PhD Economics/Investigating the effects of job search support programmes

The French government will be launching a new policy (PIC - Plan d'investissement sur les Compétences") which will dedicate substantial resources to training and counselling programmes targeting unemployed job seekers. This offers the possibility to investigate a range of open research questions that are relevant to policy makers. In particular:
- Which training programmes are most effective in helping job seekers exit unemployment and find rewarding jobs? Is the most relevant dimension of heterogeneity on the quality of training programmes, or is it more a question of assignment of the right training to the right job seekers?
- How should private training providers be incentivised to propose high-quality training? Because there is little ex-post evaluation of the outcomes of training, there is little feedback on which training should be certified (and bought) by the Public Employment Service.
- How should caseworkers be guided to match job seekers to the right training? Public Employment Service caseworkers play an important role in the allocation of the training programmes to unemployed job seekers. The criteria they currently use to decide on the allocation are not very transparent. In particular, they do not seem to use quantitative data of past training spells to inform their decision about the allocation.
- What are the general-equilibrium effects of training programmes? Training, like other active labour-market policies might push beneficiaries at the detriment of non-beneficiaries. The extent to which labour demand reacts to better trained job seekers is key to assessing aggregate effects of this policy on unemployment. Yet, there are still even fewer pieces of evidence about general equilibrium effects of training than about micro-effects. The question is key to policy-makers: if training helps individual workers, is it an effective way to fight aggregate unemployment?
As the PIC brings new resources to the actors involved in training, there is an opportunity for associate researchers to bring new evidence by launching a set of experimental and non-experimental evaluations. The exact design of the evaluation will depend on discussions with stakeholders (Public Employment Service, training providers, training funders...) but will aim at identifying causal effects, rather than just documenting correlations. Causal identification will be reached either by the design of relevant randomised controlled trials or by quasi-experiments leveraging institutional patterns (discontinuities, natural experiments...).
Most of the data used in this project will come from administrative databases relating to training, to counselling and the labour-market situations of job seekers and workers. In some instances, it might be necessary to collect primary data, to shed light on some particular aspect that cannot be looked at using administrative data. Data will be accessed from DARES and, from the UK, using the Centre d'Accès Sécurisé à Distance.

We anticipate two kinds of outcomes to this research.
- Academic outcome: Answering some of the research questions stated above will contribute to the academic literature about training unemployed workers. The outputs will take the form of working papers and academic publications in top economics outlets.
- Impact: The design of the evaluation will be decided in close association with policy makers and stakeholders. These actors will remain engaged through a Steering Committee, organised by DARES, which will follow the progress of evaluations and take stock of preliminary results. This continuous engagement will improve the chances for the results of the research to be considered in future revisions of training policies. In order to help disseminate the results, DARES will organise an impact-enhancing workshop bringing academics and policy-makers together.