Column: Republicans say Medicaid doesn’t work, so it should be cut. Here are all the ways they’re wrong

Medicaid chief Seema Verma, seen here at her Senate confirmation hearing, says the program doesn't do any good. She should look at the evidence.
(Mandel Ngan / AFP/Getty Images)

The dirtiest little secret of the Republicans’ Obamacare repeal campaign is that its genesis has nothing to do with the Affordable Care Act as such, but with a long-cherished desire to gut Medicaid, which predated the ACA by nearly a half-century.

To advance this goal, conservatives and GOP leaders have asserted consistently that Medicaid doesn’t work or even “harms its beneficiaries.” Health economist Austin Frakt of Veterans Affairs and Boston University now has done Medicaid’s defenders an important service by issuing a call to collect in one place all the claims that the program is broken or harmful, and then pointing us to research debunking those smears.

To be fair, the goal of gutting Medicaid as part of an Obamacare repeal isn’t really much of a secret. House Speaker Paul Ryan (R-Wis.) gave the game away in a videotaped discussion with National Review’s Rich Lowry in March. There he confessed, “So Medicaid, sending it back to the states, capping its growth rate, we’ve been dreaming of this since I’ve been around — since you and I were drinking at a keg.... I’ve been thinking about this stuff for a long time.”


Ryan is 47 now, so he would have been dreaming about cutting Medicaid in the early 1990s while “drinking at a keg,” perhaps at frat parties at his alma mater, Miami University of Ohio. The ACA was enacted in 2010.

So Medicaid, sending it back to the states, capping its growth rate, we’ve been dreaming of this and I were drinking at a keg.

— House Speaker Paul Ryan, R-Wisc., in conversation with Rich Lowry of the National Review

Other examples abound of Republican and conservative hostility to Medicaid. A 2013 survey of 13 state governors who opposed its expansion under the ACA found five who listed as a principal rationale that it’s a “‘broken system’ [that] harms its beneficiaries.” Health and Human Services Secretary Tom Price told a congressional hearing in March that Medicaid has “decreased people’s ability to access care.” The right-wing American Action Forum says the program is “harming those who need it most.”

We reported last week on a drive-by attack on Medicaid — wholly unsupported by the facts — launched by right-wing pundit Ben Domenech on the CBS program “Face the Nation,” and on right-wing healthcare commentator Avik Roy’s long crusade against the program.

In perhaps the most appalling example, Seema Verma, who as the administrator of HHS’ Centers for Medicare and Medicaid Services is in charge of the program, cast doubt in an op-ed last week that “Medicaid works for those it was designed to serve.” Verma based her conclusion that Medicaid had “justifiably taken a lot of heat” in recent years on three studies that have been widely questioned and even more widely misinterpreted.


None of these claims is true. So let’s look at this evidence.

Here’s a short-cut version, produced in 2011 by Frakt and co-authors Aaron Carroll, Harold Pollack and Uwe Reinhardt: “If Medicaid actually harmed health, instrumental variables studies would show that; they don’t,” they wrote in the New England Journal of Medicine. “Other complementary research, such as the RAND Health Insurance Experiment and studies of patients 65 years of age or older who were uninsured before entering Medicare, support the belief that basic public health insurance coverage improves health.”

One of the most oft-cited studies questioning the efficacy of Medicaid is a 2010 study of surgical outcomes from the University of Virginia. The 900,000 surgeries in the sample were from 2003 to 2007. The researchers found that in-hospital mortality for Medicaid patients was worse than for uninsured or privately insured patients, though lower than Medicare patients.

But as an analysis by the Milbank Memorial Fund observed, there were lots of questions about this conclusion. The Virginia researchers tried to adjust for some risk factors distinguishing the Medicaid population from the others, but they couldn’t adjust for everything.

Among the factors they acknowledged might contribute to mortality and skew the results, the Medicaid patients had the highest incidence of AIDS, and metastatic cancers, “depression, liver disease, neurologic disorders, and psychoses.” They suffered from social factors associated with poverty, including drug abuse and delayed diagnoses, and lacked support and resources for care at home.

Despite all that, it turned out that Medicaid patients actually did better than some other patients in such surgeries as lung resections, pancreatectomies, and aortic aneurysm operations, and had fewer complications in some categories. A blanket conclusion that Medicaid patients did worse simply was unwarranted.

It’s also the case, as Kevin Drum of Mother Jones points out, that mortality is an inadequate metric for assessing a healthcare program, since “the vast majority of doctor visits aren’t for life-threatening conditions.” But they can be for conditions that can profoundly affect one’s quality of life, not to mention one’s financial condition, if they go untreated. In any case, he adds, since the average age of Medicaid enrollees is 38, there won’t be much mortality in that group to begin with, so any changes are unlikely to be meaningful.

The best-known study of the effect of Medicaid coverage is the so-called Oregon Experiment, which has the best pedigree of all such studies: It was done by researchers at Harvard and MIT, including such supporters of universal healthcare as Jonathan Gruber. Medicaid critics constantly cite it as proof that having Medicaid coverage is “no better than being uninsured,” to quote Avik Roy.

The problem here is that the authors of the study disagree with that. The study, which followed newly enrolled Medicaid patients for two years, found no improvement in three markers associated with cardiac health and diabetes: cholesterol, high blood pressure and blood sugar levels. But the figures for those were not statistically significant, which makes them useless for assessing the program’s effects.

The researchers did, however, find a statistically significant reduction in the incidence of depression, a significant increase in the diagnosis of diabetes and the use of diabetes medications, and in cholesterol screening, pap smears, mammograms and other screening tests. They also found a significant “reduction in financial strain from medical costs.” Catastrophic expenses, defined as those exceeding 30% of income, were “nearly eliminated.”

These benefits can be traced directly to Medicaid coverage, and they’re not trivial. The fact that Medicaid’s critics return to the Oregon results over and over, cherry-pick a few findings and misinterpret those should tell you something. It’s that declaring Medicaid to be useless, or no better or even worse than having no insurance at all, is merely a shibboleth.

It’s an incantation that gets endlessly repeated as truth, even though empirical studies show that there’s no truth in it at all. Verma, by citing both the Virginia and the Oregon studies in her op-ed without acknowledging their limitations, turned in a shameful performance.

The congressional Republicans backing the Obamacare repeal bills that cut the meat and bones out of Medicaid to the tune of $800 billion to $1 trillion, must have some ulterior motivation. It can’t be improving its users’ health, because what they’re planning would achieve just the opposite. What could it be?

Keep up to date with Michael Hiltzik. Follow @hiltzikm on Twitter, see his Facebook page, or email

Return to Michael Hiltzik’s blog.