A Project of The Annenberg Public Policy Center

Is Medicaid Bad for Your Health?


Under the Affordable Care Act, millions of the uninsured have gained Medicaid coverage. But is Medicaid good for their health, bad for their health, or does it make no difference from being uninsured? All three claims found a political champion recently, and all refer to the same study for support.

According to Republican presidential candidate Rick Perry, the health outcomes for people on Medicaid are no better than for those without insurance.

Fellow candidate Sen. Ted Cruz went further, saying health outcomes are “markedly worse when people get on Medicaid” and that “people’s life expectancy goes down on Medicaid.”

President Obama and others in his administration, meanwhile, credit the Medicaid expansion for improving people’s health and saving thousands of lives.

The academic evidence is heavily on Obama’s side. But both the president and Perry are cherry-picking from the same study, while Cruz distorts it.

Perry and Cruz claim the expansion of Medicaid through the Affordable Care Act is a waste of money because, they say, studies show Medicaid doesn’t work. The Rand Corporation estimated that 6.5 million people who newly enrolled in Medicaid from September 2013 through February 2015 were previously uninsured. Other estimates — from both the Kaiser Family Foundation and the Urban Institute — say about 4 million additional uninsured could gain coverage if all states expanded Medicaid. There are currently 21 states that have not expanded.

In a speech at the National Press Club on July 2, Perry said, “We spend $450 billion a year on Medicaid. And yet, health outcomes for those on Medicaid are no better than those who have no health insurance at all. Instead of reforming Medicaid, the president expanded it under Obamacare.” (His remarks start at the 12:15 mark).

A few days earlier in a Yahoo News interview, Katie Couric asked Cruz what would happen to the 16.4 million who have gained insurance through the ACA if the law were repealed, as Cruz has called for.

Cruz, June 29: Well, that figure is a deceptive figure. The large chunk of that is people who have been put on Medicaid. And one of the things about Medicaid — if you look at the health outcomes on Medicaid, health outcomes are markedly worse when people get on Medicaid. Medicaid is not performing well. And in fact people’s life expectancy goes down on Medicaid. It is markedly worse.

President Barack Obama, meanwhile, claimed just the opposite, saying that those states that have expanded Medicaid have “a healthier population” than those that have not.

Obama, July 1: I will tell you the states that have taken full advantage of all the federal options available, they have an even lower uninsured rate, and a healthier population, and more people signing up for the options that are available than those states that have not taken full advantage of those options. And that’s just a fact.

Interestingly, all three base their claims, at least in part, on a single study called the Oregon Health Insurance Experiment, which was published in the New England Journal of Medicine on May 2, 2013. The study took advantage of a Medicaid expansion in Oregon that was based on lottery drawings and compared data from 6,387 adults who were able to apply for Medicaid coverage to 5,842 adults who were not selected.

The Oregon Experiment

The authors of the study concluded: “This randomized, controlled study showed that Medicaid coverage generated no significant improvements in measured physical health outcomes in the first 2 years, but it did increase use of health care services, raise rates of diabetes detection and management, lower rates of depression, and reduce financial strain.”

There was a little something for everyone in that conclusion. The White House Council of Economic Advisers, for example, highlighted the study’s findings that Medicaid coverage expanded access to medical care and improved mental health benefits.

In a June report, the CEA said that the states that have expanded Medicaid “will reduce the number of people experiencing symptoms of depression by 382,000,” and that if the other states did choose to expand Medicaid, an additional “393,000 fewer people would experience symptoms of depression.”

The Oregon study found that after two years, those covered by Medicaid had a lower probability of screening positive for depression than those who were not covered by Medicaid. The difference amounted to 9.2 percentage points, which the CEA then applies to estimates from the Urban Institute on the number gaining or not gaining insurance because of states’ decisions on Medicaid expansion.

