The Floor: Economic Package

The bill in one page

The Floor costs the federal budget about $2.2 trillion a year at full operation, 7.2% of GDP, and costs the country almost nothing in real resources, because $1.8 trillion of it is the health care bill households, employers and states already pay, moved onto one ledger and reduced by about $200 billion in overhead and prices. The genuinely new spending is $420 billion a year: housing $180 billion, food $150 billion, education $90 billion. It is paid for by contributions that replace premiums ($1.5 trillion: 7.5% of payroll from employers, 4% of income above $40,000 from households, the states' Medicaid share continued) and a progressive package of $0.7 trillion (a 28% corporate rate, higher top brackets, gains taxed at death, a cap on itemized deductions, a wealth tax above $50 million, a rebuilt IRS, a small transaction tax). The two tiers raise $2.20 trillion against a bill of $2.22 trillion; two named reserves, a carbon fee and a half-point on the household contribution, cover the margin, and a 5% VAT is the alternative to the wealth tax if the courts or politics remove it.

For a family earning $80,000 with employer coverage, the trade is a $1,600 contribution in place of about $11,000 a year in premiums and medical bills, plus a share of the $14,000 its employer stops paying, which CBO's modeling turns into wages: $9,000 to $20,000 a year better off before any rent or tuition it no longer pays. A household at $250,000 roughly breaks even. Households above $1 million, and wealth above $50 million, pay for it, along with the health industry's administrative overhead.

The evidence on work says hours fall a few percent and nobody stops; CBO's modeling of single payer puts GDP within about a point of its path either way when the financing replaces premiums rather than stacking new rates on labor, which is what this package is built to do. On the OECD's measure the United States already spends more on social provision, public and private combined, than France; the Floor changes the channel more than the amount. The transition takes eight years, revenue first, the cheap guarantees in year one, health over four years, housing last because the homes have to be built. The one flaw that must be designed out before enactment is the stacking of tapers on families just above poverty; a single Floor taper fixes it for $40–60 billion a year.

Design rules that drive the numbers

Seven choices decide what the Floor costs. Each is a design decision, not a finding, and each can be argued with; the costing below follows from them.

Universal in right, income-rated in contribution. Everyone who lives here is covered by residency, with no application to prove need. Where a guarantee has a per-person price (a home, groceries), the contribution scales with income through the tax return, the way Social Security contributions do. A household in the top half is fully covered and pays its own way; a household whose income falls is covered the same month, with no office visit. The sorting the pamphlet refuses happens in no office at all; it happens in the tax code, where it already happens to everyone.

Goods and services, not cash. Care is delivered by providers paid by one public payer; homes are delivered as leases; food as a grocery credit on a card; education as a seat. In-kind delivery keeps the inflation channel narrow, lets supply be planned, and is the point of the Floor as written: the survival question is removed, not converted into a budgeting problem.

Supply is part of the policy. Every guarantee that could raise prices is paired with a capacity program: a public building program for housing, provider capacity and price negotiation for health care, logistics and waste recovery for food, seats and instructors for college. Price pressure is answered with units, not with caps.

Federal guarantee, local delivery. The entitlement is federal and the money is federal. Delivery runs through institutions that already exist (hospitals, housing authorities, school districts, community colleges), with a direct federal channel wherever a state declines to deliver, as a dozen states declined Medicaid expansion.

Each pillar carries its own revenue. At maturity the package is pay-as-you-go: every guarantee is paired with named taxes that cover it, so no pillar is an unfunded promise. Capital spending (building homes) is borrowed and repaid from rents and the budget, the way roads and water systems are financed.

Residency is a parameter, not a loophole. The costing covers the whole Census resident population. How long someone must live here before coverage begins is a political choice; it moves the bill by a few percent either way and is costed at the margin, not buried.

One accounting convention. Figures are annual, in 2025 dollars, at full operation (year 8 of the transition). Every pillar is shown three ways: gross (what the program spends), offsets (what it replaces, public and private), and net new public spending (what must actually be raised). Where the evidence disagrees, the range is shown rather than a point.

Medical care: the largest number and the smallest new cost

Universal care moves about $1.8 trillion a year (2025 scale) of health spending from households, employers and states onto the federal ledger, and lowers what the country spends on health in total by roughly $0.2 trillion a year, with a credible range from $0.5 trillion saved to $0.2 trillion added. It is the pillar that costs the budget the most and the country the least, which is the pamphlet's "pays it twice over" in numbers.

