Complementary insurance and free-choice modality expenditure in Chile’s public insurer Fonasa: Moral hazard or selection bias?

Introduction: The Isapre system crisis, enrollee migration to the public insurer, and the expansion of complementary insurance have reshaped spending incentives in Fonasa's free-choice modality. Whether the higher spending of insured beneficiaries reflects moral hazard or selection bias is unclear. Objectives: To estimate, from the public payer's perspective, the additional free-choice-modality spending attributable to holding complementary insurance, and to decompose it into selection bias and moral hazard, between 2022 and 2025. Methods: Analytical observational study on individual-level administrative microdata (January 2022 to May 2025). Insurance was identified through its effective use. A counterfactual was built using nearest-neighbour (k = 1, without replacement) on observable covariates, and the classification error was corrected with a non-linear household-size decay model. The outcome was the percentage difference in FAM free-choice-modality spending per beneficiary, in inflation-indexed units. Results: The crude per-capita spending gap was 57.6% higher among the insured; it fell to 44.2% after matching and converged to an average treatment effect on the treated of 30.2% (95% CI: 28.2 to 32.2%) after correcting the classification error, with annual estimates ranging from 28.8% to 31.3%. Conclusions: The higher spending associated with complementary insurance reflects both risk selection and moral hazard; at least 13 percentage points are attributable to differences in the composition of the population purchasing complementary insurance. Complementary insurance is associated with an increase of approximately 30% in spending under the Free Choice Modality, after controlling for observable and classification biases.