Cutting health care “waste” is hard

The high risk of cruel cuts in the name of efficiency

I’m happy to concede that “wasteful spending” is, in fact, a real thing and that making the government more efficient is genuinely desirable, even as DOGE wreaks havoc on the American state and it becomes increasingly clear that efficiency isn’t actually the goal of Donald Trump or Elon Musk.

It’s worth thinking about what it would mean to take this seriously, especially because we may need to reconstruct many of our institutions in the future and should try to build back better. The foreign aid case strikes me as particularly clear-cut in this regard — some programs are dramatically more cost effective than others, and a sane administration would cut the ineffective ones in order to put more money behind highly effective ideas.

Everything DOGE has focused on so far is nickel-and-dime stuff, though. In fiscal terms, what really matters is defense, Social Security, and health programs. Social Security is already extremely efficient, in the sense that it’s basically cash transfers with low overhead. Defense is its own kettle of fish.

But I think it’s worth talking specifically about healthcare, where many hundreds of billions of dollars (arguably more) are spent on an incredibly wide array of programs and services. It’s obviously not the case that all of these expenditures have equally high-impact outcomes, and in any reasonable system, we would periodically review what’s actually happening and try to eliminate waste.

At the same time, the question of what waste means in a health care context is fraught.

It’s different from what it means in the context of something like the energy efficiency tax credits created by the Inflation Reduction Act. The question of how much greenhouse gas emission reduction per dollar we get from subsidizing heat pump water heaters is somewhat difficult to answer empirically, and I’m sure that people disagree about it. But conceptually, it’s pretty clear what we’re trying to measure. You can, in principle, measure the cost-effectiveness of those tax credits and compare it to the cost-effectiveness of electric car tax credits and see which one is better. Because whether or not you, personally, think it’s a good idea to spend money on reducing greenhouse gas emissions, it’s at least clear what the program is trying to do.

Health care spending, for better or worse, is much more conceptually complicated.

The specter of death panels

A lot of health care spending goes toward care for elderly people.

The way human beings work is that, notwithstanding Bryan Johnson’s best efforts, we die. According to the Social Security Administration, the life expectancy for a 90-year-old man is less than four years and for a woman of the same age, it’s less than five. As a result, the aggregate impact on the life expectancy of providing health care to people over the age of 90 is miniscule. If we categorically denied all health care to nonagenarians, how much could life expectancy possibly fall?

And it’s not just a question of money — real resources, including hospital beds and the time and labor of skilled professionals, are used in caring for the very old, and those resources could instead be put toward other health care goals.

Of course, nobody is proposing that, exactly. But there’s certainly a sense in which that kind of move would be efficient.

Budget wonks used to talk a lot about the very high level of health spending that people tend to incur during their final year of life. One influential analysis found that in 2014, fully one quarter of Medicare spending was on people in their final year of life. Other research says that’s an overstatement, but newer research seems to broadly confirm the original claim, so I think it’s probably not far off.

In other words, if the Angel of Death showed up at the Department of Health and Human Services and told the Centers for Medicare & Medicaid Services who was going to die over the next year, the government could save a ton of money by denying those people health care. Would their life spans be shortened as a result? Sure. But the maximum possible impact is quite modest, and in many cases it would be less than that. Of course, in the real world, we don’t know who is going to live and who is going to die.

But there is a kind of basic disjuncture between the practitioner’s desire to throw a lot of resources at treating people who are very sick, and a budget wonk view that this doesn’t pencil out.

Normal people reject that logic, though.

There is, from time to time, policy emphasis on the idea of improving the cost- effectiveness of end of life care, including suggestions that people should be moved more rapidly to hospice-type settings to receive palliative care rather than treated with maximum aggression.

The Obama administration got Medicare to fund end-of-life planning, with the hope that not only would many families find this service useful, but the very modest cost of the offering would “pay for itself” since some people would choose less aggressive end-of-life care options. That always seemed reasonable to me; most of my family members who’ve passed away eventually chose to prioritize what they could retain of comfort and peaceful time with family over treatment maximalism.

But in practice, of course, the upshot was years of warring over conservatives’ claims that Obama had created Death Panels that would deny treatment to the elderly to save money.

Cost effectiveness for whom?

That wasn’t true of Obama, and I don’t even expect DOGE or House Republicans to actually create death panels.

What is true, though, is that the ascendant Nietzschean strain of right-wing politics spares little interest for the welfare of vulnerable people. And it’s also true that Elon Musk tends to play fast and loose with the facts.

A third of Medicaid spending goes to long-term support services for the elderly and disabled.

I’m sure that some of that spending is wasteful or inefficient, in the sense that it’s poorly structured or implemented. Specific ideas about how to either provide better services for the same money or equivalent services for less money would be quite useful.

But there’s a risk of DOGE or Republicans looking at something like that and deciding that, as a category, they don’t think it’s very “efficient” to be spending all this money on long-term support for the elderly and disabled. After all, if you were forced to cut something, it would probably make more sense to cut this than maternal care. Medicaid covers over 40 percent of childbirths in the United States; this does not cost that much money and appears to have multi-generational benefits.

The upside of long-term support services is not nearly as big. But DOGE-world rhetoric keeps muddying the waters between two senses of efficiency: There’s “look at what the government is doing and try to find more cost-effective ways of doing it” and there’s “let’s just decide that some of the stuff the government does isn’t important.”

That question of what the government should be doing is a legitimate question for debate. But one really does need to debate it! We keep hearing right-wingers broadly categorize any spending they doesn’t happen to agree with as “waste” or “fraud.” And it’s certainly true that taking care of the elderly and disabled doesn’t necessarily get you to Mars faster.

But the United States is not a desperately poor country that lacks the financial resources to provide health care to its people. We are not on the verge of bankruptcy.

