Senators supporting deep cuts in Medicaid may find that they are out of step with their low-income constituents, who report increased support for the Affordable Care Act (ACA) across party lines in states that expanded Medicaid.

In a new working paper, Daniel Hopkins and Kalind Parish compare public opinions of the ACA in states that did and did not expand Medicaid, in a population most likely to benefit from the expansion. It turns out that direct experience trumps partisanship, and may reduce the difference in ACA support between Democrats and Republicans by up to 50%.

The authors used data from a rolling Kaiser Family Foundation (KFF) survey on public views of the ACA, which included more than 63,000 adults who were interviewed in 67 surveys between 2010 and 2016. They focused on the nearly 17,000 under 65 who reported a family income of less than $40,000 (just above the income cap for Medicaid eligibility for a family of five in expansion states). Using a differences-in-differences analysis, they compared attitude shifts in the 31 states expanding Medicaid with the 19 states that did not.

First, they document that the share of respondents receiving Medicaid did increase by 5.7 percentage points in expansion states. Adjusting for many sociodemographic factors, they find a mid-sized increase in support for the ACA among low-income people in expansion states, and no such effect among higher-income respondents. On a four-point scale, low-income people move 11% of the way from “somewhat unfavorable” to “somewhat favorable” views of the ACA (and most of them are not on Medicaid). This suggests that people who directly benefit from a policy develop more favorable opinions of it, something political scientists call a policy feedback. The following charts from the paper illustrate the study results (although they might take a bit of study to fully digest):

Research has shown that personal experiences with the ACA and Medicare – positive and negative – correlate with attitudes toward the ACA. Hopkins and Parish had reason to question whether this would hold true for the Medicaid expansion, given the strength of pre-existing attitudes and partisanship about the program, and its stigma as “welfare.” Given that most Medicaid coverage comes through private, third-party payers, it is also possible that Medicaid enrollees might be unaware that they’re on Medicaid, which would limit the extent of the policy feedback. Prior studies show greater underreporting of Medicaid coverage than of private insurance.

This study demonstrates that Medicaid expansion had discernible, positive effects on public opinion of the ACA among low-income Americans, effects that grew over time. It likely induced much larger effects among the fraction of that population who actually received the coverage.

The Better Care Reconciliation Act of 2017, the Republican-led plan now before the Senate, proposes severe Medicaid cuts that may be greater than the $800 billion in Medicaid cuts proposed by the House. The Senate version has a slower phaseout of the Medicaid expansion, but might cut even more from Medicaid over a greater timeline; the Congressional Budget Office will weigh in on that some time next week. But both plans represent more than a rollback of the ACA’s Medicaid expansion; they restructure  the entire 50-year old program, which will result in millions of low-income people losing Medicaid coverage. With evidence that most Medicaid beneficiaries are satisfied with their coverage, coverage that would cost 25% more if they had private insurance, we question the wisdom of “fixing” the ACA by radical cuts to Medicaid. Why not start with the proposition:

If you like your Medicaid, you can keep it.