Jonathan Rossi (Erasmus University Rotterdam)
Pierre Koning, Paul Muller and Carlos Riumallo Herl
Existing research documents wage scarring after job loss, but little is known about the long-term effects of sickness leave. We study how sickness leave duration affects subsequent labor market outcomes for temporary workers in the Netherlands. Using administrative population-wide data on temporary workers who called in sick between 2010 and 2015, we first estimate descriptive event studies matching sick workers to observationally equivalent controls. We find that even short sickness spells are persistently associated with lower employment, wages, hours and permanent contract probability. Because the duration of sick leave is mechanically correlated with health, these estimates cannot be interpreted as causal. To isolate causal effects, we exploit a 2013 reform introducing a health assessment after 12 months of sickness leave, which substantially shortened sickness spells. Implementing an augmented regression discontinuity in time design, we find that shorter spells increase employment by about five percentage points for up to nine years, accompanied by reduced reliance on sickness and disability benefits. Conditional on employment, there is no effect on wages, hours, contract type, or sectoral mobility.