Medicare and Medicaid Fail a Basic Scientific Test

I know there are those who claim American medicine treats symptoms not the underlying lifestyle choices that promote good health.

OMG, what a bunch of crap! Sure people could eat right, stop downing alcohol and smoking hooch, perhaps sleep a bit more and throttle back stress....boring. Grow a pair and quit complaining.

Medicare and Medicaid Fail a Basic Scientific Test

The programs don’t do what they claim to. If they were drugs, regulators would pull them from the market.

By Michael F. Cannon, WSJ

July 21, 2025 5:47 pm ET

Critics of the One Big Beautiful Bill Act claim that, because it would slow the growth of federal Medicaid spending from 4.5% to 2.7% annually, thousands will die. Yet evidence for this is wanting. If Medicaid were a drug, the federal government wouldn’t approve it—and could penalize its salesmen with prison time for claiming it saves lives.

When President Lyndon Johnson signed Medicare and Medicaid into law in 1965, he announced that one of the goals was “to improve the health of all Americans.” Around the same time, Congress barred pharmaceutical companies from introducing or making health claims about new drugs absent “substantial evidence” that “the drug will have the effect it purports,” which the Food and Drug Administration defined as at least one, but usually two, successful randomized, controlled trials.

Such trials are the gold-standard of social-science research. They control for more variables than any other type of study, and therefore better distinguish the effect of a drug from other influences. They are costly and time-consuming. The FDA traditionally requires two successful trials, with only narrow exceptions at Congress’ insistence.

Yet Congress has never demanded a randomized, controlled trial to investigate whether Medicare and Medicaid improve health, much less conditioned their existence on two successful trials. One might think such measures are unnecessary. Subsidizing medicine delivers more health, right?

Not necessarily. Medicare and Medicaid impose rules that reduce healthcare quality. They increase prices for private-sector medical care and health insurance. They require taxes that reduce incomes, financial security and potentially economic growth. They have subsidy phaseouts that discourage upward mobility. Any health gains would have to overcome the health losses these factors introduce. It isn’t even clear whether the additional medical care that subsidies purchase improves health.

Multiple randomized, controlled trials examining subsidized medical care have failed to find health improvements. The Oregon Health Insurance Experiment measured the effects of giving Medicaid to the lowest earners among the able-bodied adults in ObamaCare’s Medicaid expansion population. Like other randomized trials, it found that subsidies led to additional medical care but no discernible improvements in physical health.

Failing to detect a relationship isn’t the same as proving that a relationship doesn’t exist. Many non-randomized studies that compare people with Medicare or Medicaid with people without it suggest those programs do save lives. On the other hand, the Oregon study’s more reliable design found that some non-randomized studies likely overstate Medicaid’s potential health effects, whether due to random chance or because they compare apples to oranges. Still other non-randomized trials cast doubt on Medicare and Medicaid’s effects on health, including one that found that Medicare had no discernible effect on elderly mortality, even after spending 10 years and $515 billion subsidizing healthcare for seniors.

The Oregon study’s null results mean that Medicaid failed the test the federal government uses to determine whether new elixirs save lives. Medicaid would not qualify for the narrow exceptions the FDA makes to its rigorous standards; policymakers have alternatives to Medicaid and randomized, controlled trials of Medicaid are feasible.

The Oregon study’s other findings are no help. Medicaid recipients got more medical care, but that’s an input, not an endpoint. It found that receiving subsidies improved financial security but didn’t capture the effects of taxes and other program features on financial security. The study found improvements in self-reported mental health, but that’s not a slam-dunk. Since two-thirds of those improvements occurred before enrollees even received care, the FDA might want further study. This is like judging a drug cocktail by the effect of one ingredient. Were the FDA to approve Medicaid because it improves some measure that correlates with health, it would eventually revoke approval because no one is conducting follow-up trials.

The lesson of Oregon is that we don’t know whether the additional care that Medicare and Medicaid purchase improves health, much less the overall health effects of those programs. For that, we need further, larger and longer randomized, controlled trials.

Since 1965, Medicare and Medicaid have spent $45 trillion without substantial evidence of effectiveness. If we care about patients wasting millions on drugs that don’t work, we should care about the nation potentially wasting trillions on subsidies whose efficacy remains unclear.

Mr. Cannon is director of health policy studies at the Cato Institute.