Measuring cause-specific tolerance for health inequalities using the social marginal rate of substitution.

OBJECTIVES: We examine if tolerance for a fixed health inequality varies by whether it is caused by differences in: household income, freely-chosen lifestyles, or genetics. Distributional Cost Effectiveness Analysis (DCEA) uses equity weights to reflect societal preferences over improving total health and reducing health inequality, and cause-specific tolerance can ground DCEA in empirical preferences. METHODS: Using Benefit Trade-Off (BTO) tasks in an unassisted online survey, respondents recruited from the Spanish public (n=497) chose between hypothetical options that traded total health for reduced health inequality between two equally sized population halves. For a given inequality in healthy life years, we use the social marginal rate of substitution (SMRS) as a direct measure of tolerance. We elicited indifference points (IPs), compared paired IP distributions non-parametrically, and estimated implied SMRS with individual fixed-effects regressions. RESULTS: Most respondents traded total health for equality with a sizeable non-trading minority. Implied mean SMRS by cause were 3.60 (Income), 3.25 (Genetics), and 1.75 (Lifestyles). Income-Lifestyles and Genetics-Lifestyles differences were significant (p<0.001); Income-Genetics difference was not (p=0.512). Results were robust to scenario order and to re-coding IPs for non-traders. CONCLUSIONS: Tolerance for a given health inequality depends on its perceived cause: tolerance was higher when inequality was caused by lifestyles than income or genetics, between which no significant difference was found. Thus, "natural" (genetics-caused) inequalities are not viewed as more legitimate than income-caused ones. Cause-specific SMRS can help inform the calibration of DCEA equity weights to capture not only the extent of inequality but also its cause.