This paper is based on a randomized experiment conducted in order to understand the low takeup rate of a complementary health-insurance voucher program for the poorest in France. We explore two of the main hypotheses put forward to explain low enrollment: difficulties in accessing information...