A focus on the HI Trust Fund¹ misses the bigger problem that much of Medicare (including physician visits and prescription drug benefits) is at least partly financed through general revenue funds. Funding from nondedicated revenue sources like payroll taxes, premiums, and taxes on Social Security benefits cover just over half of outlays, necessitating transfers from the general fund. These expenditures strain the federal budget and add to the country’s fiscal challenges. Medicare’s trustees have issued a determination of excess general revenue Medicare funding for 10 consecutive years and 18 of the last 21 years.

Outlays for Part A are around ninety percent funded by payroll taxes. Outlays for Parts B and D are only about 25 percent funded from premiums paid by enrollees. The rest comes from taxpayers, and if eliminated, would almost completely fix our deficits going forward:

From 2026 to 2035, Medicare’s general revenue contribution from nondedicated financing sources is projected to increase from $639 billion to $1.32 trillion, and total almost $10 trillion over 10 years (in nominal dollars). The bulk of the general revenue funding requirement comes from Medicare Part B (71%), where taxpayers will kick in $451 billion in 2026 and $944 billion in 2035. (Analyses were done by the authors based on Tables III.B4 [HI], III.C4 [Part B], and III.D3 [Part D] in the Trustees Report cited above.)

Increase dedicated revenues for Parts B and D (and to some extent Part A), and you do most of the work to prevent our ensuing fiscal calamity. Reduce outlays at the top end and you get most of the same effect. But that’s not a popular avenue of reform.

So it was nice to see this tweet and ensuing Washington Post editorial from Dominic Pino:

I’ll have more to say soon about the budgetary options within Medicare to shrink the general revenue draw on the budget (Hoover just hosted a terrific Medicare budget options conference), but one key part is going to involve seniors paying more in the form of higher premiums or higher cost sharing.

The claims are processed by a small handful of private companies called Medicare Administrative Contractors that have regional, state-based monopolies for claims processing. You’ve never heard of yours. You also don’t have a say in which one you use.

Brian Miller and I are joined by Karen Jackson, former Deputy Chief Operating Officer at CMS. Karen is one of the few people who can credibly say, “CMS does a good job, but it can also do better.” She’s now a SHIP counselor and sits on several healthcare boards, so she sees the beneficiary end of it too.

We discuss Medicare Administrative Contractors (MACs) and their paper (with Luke Anderson), “The Forgotten Opportunity: Improving Traditional Medicare Operations.”

They suggest that beneficiaries should be able to choose their contractor, which could only happen if the regional monopolies had to allow additional contractors to process claims. It’s a competition idea for the half of Medicare where there is currently no competition at all.

What’s in the episode

[1:44] Karen’s path from presidential management intern to Deputy COO and how Medicare contracting reform got tucked into the 2003 prescription drug law.

[4:12] Medicare history and how the Blue Cross/Blue Shield plans originally ended up running so much of the program.

[6:06] Brian compares MACs to the third-party administrator his state health plan board hires. Same structure, about a hundred times smaller.

[7:41] Medicare is around $1.1 trillion. Half goes out as capitated payments to Medicare Advantage plans. The other half runs through fee-for-service.

[9:51] Who actually cuts the check? Not CMS. Treasury does, after every payment clears what was once the largest general ledger accounting system in the world.

[11:00] Why your claim can’t be paid the day you leave the office. Authorization can happen fast, but other considerations slow it down, including the cash flow management across the whole federal government and the 14-day payment wait time.

[13:30] Congress told CMS to pay the right amount on the right day and never asked it to have a payment strategy.

[15:27] What a Medicare Administrative Contractor (MAC) really is.

[16:48] Medicare Administrative Contractors, explained. Provider enrollment, claims intake, state-grouped jurisdictions, separate specialty contractors for home health and medical equipment, and the local edits that let a contractor in California behave differently than one in Florida.

[18:41] MACs have responsibility for specific jurisdictions, done at the state level. Roughly 99.99 percent of claims arrive electronically. The last hundredth of a percent is really expensive.

[21:10] Is CMS running Medicare, or are the contractors? Karen’s answer is more interesting than yes or no.

[22:31] Brian asks about the structural trap. A half-trillion-dollar safety net program with a sole-source operational contract means anything you break, you break for real. Karen makes the case for why CMS is right to be careful.

[27:11] “It should just work, because when it doesn’t work, it’s kind of catastrophic.”

[28:56] Why nobody at CMS is touching this right now. Provider tax thresholds, Medicaid work requirements, the inpatient-only list, site-neutral payment, 340B edits. The operational staff are busy.

[33:18] Karen’s proposal. Let a beneficiary actually talk to the entity adjudicating their claim instead of bouncing between offices.

[35:48] Let beneficiaries elect a contractor every year instead of being assigned one by geography.

[37:19] How you’d pay for it. Fold the program integrity contractors back into the MACs, then return the savings to beneficiaries three ways. A Part B premium rebate, higher payments to rural hospitals and hard-to-find specialists, or coverage for new technology where there’s no national coverage determination yet.

[38:43] What could CMS do administratively today, and what would take an act of Congress?

[43:14] Could you hand this to AI? Karen’s answer is about traceability and trust.

[46:23] There is no trade association for beneficiaries who want a better claims processor. Who would champion their proposed change?

Think you have the solution for Medicare’s budget problems? Not so fast! There’s a gauntlet standing in your way ready to say no.

My colleague Brian Miller walks us through every committee, every lettered agency, and every interest group that will weigh in on your idea before it can go into effect.

What’s in the episode

[2:05] Why talk Medicare: A trillion dollars a year, but 55 percent of it is paid for with general fund revenue. That’s the share not paid for by payroll taxes or premiums.

[4:03] The committees of jurisdiction for Medicare.

[7:01] The differences between the fee-for-service (FFS) and Medicare Advantage approaches. I like Brian’s pizza shop analogy.

[8:48] Medicare started with FFS. MA was a reaction and provides delegated flexibility.

[12:36] Inside CMS and the alphabet soup of agencies that all get a say.

[15:16] A tour of agencies that weigh in on healthcare policy, from CBO to OMB.

[17:40] The major interest groups and why healthcare has way more than 584 vetoes.

[21:40] Who sets the prices for the government? Plus, the RUC’s role.

[24:06] How often do you think prices should be allowed to change in healthcare markets?

[25:42] The “Jupiter-like gravity” of fee-for-service, and why the patient experience remains terrible.

[30:21] The neat thing Utah is doing with automated prescription refills.

[33:42] The case against central planned price-setting and why prices should change more than once a year.

The results are two-fold:

1) There aren’t that many ships built or operated that qualify, raising prices of critical supplies or preventing them from being transported entirely.

2) Non-compliant foreign ships do weird things to deliver goods, like leaving a U.S. port, going abroad, then coming back to a nearby U.S. port.

I’ve written before about how when you see public policy problems, the way to fix them is to to look for the government rule that says you’re not allowed to fix them.

In response to fuel supply interruptions from the Iran conflict, the Department of Homeland Security under the Trump Administration issued a Jones Act waiver to allow foreign-flagged vessles to move oil, LNG, fertilizer, coal, and other critical commodities for 60 days in March 2026 (later extended to August 16th), leading to a nice natural experiment.

What happens when you allow ships to deliver fuel to people who need them without artificial restrictions?

Here’s Colin Grabow of CATO with early results:

The results are even better when compared annual average to annual average:

That’s how much economic activity that was held back by the government telling the market it could only work one way, when better alternatives were readily available.

Colin and others at CATO are running the Jones Act Waiver Tracker and the Jones Act Tanker Tracker, which are demonstrating that even with the wavier, the existing Jones Act tankers are still operating.

Some more of their highlights:

The extended waiver doesn’t last forever. Here’s hoping that the political pressure to keep prices lower and critical resources flowing is enough to continue extending it.

Or even better: These positive results get to Congress to repeal the law entirely.

It once was…but not anymore.

For more, read the full blog post over at the Hoover Institution’s Freedom Frequency Substack.

Howdy folks. Today I’m joined by Brian Miller, our new visiting fellow at the Hoover Institution, to talk about the Affordable Care Act’s effective ban on new physician-owned hospitals.

