Can sickness absence be affected by information meetings? : evidence from a social experiment

During the last decade several empirical studies have stressed the importance of norms and social interactions for explaining sickness absence behavior. In this context public discussions about the intentions of the insurance, and of the rights and duties of the receivers, may be important for reducing the sickness absence. In this paper we study whether information meetings about the Swedish sickness insurance affect the length of sickness absence spells. The study is based on experimental data on individuals with weak labor market attachments. The displacement of when the call to the meeting was sent out was randomized. Comparing the survival functions of those called immediately with those whose calls were delayed (by about 30 days) makes it possible to study whether the length of sickness absence is affected by receiving the call earlier. The result suggests that the length is reduced by, on average, 20 percent. In the long term (12 months later) there is no effect of the information meeting. This suggests that attendance to the information meeting does not change individuals' long-term behavior. -- monitoring ; moral hazard ; public social insurance ; survival analysis ; instrumental variables