Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
This paper presents the results of 20 randomized experiments aimed at understanding the low take-up of in-patient health insurance observed in developing countries. Take-up does not increase when participants receive information about the product, or an assistance to register, or small subsidies of 2, 10, or 30%. Take-up does not increase when the same information is provided by local respected community leaders, when participants are offered an in-kind gift (a chicken) if they register, when participants are offered the possibility to contribute lower and more frequent payments, or the possibility to pay by cellphone. A full subsidy generates a mere 45% take-up (with no retention after one year). In contrast to these low take-up rates, presenting the same information without any subsidies to existing informal groups raises take-up to 12% (still 7% after one year), as well as trust and knowledge of the product. Social networks play a major role in the adoption of health insurance. This paper provides a cost-effective way to increase take-up of health insurance, while subsidies are found to be largely ineffective at raising take-up in the long run.