Brian is an associate professor of medicine at Johns Hopkins, a MedPAC commissioner, and the vice chair of the North Carolina State Health Plan board. He’s a practicing hospitalist who has also done tours at CMS, the FTC, and the FDA.

The existing ban on physician-owned hospitals from billing Medicare patients

On this episode, Brian and I discuss the ban on new physician owned hospitals (POHs) that was put in place in the Affordable Care Act. Section 6001 of the ACA prevents new POHs from billing Medicare services. Why technically new POHs are allowed to open and charge commercial or non-Medicare patients, the lack of Medicare patients has effectively frozen new supply of physician-owned facilities.

Case in point: There were roughly 250 POHs operating when the ACA passed in 2010. There are still roughly 250 today.

What’s the case for the ban?

The ban was pushed by the hospital industry as a way to prevent competition for its services. The hospital industry’s argument went like this: physicians, if turned loose to own hospitals, physicians would cherry-pick the healthiest and best-insured patients, stand up boutique specialty facilities without emergency rooms or labor and delivery, and starve community hospitals of the volume that cross-subsidizes their unprofitable services.

Why it doesn’t hold up

Brian walks through the argument piece by piece, and very little of it survives contact with the data.

• Cherry-picking is a risk-adjustment problem, not a business-model problem. If the patient mix at one facility is sicker, pay that facility more. Medicare does this constantly. We chose to ban a business model instead of fixing a payment formula.

• Half of physician-owned hospitals were full-service community hospitals with emergency rooms, OB-GYN, general surgery, primary care, the works. The other half were specialty hospitals, and even there the cardiac POHs are operationally full-service (you can’t really cherry-pick when people walking into your ER are having a heart attack).

• Specialization is a feature of competitive markets, not a flaw. And it’s allowed for everyone except physicians. Imagine if we didn’t let lawyers own their law firms. Or workers own parts of their own businesses.

• Post-ACA POHs that opened without Medicare went bankrupt. A 2016 Health Affairs study tracked POHs that tried to operate after the ban. Only using commercial patients wasn’t enough. After all, Medicare patients are an enormous part of the market for medical care.

• The cross-subsidy argument is, in Brian’s words, “absolute hogwash.” Running a business on the premise that one profitable line covers a permanently unprofitable line isn’t running a business, it’s running a fiefdom. Hospitals can and should make every line of service viable on its own. And if they can’t, allow competition elsewhere to provide the service.

Ultimately, the ban on new physician-owned hospitals is an anti-competitive measure that has further consolidated the market and driven up prices for everyone.

What Congress (or CMS) could do to fix it

Brian lays out a menu of policy fixes, ranked roughly from modest to ambitious:

1. CMS rule making to expand the existing “high Medicaid need” exception. Real but small, and reversible by the next administration.

2. A rural carve-out that would repeal the ban in geographies with a minimum drive time to the nearest hospital.

3. A joint-venture carve-out that would allow physician-owned specialty hospitals if they include a non-physician hospital partner.

4. A community-hospital carve-out that repeals the ban for physician-owned hospitals that meet the Medicare definition of a full-service hospital (24/7, ER, etc.).

5. A full repeal of Section 6001, optionally paired with a small Medicare rate haircut to address residual risk-adjustment concerns.

6. Stark Law reform in managed care settings. This would let physicians self-refer to their own facilities when there’s already utilization review and prior authorization to police induced demand.

The bipartisan case for allowing new physician-owned hospitals

The fun thing about this issue is that it scans differently to different audiences. To a market-oriented Republican it means more entry, more competition, lower prices, and more small-business formation. To a more progressive office, it’s that workers (the doctors, nurses, and pharmacists actually delivering the care) should be able to own a piece of the business.

The cost of inaction, meanwhile, has been a decade-plus of hospital consolidation. With ownership off the table, independent physicians have one realistic exit: sell their practices to the big hospitals. The hospital systems that lobbied to ban physician ownership are the same systems that have absorbed those practices and raised prices on the way.

Ultimately, this is yet another case of politicians trying to lower prices by centrally planning the market, but ultimately just artificially stifling supply that would otherwise be happily provided. The results are obvious.

What’s in the episode

[0:16] Introducing Brian Miller: New Hoover visiting fellow, associate professor at Johns Hopkins, MedPAC Commissioner, and NC State Health Plan vice chair

[2:40] What Section 6001 of the ACA actually does and why it ended up in the bill

[4:23] The cherry-picking argument and why risk adjustment is the right tool

[5:51] A tour of the physician-owned hospital market: community POHs vs. cardiac vs. orthopedic vs. general surgical specialty hospitals

[8:15] Why specialty hospitals deliver higher quality at lower cost

[10:05] What happened to the POHs that tried to operate without Medicare after the ban

[14:35] The market already segments, just not under physician ownership

[15:49] Why the cross-subsidy argument is “absolute hogwash”

[25:38] Policy menu: what CMS could do via rulemaking, and why it’s not enough

[27:14] Statutory options: rural carve-outs, joint ventures, community-hospital carve-outs, full repeal

[28:54] Stark Law reform in managed care settings

[32:01] Is this a partisan issue? (Less than you’d think…)

[34:11] How the ban accelerated hospital consolidation and the corporate employment of physicians

[37:30] What happens after repeal? Timelines, joint ventures, and why incumbents should welcome the competition

Send us your feedback or policy questions by emailing feedback [at] 584vetoes.com or finding me on Twitter @TomVChurch and Brian at @DrBrian4Health

In other words, someone who earns $60,000 a year (wage-indexed over their career) receives $3.44 in benefits for every dollar in Medicare payroll taxes they paid. Medicare recipients break even on what they pay into Medicare once they average $230,000 a year in earnings.

About 95 percent of people who receive Medicare get more out of the program than they paid into it. Is it any wonder why there’s a structural deficit there?

Full column here.

But coverage doesn’t equal access. Medi-Cal patients face long wait times, fragmented care, and too often end up in emergency rooms for problems that could have been handled by a primary care doctor if one were available.

States have the power to experiment through Section 1115 Medicaid demonstration waivers. These waivers let states test other approaches while remaining budget-neutral to the federal government.

We believe there should be a lot more experimentation among the states. Even ideas that seem unconventional deserve a shot if they’re fiscally responsible and could improve access or the quality of care.

In that spirit, we submitted to the California Health Care Foundation as a part of their recent “Call for Bold Ideas to Strengthen and Preserve Medi-Cal.” They didn’t select it for further development, but we still think it’s an idea worth exploring.

Here’s the Idea

Allow Medi-Cal enrollees to redirect a small portion of the capitated payments made on their behalf - about $80 per month - to a Direct Primary Care (DPC) membership. Do it through a pre-paid debit card with as few strings attached as possible.

DPC practices operate outside the traditional insurance billing model. For a flat monthly fee, patients get direct access to a physician - often via text or email, with same-day appointments, and 24/7 availability for non-emergency issues. DPC doctors typically have far fewer patients (around 500, compared to thousands in hospital systems), which means more time and attention per patient.

Under our proposal, Medi-Cal managed care organizations (MCOs) would continue to handle specialty care, hospital services, prescriptions, and behavioral health. But responsibility for primary care would shift to DPC practices that could offer the kind of access Medi-Cal patients currently lack.

The pilot would operate as a voluntary lottery, run for three years, and be rigorously evaluated on access, utilization, cost, health outcomes, and patient satisfaction. If it fails, it ends cleanly. If it works, it offers a scalable model for improving primary care access in Medicaid.

The Way It Could Save Money

Let’s use some round numbers. If the state pays an MCO $8,000 a year per person to provide them with Medi-Cal coverage, the MCO then budgets out where that money is expected to be spent. Let’s say $1,000 of it goes to primary care and $6,000 of it goes to other spending like inpatient and non-primary outpatient spending.

But if that enrollee were given $80 a month ($960 a year) to enroll in a DPC membership and had their primary care needs taken care of that way, would they end up using less than $7,000 in other Medicaid-covered services?

After all, the way that DPC saves money is by reducing unnecessary utilization of higher cost settings like emergency rooms or urgent care centers (PDF).

Why This Matters

Medicaid’s fiscal challenges won’t be solved by a single policy. They’ll be solved by states trying many different approaches and learning what works. That requires a willingness to experiment, and a willingness to share ideas even when they don’t get picked.

The full proposal is below for anyone interested in the details.