The CEA makes similar calculations using the Oregon study’s finding that those with Medicaid coverage were more likely to self-report that they were in good, very good or excellent health (as opposed to fair or poor health). About 572,000 additional people in Medicaid-expansion states, and another 556,000 in other states, should they decide to join the expansion, would achieve that outcome, CEA said. The CEA makes other calculations on increased usage of preventive screenings by those gaining Medicaid coverage, another finding of the Oregon study.

The study also provides some backing for Perry’s claim that “health outcomes for those on Medicaid are no better than those who have no health insurance at all.” But the study wasn’t as all-encompassing as Perry suggests.

As we noted in an article last year, the study measured only three health indicators — blood-pressure, cholesterol and glycated hemoglobin levels (which measure diabetic blood sugar control) — and only over a two-year period. We pointed out then that there could be other improvements that the study didn’t attempt to measure, or that could show up once patients are covered for longer than two years.

There was nothing in the study to back up Cruz’s claim that Medicaid caused people to fare worse than if they remained uninsured, even though his campaign pointed to the Oregon study when we asked for support for his claims.

One of the authors of the study, Amy Finkelstein, told us via email that Cruz’s statement was “really not a fair characterization of the results from the Oregon Health Insurance Experiment.”

“Our randomized evaluation finds that Medicaid improves mental health (i.e. reduces depression – by 9 percentage points or 30 percent),” Finkelstein said. “We find no statistically discernible effects of Medicaid on physical health measures or mortality. But that means we don’t have evidence that Medicaid has an impact; it does not mean that we have evidence that Medicaid is bad for people’s health.”

Nor was the Oregon study the only word on this issue in the academic community.

Other Evidence

A study published on May 6, 2014, in the Annals of Internal Medicine compared mortality rates before and after the health care overhaul in Massachusetts to a control group with similar demographics and economic conditions in other states and concluded, “Health reform in Massachusetts was associated with significant reductions in all-cause mortality and deaths from causes amenable to health care.”

“This is an important piece of the puzzle,” Katherine Baicker, a professor of health economics at the Harvard School of Public Health and a coauthor of both the Oregon and Massachusetts studies, told the New York Times. “Putting the evidence together paints a very strong picture that expanding insurance substantially improves the well-being of people who get it.”

Another study published in the New England Journal of Medicine in 2012 — also coauthored by Baicker, a former member of President George W. Bush’s Council of Economic Advisers  — compared several states that substantially expanded Medicaid (before the ACA) to neighboring states that did not expand Medicaid and concluded, “State Medicaid expansions to cover low-income adults were significantly associated with reduced mortality as well as improved coverage, access to care, and self-reported health.”

A 2013 report from the nonpartisan Kaiser Family Foundation looking at the breadth of academic study — including the Oregon study — concluded simply that “[h]aving Medicaid is much better than being uninsured.”

Kaiser Commission on Medicaid and the Uninsured, 2013: Consistently, research indicates that people with Medicaid coverage fare much better than their uninsured counterparts on diverse measures of access to care, utilization, and unmet need. …

A large body of studies over several decades provides consistent, strong evidence that Medicaid coverage lowers financial barriers to access for low-income uninsured people and increases their likelihood of having a usual source of care, translating into increased use of preventive, primary, and other care, and improvement in some measures of health. Furthermore, despite the poorer health and the socioeconomic disadvantages of the low-income population it serves, Medicaid has been shown to meet demanding benchmarks on important measures of access, utilization, and quality of care. This evidence provides a solid empirical foundation for the ACA expansion of Medicaid eligibility to millions of currently uninsured adults, and individuals and communities affected by the Medicaid expansion can be expected to benefit significantly.

Obama’s Council of Economic Advisers points to the 2014 Massachusetts study and the 2012 Medicaid study to claim that the risk of death will be lower in states that have adopted the ACA expansion. The Massachusetts study found a 2.9 percent reduction in mortality rates for Massachusetts counties from the four years before the 2006 health care overhaul to the three years after, as compared with other states’ counties. The 2012 Medicaid study examined five-year periods before and after state Medicaid expansions from 1997 through 2007 and found a 6.1 percent reduction in adults’ mortality rates in those states, compared with adults in neighboring, nonexpanding states.

