Predicting long-term unemployment in Portugal
Fellows: Jordan Kupersmith, Laura Moraes, Pranjal Bajaj and Ruqian Chen
Technical Mentor: Nuno Brás
Junior Technical Mentors: Iñigo Martínez
Project Manager: Lénia Mestrinho
- IEFP (Instituto de Emprego e Formação Profissional) | Key Partner
- EAPN (European Anti-Poverty Network)
Institute for Employment and Vocational Training (IEFP) is the Portuguese government employment agency whose mission is to foster quality job creation and combat unemployment. The agency does so by implementing active employment policies, including vocational training.
EAPN Portugal, established in 1991, is the national chapter of the largest European network of national, regional, and local networks involving anti-poverty non-governmental organisations (NGOs), grass-root groups, and European organisations active in the fight against poverty and social exclusion. The organisation’s historical and institutional knowledge will be invaluable during the implementation phase of the policy recommendations which will have emerged after the summer.
The purpose of the project is to:
- Better identify individuals at high risk of long-term unemployment (LTU);
- Support more efficient allocation of IEFP’s resources to respond to the needs of unemployed individuals.
The IEFP’s current model is based on risk categories (high-, medium-, or low-risk) assigned to each applicant when registering at IEFP. This project aims at improving the existing system by creating a model that dynamically updates an individual’s score over time, and better identifies high-risk individuals (as is appropriate to the needs of IEFP and partner institutions).
Providing a dynamic score for LTU would enable the IEFP to strategically prioritize which support to provide, and how to optimize the delivery of services to those who need it most. By identifying which individual is more in need of support, and which individual is more likely to be responsive, the IEFP will be able to more strategically offer their services. Also, the dynamic properties of the model will better estimate partnering organizations’ needs, and allow for better targeting of intensive campaigns that might decrease job seekers’ chances of falling under the “unemployment spell” (i.e. the longer the unemployment period, the lower the chance to find a job).
Data provided by project partners include characteristics of job seekers and interactions with IEFP.