Allow Medi-Cal Enrollees to Use a Portion of Capitated Payments for Direct Primary Care (DPC) Memberships

1. Authors

1. Tom Church, MPP, Policy Fellow, Hoover Institution.

2. Anthony DiGiorgio, DO, MHA, Department of Neurological Surgery, University of California, San Francisco; Philip R. Lee Institute for Health Policy Studies, University of California, San Francisco.

2. The Idea

Create a pilot program that allows a set of Medi-Cal enrollees to redirect a small, fixed portion of their capitated Medi-Cal spending to a Direct Primary Care (DPC) membership. Managed care organizations (MCOs) would remain responsible for all other non-primary care and retain the rest of the capitated payment. But enrollees who use funds to obtain a DPC membership would gain access to greater primary care coverage, allowing better access and quality of care without increasing costs to the state or federal government.

Instead of the state sending around $8,000 a year to an MCO to pay for a Medi-Cal enrollee, the pilot program could send $7,040 to the MCO and provide up to $960 ($80 a month) to the enrollee to pay for the DPC membership.

The goals are to increase access to primary care and lower wait times by unbundling primary care from costly hospital-centric systems, all while driving long-term spending lower. Fiscal savings would occur if the cost of non-primary care capitated payments handled by the MCOs were to fall because of changes to utilization and better health management. This Section 1115 Waiver program would operate as a voluntary lottery, run for three years, and be evaluated on access, utilization, cost offsets, health outcomes, and patient satisfaction.

3. The Problem Being Addressed

For many enrollees, primary care is inaccessible even if it is nominally “covered.” That leads routine issues to eventually escalate into urgent or emergency matters.

By using capitated payments, the state limits how much it spends a year providing medical coverage. But when money is limited, as it is in Medi-Cal, patients pay with their time. This leads to relatively long wait times for appointments, fewer providers accepting Medi-Cal patients, fragmented care with little continuity between patients and their physician, and an overreliance by patients on hospital systems and emergency departments for care that could be provided in a lower cost setting.

Direct primary care memberships are characterized by timely and direct access to the physician without any required insurance billing. They typically cover preventative care, vaccinations, care coordination, basic labs, and discounted prescriptions. DPC physicians typically have far fewer patients (~500 instead of thousands in hospital systems), respond 24/7 to text or email, and are available after-hours as a substitute for non-emergency care currently provided at late-night ER visits when no other facilities are open.

4. The Benefits

Enrollees would experience dramatically increased access to primary care, continuity of their care with a single clinician or practice, and reduced reliance on urgent care and emergency departments. The Medi-Cal program would experience lower hospital-billed primary care, better chronic disease management, and greater capacity to treat higher-priority patients. Physicians would see reduced administrative burdens, billing complexity, and higher satisfaction. Populations that require frequent monitoring or non-emergency visits would particularly benefit from a more active physician-to-patient relationship.

5. The Viability

This proposal is both administratively and fiscally viable because it operates entirely within Medi-Cal’s existing managed care structure and uses a fixed, capped reallocation of current spending. DPC practices require no additional state permitting, direction, or funding.

One limit on this program will initially come from the number of physicians who are available to run DPC memberships. That limit availability will allow for a lottery to be run if more Medi-Cal enrollees enroll than there are DPC practices available.

To prevent double payment and gaming, participating DPC providers would be required not to bill Medi-Cal fee-for-service for primary care services covered by the DPC membership and be prohibited from additional membership tiers or add-on fees for Medi-Cal enrollees. Managed care plans would remain fully responsible for specialty care, hospital services, prescription drugs, and behavioral health. DPC providers would function as front-line primary care, not as substitute insurers.

As a Section 1115 Waiver, this experiment would meet all four required criteria: It aligns with Medicaid’s objectives, it tests a policy hypothesis, it is budget neutral to the federal government, and it would lead to a clear and credible evaluation of the experiment.

If the pilot fails to improve access or worsens total costs, it can be terminated cleanly. If it succeeds, it offers a scalable way to improve primary care access while reducing reliance on hospital-based delivery systems.

From 2014 to 2025, the number of ACA enrollees who opted for HSA-eligible plans declined from 8 percent to just 2 percent. The steep decline was due to a technical (and likely unintended) difference in indexing between HSA-compliant plans and ACA plans. In short, maximum-out-of-pocket (MOOPs) limits rose faster among ACA plans than HSA-compliant plans. That meant that the number of HSA-eligible ACA plans was quickly approaching zero; from 2014 to 2025, the share of eligible plans fell from 11 percent to 2 percent.

The red bar in our plot represents quite the change. If we assume the same share of enrollment in catastrophic and bronze-level ACA plans as in 2025, then the number of enrollees in HSA-eligible plans will jump to over 30 percent. The change means more than 5 million Americans now have greater access to HSAs. It could be even more if the reform encourages individuals to switch from higher-premium silver plans to a bronze plan.

That’s from Danny Heil and me at the Hoover Institution’s Freedom Frequency Substack. Link here.

I saw Alex Berenson post this thread:

$50,000 a year for premiums is too much. The median family income is just above $100,000, meaning they’d spend half of their income just to have health insurance.

How in the world did insurance get so expensive?

Aside from the fact that I don’t believe $50,000 plans are at all the norm in ACA individual exchanges (although $30,000 family plans are quite common), I can think of two contributing reasons:

1. The risk pool isn’t very healthy, because a healthy risk pool lowers required premiums for everyone.

2. Insurers cover every single medical expense, which means they have to raise premiums to pay out expected claims.

Let’s dive into these two ideas.

1. A Risk Pool Without Lots of Healthy People Leads to Expensive Insurance

Let’s run some numbers to see why.

Example 1: Really healthy Risk Pool (90% Healthy)

Imagine there are ten people who buy insurance in a risk pool. None of them expect to be sick and all are generally healthy. Nine of them have low-cost years and end up spending $1,000 on medical care. The tenth has a bad year and has medical costs of $11,000. How much should should an insurer charge them for premiums at the beginning of the year?

In our overly simplified example, premiums come to $2,000 per person.

Example 2: Unhealthy Risk Pool (60% Healthy)

Let’s make more people have medical events. Now forty percent of the risk pool has costs that total $11,000 in a year.

An insurance company would rightly increase premiums by $3,000 (up 150%) in order to insure this simplified risk pool.

Example 3: Catastrophic Year Risk Pool ($100,000 Covered Expense)

Let’s go back to a healthy risk pool where ninety percent of covered individuals only have $1,000 in medical costs. But one person requires a life-saving procedure or medicine and has costs of $100,000. What happens to premiums then?

Now we start to see how insurance can get so expensive and why it is crucial that as many healthy people as possible buy health insurance.

You can also start to see why healthy people (who tend to be younger) aren’t as interested in purchasing health insurance. Paying $9,900 more than their annual cost in this example as opposed to $1,000 more than their annual cost will make more of them consider forgoing insurance in the first place.

Lastly, let’s see how uncertainty about expenses increases premium costs, because insurers need to build in the possibility of losing a lot of money.

Example 4: A More Real Life Example of How Premiums Are Determined

Let’s take our existing three scenarios and find out what premiums should look like if they’re all equally likely to happen.

This is the most interesting of the results, because you can infer a lot about premium pricing.

If an insurer charges annual premiums of $5,967 to these ten people, they could range from making $39,667 in example one, to making $9,667 in example two, to losing $49,333 in example three. On average, it’s a 16 percent profit margin. (UnitedHealth’s gross, operating, and net margins from 2011 to 2025 are here.)

What would you do if you were an insurance company, knowing that if you priced things wrong you could operate deeply in the red? My guess is that you overweight the possibility that expenses are high and then charge higher premiums.

If instead of equal probabilities you instead assume likelihoods of 25 percent, 25 percent, and 50 percent for examples 1, 2, and 3, then required premiums jump to $7,200 a year.

Note that these numbers are purely illustrative - I picked them out of thin air. But I think they do a good job of getting people to wrap their heads around insurance pricing. And a sample size of ten means the variance is huge. Once you start running numbers based on previous medical claims with tens of thousands or hundreds of thousands of people, you can see why insurance companies’ profit margins don’t swing by enormous amounts.

