Support for Medicaid and Remaining Disparities

Background: The Patient Protection and Affordable Care Act (ACA) gives states the option—with very generous federal financing—to provide coverage to all low-income Americans. After 2014, new Medicaid enrollees will makeup more than 25% of total enrollment in many states that choose to expand coverage. While the ACA Medicaid expansion is hugely significant, Medicaid has been expanding for the last 3 decades. It was our nation’s largest health insurance program a decade before the ACA passed. In 2011, 51% of the public said they had a close connection to the program: either s/he had been on Medicaid (21%) or had a family member or close friend on the program (30%). This raises an important question: as Medicaid expands and more people are connected to the program, are they more likely to support Medicaid expansions? This project is important to social work knowledge base on whether expansions of means-tested targeted programs can translate into more political demands thereby de-stigmatizing the program. In particular, we ask the following research questions: (1) As connection to Medicaid increases over time, do levels of support for Medicaid increase? (2) Do different types of connections—self, adult elderly parents, or children—produce different levels of support? And (3) are there variations across the American states in the public’s type of connection to Medicaid and their level of support?

Methods: We use three datasets from the Roper Center for Public Opinion Research database that include respondents’ information in connections and supports toward Medicaid in 1995, 2005 and 2011. The aggregated total sample size across three datasets is 2,885. We use multinomial and simple logistical regression analyses on the Medicaid support variables to analyze the relationship between Medicaid connection and support. In addition, we estimate state-level public Medicaid support by using Multilevel Regression and Poststratification (MRP) method. To compensate the limited connection type information in 1995 and 2005 datasets, we triangulate findings in multiple ways to determine if the results are robust.

Results: First, as connections to Medicaid increase over time, support for Medicaid also increases. While no relationship between connection and support was observed in 1995, the association between Medicaid connections and support become marginally significant (p<0.1) in 2005 and 2011. Second, people connected to Medicaid through self and family—particularly through parents—are more likely to support Medicaid (p<0.05) controlling for other factors. Third, people in Northeastern, Midwestern, and Pacific states are more likely to be connected to Medicaid through their parents and strongly supportive of Medicaid; whereas people in Southern and Mountain states are more likely to be connected to Medicaid through their children’s coverage and there is no relationship between this connection and support for Medicaid.  

Conclusions and Implications: The results suggest that Medicaid’s expansion over time has created more public support for the program due not to altruism but people’s own connection to the program. The findings raise important questions about disparities in public support across population groups within Medicaid and across the states.