Utilization without Valuation: Evidence from a Randomized Trial of Health Insurance in Cameroon

A common rationale for subsidised insurance expansions is that coverage teaches users the value of financial protection, generating durable demand once subsidies end. This logic may fail if experience reveals non-monetary access frictions, including administrative barriers, and stigma at the point of care. We test this in a randomized trial among 1147 women in commercial and transactional sex in Yaounde ---a poor and highly discriminated-against group with health needs and no prior insurance exposure. We randomly assigned 579 women to one year of free comprehensive insurance covering themselves and up to six nominated dependents---2,428 individuals---with 568 in a control arm; endline retention was 70.3% (N=806). We find that free insurance raised care-seeking by 11 percentage points but left willingness to pay unchanged (USD~0.06, SE~3.63); mean willingness to pay (USD~63.53) falls well short of actuarial cost (USD~109.46), implying a per-enrollee subsidy of USD~45.93. Heterogeneity analyses suggest that this null effect masks opposing responses: valuation increases among participants with greater baseline exposure to formal-care costs, but falls among those reporting higher interpersonal discrimination. Both margins are prevalent in the study population: recent formal-care use is limited for many participants, weakening the financial-protection signal from insurance experience, while discrimination in health care is widespread. The findings are consistent with a framework in which experience reveals two competing features of the formal-care encounter---the financial benefit of avoided out-of-pocket spending and the non-monetary frictions of using insured care. For marginalised populations, subsidies and friction-reduction interventions are likely to operate as complements rather than substitutes.