Of course, we’ve been operating under the assumption that insurance only covers expenses that need to be insured. There’s another reason why health insurance has gotten more expensive.

2. Health Insurance That Covers More Services Leads to More Expensive Insurance

Almost every medical expense goes through health insurance, but it shouldn’t.

One of the complaints among those of us who like markets and less government intervention in health care is that health insurance doesn’t act like insurance anymore. It functions as pre-payment for all medical care.

Insurers are the intermediary for almost every medical event or coverage, and it leads to a situation where patients never see prices and end up paying much more for the sum of their medical care than they otherwise would.

The ACA added the ten categories of Essential Health Benefits to all individual and small group plans. It also basically killed the catastrophic health insurance market. It told insurers that they would have to cover more and more services if they wanted to offer insurance at all. As a result, premiums are higher because insurers are on the hook for more possible costs.

As an example, let’s say there are two insurance plans. The first operates like insurance plans are required to operate now: It covers routine, predictable, and low-cost care like preventive visits, simple blood panels, ongoing medication, common outpatient procedures, and many other mandated services.

The second operates like car insurance: Routine, predictable and low-cost care is paid out of pocket and not through insurance. It’s the equivalent of paying for tanks of gas, oil changes, and tire rotations out of pocket but having insurance for when your car gets hit by another car.

Which one is going to have higher premiums? Obviously the first plan. Covered expenses are expected to be higher, so premiums will necessarily be higher. And since premiums are higher, expected out-of-pocket spending is lower, leading to less price sensitivity.

Your objection is probably, “Yes, but if you remove coverage, then people are going to face higher bills.” I would contend that’s not true if what is removed are the things that shouldn’t be covered by insurance in the first place, and that people will naturally shop for better than a third party payer would.

Not all preventative measures save money. In fact, the ones that do are the ones that insurance companies are going to voluntarily cover if they’re not required to in the first place. If we mandate that everyone should receive annual MRIs to check for early tumors, we’d certainly find more early cancers. But resources are scarce and it seems entirely possible (based on how expensive MRIs are) that the cost would outweigh the benefits of early detection from a cost perspective.

Another thought experiment, brought to you by Anthony DiGiorgio: How much more expensive would your car insurance be if it covered gasoline, oil changes, and new tires? And how long would the wait be to get new tires when you got a flat and needed a new one? If the government, or your insurance company, were the ones handling payment, there is no way that the average tire cost would go down.

Takeaways:

Our goal should be a functioning, competitive health insurance market with lots of options, really broad risk pools, and coverage for medical expenses where insurance makes sense.

Subsidizing insurance companies (which makes sick people appear healthy, because they cost less) is one way to drive down the cost of premiums. Another is to broaden the risk pools. And another is to reform health insurance so that it doesn’t cover the equivalent of tires and oil changes.

Here’s one tweet that’s thematic of my timeline the last few days:

That leads me to two main questions:

1. If people were to go uninsured but set aside their premiums and annual deductibles or out-of-pocket maximums each year, how would they do under various medical event history assumptions?

2. What would a health insurance system that actually functions as insurance look like?

1. What would happen if instead of buying health insurance, you set aside annual premiums plus your deductible every year and paid out of pocket?

To be clear, I’m not advocating this. I’ve just always wanted to run the numbers. Is there a world in which it makes sense to avoid insurance all together and just pay out of pocket? Even when medical events get expensive? How often and how large do the insurance shocks have to be to be worth it to buy insurance?

Let’s stipulate a simple starting point:

• I’m mapping out a 26 year old for 35 years to age 60. To keep it simple, I’m going to assume they stay single the whole time. They work at a large firm that offers employer-sponsored health insurance.

• Premiums keep increasing at 6% a year. Deductibles and out-of-pocket maxes also go up 6% a year. Investments go up 4% to account for a conservative asset mix.

• This Commonwealth fund report (PDF) has premium costs in 2023 for single and families, split by employee and employer contribution. It also has deductibles. I’m going to assume out-of-pocket max in year one is $5,456 per KFF.

• We’ll waive our hands a bit and say that the employer-side contribution could be negotiated to be given to the employee instead of being covered by insurance. Granted, it should be reduced a bit by the percent of the tax advantage, but we’re keeping it simple.

You may object and say, “A cancer diagnosis costs a million dollars, which would wipe out any fund.” Correct! I’m not advocating people go without insurance. But what matters for running this thought experiment is the average expense per year.

What I’m showing you is why health insurance, in its current format, looks like a bad deal to so many people. The unanswered question is how many people?

Let’s get into three scenarios, from unlikely to more likely.:

1. Unlikely: Medical expenses each year total 75% of the individual’s deductible.

2. More reasonable: Medical expenses each year total the out-of-pocket max.

3. Relatively reasonable: Medical expenses each year total the full deductible, but every five years they total five times the out-of-pocket max as a result of a major medical event.

Over a lifetime, this is not a very realistic scenario. People have big medical expenses every once in a while.

Not surprisingly, this individual would have been vastly better off never purchasing insurance. After all, they pay a premium but never reach the cost-sharing portion where their insurer handles some of the cost.

You’ll need to hit “+ Show 15 more” to see the ending investment balance result, but it’s over a million dollars. That’d be quite a balance to take into your Medicare years.

Unsurprisingly, this is also a scenario where insurance is much more costly than being covered. An insurance company would typically handle some of the cost sharing above the deductible, but it’s still less than the premium and deductible combined.

This is the more realistic scenario. (I know it fits my medical history: Three somewhat routine surgeries in fifteen years.) You can play around with the numbers to get to a net zero investment balance or put together a few major expense years close together to make it go negative, but the numbers are illustrative.

2. What would health insurance that actually functions like insurance look like?

Let’s remind ourselves what insurance is supposed to be for: Large, unexpected, costly events. You pay premiums to cover those costs and pay out of pocket for routine, inexpensive, and predictable costs.

Car insurance is the classic example: You pay insurance for when you get into a major accident. You don’t use your insurance to pay for routine and expected costs like gas, oil changes, and tire rotations.

Let’s also remind ourselves how health insurance differs from normal insurance: It covers routine, inexpensive, and predictable costs. In fact, health insurance is mandated to cover those things (see: the Affordable Care Act). Routine doctor visits, laboratory services, preventative wellness visits, and vaccinations are mandated as Essential Health Benefits (EHBs) that all individual and small group health insurance plans must cover. Various preventative measures have no deductible to enrollees.

The unpopular truth is that health insurance in America covers too many things. Health insurance is largely about pre-payment of care. All health care shouldn’t go through health insurance. Premiums and deductibles are high because everything goes through insurance.

What would a health insurance system that looks like car or home insurance look like? I’d say one with much lower premiums, higher out-of-pocket cost-sharing, and everyday expenses like routine doctor visits, regular lab panels, vaccinations, and generic prescriptions (high blood pressure, statins) handled directly and not covered by insurance.

More health insurance would be catastrophic in nature. The above examples demonstrate a person going without health insurance and saving the money instead. How about instead of going without, they purchase catastrophic insurance, pay much lower premiums, and pay up to a higher out-of-pocket max?

By necessity, there would be less third party payment and much more direct billing to consumers. As scary as that sounds for people, it’s the way that would actually harness the tried-and-true competitive pressure that leads to lower prices and higher quality.

I know I’d rather pay for a Direct Primary Care membership, have catastrophic insurance that covers medical spending above a certain dollar value, and contribute to an HSA every year in the meantime while I’m healthy.

This sort of system requires a more permissive health insurance market that allows insurance companies to remove large parts of mandated coverage. After all, why mandate what type of coverage is needed? Why not just set allowable dollar values and let people get whatever coverage works for them that year?

When you sign up for car insurance, you get a menu from the insurer that allows you to change your coverage amount for various categories (comprehensive, collision, bodily injury) and see how it changes your premium in real time. I’d love to have much more control over health insurance options. And I’d have to think insurance companies would love to offer them.

In short, more sensible health insurance would:

• Cover less low level medical procedures, office visits, and prescriptions

• Function more like catastrophic insurance that kicks in above a high dollar threshold

• Have relatively low premiums and encourage dedicated saving for health expenses

• Be customizable to a much greater extent than current options

Nearing the Breaking Point

The numbers above are just for a single individual. The magnitudes are greater when running them for family plans, and they get sillier when you start to use the premiums charged to ACA individual market plans.

