The demand for insurance under limited trust: Evidence from a field experiment in Kenya∗

In spite of strong theoretical reasons to believe in the welfare-enhancing value of microinsurance products, demand for such products to date has been disappointingly low across a range of developing countries. In this paper we investigate the role of trust in the demand for indemnity insurance. First, we develop a theoretical model of insurance demand under limited trust to derive predictions for the way trust, risk aversion, and insurance premiums interact. Second, we test these predictions using field and laboratory-experimental data from a randomized controlled trial introducing a composite health insurance product to tea farmers in Kenya. Consistent with the theory, we find that not only low trust but also risk aversion is negatively associated with insurance demand, and that individuals with low trust are more responsive to experimental variation in premium costs. Third, we combine take-up decisions with subjective probability distributions for health costs to structurally estimate the model. Structural estimates reveal that choices are consistent with pessimistic and heterogeneous beliefs about the probability of insurance payouts for indemnified events. These estimates allow us to calculate welfare losses relative to counterfactual insurance products that are (perceived as) fully credible: expected losses from foregone insurance due to low trust exceed 31 percent of premium costs. Our results suggest that limited trust is an important barrier to the adoption of insurance, particularly among poor and risk-averse households who stand to benefit the most from such financial products. ∗We thank Michal Matul (ILO), Ann Kamau and David Ronoh (CIC Kenya), Therese Sandmark (SCC), and Edward Kinyungu and Morris Njagi (Wananchi SACCO) for their support in the design and implementation of the project. Job Harms, Naureen Karachiwalla, and Felix Schmieding provided excellent research assistance. We are grateful to George Akerlof, Abigail Barr, Lori Beaman, Martin Browning, Gharad Bryan, Daniel Clarke, Mark Rosenzweig, and John Rust for helpful comments. Financial support from the ILO Microinsurance Innovation Facility is gratefully acknowledged. †University of Oxford ‡VU University Amsterdam §Georgetown University