Factors associated with differences in mortality and self-reported health across states in the United States

Objective: Recent studies indicate continuing health disparities across geographic units in the US. This paper provides updated estimates of the association between socioeconomic factors and population health using a new state-level dataset and panel econometric methods that account for state-specific effects and autoregressive error structure. Methods: Data from multiple sources for the 50 US states and the District of Columbia are merged. The dependent variables are age-adjusted all-cause mortality, self-assessed health status, and number of healthy days. Panel econometric models are used to accommodate state-specific unobserved factors and to incorporate autoregressive random disturbances to provide consistent and robust estimates. Results: A 1-unit increase in the number of physicians per 1000 population is associated with a reduction in mortality by 30/100,000. The effects of physician-to-population ratio on self-reported health measures are mixed. Socioeconomic, demographic, as well as the prevalence of smoking and obesity have varying effects on mortality and self-reported measures of health. Conclusions: The new estimate of the association between physician supply and lower mortality suggests continuing efforts to assess the need for policies and incentives to induce physician labor supply in underserved states. Strategies and policies to reduce health disparities should address social, economic and individual risk factors.