The CEA report says that because there are uncertainties with this type of data and “in the interest of being conservative,” it uses the smaller effect found in the Massachusetts study to estimate “5,200 fewer people would die each year” if all states expanded Medicaid and that 5,000 fewer deaths per year will occur in states that have expanded the insurance program.

That Massachusetts study said it found an “absolute decrease of 8.2 deaths per 100,000 adults” and that the changes it found were “larger in counties with lower household incomes and higher prereform uninsured rates.”

The president’s claim of a “healthier population” in expansion states is vague, and the CEA analysis shows relatively small numbers for improved health outcomes compared with the millions who have gained or stand to gain Medicaid coverage under the ACA. But past studies do show some improvement in mental health, self-reported health status and mortality rates for those gaining Medicaid coverage. It remains to be seen what health impacts expansion states actually will experience.

These are only estimates, extrapolated from other studies. The CEA notes several times in its report that there are limitations to this type of research and analysis. The evidence available from studies of past policy changes “is necessarily an imperfect guide to the future, and the actual effects of Medicaid expansion under the Affordable Care Act could be larger or smaller than the estimates presented herein,” it says.

Cruz’s Previous Claims

This isn’t the first time Cruz has made claims about Medicaid increasing the risk of death. In his first-ever speech from the Senate floor, Cruz claimed that studies showed people on Medicaid were “more likely to die” and had “worse health care outcomes” than those who were uninsured (starting at the 11:10 mark).

Cruz, Sept. 18, 2013: Now, the data demonstrate that Medicaid beneficiaries face worse health outcomes than just about anybody else in the marketplace. In 2010, the Annals of Surgery issued a landmark study that examined the outcomes from nearly 900,000 individuals undergoing surgery from 2003 to 2007. The conclusion of that study was that Medicaid patients were almost twice as likely to die as those with private insurance. Medicaid patients, their hospital stays were 42 percent longer and cost 26 percent more. Even more striking, Medicaid patients, when compared to people without health insurance, people who were uninsured, Medicaid patients were 13 percent more likely to die. And they stayed in the hospital for 50 percent longer and cost 20 percent more.

In 2011, Johns Hopkins did a study of patients undergoing lung transplantations. And their conclusions were very much the same. They found that Medicaid patients were 8.1 percent less likely to be alive 10 years after the transplant compared with those with private insurance and also compared to those without any insurance at all. Overall, the Johns Hopkins study found that Medicaid patients faced a 29-percent greater risk of death, and yet Obamacare is moving more and more of the economically disadvantaged onto Medicaid, which subjects them to those worse health care outcomes.

It’s true that in the two studies Cruz referenced — one published in the Annals of Surgery and the other out of Johns Hopkins — patients on Medicaid fared worse than those who were uninsured on some measures. But neither suggests they fared worse because the patients were on Medicaid.

Rather, it has been shown that patients on Medicaid are, in general, poorer and sicker than those without insurance, according to the Kaiser Family Foundation.

KFF, 2013: Because of Medicaid’s eligibility criteria and the strong correlation between poverty and poor health and disability, Medicaid beneficiaries are poorer and have a poorer health profile compared with both the privately insured and the uninsured. This is true even within the low-income population, as Figure 2 illustrates for adults. The distinctly higher rates of poverty, chronic illness, and disability in the Medicaid population are important to bear in mind when considering the evidence on Medicaid’s impact.

Irving Kron, a coauthor of one of the studies cited by Cruz, said the Texas senator’s statement is incorrect.

“Both Medicaid and uninsured did equally poorly,” Kron told us in an email. “However the uninsured were healthier than Medicaid and should have done better than Medicaid. They did not because they had more emergency operations presumably due to lack of resources or access. The Medicaid patients were the sickest of all groups.”

Kron said the purpose of the studies was to show whether payer status is a predictor of risks of surgery faced by patients. The “purpose of the studies was not to show, and does not show, that the Medicaid program negatively affects patients’ health,” he said.

 — Robert Farley and Lori Robertson