Comment déterminer le salaire de référence des chômeurs indemnisés ?

The wage to be replaced is a key parameter to providing optimal unemployment insurance. In France, the so called "reference-wage" rules used until 2019 created large inequalities, between unemployed with splintered employment backgrounds and others. Indeed the insurer paid benefits implying monthly replacement well above 100%, to more than a fifth of eligible unemployed, namely those with low work intensity during the qualifying period. This setting subsidized employers using short-term contracts, contributing to the steady increase of these labor contracts, and of related insurance spending. To fully understand this complex problem, we break it down. It's clear that what is commonly viewed as a reference wage problem, is in fact mainly a by-product of eligibility rules. Looking for a better control the reference wage, without having properly controlled eligibility, as well as how an unemployed built his employment history, necessarily leads to a sub-optimal formula, and misleading incentives. There is no magic formula to compute the reference wage. First of all, eligibility rules split non-eligible unemployed that have a zero replacement rate and those eligible to unemployment benefits. Then, for the latter, the regularity with which rights were acquired should be taken into account, as well as the involuntary nature of unemployment spells between contracts into fragmented employment histories. Being linked to behavior, or contributions, such criteria are much better than meaningless parameters only aimed at biasing the reference wage computation. Finally, the reference wage formula is somewhere half way from two imperfect ends: daily wage for only the days worked, which prevailed during decades, or the average wage over the qualifying period which was to replace it in 2019.