Ownership: ban the dividend, not the owner

This is a beta: early, unfinished, and open to being wrong. If you think something here is wrong or would fail in practice, tell us.

This is an appendix to the full plan. It is the detail behind one part of it, kept whole for readers who want the argument rather than the conclusion.

Figures below are computed from OECD hospital-bed and hospital-count series, latest year per country, mostly 2023–24.

The finding: ownership is not what America gets wrong

Both patterns work. Japan's hospitals are mostly small, privately owned institutions; South Korea's are larger private groups. What matters is not the pattern of ownership but whether the system imposes the conditions this plan describes: no profit distribution, mandatory participation at the public price, and open books to verify both. Running the OECD ownership numbers, single-payer systems operate through privately owned providers as often as through public ones.

Country	For-profit beds	Public beds	Beds / 1,000	Hospitals / million	Health score	$/person
South Korea	0.0%	9.4%	12.6	81.9	95.3	$5,081
Japan	n/r	27.8%	12.6	65.8	98.1	$5,365
Netherlands	~0% (100% nonprofit)	~0%	2.3	41.2	87.2	$8,195
Belgium	~0% (73.6% nonprofit)	26.4%	5.3	13.6	89.4	$7,489
Israel	13.2%	64.4%	3.0	8.6	94.2	$4,033
United States	17.4%	21.1%	2.7	17.9	69.9	$13,473
Spain	18.5%	69.3%	2.8	15.3	95.4	$4,935
France	24.5%	61.0%	5.4	43.1	88.7	$6,868
Germany	32.1%	39.5%	7.6	35.4	86.2	$8,826
Italy	34.6%	65.4%	3.0	18.0	91.8	$4,993

Caveat on the bed columns. Japan and Korea count long-term care and psychiatric beds inside hospitals; the United States counts the equivalent capacity as nursing-home and skilled-nursing beds, which this OECD series excludes. Add US nursing home beds back and the gap narrows substantially. The comparison that matters is curative and acute beds only, and the OECD bed series used here has no such split. The CARE_TYPE column in the OECD bed series is empty for every row; the OECD curative-care bed series must be checked before quoting 4.5× anywhere public. The same applies to hospital counts: Japan classifies any facility with 20+ beds as a hospital, so much of its 65.8-per-million is what an American would call a clinic, while the US count excludes thousands of ambulatory surgery and urgent care centers. The capacity gap is real; this particular multiple is not yet defensible on available data. The unambiguous access measures (obstetric-desert counties, ED boarding hours, days to third-next-available appointment, psychiatric bed availability) are stronger evidence and cannot be dismissed as a definitional artifact.

And a second, harder finding: the bed ratio does not predict health outcomes at all. Across the 41 countries in this dataset, the correlation between hospital beds per 1,000 and the health score is +0.07, and with life expectancy −0.10, statistically nothing. The US ranks 30th of 41 on beds: low, but not an outlier. Italy (3.0), Spain (2.9), Canada (2.5), and the Netherlands (2.3) all sit at or below America and achieve far better outcomes for far less money. The plan's own data does not support "America is short of beds" as a causal story.

The capacity argument is about distribution, not the national ratio. Over a third of US counties have no obstetric care. More than 100 rural hospitals have closed since 2010. Psychiatric beds fell from ~560,000 to ~35,000 with no community replacement. Emergency departments board patients for days. Those are unambiguous and locally verifiable, and cannot be argued away as a definitional artifact of how different countries classify hospital beds versus nursing-home or long-term-care beds. Rebuild the capacity section of the plan's evidence on those and drop the 4.5× multiple.

The for-profit column in the ownership table tells the same story. America is not an outlier. At 17.4% investor-owned beds the US sits in the middle of the pack, below Italy (34.6%), Germany (32.1%), Greece (31.6%), France (24.5%), Norway (19.2%), and Spain (18.5%). Italy has twice America's for-profit share, scores 91.8, and spends $4,993. Whatever is wrong with American health care, "too many for-profit hospitals" does not distinguish it from countries that work.

