Abstract
This research explores the relationship between episodes of contaminated drinking water and health care expenditures in the United States. The analysis relies on panel data from the 48 contiguous states from 2000 to 2011. We use the population served by public water systems that violate health-based standards of the Safe Drinking Water Act as a proxy for contaminated drinking water. We estimate spatial and non-spatial models and control for factors that may affect per capita health care expenditures including variables that reflect air quality violations along with ability to pay plus demand for and supply of health care services. The results from a Spatial Durbin Model indicate that a 1% decrease in the annual percentage of population exposed to drinking water quality violations is associated with reductions in in-state and regional effects equal to 0.005% and 0.035% of per capita health care expenditures, respectively. While relatively small on a per capita basis, drinking water violations have a larger impact on health care expenditures than air quality violations (whose effects are not statistically different from zero). However, compared to other factors, such as Medicare enrollment and income, the impact of these violations on health care expenditures is small. We find that the cumulative regional health care expenditure impacts from drinking water violations are substantially greater than in-state impacts. Thus, a regional approach is recommended to addressing drinking water quality improvements.