Place as a predictor of health insurance coverage: A multivariate analysis of counties in the United States.

This study assessed the importance of county characteristics in explaining county-level variations in health insurance coverage. Using public databases from 2008 to 2012, we studied 3112 counties in the United States. Rates of uninsurance ranged widely from 3% to 53%. Multivariate analysis suggested that poverty, unemployment, Republican voting, and percentages of Hispanic and American Indian/Alaskan Native residents in a county were significant predictors of uninsurance rates. The associations between uninsurance rates and both race/ethnicity and poverty varied significantly between metropolitan and non-metropolitan counties. Collaborative actions by the federal, tribal, state, and county governments are needed to promote coverage and access to care.