The last three columns do distinguish the US from the countries that work. The United States has 2.7 hospital beds per 1,000 people and 17.9 hospitals per million. Japan and Korea, the two best-performing systems on Earth, and the two most privately owned, have 12.6 beds per 1,000, roughly four and a half times as many, and three to four times as many hospitals per capita. America's hospital sector is too small and too consolidated. (cut)

Korea runs the most privately-owned hospital system in the developed world (90.6% private, and only 9.4% public, less public than America) and buys a 95.3 health score for $5,081. (cut)

What those countries actually do instead

They do have explicit policies, and they are more interesting than ownership rules:

1. They ban the dividend, not the owner. Korea's 0.0% for-profit share is not a market outcome. Korea prohibits investor-owned hospitals outright; institutional providers must be non-profit corporations, school foundations, or government. Japan requires hospitals to be run by physicians or by non-profit medical corporations (医療法人) that may not distribute profits. The Netherlands and Belgium do the same. You may own a hospital. You may not pay a dividend on it. That is the rule in the most privately-owned successful systems, a far lighter touch than public ownership.
2. One public price, and you cannot opt out. Korea's mandatory designation rule requires every licensed provider to accept national insurance patients at national insurance prices. Japan's Chuikyo fee schedule applies to everyone. The private provider competes on quality and convenience, never on price, and is required to accept the public patient.
3. The state decides where capacity goes. Japan's prefectural health plans set regional bed quotas; you cannot open beds in a region that has met its quota, and public capital fills the regions that haven't. Japan ended up with 65.8 hospitals per million distributed across the country; America ended up with 17.9, clustered in profitable metros.

Germany is the useful counter-case. It is the one comparator that does permit large for-profit hospital chains, at 32.1% of beds, and it is the most expensive system in Europe at $8,826. That is suggestive rather than causal, but the direction of drift in the one for-profit-friendly European system points the same way.

The revised recommendation: conditions, not conversion

The ownership policy becomes "let them stay private under three rules" rather than "convert them to public ownership." The evidence supports this better.

The three conditions of participation:

• No profit distribution by institutional providers. Hospitals, clinics, physician groups, dialysis, home health, nursing homes: you may be private, you may be well run, and you may pay your executives and reinvest your surplus. You may not pay dividends or make distributions to shareholders. 61.5% of American hospital beds are already non-profit and comply on day one. The rule falls on the 17.4% investor-owned segment, the physician roll-ups (large corporate entities, often backed by private equity, that acquire independent physician practices and consolidate them under single ownership), and private-equity-backed providers more broadly, which is exactly where the evidence of harm is concentrated.
• One price, participation as a condition of billing the payer, no balance billing, and a real, legal, unattractive exit that a provider may actually take. (cut) Saying "any licensed provider must accept the public schedule" would likely have lost in court: a compelled-participation mandate tied to licensure raises takings and compelled-speech challenges that a voluntary billing condition does not. Participation is conditioned on billing the payer, never on holding a license: a provider who does not wish to accept the public price may decline to bill the public payer and treat only private-pay patients, which is the "real, legal, unattractive exit" mentioned above. See the payer section of the plan's note on federal provider eligibility: conditioning participation on state licensure lets a hostile state legislature restrict the licensure of clinics, nurse practitioners, or foreign-trained physicians and thereby determine whom the national system may pay.
• Active anti-consolidation, plus public capital where the market won't build. Merger review with teeth and court-ordered divestiture of the roll-ups that used market power to raise prices (the same private-equity-backed consolidations described in the plan's opening), paired with the capacity section of the plan's Hill-Burton-modeled capital program aimed at the places private capital has left unserved. (cut)

What this preserves: private hospitals, private practices, physician ownership, and competition on quality, with no mass expropriation fight. What it removes: the ability to extract.

Where "treat them all the same" is right — and the one place it isn't

The formulation: treat public, non-profit, and for-profit identically, fix the prices, let nobody opt out. That is right for payment and access: one schedule, one set of obligations, no ownership carve-outs in either direction. A public hospital gets no price advantage; a for-profit gets no escape hatch.