It’s not that health insurance should be a winning proposition for everyone. By definition insurance is on average net negative for people buying it. But when it’s extremely net negative for so many people, then there are big improvements to make.

Insurance should protect against risk, not guarantee consumption through pre-payment. It seems like we’re getting to a breaking point where health insurance, as it currently works in the United States, is due for a major overhaul.

The recent changeover from a Democrat to a Republican president shows how partisan affiliation leads people to ignore any principles they believe they hold in order to justify their side’s policies. It happened when Biden took over in 2021 and it’s happening now that Trump is president again. Just look at how Republicans are treating tariffs.

Charlie Munger called this the “Liking/Loving Tendency” in his amazing essay “The Psychology of Human Misjudgment.”

We excuse mistakes from those we like and we magnify the flaws in those we don’t. That’s a mental shortcut that leads us astray from time to time.

For most voters, aligning with your party is probably rational. There are too many issues in politics to learn about and life is busy. But if you’re actually going to be involved in the political debate and want to remain internally consistent, you’re going to run into trouble if you outsource your thinking to a political party or politician.

Hence, declare your allegiance to principles, not people or parties.

Writing out your principles helps to inoculate you from being swayed by the issue of the day. It encourages accountability and gives you an easy way to judge issues as they come up that you’re susceptible to being biased toward.

That does mean that you’ve got to make clear what your principles are. While not an exhaustive list, I’ll provide a few of mine.

Big Picture Principles

Personal liberty stems from negative rights, not positive rights. Our individual rights are “rights from” government interference, not “rights to" goods or services. Why? Because the “right to” something generally requires others to give us something that has been taken from them without their consent. Approaching rights this way is more likely to restrain the government from overreach. (See Isaiah Berlin’s two concepts of liberty for more.)

The worst things in history that happened due to humankind were done by governments or groups with a monopoly on the use of force. That is why restricting government power is so important. “Small government” is not a call for the government to leave people out on the street. It is a historically-inspired preventative measure to stop the government from doing terrible things to people.

The rule of law is paramount. The rule of law is when the law applies to everyone equally, no matter how powerful or connected they are or whose side they’re on. You never know who is going to be in power, so you want a system that restrains your worst enemies from abusing that power. Even if it means restricting your side when you’re in the majority.

Free markets lead to prosperity. We should prioritize economic growth above most other priorities. Mandates, restrictions, and legal hurdles often reduce the rate of economic growth. Free-market fueled economic growth is the greatest anti-poverty program in history. Nothing comes close, so slowing down economic growth should be weighed against the people it will save from poverty in the future.

The end goal of foreign policy is for all conflicts to be resolved peacefully, which is done by increasing the costs (implicit or explicit) of resolving conflicts using force. War is politics by other means. Countries go to war because they believe they can get what they want at an acceptable cost. “Peace through strength” is about raising the cost to would-be aggressors so much that they determine it’s not worth it to fight, and instead defer to non-violent means to achieve their goals.

Procedural justice leads to a fairer system marked by the rule of law, even if it means some individual outcomes don’t seem fair. Innocent until proven guilty is an offshoot of this principle. Someone who broke the law but wasn’t read their Miranda Rights should be let go, because it leads to a system where police have to follow the law to detain people. That leads to trust in the system and it prevents abuses of power. This is also why Pierson v. Post is a key reference point in my mind. The long-term effects of a decision need to be weighed against the “fairness” of one case.

Allowing dissent and organization by opposition parties is crucial to the survival of liberal democracies. The extent to which a country protects minority rights against the tyranny of the majority is another key indicator of a liberal democracy.

Economic Principles

Three lessons to start. First, economic goods are scarce. Goods and services are finite.

Second, everything has a trade-off. As Thomas Sowell says, “There are no solutions, only trade-offs.” If you use your money to buy one thing, you can’t use it to buy something else. If you spend your time on something, you can’t spend that same time on something else.

Third, prices are the best way to determine how to allocate scarce goods in the face of trade-offs. Prices are a signal wrapped up in an incentive. They help organize the economy, which is unfathomably complex. Prices are determined by supply and demand, and can be monetary in nature or not. Effort is a price, in a way, which is why a person’s utility function includes all sorts of non-monetary values that nonetheless reflect the price of that thing.

An important corollary is: “Things go badly when you mess with prices.” Any time you mess with the price mechanism, you get an inefficient outcome. Supply and demand can be artificially changed by imposing mandates like price ceilings or floors. Messing with the price mechanism is like fighting against gravity when all you want to do is roll a ball downhill. Just let it go.

Competition is the only force that reliably leads to lower prices and better quality. Competitive markets improve and innovate because they have to in order to survive. Markets without competition stagnate are the result of legal barriers to entry. There are too many examples to count.

Comparative advantage and trade lead to positive-sum growth. Trade is good. Free trade is better. Voluntary exchange is obviously good because it can only happen if both sides believe they’ll be better off. Comparative advantage really clicks when you see the numbers for yourself.

“Capitalism” is just the name we give to the answer to this question: Should economies allocate scarce goods through voluntary exchange or central planning? Blaming “capitalism” for the ills of society is historically illiterate and conflates violations of the rule of law (and rent seeking) with the positives from competition and voluntary exchange. On one side is socialism and central planning. On the other is capitalism and voluntary exchange. The historical record has plenty of examples of both.

There is good inequality and there is bad inequality. Fair doesn’t mean equal. Gains earned from personal success in a free market aren’t a bad thing. Steph Curry makes a lot more than a benchwarmer on the same team, and we all know why. But making money because you’re politically connected and can use the government to prevent competition is rent seeking and should be eliminated as much as possible.

Rules of Thumb For Public Policy

Every political program inevitably expands to cover individuals or benefits that it was not originally intended to cover. See: Cogan’s Equally Worthy Law. Before you support the creation of a new program, requirement, or benefit, it is worth first asking if you would support it if it were ten times as costly and benefitted people you think would never receive it.

Director’s Law is closely related: Public programs are designed primarily to benefit the middle class, but are paid for by the poor and the rich. Government programs rarely shrink.

Things go wrong when people create costs they don’t have to pay for. Obvious and visible everywhere, once you know it. Creating costs without being responsible for them leads to inefficient allocations of time, resources, or money. In every day life, it’s the manager who tells his superior that some work will be done and then has their subordinates work late, only to not need what was done in the morning. Look for wastes of people’s time, and you’ll often find other people creating costs they don’t have to pay for.

When in doubt, be on the side of the consumer. You’ll be right about 90 percent of the time. Don’t know what to think about tariffs? They raise prices for consumers, which means they’re probably not good. Should we replace an old gas station with a 60 unit apartment complex? Seems like a good way to offer people more places to live. Are charter schools a good idea? Well, why would we stop parents from sending their children to the school of their choice? Do what’s good for the consumer. Objections often handled here.

Make it easier to add supply. Don’t subsidize demand. Subsidizing demand exacerbates the problem. In response to high housing prices, some politicians have proposed $10,000 benefits for first-time homebuyers. Take a quick guess as to what would happen to the price of houses if that program went through? Adding supply is most often prevented by government rules that protect incumbents. Environmental reviews and permits that take years lead to “affordable housing” that costs a million dollars a unit.

“Make progress by letting people do things” is closely related.

If you want the government to help people, direct subsidies are the most efficient way to help them. Congress prefers instead to use cross-subsidies, where they mandate lower prices for one group but have to prevent competition by others if they want to do it without spending any money. As John Cochrane would say, do it on budget.

Trust people who are willing to criticize their own side. I loved Hitchens’ telling of John O’Sullivan’s joke: “If you hear the Pope say he believes in God, you think well, the Pope’s doing his job again today. If you hear the Pope saying he’s really begun to doubt the existence of God, you begin to think he might be onto something.” Right now, those who are right-of-center and willing to criticize the administration’s actions have a chance to increase their trustworthiness. Few are succeeding. It will be the left-of-center’s turn the next time a Democrat is president.

Apply the Golden Rule liberally. It’s probably the one I see most violated. If you would react differently to something if the other side did it than how you’re reacting when your side does it, you know you’ve gone astray.

