What every plan leaves out

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.

These are the real ones, in rough order of how badly they'd hurt us if omitted.

1. Long-term care — the biggest gap in American life

Medicare does not cover it. Medicaid covers it only after you have been impoverished. Around 70% of 65-year-olds will need it, at $100,000+ a year. This is the single largest uninsured risk an American family faces, and no serious plan can omit it. Germany added a mandatory long-term care pillar in 1995; Japan added one in 2000. Both work. Build it as a full benefit (home care first, institutional care when needed), which is only physically possible if the professions section of the plan's direct-care workforce plan is real.

2. Mental health and addiction — a capacity problem, not a coverage problem

Parity, the legal requirement that insurers cover mental health and addiction on the same terms as physical illness, has been federal law since 2008 and is not enforced. Ghost networks, provider directories listing clinicians who are not accepting new patients, are pervasive; a large share of psychiatrists don't take insurance at all, so an insurance card buys access to a directory, not to a doctor. Overdose deaths run in the tens of thousands a year.

America went from roughly 560,000 state psychiatric beds in 1955 to around 35,000, and never built the community system that was supposed to replace them. The 1963 Community Mental Health Act was passed and never funded. The consequence is that the largest mental health facilities in the United States are county jails. Fixing it requires its own build-out (crisis centers, community mental health centers, psychiatric beds, and a trained workforce to staff them), not a line in a benefits list.

3. Dental, vision, hearing

Bernie includes these and is right to. Dental deserts, counties with no dentist at all, are worse than medical ones, and dental disease is the leading cause of preventable emergency room visits, for a condition a $200 filling would have solved. State dental boards dominated by practicing dentists have blocked competing dental therapists without the active state supervision that North Carolina State Board of Dental Examiners v. FTC requires for antitrust immunity.

4. Public health, which is not the same thing as medical care

Public health is a tiny share of health spending and has the highest return of any of it. Local health departments lost much of their workforce over the 2010s and were then hollowed out further by the politics of the pandemic. Rebuild it, and rebuild it with legal independence, so the next emergency isn't run by whoever is politically convenient.

The largest determinants of health are not medical at all: housing, food, air, cars, guns, work. Fund what's inside health policy's reach, but don't let it swallow the plan. A health care plan that becomes a plan for everything will not deliver on any of it.

5. Medical debt — cancel it, and pay back the people who paid

About 41% of American adults carry medical or dental debt, roughly 106 million people: bills past due, bills being paid off directly to a provider, debt sold to a collection agency, or debt sitting on a credit card. Medical debt trades on the secondary market at pennies on the dollar, so canceling all of it is cheap, hundreds of millions to low billions of dollars for a face value in the hundreds of billions. It is among the least expensive measures in the plan. It can be done on day one, before the rest is even built.

The objection that has to be answered: what about the people who paid?

Cancel debt and you anger everyone who worked two jobs and paid theirs off. That is what happened to student debt cancellation, and the grievance was potent because it was legitimate.

Two things are true about it.

The medical version of that backlash has been tested at real scale and it did not appear. States, counties, and cities have used American Rescue Plan funds to cancel an estimated $7 billion of medical debt for up to nearly 3 million Americans, through Undue Medical Debt (formerly RIP Medical Debt), Cook County, Toledo, New York City, and dozens of others, including deeply Republican jurisdictions. There is no recorded backlash from people who had paid theirs off. The polling is the same: 81% support banning wage garnishment for medical debt, including strong majorities of both parties, and 84% say government is responsible for making health care affordable.

The reason is structural. Medical debt is not student debt. You did not choose it, you did not apply for it, and you kept no asset at the end of it. "I paid off my degree and I still have the degree" is a real grievance. "I paid off my heart attack" is not a thing anyone wants to have been. The counterfactual to being forgiven a medical debt is not a free education, it is not having got sick.

Second, do it anyway: the objection is cheap to answer and dangerous to leave unanswered. Being right about the polling, that medical-debt cancellation enjoys overwhelming public support and has triggered no measurable backlash at state or local scale, is not a reason to hand the opposition a free grievance with a real human face on it. The plan pays back the people who paid.

The medical debt fairness refund

Medical debt payments made in the seven years before enactment are refunded, dollar for dollar, capped at $10,000 per person.

Four design choices:

① Dollar-for-dollar, not a flat check. A flat $10,000 to anyone who ever paid produces an immediate absurdity: a person who paid down $3,000 receives $10,000, while a person whose $5,000 of live debt is canceled receives only $5,000 in value. Refunding what people actually paid is the only version that is internally fair. The $10,000 becomes a cap rather than a grant.

② A seven-year lookback, not a lifetime. Seven years is not arbitrary: it matches the Fair Credit Reporting Act's retention period for collection accounts, the maximum time a debt in collections may appear on a credit report, and, conveniently, the window in which payment records reliably exist. A lifetime lookback cannot be documented and so collapses into self-attestation, which is disqualifying (see design choice ④ below). It also captures the people whose grievance is real and current: those still paying, and those who just finished.

③ Automatic wherever the records exist; an application only where they do not. Active payment plans, collection accounts, and provider accounts-receivable are all documented in credit files. Pay those automatically, with no form at all, which is cheaper to administer and puts the "legible relief" principle into practice. Save the claim process for older self-financed debt.

