This project examines the role of algorithmic tools in the job placement activity of France Travail’s counselors, through experimental and quasi-experimental analyses. It is based on administrative data built from internal France Travail databases, securely matched with external sources such as FORCE, DSN, DADS, D@ccord, and the JOCAS database (webscraped by Dares). The project aims to evaluate the impact of algorithmic recommendations, with or without counselor involvement, on sustainable return to employment. It also compares the effectiveness of algorithmic targeting versus human decisions in orienting jobseekers. The project is carried out in close collaboration with France Travail’s DSEE and forms part of the Dares’ call for projects on the “Prefigurative Evaluation of the France Travail Reform.”

---

---