Limits on government power are important because anything you let your side do, your opposition will take even farther. And that’s probably not a world you want to live in. Many smart and earnest Democrats are in favor of single payer healthcare or Medicare for All. I don’t think they would enjoy having President Trump and RFK Jr. in charge of the entire healthcare system in America. That alone is probably worth it to them to hold back.

Falsifiability is the way to find the truth. The only way to reliably discover the truth is to make falsifiable tests, run those tests, and iterate with new falsifiable tests. Constantly ask, “How would I know if I were wrong?”

No final say, no personal authority. Jonathan Rauch’s book The Constitution of Knowledge introduced me to Charles Sander Pierce’s great maxim: No final say, no personal authority. It means: (1) No one group gets to declare questions about a topic off limits or completely settled, and (2) The truth of a claim cannot depend on the authority of the person making the claim.

The existence of a possible negative consequence is not a sufficient reason to oppose a policy. “It could lead to a bad thing” is a starting point of an argument, not an ending of one, as I wrote here.

There are so many more principles to cover, especially within major topics like health care and immigration, but I’d say this list a good start.

Everyone likes to believe they are fighting for their principles. But the real test occurs when those principles put you at odds with your own side.

Principles, not people or parties.

But the existence of a possible negative consequence is not a sufficient reason to oppose a policy.

Its a starting point, but not the end of the conversation.

In the housing market, you’ll hear opposition along the lines of: “This new taller building is not in keeping with the character of the neighborhood.” The harm: The neighbors don’t like it and come up with flimsy excuses like “people could drop beer bottles off a balcony.” Meanwhile, preventing new supply means prices keep going up.

The port worker unions argue that automating our ports would lead to job losses, therefore we shouldn’t do it. Meanwhile, American ports are some of the costliest and least productive in the world.

Proponents of tariffs say, “Free trade hurts American manufacturing.” Meanwhile, protectionist tariffs increase costs and slow down the economy.

Teachers unions don’t like competition from charter schools and say, “Students that go to charter schools pull dollars away from public schools, making them worse off.” Meanwhile, parents desperately want to send their children to the best school available—or just away from terrible public schools.

Opponents of legal immigration say “More foreign workers means lower domestic wages.” Meanwhile foreign talent fuels start ups, economic growth, and ultimately many more jobs.

Healthcare is rife with examples. Hospitals argue against site-neutral payments in Medicare because it would “jeopardize access to care and services for patients.” Meanwhile Medicare overpays for tons of services done in hospitals that can be done just as safely in ambulatory surgery centers or physicians’ offices for lower cost.

Rep. Greg Murphy (R-NC) recently set off a firestorm when he suggested foreign medical graduates should be used to fill the shortage of primary care physicians at rural hospitals. The reaction was to call instead for more domestically trained physicians and general worries about patient safety based on lower training standards in foreign medical schools.

(The corollary to this worry is, “Your concerns can be handled in the legislation that authorizes the thing you’re worried about.” But that’s something to cover in a future post.)

What are you optimizing for? Maximize the effect on that priority.

Everything has a trade-off: Every decision, every policy option, every regulation. Whenever you’re trying to address a policy problem, you have to decide what you’re going to prioritize over everything else. Then you measure policy against doing the thing that works for that priority.

If the main problem you’re solving for in housing is that it’s too expensive, your main priority and measure of success is lowering the cost of housing. How do you get that? You increase supply like Austin, Raleigh, Dallas, or Houston. Whatever you do, don’t prevent housing like San Francisco, Los Angeles, or San Jose. Existing homeowners and NIMBY types aren’t going to like it. But we should be optimizing for everyone’s welfare, not just existing homeowners.

If the main problem is that our ports are inefficient and expensive, and automated ports are more efficient and less expensive, then do the thing that works and automate the ports. Even if it leads to the negative consequence of fewer longshoremen. We don’t want to optimize for the number of longshoremen jobs. We want to optimize on the lowest cost way to import and export at our ports.

If the main problem is that American manufacturing is declining, you shouldn’t increase the cost of inputs on American manufacturing (as tariffs are doing). In fact, you should do everything you can to lower the cost of inputs to make manufacturing more competitive.

If the main problem is that student achievement isn’t increasing rapidly enough, then don’t keep throwing good money after bad. We’re optimizing for improvements to numeracy and literacy. Do the thing that works, even if teachers unions aren’t on board.

If the problem you’re trying to fix is that there aren’t enough primary care physicians (or physicians in general), then do whatever it takes to increase the supply of qualified doctors. Even if that means making it easier for those who trained or worked overseas to work here while you fix the current system. We’re optimizing for lower costs and lower wait times for everyday Americans.

To make a decision, you need an honest accounting of the trade-offs of both options. Option A: Keep doing what we’re doing. Option B: Make this policy change. For each option, list the benefits and the costs. If the ratio of the benefits > costs is higher for option B, then move forward. If not, stick with option A.

The acknowledgement that some people will be made worse off is a nod to Kaldor-Hicks efficiency

A reminder: Pareto efficiency is where a change is made and everyone is either better off or no worse off. In the real world, I’d say that’s rarely possible in politics.

Kaldor-Hicks efficiency is where a change makes most people better off, who can use those gains to compensate those who are made worse off. It feels like the real world application of the ideal policy improvement that would happen with Pareto efficiency: A change happens where some people are made better off and no one is made worse off.

How do you actually make progress in the real world? I think it helps to identify the policy issues where 10 (or 1!) percent of people are benefitting at the cost of the other ~90%. Leaving the 10% to benefit while the 90% live in a sub-optimal system is folly. Improve the system in the long-run for everyone else.

Every reform has losers. We need to accept small and fixable harms to make the system better, and get over the fear of possible negative consequences that freeze us into even worse long-run policy.

It would have been even larger, except Congress offset a portion of the bill’s budget effect by slowing down the growth in Medicaid spending. Among other Medicaid reforms, Congress enacted new limits for state provider taxes, work requirements for able-bodied adults, and more frequent redetermination requirements. CBO projected these reforms would reduce Medicaid spending by $910 billion over the next decade.1 Combined with other reforms to federal health programs, federal health care outlays would fall by more than a trillion dollars over the next 10 years.

A trillion dollars is hard to comprehend. It certainly seems like a big number. Recent news coverage implies it will have a significant effect on America’s health care system. Just last week, the Wall Street Journal noted that health care jobs were a “rare bright spot” in the recent jobs report, but their subtitle warned that “One danger: Coming Medicaid cuts pose a threat to health services.” The major medical coalitions—from the American Hospital Association to the American Medical Association—have similarly warned about the dire consequences of OBBBA’s health care changes.

So how big is a trillion dollars? In almost all cases, the answer to the question would be “a lot.” But the answer is more complicated when it comes to nation’s health care budget.

When Is a Cut Not a Cut?

The trillion dollar cut is from CBO’s January 2025 current law baseline. That baseline reflects CBO’s best guess on how Medicaid will grow if there are no changes to current law. CBO had projected in January 2025 that Medicaid spending would rise by 50 percent from 2025 to 2034 (or 23 percent after adjusting for inflation). The figure below shows Medicaid spending with and without the OBBBA.2 The area between the lines is the $910 billion cut that has garnered dire warnings about the future of American medicine.

The definition of “cut” itself is debatable. One option is to measure it as the inflation-adjusted change from CBO’s baseline. Because Congress backloaded the reforms (fewer than two percent take effect in 2025–26), the cut in 2025 inflation-adjusted dollars is only $792 billion.3

But even that framing is misleading. A “cut” usually means getting less tomorrow than today. Medicaid spending keeps rising year-after-year, even after the “cut.” Nominal Medicaid spending grows by a total of $740 billion over 10 years after OBBBA. Perhaps more interesting, once adjusted for inflation, real Medicaid spending is essentially flat, with cumulative growth of just under $20 billion.

Importantly, the base year for this “cut” is unusually high: Medicaid spending ballooned during the pandemic, rising 28 percent faster than inflation from 2019 to 2025—10 percentage points more than CBO had projected in January 2020.

The Effect of Medicaid “Cuts” on Total Health Care Spending

Medicaid is a large government program—it comes in fifth in spending after Social Security, interest payments, Medicare, and defense spending. But Americans spend trillions on health care. In 2024, Medicaid accounted for 17.6 percent of national health expenditures.4 That means that if Medicaid spending were cut by 10 percent, it would only reduce national expenditures by 1.76 percent.

