic transparency, administrative law

Description

Parcoursup is France's national platform for admissions to higher education. Under the 2018 ORE law, each programme's examination committee ranks applications, and institutions may use local algorithmic tools to help. Candidates could only ask, after a decision, how their own application was assessed.
The student union UNEF challenged this limit. On 3 April 2020, in decision 2020-834 QPC (referred by the Conseil d'État on 15 January 2020), the Constitutional Council upheld the candidate-only access rule but attached a reservation. Once the national pre-enrolment procedure ends, denying third parties any access to the criteria actually applied would disproportionately limit the right of access to administrative documents (Article 15 of the 1789 Declaration). Institutions must therefore publish a report on the criteria used, stating whether and how algorithms were used, while respecting candidates' privacy. The Council accepted that secrecy of pedagogical deliberations is a general-interest aim, noted that algorithms are optional and that no admission may rest solely on one, and required human review.
The same period saw the Cour des comptes (27 February 2020, report of about 200 pages with 15 recommendations) criticise a lack of transparency and increasingly automated, sometimes questionable ranking criteria.
Evidence of effect is mixed. The ruling created a clear legal standard that other public algorithms can borrow. But a written Senate question of 15 June 2023 (n°07314, lapsed without an answer) states that universities still refuse to disclose local algorithms and that the CNIL and the Défenseur des droits recommend publication. No independent measurement of compliance rates was found, so impact is rated cautiously.

Read the full analysis: https://www.conseil-constitutionnel.fr/decision/2020/2020834QPC.htm

Implementation

Implementation detail (cost, timeline, staffing, conditions for success) is not yet available for this practice.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful