algorithms at the heart of the relationship between job seekers and their advisors

Four hundred. This is, on average, the number of unemployed “close to employment” taken care of by an agent of the French public employment service, according to a report from the General Inspectorate of Finance (IGF) and the General Inspectorate of Social Affairs (IGAS) published in February 2023. When it is acts as applicants “distant from employment”the figure drops to 150, but the average portfolio of a France Travail advisor (formerly Pôle emploi) remains between 350 and 400.

In Denmark and Germany, where the unemployment rate is 4.5% and 3% respectively in 2022, the number of unemployed monitored falls to 100 and 115. This report puts forward the hypothesis that the higher the unemployment is in a country , the more the portfolio of public employment service advisors is.

In France, the digitization of these numerous files has revolutionized the profession of advisors. In his contribution to the scientific mediation project “What do we know about work? ” of Interdisciplinary laboratory for public policy evaluation (Liepp), broadcast in collaboration with Presses de Sciences Po on the site’s Employment channel Lemonde.frsociologist Didier Demazière looks back on this transition.

Contradictions

It highlights two types of appropriation of new tools.

First of all, faced with the growth in unemployment, digitalization has made it possible to improve the efficiency of agents, by allowing them to go more quickly to complete the files of citizens, but also by helping them to make decisions. During interviews, advisors have at their disposal a series of key steps suggested by their software: update the situation, take stock of the job search process, define the objectives to be achieved, prescribe support, training, etc.

Later than in other countries, profiling methods based on the needs of the unemployed appear. In terms of “scores calculated for each job seeker on their professional project, their job search, the labor market situation or even peripheral obstacles”the algorithm suggests a personalized diagnosis of the job seeker, to facilitate the interpretation of France Travail agents.

Also read the survey: Article reserved for our subscribers From ChatGPT to Midjourney, generative artificial intelligence is taking hold in businesses

But, for some advisors, the scale taken by these suggestions is seen as an encroachment on their autonomy and skills. The first contradictions appear: the questions proposed by the screen make the interview easier, but prevent a real dialogue with the job seeker… Which would nevertheless make it possible to obtain better quality information, so that its algorithmic profiling is more interesting.

You have 30% of this article left to read. The rest is reserved for subscribers.

source site-30