Asking About Subsidies for Health Insurance Premiums in Surveys: Does Question Wording Matter?

Abstract

The Affordable Care Act (ACA) introduced the “marketplace” — a portal for enrollment in both private and public health insurance coverage (e.g., Medicaid). Private coverage on the marketplace is purchased directly from insurance companies, and subsidies are available for those who meet income guidelines. Federal surveys, including the Current Population Survey (CPS), the Medical Expenditure Panel Survey (MEPS), and the National Health Interview Survey (NHIS), adapted their questions about health insurance for the new marketplace and added questions about subsidies. We evaluate whether the wording used in these different versions of the subsidy question yields equivalent estimates of subsidy receipt and levels of missing data. Differences in these measures could mean that subsidy estimates are sensitive to the question wording, which could affect inference about the effects of the ACA. We use data from experiments in the Health Reform Monitoring Survey (HRMS) in which adults were randomly assigned to one of three subsidy questions. Due to the many complexities of the marketplace, establishing the universe of those enrolled is an imprecise process. Thus, we employ a conservative method to identify respondents who report coverage that could be a private marketplace plan. Within this group, we examine reporting among various income subgroups that approximate eligibility for marketplace subsidies. The three questions yielded equivalent estimates of subsidy receipt and missing data overall and across all income groups examined. Findings suggest that although the wording of the subsidy question varies across surveys, the CPS/MEPS subsidy question yields results equivalent to the NHIS and the HRMS.