Under a fixed schedule, four extraction routes remain, and three of them are already controlled elsewhere in this plan:

Extraction route	Controlled by
Charging more	The fee schedule. Solved.
Doing more (volume, intensity)	Global budgets rather than fee-for-service. Solved.
Choosing profitable patients	Mandatory participation, no balance billing, case-mix adjustment. Solved.
Cutting quality and stripping assets	Nothing above touches this.

That fourth row (cutting quality and stripping assets, the one extraction route not controlled by the fee schedule, global budgets, or mandatory participation) is the argument for ownership rules, a narrower argument than the one the draft started with. The American case that proves it is Steward Health Care: a private equity owner executed a $1.25 billion sale-leaseback of the hospitals' own real estate, extracted the proceeds, and saddled the operating company with rent it could not pay. The system collapsed in 2024. The operating company could not cover the rent, and community hospitals that depended on it closed. No fee schedule in the world would have prevented that. The extraction worked through the sale-leaseback itself: the PE owner sold the hospitals' real estate out from under the operating company, pocketed the proceeds, and left the operator owing rent it could never cover. (cut)

So the narrow, defensible rules are about behaviors, not ownership form, which makes them enforceable against a non-profit behaving badly too:

• No sale-leaseback or encumbrance of core clinical real estate without regulatory approval.
• No dividend recapitalizations or debt-funded distributions by an operating provider.
• Minimum staffing ratios (already in the professions section of the plan), which is where quality-cutting actually shows up.
• Open books — full public financial reporting, including of related-party transactions, which is how the extraction is usually hidden.

This is a better answer than "ban for-profits" because it is narrower, survives a court, and catches the non-profit systems that behave identically. The profit-distribution ban remains the belt-and-braces option, and Korea and Japan both use it. But for the lightest touch that still works, the four rules above plus the fee schedule are probably enough.

But keep public ownership as a real instrument, for three specific jobs

Conditions are the general rule. Public ownership still does things conditions cannot:

1. Reopening what closed. Where a rural hospital shut and no private operator will reopen it, the public builds and runs it. Nobody else will.
2. A public benchmark in every region. A publicly run system sets a floor on quality and a reference on cost, the way public universities do. The VA already plays this role and plays it better than its reputation.
3. The worst actors. Where fraud enforcement produces a structural remedy, such as a court-ordered divestiture or receivership following asset-stripping or patient-safety violations, public or community ownership is where the asset lands.

That is a target more like 35–45% public beds (roughly Germany's level, up from 21.1%) reached mostly by building new public capacity under the capacity section of the plan rather than by taking existing capacity. Growing the public share by growing the pie is a different political proposition from confiscation, and it gets to the same number.

The legal ladder, for the cases where conversion is still needed

Forced divestiture based on documented fraud or patient-safety violations is the weakest legal path (proving systemic fraud case by case is slow and uncertain even when the facts are damning) but it is the strongest political one. Conditioning participation in the public payer on meeting the ownership and financial rules described above is stronger: a provider who will not comply simply loses its right to bill the system, and forced divestiture becomes a last resort rather than a first move. Ranked by legal solidity:

1. Conditions of participation — the strongest tool, and it isn't a taking (recommended)

Once there is a single payer, Congress can make the conditions described above (no profit distribution, one public price with mandatory participation, open books) a condition of participation. To bill the national system, an institutional provider must make no distributions to shareholders, accept the public price schedule, keep open books, cap executive compensation, and serve its whole service area. This is the same legal mechanism whether the ask is "become a nonprofit" or "become public": the provider is not being seized, it is being told what the government requires in exchange for paying it. The softer version, conditions on behavior rather than forced conversion of ownership, is just as strong legally and far easier to defend politically.