What it is. One public payer covers every resident for hospital, physician, drugs, mental health, maternity, dental, vision and hearing care, with no premiums, deductibles or copays. Providers are paid at rates between Medicare's and today's private average (hospitals near Medicare plus 10%), hospitals move to global budgets, and drug prices are negotiated. Long-term services and supports come in from year six. In CBO's terms this sits between its Options 3 and 4 (lower cost sharing, no long-term care) with cost sharing set to zero, and becomes its Option 5 when long-term care is added.

The arithmetic. CBO's 2020 analysis projected 2030 health spending of $6.63 trillion under current law, with federal subsidies of $2.82 trillion. Its Options 3 and 4 raise federal subsidies by $1.77 and $2.40 trillion while cutting total national health spending by $650 and $42 billion; payers' administrative spending falls about $410 billion in every option, and out-of-pocket spending falls by $306–327 billion of the $721 billion projected. Floor Health, at the midpoint with cost sharing at zero and dental added, raises federal spending by about $2.2 trillion and lowers total spending by about $250 billion in 2030 terms. Scaled back to 2024's $5.3 trillion of spending, that is the $1.8 trillion and $0.2 trillion above. Long-term care later adds about $0.5 trillion to the federal figure and turns the total from a saving into a small cost, which is Option 5's result and the reason it is phased.

Where the money already is. In 2024 the country spent $5.28 trillion on health, 18.0% of GDP and $15,474 a person: households carried 28% of it ($557 billion of that out of pocket), private business 18%, states and localities 16%, the federal government 31%. Under Floor Health the household premium and the medical bill go to zero, the employer premium goes to zero, and the states' Medicaid share ($335 billion in 2024) continues as a maintenance-of-effort payment to the national plan. Those three streams, about $2.4 trillion a year, are what the contributions in the funding section replace. The federal government already runs the largest share: Medicare's net outlays were $988 billion in fiscal 2025 and federal Medicaid $627 billion, on top of the tax exclusion for employer coverage, which CBO values at $471 billion in 2026 counting income and payroll taxes.

Why the total falls. Administration is 34% of U.S. health spending, $812 billion in 2017 against 17% in Canada; U.S. insurers' overhead runs $844 a person against $146. Traditional Medicare administers care at about 2% of spending and private insurers at 12%, the ratio CBO builds on. Prices are the other half: the OECD's 2024 figures put U.S. spending at $14,885 a person against $9,365 in Germany and $7,367 in France, for a life expectancy of 79.0 against the OECD's 81.1. Independent models find larger savings than CBO: the PERI team's 9.6%, the Lancet model's 13% and 68,000 fewer deaths a year. CBO is the conservative estimate and the one this package uses.

The baseline is getting worse, which raises the gross bill and not the net. The July 2025 reconciliation law cuts federal Medicaid by $911 billion over ten years and leaves 10 million more people uninsured by 2034; the enhanced marketplace credits expired at the end of 2025 and add 4.2 million more. CBO now expects 30 million uninsured in 2026, 37 million by 2036. Every one of them is new coverage for Floor Health to pay for, so the federal line is larger against today's baseline than against 2020's; the country's total spending is what it is either way.

Transition. Four years. Year one: drug price negotiation across the market, and the uninsured enrolled automatically in Medicare-administered coverage. Year two: Medicare for everyone under 26 and over 55. Year three: everyone. Year four: cost sharing to zero, dental, vision and hearing in. Two costs belong to the transition and not to the steady state. About 1.8 million people work in insurance and provider billing (PERI's estimate: roughly 800,000 in insurers, a million in provider administration); a just-transition fund on PERI's terms, a year of wage replacement plus $10,000 for retraining and $10,000 for relocation, is about $60 billion spread over the phase-in, and the Floor itself is what those households fall back on. Hospitals that live on private rates need a corridor: a five-year stabilization fund of about $20 billion a year for rural and safety-net hospitals, which is cheaper than the closures it prevents.

Housing: the largest new cost

The housing guarantee costs about $180 billion a year in net new public spending at full operation, range $140–230 billion. That is less than the roughly $300 billion a year the tax code already spends on homeowners, which is the comparison the pamphlet's second page should carry.

