The short-term effect of a "one-day waiting period" on sickness absence, perceived health, and health care use

Mélina Hillion (Insee-Dese – Département des études économiques – Division « Marché et entreprises », PSE), Édouard Maugendre (DREES, DEPP)

Documents de travail
No 2023-06
Paru le :Paru le08/03/2023
Mélina Hillion (Insee-Dese – Département des études économiques – Division « Marché et entreprises », PSE), Édouard Maugendre (DREES, DEPP)
Documents de travail No 2023-06- March 2023

Since January 2018, as was the case from January 2012 to December 2013, the first day of ordinary sick leave is no longer compensated for French civil servants. Using comprehensive administrative data from the Ministry of Education for the period 2006-2019, this study evaluates the impact of this "waiting day" on the absenteeism of public education employees, who represent about 16% of civil servants (and 37% of state civil servants). We show that, on average, the waiting day is associated with a 23% reduction in sick leave episodes and a 6% reduction in their cumulative annual duration. The effect decreases with the duration of the absence episode, with the shortest sick leave episodes being most affected (-44% for episodes lasting exactly one day, -27% for episodes lasting 2-3 days). We show that women, the least qualified employees and those working in priority teaching areas are more likely to be absent, including when a waiting day is applied, and are therefore the most financially penalised by this measure.

Second, we examine the impact of this reform on the perceived health and use of health care by civil servants. We compare all civil servants with private sector employees (double difference strategy) using the Labour Force Survey (Insee, period 2013-2019) and the Working Conditions Survey (Dares, 2013 and 2016) matched with health insurance data (Cnam, period 2012-2017). We show that the waiting day had no significant effect on the perception of general health status, medical visits (general practitioners and specialists), drug dispensing and hospitalisations. The available data do not allow us to examine the productivity of the working days "generated" by the waiting day, nor to draw conclusions on individual productivity and the performance of administrations.