• This is not a seizure and requires no payment — it is a condition on doing business with a customer. Medicare's existing Conditions of Participation already govern enormous amounts of hospital behavior and have never been seriously challenged as takings.
• The alternative is real and unattractive, and must be genuinely available, not a formality. A clinician may leave the public system entirely, on the model of the existing Medicare private-contracting rule (42 U.S.C. § 1395a(b)): opt out, and forgo the program for two years. Very few will. That is the point. It must be the result of the door being unattractive rather than the door being fake. The legal position rests on taking part being a real choice.
• Opt-out is an individual election; any coordinated group boycott of the public system would constitute a concerted refusal to deal among competing economic actors, a per se antitrust violation. The statute should name the FTC as enforcement authority. A coordinated campaign to get one high-margin specialty to opt out en masse in year one, so that an instant boutique tier appears and a "two thousand doctors quit" headline follows, is the most likely opposition play, and existing law already answers it if the statute says so.
• Legal risk to manage: the unconstitutional-conditions doctrine (the principle that the government may not condition a benefit on giving up a constitutional right), and NFIB v. Sebelius (2012), which found the ACA's Medicaid expansion unconstitutionally coercive because it threatened states with the loss of all existing Medicaid funding if they refused the new expansion. But NFIB protects states under the spending power and federalism; it says nothing about conditions on private providers' choosing to take part in a federal payment program, which are routine, and the existing Conditions of Participation described above are the standing precedent. Reduce the remaining risk with a long transition (5–7 years) and compensation for stranded capital, meaning that providers whose business model depends on profit distribution or balance billing get years of advance notice and support for the conversion, rather than an overnight mandate.

The cheapest fix in the plan.

Everything in this Part is protected by one doctrine: that participation is voluntary. An unbroken line of cases holds that where a provider chooses to join a price-regulated federal program, rate-setting and conditions cannot be a taking: Garelick v. Sullivan (2d Cir. 1993) ("Economic hardship is not equivalent to legal compulsion"), Franklin Memorial Hospital v. Harvey (1st Cir. 2009), Baker County Medical Services (11th Cir. 2014), Burditt v. HHS (5th Cir. 1991). The capstone is recent and decisive: the Second Circuit in Boehringer Ingelheim v. HHS (Aug. 2025) held that "the choice to participate in a voluntary government program does not become involuntary simply because the alternatives to participation appear to entail worse, even substantially worse, economic outcomes" — and the Supreme Court denied certiorari in all six pharma challenges to Medicare drug-price negotiation on May 18, 2026.

It is easily thrown away by careless drafting. "Any licensed provider must accept the public schedule." "Exit is not a real option." "Converts or liquidates." Judge Hardiman dissented in the Third Circuit on precisely the ground that opting out was economically impossible, and a careless draft writes his brief for him, in the proponents' own document. Once participation is compelled and the alternative market is abolished by statute, the 40-year answer evaporates and every rate becomes reviewable as a public-utility rate case under Duquesne Light v. Barasch (1989).

The fix is in the drafting:

1. Delete the sentences. They are discovery exhibits, not policy.
2. Condition participation on billing the payer, never on licensure. Keep a real, legal, deliberately unattractive exit modeled on Medicare's private-contracting rule (§ 1395a(b)): a clinician who opts out cannot bill the program for two years. The door almost nobody uses is what keeps the program voluntary under existing caselaw. Korea's mandatory-designation rule works because Korea has no Takings Clause of this kind; import the result, not the mechanism.
3. Build a Duquesne-proof record into the statute anyway: require that the schedule let an efficiently operated provider recover reasonable operating costs plus a reasonable return on prudently invested capital, with an administrative confiscatory-rate appeal. At this plan's spending level no rate is confiscatory in fact, so the provision costs nothing and moots the main claim against us.
4. Draft "may not refuse public patients" as a non-discrimination duty for facilities open to the public, not as a right of entry. PruneYard supports the first; Cedar Point Nursery (2021) cuts against the second.

(cut)

2. Fraud enforcement — the political rationale and the stick

The politics here are sound: recovered health care fraud runs in the low billions a year through enforcement actions, against credible estimates that total fraud losses run $150–500 billion annually. Enforcement is nowhere near proportionate to the conduct.