We do need some deficit reduction. I’ve scolded progressives on both the environmental and care economy fronts that we need to ditch the depression economics mindset and realize the country no longer needs a New Deal-style program.

But what Trump, Musk, and Mike Johnson are cooking up isn’t a deficit reduction program — it’s cutting spending to partially offset a gigantic multi-trillion dollar tax cut. Whether it makes sense to do that on the backs of the most vulnerable people in society is ultimately a question of values, not efficiency.

Even good cuts are tough

There is, of course, plenty in the health care system that’s worth watchdogging. Progressives have been complaining for years that Medicare Advantage — essentially a “private option” for Medicare recipients — is bilking the taxpayers to benefit for-profit insurance companies. Green eyeshade budget wonks increasingly agree with them. The Committee for a Responsible Federal Budget thinks there will probably about $100 billion per year in overpayments to Medicare Advantage plans over the next decade.

To simplify this, the way it works is that when MA plans get paid, the amount of the payment is determined by the health status of the person getting coverage.

Evaluating health is complicated, multidimensional, and involves judgment calls. The risk-adjustment process gives insurers strong incentives to systematically skew every possible judgment call toward negative assessment of their enrollees’ health status. This kind of “upcoding” is abusive, it costs the government a bunch of money, and we should absolutely crack down on it. And one way you can tell the Trump administration isn’t really on the level is that the person picked to run health policy at the Office of Management and Budget, Don Dempsey, worked at the trade association for Medicare Advantage plans before joining the White House.

It also has to be said, though, that even a “good” cut like this isn’t a case of pure efficiency.

When Medicare Advantage plans bilk the taxpayer, they do so for the sake of shareholder value and executives’ salaries. But they do need to get patients to sign up for the plans before they can upcode them. That’s accomplished in part with marketing, but in part with goodies. Medicare Advantage plans often cover services like dental or optical care that aren’t in the core Medicare package. So it’s not like this is a magic wand for saving money without harming patients.

Republicans, at any rate, are fanatically committed to defending that particular category of health care waste.

Another thing budget wonks hate that Republicans are more likely to target is Medicaid provider taxes. Medicaid is a joint state-federal program, in which the amount of federal dollars a state gets is, in part, a function of how much of its own money the state kicks in. States levy taxes on the providers of Medicaid financed services, and then put that money into their state Medicaid funding stream. On its own this is pointless — states are extracting money from providers and then giving it back to them. But because the federal government matches state spending, you can materialize extra federal money “for free” this way. Obviously, if the federal government allowed that to go totally unchecked, it would wreck the whole structure of the program, so there are safeguards and limitations, but also proposals to make those limitations stricter.

It’s absolutely true that these tactics are a kind of shady gimmick that, from a strict budget management perspective, probably shouldn’t be allowed. And cracking down on them definitely reduces federal spending. But that doesn’t mean that this is some kind of immaculate spending cut where pure waste vanishes — states will need to cut their Medicaid offerings.

Health care is weird

If you argue about health care on the internet long enough, you’ll run into a smart, well-informed conservative who brings up the Oregon Medicaid experiment in which a lottery-style enrollment provided an ideal opportunity to get causal evidence on the impact of Medicaid. The punchline here is that while Medicaid significantly improved patient’s self-assessed health, this was not reflected in objective measures of physical health that researchers could make. A lot of conservatives take this to prove that health insurance coverage is basically pointless virtue signaling and we could just cut this program willy-willy with no real negative impact.

But I find it striking that none of them really seem interested in running this “what if you just didn’t get routine health care” experiment on their own kids.

My 10-year-old child has, fortunately, never had any major health problems. Nonetheless, he’s had plenty of routine visits to his pediatrician. He’s had antibiotics for strep throat and an ear infection. If he has a fever, we give him ibuprofen. If we didn’t give him medicine when he had a fever, would he die? Almost certainly not. By the same token, though antibiotics unquestionably save lives in the aggregate, the odds that any given strep infection would be lethal if left untreated are actually very low. When he was a baby, we had a lot of pediatrician consultations that had the broad tenor of, “What’s up with this weird rash?” The answer was usually that nothing in particular was up with it.

I had a weird spot on my arm a few years ago. A doctor cut it out, sent it to a lab, and the lab said it was benign. In my 20s, I had a painful situation in my jaw that I went off-insurance to get treated quickly rather than waiting weeks in pain. I was especially alarmed about this because, again, the dentist told me he thought it might be cancer. I spent a bunch of money and learned that it was not cancer. But I was pretty glad to be pain-free and to have the peace of mind of knowing that it wasn’t cancer rather than waiting weeks.

Which is just to say that it’s probably true that most health care is ex-post pointless if judged from a metric of whether you died. And, of course, under the status quo, hospitals need to provide indigent care to people facing acute emergencies anyway.

But I’m deeply concerned that a bunch of right-wingers who don’t really care about the poor, the elderly, or the disabled are going to start running around, waving the headline finding of one big study and decide that all kinds of health care for the vulnerable is “waste” and should be done away with.

But we have other studies that do show a meaningful impact on mortality, as well as long-term better economic outcomes for covered kids and reductions in crime. We know that normal people like to purchase health care services in situations where the risk of death due to untreated illness is objectively low, and that poor people report large gains in subjective well-being when they get coverage. We also know that Medicaid is an extremely thrifty program that provides services at lower per unit prices than Medicare (and much lower than private health insurance). There are a lot of valid questions one can ask about the cost-effectiveness of different medical interventions, but the biggest drivers of “inefficiency” are fundamental ethical commitments to provide care to the elderly. And we’re running a huge risk of careless people using concerns about waste as cover for inflicting serious harms.