Income Inequality and Self-Reported Health: A Difference-in-Differences Study.

BACKGROUND: Health disparities in the United States (US) are closely linked to income inequality. While many studies have reported associations between income and health, causal evidence remains limited. OBJECTIVE: To estimate the causal effect of income-equalizing state policies, such as minimum wage increases, Medicaid expansion, and Earned Income Tax Credit (EITC) adjustments, on adult self-rated health using a difference-in-differences (DiD) framework. METHODS: Using the Centers for Disease Control and Prevention (CDC) Behavioral Risk Factor Surveillance System (BRFSS) data from 2018 to 2023, a 2-way fixed-effects DiD model was employed to compare changes in the proportion of adults reporting fair or poor health between states that implemented income-related policies and those that did not. The covariates included the demographic and economic characteristics of the American Community Survey. Robustness checks included event study analyses, placebo tests, and models with state-specific linear trends. RESULTS: In baseline difference-in-differences models, policy adoption was linked to a -0.00403 (SE = 0.00141, P = .006) change in the likelihood of reporting fair or poor health, representing a 0.4 percentage-point decrease compared to control states; however, place-study diagnostics showed a significant pre-policy trend violation (F = 47.24, P < .001), which challenged the parallel-trends assumption. After adjusting for state-specific linear time trends, the estimated effects were both statistically and practically null. Placebo models with randomized policy dates produced null estimates, confirming robustness. CONCLUSIONS: The observed improvements in self-reported health in baseline models were not robust to trend-adjusted specifications and likely reflected the underlying pre-policy trends. These findings underscore the importance of rigorous diagnostic testing in quasi-experimental evaluations of policy effects.