• The False Claims Act (31 U.S.C. § 3729) carries treble damages plus a per-claim penalty. For an entity submitting millions of claims, exposure can exceed enterprise value, which is exactly the leverage that produces a negotiated structural remedy.
• Exclusion authority (42 U.S.C. § 1320a-7) lets HHS bar an entity from Medicare and Medicaid. For a provider whose revenue is half or more government money, exclusion is fatal. Under a single payer, where the public system is the only payer, it is fatal without exception. Exclusion is therefore the credible threat behind the conditions-of-participation framework described above: accept the rules, or lose the ability to bill the only payer that exists.
• Corporate Integrity Agreements, the binding compliance and oversight terms HHS negotiates with providers as part of fraud settlements, already impose governance conditions; there is precedent for escalating to structural remedies such as forced divestiture or changes in corporate control.

Build the enforcement title around exclusion and recoupment, not around penalties. In SEC v. Jarkesy (2024) the Supreme Court held that when the government seeks civil penalties for fraud, the Seventh Amendment entitles the defendant to a jury trial in an Article III court, because the claim is "in the nature of" common-law fraud. That exposes the civil monetary penalties currently imposed by HHS administrative tribunals, the fast, cheap stick this section was assuming, to a requirement of full jury trial in federal court. Each case would need a jury empaneled and a docket slot in a district court, so penalties become far slower and more expensive to pursue. Exclusion (barring a provider from billing the public payer) and recoupment (recovering overpayments already made) are not common-law analogues and sit on far safer "public rights" ground, meaning they can be imposed administratively without a jury. Fortunate, because exclusion is the stronger instrument. Route punitive penalties to the courts and treat them as the slow path; make exclusion and recoupment the fast one.

But be clear-eyed: fraud enforcement produces settlements and monitorships, not nationalization. Courts will not hand the government a hospital chain because it upcoded. Use fraud as the moral case for why these institutions forfeited the presumption that they should be trusted with public money, and as the specific instrument against the worst individual actors. Don't build the architecture on it.

3. Antitrust and structural divestiture

This plank has fresh, enacted, bipartisan federal backing. The health insurance industry's antitrust exemption under McCarran-Ferguson was already repealed (the Competitive Health Insurance Reform Act, Pub. L. 116-327, signed January 13, 2021). The repeal applied the antitrust laws to the business of health insurance, with only narrow safe harbors for historical loss data. Congress has already voted, on a bipartisan basis, that competition law should reach this industry.

Prices rise sharply after hospital mergers in concentrated markets, one of the best-documented price effects in health economics. Divestiture, a court-ordered breakup of a combined entity, is a standard Clayton Act § 7 remedy, and divestiture to a public entity or a new community nonprofit is available. This opens a path to unwind the last two decades of hospital and provider-group roll-ups on well-trodden legal ground.

4. State charitable-trust and licensure authority

Most American hospitals are already nominally nonprofit, and many have drifted far from charitable purpose while keeping the tax exemption. State attorneys general already have authority over charitable assets, including the power to block conversions, impose conditions, and enforce charitable purpose. Hospital licensure and certificate-of-need are state powers. A cooperating state can convert governance without any federal action at all, which makes this the natural place for a blue-state proof of concept before the federal fight.

5. Eminent domain — narrow use only

The Fifth Amendment permits it, and Kelo v. New London (2005) reads "public use" broadly. But it requires just compensation, which makes it expensive at scale. Reserve it for specific facilities the public needs reopened, where a closed hospital's market value is near zero and buying it outright is cheaper than building.

A recommendation on language

The word "nationalization" is wrong here. It is not what the plan does. The one-line description: hospitals may be private, and hospitals may not distribute profits to shareholders. That is Korea's rule and Japan's rule, in two of the most privately-owned and best-performing systems on Earth, and it is a far easier sentence to defend than anything involving the word seizure.

And the offensive framing: we are not proposing anything Japan and South Korea haven't already done. They have four times as many hospital beds per person as we do and better health outcomes than anyone in the world, and they pay 40% of what we pay. Their hospitals are private. Their hospitals just can't pay dividends or turn a patient away, and they can't set their own prices.