Adverse health shocks, social insurance and household consumption: evidence from Indonesia's Askeskin program.

This study examines the efficacy of Askeskin, a subsidized social health insurance targeted towards poor households and informal sector workers in Indonesia, in mitigating the impact of adverse health shocks on household consumption. To overcome selection bias from non-experimental nature of Askeskin enrolment, I use a robust estimation strategy, where outcome regressions are run on a propensity score-based matching sample. Using longitudinal data from the Indonesia Family Life Survey, this study finds that uninsured households facing extreme health health shocks experience a 1.3% point loss in growth in food and 2% point loss in non-food consumption growth. Importantly, households having Askeskin insurance, are fully insured in terms of food and medical consumption. But non-food spending, a discretionary component, is not insured fully resulting in a 1.2% point fall in consumption growth rate, despite Askeskin. This result is robust to a battery of sensitivity and robustness checks, including alternate definition of health shocks. Further, I investigate whether the Askeskin program simply displaced informal, community-based mechanisms of risk sharing. No crowd out effect is observed and informal risk-sharing coexists with Askeskin.