by Ohto Kanninen
Policy makers need to understand the nature of risks individuals face to make informed decisions on public policy, especially social insurance. In this paper, we study empirically how sickness plays a role in the association between well-being and income to draw conclusions on optimal policy. We show that those who are on sick leave exhibit a significantly different association between well-being and income. In particular, in the sick population, the levels of measured well-being are lower at all incomes and higher incomes are associated with higher increases in well-being due to an observed fixed cost of sickness. These observations together imply that risk is domain-specific, i.e., each policy tool needs to be adjust to the specificities of risk in the domain that the policy tool targets. Such domains are sickness, unemployment or disability. We apply our findings to study the optimal policy in sickness insurance and find that optimal replacement rates should be non-linear and decreasing with income.