The guarantee is one promise delivered by three instruments. Nobody sleeps outside or in a shelter who wants a home (Housing First). Nobody pays more than 30% of income for a modest home (the rent guarantee). And the homes exist to be guaranteed (a public building program). The order matters: the third instrument is what keeps the second from being a rent subsidy that landlords capture.

Housing First. On one night in January 2025, 745,652 people were homeless, the first fall since 2016, but the count of chronically homeless individuals, 155,750, was a record. More than 1.3 million people used a shelter at some point in 2022, so the yearly flow is roughly double the nightly count. Permanent supportive housing costs about $20,100 a household a year and rapid rehousing about $8,500; the National Alliance to End Homelessness put the cost of offering Housing First to every shelter household at $9.6 billion a year, and the figure here adds the unsheltered and outreach. Houston cut homelessness 62% in a decade this way. The offsets are real but should not be sold as free money: the Community Guide's review found $1.80 of public benefit per dollar in U.S. studies, while RAND's 2025 review notes that the most rigorous designs find smaller offsets. The money saved lands in county hospitals and jails, not in the program's budget, so the bill below counts the full cost. Finland is the other warning: after eleven years of decline its homelessness rose 20% in 2025 once funding was cut. A guarantee has to stay funded to stay a guarantee.

The rent guarantee. Today 22.7 million renter households (49%) pay more than 30% of income for housing and 12.1 million pay more than half; 13.8 million income-eligible households get no assistance, because vouchers are a lottery that reaches about one eligible household in four. The Floor makes the voucher an entitlement: any resident household whose rent for a modest home (HUD's fair market rent, $1,465 a month for one bedroom on the 2025 national housing wage) exceeds 30% of income receives the difference, set from tax data, portable, with landlords required to take it. Fully funding vouchers for every household now eligible would add about $118 billion a year (Urban Institute, 2023) and lift 6.4 million people out of poverty; the line below adds a thinner tier for moderate-income renters up to 80% of area median. Rents are a dial here too: the 3.87 million homes vacant and for rent in the third quarter of 2026 (a 7.6% rental vacancy rate, up sharply from 2022) mean many markets can absorb the demand now, and the hot ones are exactly where the building program goes first.

The building program. Every serious estimate of the shortage lands between 3.7 and 4.7 million homes (Freddie Mac 3.7M, Zillow 4.7M, Up for Growth 3.8M, Realtor.com 4.0M). A federal Housing Development Authority, on the design of the 2024 Homes Act ($30 billion a year, a revolving fund, up to 1.3 million homes, rents capped by income), is sized here to 2 million permanently affordable, mixed-income homes over ten years: 200,000 a year, about 13% on top of the 1.5 million homes completed in 2025. At $250,000–350,000 a home all-in (Texas builds market-rate apartments near $150,000 a unit, California near $430,000), that is $50–70 billion a year of construction, financed by the Authority's bonds and repaid from rents; the budget carries a $30 billion yearly capital grant for the deeply affordable share plus $5 billion for repairing the public housing that exists. Vienna, the model, houses more than 60% of its residents in municipal or limited-profit homes. As these homes come online with rents set by income, the rent guarantee's subsidy bill falls; the two lines trade against each other over the decade.

What the tax code already does for housing. The mortgage interest deduction cost $52.6 billion in tax year 2025 (up from $25.6 billion after the 2025 law raised the SALT cap); the property-tax deduction about $29 billion a year and the exclusion of home-sale gains about $63 billion a year on JCT's 2025–29 estimates; the untaxed imputed rent of owner-occupied homes $151.6 billion (Treasury, fiscal 2025). Call it $300 billion a year for people who own, against $38.4 billion for vouchers. The Floor's housing bill fits inside the owners' subsidy with room to spare.

Capacity. Construction employed 8.36 million people in September 2026, and the industry says it needs 349,000 more in 2026 before any new program. At the industry's rule of thumb (about 3,450 jobs per $1 billion), 200,000 homes a year needs 170,000–240,000 workers, which is why the program ramps over four years, why the education pillar's apprenticeship fund points at the trades, and why tariffs that add about $10,900 to a home are a housing policy too.

Food: a card every resident holds

The food guarantee costs about $150 billion a year in net new public spending, on top of the $103 billion SNAP spent in fiscal 2025. The range is $120–190 billion, and the whole spread comes from one dial: where the credit tapers to zero.