④ Keep PPP's ease. Do not keep PPP's attestation. The Paycheck Protection Program (PPP), the COVID-era small-business lending program, got one part right (a simple form and fast money) and one part dangerously wrong. The SBA Inspector General estimates more than $200 billion in potentially fraudulent COVID relief lending, including roughly $64 billion in PPP specifically, and PPP loan guaranty purchases carried a 49.1% improper payment rate in FY2023. A self-attested $10,000 federal payment to anyone who states they once paid a medical bill would be the most fraud-attractive program ever legislated. "The government is mailing $10,000 checks to anyone who fills in a form" would define this plan for a decade. Matching against records first and paying automatically where possible closes the largest fraud surface.

What it costs

Version	Rough one-off cost	Why
Cancellation alone (the original plan)	low billions	Debt bought on the secondary market at pennies on the dollar
Recommended: 7-year refund, dollar-for-dollar, $10,000 cap	$100–250B	Bounded window, proportional, capped, take-up well under 100%
Lifetime lookback, flat $10,000 per person	$1.4–1.8 trillion	If 41% of adults hold medical debt today, the share who have ever paid some down is plausibly 55–70% of ~258 million adults — call it 145–180 million people — at $10,000 each

The recommendation departs from the simplest version of the idea on one point. The lifetime-flat-check version might be estimated in conversation at "a few billion." It is closer to one and a half trillion dollars, one-off, landing in exactly the deficit years the financing plan already names as this plan's real vulnerability (the early years when savings have not yet ramped up but new spending has), roughly a third of a full year of total national health spending, larger than every other line in the cost section of the plan combined, spent to pre-empt a backlash that $7 billion of real-world medical debt cancellation has so far failed to produce. The recommended version delivers the same promise, if you paid, you get it back, at roughly a tenth of the cost and with a fraud surface that can be defended on television. The lifetime version is a coherent choice; it just has to be priced in the cost section of the plan as a trillion-dollar line and financed deliberately, not discovered later.

One drafting warning

Barring medical debt from credit reports cannot be done by rule. The CFPB finalized that rule on 7 January 2025; the Eastern District of Texas vacated it on 11 July 2025 in Cornerstone Credit Union League v. CFPB, holding that it exceeded the Bureau's statutory authority under the Fair Credit Reporting Act, and further that FCRA expressly preempts state laws attempting the same thing, which closes the state workaround. Medical debt is back on credit reports as of 2026. This must be a statutory amendment to FCRA in the Relief Act itself. The regulatory route has already been tried and lost; drafting it as a rulemaking would hand the opposition a ready-made injunction in month two.

6. Malpractice — buy the doctors' support with it

Malpractice is a modest direct cost but a large psychological one for clinicians. Replace it with public indemnity for clinicians (the government assumes malpractice liability for participating providers, eliminating their premium costs) plus a no-fault injury compensation scheme for patients, on the model of New Zealand's ACC (Accident Compensation Corporation, New Zealand's universal no-fault injury scheme covering all accidents including medical injury) and Sweden's patient-injury insurance system, where any patient harmed by medical treatment applies to a public body for compensation without having to prove negligence in court. Injured patients are compensated faster and more often. Doctors stop practicing defensively. The money currently spent on litigation goes to care. Injured patients are compensated faster because they need only show that treatment caused harm, not that a doctor was negligent, and organized medicine moves from opposition to neutrality once the malpractice premium, which can run tens of thousands of dollars a year per doctor, disappears from their practice expenses.

7. The transition's human cost — now handled in full in the insurance-industry section of the plan

This plan eliminates 1.5–2.5 million jobs in insurance and in provider billing. That is the point of it. Bernie's bill sets aside a small transition fund for a few years; it is not enough. The insurance-industry section covers what happens to the companies four ways and the Health Workers Transition Guarantee. The one line to carry from it: a billing coder in Dayton will decide this election, and we can afford to be extraordinarily generous to her because her job is where the money is.

8. Where the employer money goes

Employer-sponsored insurance covers around 165 million people at roughly $25,000 a year for a family, well over a trillion dollars, and it is hidden compensation that workers earned and never see. The financing has to capture that money (through an employer contribution replacing the premium, a payroll tax, or both) or the budget does not close. Every Medicare for All debate has foundered here.

9. Abortion and immigration — answer, don't dodge

Both will be asked, and both will define the plan's coverage for a week. Evasion looks weak. Cover everyone physically present in the United States, regardless of immigration status. Cover the full range of reproductive health care including abortion. Say it once, plainly, and then decline to relitigate it. Repealing the Hyde Amendment, the longstanding rider that bars federal funds from paying for most abortions, is a necessary component and should be stated as such rather than discovered later.

10. Federalism and the courts

The plan has to survive hostile states and a hostile judiciary. Build it as a direct federal benefit to individuals wherever possible rather than as conditions on state programs, because an individual federal entitlement, like Social Security, can be enforced by the beneficiary in federal court without depending on a governor's cooperation, whereas a program that flows through state governments can be blocked or starved by a hostile state, as happened when 14 states refused the ACA's Medicaid expansion after NFIB v. Sebelius. (cut) Assume every major provision is challenged, and write severability into every title.