To be fair, the OBBBA contained other changes to federal health programs. If we include those changes and assume that states reduce their Medicaid spending in proportion to the federal changes, OBBBA will lower projected total health expenditures by $1.4 trillion over the next 10 years after adjusting for inflation.5 That is a 2.1 percent reduction from the $64.8 trillion that Americans were expected to spend on health care over the next decade.6 The figure below shows the change in projected health spending with the OBBBA. Be prepared to squint to see the effect.

But, as we discussed above, a cut typically means something will be smaller tomorrow than today. That certainly isn’t the case with national health expenditures. Without the OBBBA, real health expenditures would have grown by 2.8 percent; with the OBBBA, the growth rate will instead be 2.4 percent.

Even this likely overstates the change in spending. We are assuming the loss in government spending isn’t made up for by increased spending by individuals or employers. That’s an aggressive assumption. In 2023, states were required to redetermine eligibility for existing Medicaid recipients that had been allowed to stay on the rolls during the pandemic. At the time, the CBO had estimated that of the 15.5 million that were set to lose coverage, 7.8 million would obtain or maintain employment-based coverage. In short, the 2.1 percent change in 10-year national health expenditures is likely an overestimate.

Hospitals and Providers Do Face Budget Threats

Some hospitals and providers still have legitimate budget concerns. Rural hospitals and providers with thin margins might be rightfully concerned above their ability to maintain services. The OBBBA tried to address some of these concerns with a new rural health fund. Whether lawmakers did enough (or designed it well enough) remains to be seen.

But if Americans are worried about a 2.1 percent reduction in projected health spending under OBBBA, they should be far more concerned about reforms proposed on the left. Plans like Medicare for All or public option plans would require hospital and physician payments fall to rates closer to Medicare levels. In fact, CBO estimated in 2020 that Medicare for All could lower long-term national health expenditures by more than 10 percent—far larger than anything OBBBA imposed.

What isn’t clear is which Medicare reforms we should pursue to achieve meaningful budget savings. There is no free lunch with the program. Sure, there is waste, fraud, and abuse in the program, but there isn’t that much. Instead, significant reforms will have large political, economic, and health consequences for recipients, providers, and insurers.

Who Should Face the Costs of Medicare Reforms?

Even just wrapping your head around possible Medicare reforms is hard because there are so many players involved and so many parts to Medicare. Contrast that with Social Security, where the options boil down to “Cut benefits, limit eligibility, raise taxes, or some combination.”

Within Medicare, making changes to Part A (hospital coverage) doesn’t necessarily fix structural issues in Part B (doctor’s visits), Part D (prescription drugs), or arrangements within Medicare Advantage (supplemental coverage). Without many opportunities for a free lunch, reducing federal Medicare spending ultimately means shifting costs from the government to someone else. Medicare is complex, but there aren’t that many groups to choose from. The big groups are:

1. Enrollees

2. Providers (e.g., physicians, nurses)

3. Hospitals

4. Insurers

5. Drug companies and device makers

Congress could try to focus reforms on any one of these groups, but that doesn’t mean that group will necessarily bear the burden of the reform. There’s a long history of the government trying to shift costs to one group only to find that the economics meant someone else would be paying for it.

For example, there’s a growing effort to reform Medicare Advantage (MA). MA providers have been accused of “upcoding” patients. They diagnose recipients with more conditions—perhaps even imaginary maladies—to extract bigger payments from Medicare. The Wall Street Journal practically has a regular feature discussing how United Healthcare, Humana, and other MA providers have collected billions in extra payments through these schemes.

Stopping these MA payments schemes would save Medicare billions in dollars. But the billions wouldn’t necessarily come from insurers’ profits. The competitive bidding process in MA means that, over the long-term, insurers must pass on the extra payments to recipients in the form of lower cost-sharing requirements, extra services, or other benefits. In fact, the Wall Street Journal recently noted that efforts to restrain MA spending growth have already led insurers to leave certain markets—leaving Medicare recipients with fewer options.

Similarly, the Inflation Reduction Act (IRA) featured several efforts to reduce Medicare drug spending. This included a $2,000 maximum out-of-pocket cap on prescription drug cost sharing in Medicare Part D. Spending above that limit was to be covered primarily by insurers. Unsurprisingly, insurers didn’t simply eat the cost and pay for it. They raised premiums. Wanting to avoid substantial premium increases in an election year, the Biden Administration ultimately provided insurers an extra $15 per enrollee each month.

The IRA also tried to reduce drug costs by directly negotiating drug prices. The aim was to reduce costs for the government and recipients by paying drug makers less. Danny’s research with John Cogan and Casey Mulligan, however, shows drug makers are likely to try to mitigate the price changes by raising launch prices. More than half of the promised savings from the IRA’s negotiation program could prove illusory as a result.

Now, there are legitimate political reasons why these reforms might still make for good spending reforms. In the case of MA, there would likely be a short-term effect on insurers. And over the long-term, recipients might not realize that the MA reforms limiting upcoding are responsible for less generous MA plans. Letting lawmakers hide behind complicated reforms where the long-term effects aren’t obvious isn’t necessarily a bad thing!

Thinking About Big Picture Changes

Like we said, picturing fixes to Social Security and who prefers them is relatively easy. Democrats propose increases on the tax side (generally for high-income earners) and Republicans favor slowing the growth rate of future benefits. Seniors and the AARP are supportive of any increases to benefits.

But what about Medicare? We’re mostly focused on cost-saving options that either reduce payments to various healthcare providers (like doctors, hospitals, insurers, and drug makers), increase cost-sharing requirements for enrollees, or limiting what services providers and insurers are allowed to offer. The key questions for each reform will depend on the group:

• Enrollees: How will the reform affect their premiums and how much are they going to pay out of pocket?

• Providers and Hospitals: How much will the federal government pay them for each service provided? Who will be allowed to provide care and in what form?

• Insurers: How much will Medicare pay them, what strings will be attached, and how do the government and insurers handle the inherent risks of enrolling particularly costly participants?

• Drug companies and medical device makers: What will they be allowed to charge enrollees, what drugs will be covered, and what role will the federal government play in determining what constitutes a reasonable price?

Some Lunches Could Come with a Free Dessert

There might not be free lunches within Medicare (or even low-hanging fruit), but some reforms might do more than merely shift costs. The best reforms are the ones that improve incentives in the program. If a reform reduces utilization without producing worse outcomes, that’s a clear win (although even then some providers and hospitals might not appreciate it). A few examples:

Shifting costs to recipients through more cost-sharing requirements may create incentives to reduce consumption of low-valued treatments. This isn’t merely hypothetical. Since the RAND Health Experiment in the 1970s, research has consistently shown that when patients face extra cost-sharing requirements they change their health consumption.

Likewise, revising payment methods to providers or hospitals could encourage more efficient care delivery. In the 1983, for example, Medicare replaced its open-ended inpatient reimbursement system with its inpatient prospective payment system (IPPS). Prior to the reform, hospital payments were based on the actual cost incurred, giving hospitals little incentive to contain costs. IPPS replaced the open-ended reimbursement system with fixed payments based on a patient’s diagnosis group. This change gave hospitals less incentive to provide unnecessary services. The results were dramatic. As shown in the figure below, the length of hospitals stays for Medicare recipients fell immediately.

A Framework for Assessing Reforms

With these ideas in mind, in subsequent posts we’ll assess specific Medicare reforms. We’ll ask a few questions for each reform.

• What is the effect on the federal budget?

• Who will ultimately pay for it?

• Does it improve incentives?

• Is it politically possible?

• Is the reform sustainable?

The last one is important. Over the last 40 years, Congress has implemented several cost-saving reforms to federal health spending. Many proved fleeting once the public learned of their consequences. Think the ACA’s Independent Payment Advisory Board (IPAB), the ACA’s “Cadillac tax,” the 1990s-era Sustainable Growth Rate (SGR) “doc fix,” 2015 MACRA physician fee reforms, and the Medicare Catastrophic Coverage Act of 1988.

The budget math doesn’t work if we don’t reduce Medicare spending by trillions in the next decade. Every option is unpopular with powerful interest groups. But that shouldn’t stop us from being prepared when a fiscal crisis inevitably forces the hands of Congress and the president.

