Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We use a randomized experiment in Nairobi to compare the effect of free health insurance to an unconditional cash transfer of the same value and a control group. Despite high baseline rates of injury and illness, the median insurance taker does not use the insurance. We observe no significant effects of either insurance or cash on economic outcomes, self-reported health, and healthcare utilization. We find some evidence that the provision of health insurance reduced levels of self-reported stress and the stress hormone cortisol relative to cash and control. This result suggests that insurance may have a “peace of mind” effect, although the most conservative bounds for attrition and multiple inference correction render it statistically insignificant. Together, our results suggest that health insurance may reduce stress in our setting, but its benefits are otherwise limited.