A randomized experiment of issue framing and voter support of tax increases for health insurance expansion

ObjectiveTo assess the effect of issue framing on voter support of tax increases for health insurance expansion.MethodsDuring October 2008, a random sample of registered voters (n = 1203) were randomized to a control and two different ‘framing’ groups prior to being asked about their support for tax increases. The ‘framing’ groups listened to one of two statements: one emphasized the externalities or negative effects of the uninsured on the insured, and the other raised racial and ethnic disparities in health insurance coverage as a problem. All groups were asked the same questions: would they support tax increases to provide adequate and reliable health insurance for three groups, (1) all American citizens, (2) all children, irrespective of citizenship, and (3) all military veterans.ResultsSupport for tax increases varied substantially depending on which group benefited from the expansion. Consensus on coverage for military veterans was highest (83.3%), followed by all children, irrespective of citizenship (64.7%), and all American citizens (60.1%). There was no statistically significant difference between voter support in the ‘framing’ and control groups or between the two frames. In multivariable analyses, political party affiliation was the strongest predictor of support.ConclusionsVoters agree on the need for coverage of military veterans, but are less united on the coverage of all children and American citizens. Framing was less important than party affiliation, suggesting that voters consider coverage expansions and related tax increases in terms of the characteristics of the targeted group, and their own personal political views and values rather than the broader impact of maintaining the status quo.