The federal budget is on a path to perpetually growing deficits and debt, leading to slower growth and future fiscal crises. The math is simple: Federal debt is 97 percent of GDP—up from 36 percent of GDP 20 years ago in 20051—and rising annual deficits averaging 6 percent will lead to even more unsustainable debt for the foreseeable future.

For decades, long-term federal tax revenue averaged around 17 percent of GDP and federal outlays averaged around 21 percent of GDP. The difference, while leading to annual deficits of around 3 percent and higher nominal debt, was manageable when the economy was able to grow at 3 percent and hold debt-to-GDP at a roughly constant rate.

Even with tax cut extensions in One Big Beautiful Bill Act (OBBBA), revenues are expected to be 16.8 percent of GDP over the next decade. That is not far off from the thirty-year average of 17.2 percent from 1995 to 2024. Meanwhile, spending is expected to average 24.7 percent over the next decade—3.6 percentage points above the average of the last 30 years.2

Spending, not revenue, is our main problem.

Health care accounts for a large share of the spending increase. Over the last 30 years, spending on major health care programs averaged 4.4 percent of GDP. But for the next decade, it will account for 5.9 percent. And that’s after accounting for the OBBBA slowdown on Medicaid and other healthcare spending.

To put it another way, 42 percent of the projected increase in federal spending is due to health care programs. The effect on the budget is even worse if you account for health-related tax expenditures that don’t show up on budget.

There are four categories where the federal government can save money in health care: Medicare, Medicaid, the employer-sponsored market, and the ACA’s individual exchanges. Of the four, Medicare is growing the fastest and leading to the largest growth in federal borrowing.

Medicare spending needs to slow down

Federal outlays on Medicare are almost $1 trillion a year. If you include premiums paid by enrollees, annual Medicare expenditures account for 3.8 percent of GDP, which is nearly a percentage point more than is spent on national defense.3

Medicare’s three parts cover hospital visits (Part A), doctor’s visits (Part B), and prescription drugs (Part D). In 2024, they had expenditures of $424 billion, $543 billion, and $149 billion.

If Medicare is not reformed soon, two problems will occur. First, Part A’s HI Trust Fund will go bankrupt by 2033, leading to cuts to healthcare providers and downstream effects on elderly patients.4 Second, the strain on the federal budget will grow. While Part A’s HI trust fund is financed through payroll taxes, Part B and Part D outlays come mostly from general fund taxes. As expenditures grow, more and more general fund revenue will go to fund Medicare enrollees, leading to dramatic increases in federal borrowing, higher interest rates, and potentially lower economic growth.

Medicare is heavily subsidized by taxpayers

Only about half of Medicare’s outlays are covered by payroll taxes or premiums. For every dollar spent on Medicare, more than half comes from general fund sources (Table II.B1 on the Medicare Trustees Report, if you’re interested). In 2024, Medicare received $498.7 billion of “government contributions.”5 In an era of large deficits, you can think of Medicare as responsible for half a trillion in federal annual borrowing.6

Payroll taxes are specifically used to pay for Part A expenses and the HI trust fund, covering 95 percent of its outlays, but none are used for Parts B or D. Instead, premiums are charged to some enrollees for Part B and Part D coverage, but they only covered, respectively, 25 percent and 13 percent of those two programs’ expenditures.

Government contributions (also known as general fund transfers) represent 72 percent of Part B’s revenue and 75 percent of Part D’s revenue. In short, three-quarters of an enrollee’s Part B benefit and Part D benefit is subsidized by taxpayers.7

In a future survey, we intend to ask seniors on Medicare what percent of their coverage they believe is subsidized. Should be illustrative.

How much can we restrict the growth in Medicare spending and limit future payroll tax increases?

From 2026 to 2035, Medicare is expected to spend $16.2 trillion. Premiums will cover $2.8 trillion of the spending and HI payroll taxes will account for $5.1 trillion. Almost all of the remaining amount will be financed by “government contributions” and therefore federal borrowing. That means Medicare is set to add over $8 trillion to the federal debt in ten years. So how much can Medicare spending be slowed down in the next ten years?

Reducing outlays by $1 trillion over ten years would lower net Medicare spending by 7.5 percent. It could occur by reducing net spending by an additional $18.18 billion year-after-year—$18.18 billion the first year, then $36.36 billion the next, and so on. Net spending would be 10.7 percent smaller in 2035 than currently projected, but still 66 percent larger in nominal dollars than today’s spending level.

Read that again: Medicare is growing so fast that even a $1 trillion “cut” over ten years leaves the average year-over-year increase in Medicare spending at 5.7 percent.

The policy options

Setting budget cut goals of $1 trillion or $2 trillion over ten years is a big-picture, macro-level starting point. To make it happen would mean making targeted reductions to individual parts of Medicare spending or making changes to enrollees’ premiums and their required cost-sharing limits.

Some common policy options that could save tens or hundreds of billions of dollars include:

1. Implementing widespread site neutral payments

2. Adjusting Medicare Advantage payment rates and benchmarks

3. Limiting the ability of hospitals to abuse and gamify 340B payments

4. Increasing means testing by broadening income-related premiums for Part B

5. Modernizing cost-sharing and revamp Medigap out-of-pocket limits

6. Increasing the Medicare eligibility age to match Social Security’s full-retirement age

Some of those items have already gained bipartisan support. Others have universal bipartisan opposition. We like some of these ideas more than others, but a mix of the reforms could limit Medicare’s untenable growth rate. In future posts, we’ll look at these options in more depth.

The House version voted to keep nearly all of the 2017 Tax Cuts and Jobs Act in place, sprinkled in some spending cuts to Medicaid, and shipped H.R. 1 to the upper chamber. The Senate promptly ripped out every HSA tweak, sweetened the pot with permanent full expensing for business investment, and punted the rest back across the rotunda. Meanwhile the CBO’s dynamic score says the package raises debt more once growth and higher interest rates are both tallied, meaning it doesn’t come close to paying for itself.

What looks good?

• Medicaid provider-tax cap: the House’s 6 % limit and the Senate’s tougher 3.5 % version would finally crack down on states’ favorite Medicaid gimmick.

• Full expensing: letting firms deduct new equipment upfront is one of the few truly pro-growth ideas.

• HSA/ACA fix & direct-primary-care clause in the House version: solves a regulatory glitch so exchange customers with sky-high deductibles can finally open HSAs, and lets everyone use HSA dollars for a DPC membership.

What doesn’t?

“No tax on tips,” “no tax on overtime,” and a grab-bag of new credits look great in press releases but blow holes in the base and lead to marginal-rate cliffs.

We also talk about the “16 million people will lose coverage” estimate going around. Roughly 5 million are able-bodied adults who would rather skip 80 hours a month of job-search or training, another 1½ million are undocumented enrollees that some states currently finance, and the rest fail new eligibility or verification checks. That’s a far cry from mass expulsions of vulnerable patients.

What’s missing?

We discuss a handful of Medicare changes that are omitted but will need to be addressed eventually.

• Medicare site-neutral payments.

• Medicare Advantage “up-coding” reforms and a bigger bite of Part B & D premiums for high-earners.

• Small but important changes to federal-worker pension contributions, SNAP sanity checks, and other small entitlements that quietly add up.

Veto of the Week: Grover Cleveland’s 1886 veto of a special pension boost for Civil War veteran John Taylor, because not every benefit needs an increase. A neat parallel to today’s senior programs that have quietly become very generous.

What’s in the Episode

[02:06] - House vs. Senate: Key Differences in the Bill

[05:59] - Economic Implications of Tax Cuts

[09:56] - Spending Cuts and Medicaid Reforms

[12:48] - Medicaid Provider Tax Cap

[19:30] - How Many People Will Lose Health Insurance?

[24:45] - Discussing HSA Changes - Expand Enrollment

[27:30] - Direct Primary Care and Health Savings Accounts

[31:00] - No Tax on Tips, No Tax on Overtime

[35:07] - Addressing Medicare and Missing Spending Cuts

[42:45] - Non-Medicare Spending Cuts

[45:53] - Veto of the Week: "The rate he was receiving was commensurate with the degree of his disability, a board of surgeons having reported that he was receiving a liberal rating."