What it is. Every resident holds a Floor card, usable for groceries and, for people without a kitchen, prepared meals. The monthly credit is USDA's Thrifty Food Plan for the household's size: $1,024 a month for a family of four in September 2026, about $256 a person. It is full for households at or below the poverty line, tapers to zero at 250% of poverty, and is set from the latest tax return, with an immediate reset when someone reports an income drop and reconciliation at tax time. No application, no work test, no asset test, no time limit. Sending every resident the full credit and recovering it through the income tax would cost the Treasury the same net amount; the taper is the same arithmetic with smaller gross flows.

Why it costs more per person than SNAP. SNAP paid an average of $189 a person a month in fiscal 2025 because benefits are cut by 30 cents for every dollar of counted income. The Floor pays the full plan to the poor and tapers above them. The starting line is also lower than it was: the July 2025 reconciliation law extended work requirements to age 64, shifted benefit costs to states from 2028 and cut $187 billion over ten years, and participation had already fallen by about 4.7 million people by March 2026.

The population shares come from KFF's 2024 distribution by poverty level applied to the Census's 341.8 million residents (July 2025). Tapering to zero at 200% of poverty instead of 250% takes the net to about $120 billion; tapering at 300% takes it to about $190 billion.

School meals. Federal lunch and breakfast programs cost $25.0 billion in fiscal 2025 and already serve 73% of lunches free or reduced-price; nine states have made meals free for every child. Covering the remaining paid meals nationally adds roughly $13 billion a year, close to the $11 billion the 2024 federal bill was priced at and well under the $30 billion sometimes quoted.

The waste side of the pillar. The U.S. generated 70 million tons of surplus food in 2024, 29% of the supply, worth $380 billion, while fewer than 13% of donatable surplus was donated. ReFED's full menu of fixes needs about $14 billion a year of investment for a decade and returns roughly $73 billion a year, mostly to businesses and households rather than the Treasury. So recovery does not fund the credit; it is this pillar's supply program, and the public share (grants for donation logistics and organics capacity, a date-label standard that costs almost nothing) is the $4 billion line above.

The gap being closed. In 2024, 47.9 million people lived in food-insecure households, 13.7% of households and the highest rate since 2014; 14.1 million of them were children. USDA ended that survey in 2025, so the Floor has to carry its own measurement: a yearly food-security survey is a $20 million line that keeps the guarantee honest.

Education: tuition-free public college and training

Tuition- and fee-free public college plus a training guarantee cost about $90 billion a year in net new public spending, range $70–110 billion. It is the cheapest pillar, and it is cheap for a reason the pamphlet already states: the Floor's other three guarantees pay for the room, the food and the doctor, so this one only has to pay for the seat.

What it is. Public two- and four-year institutions charge residents no tuition or fees, first dollar, with no income cap. The federal government pays each institution a per-student amount in place of the tuition it now collects from residents, and states keep their appropriations as a condition of the money: $130.7 billion in fiscal 2025, which the guarantee locks in rather than replaces. Pell continues ($38 billion, 7.3 million students) but as a living-cost grant, no longer a tuition grant. Training is the same guarantee at a shorter length: free credentials and short programs at community colleges (the Workforce Pell that began in July 2026 covers 150- to 599-hour programs; the Floor makes it universal and unconditional), plus an apprenticeship fund sized to triple the 702,000 active apprentices and roughly triple WIOA's $3.2 billion in adult, dislocated-worker and youth training.

Georgetown's Center on Education and the Workforce priced a first-dollar, no-income-cap design at $58.2 billion in its first year in 2016 dollars, about $77 billion today, which is where the $78 billion line sits. Their stronger claim, that the added tax revenue from more graduates exceeds the program's annual cost within ten years, held for the income-capped design they modeled; it is the best pays-for-itself claim in the whole package, and it is a model result, not an observation.

What the Floor makes unnecessary. Proposals for debt-free college run to $75 billion (2016 dollars) because they must cover living costs. Here those costs are the housing and food pillars' job, so the education guarantee stays close to the price of tuition itself. The $1.7 trillion owed by 42.3 million borrowers is the stock that this flow stops growing; what to do with the stock is a separate decision and is not in this bill.

Not in the base. The pamphlet's pillar runs from public K–12 through college. Extending it down to early childhood is the obvious next guarantee: universal pre-K for 3- and 4-year-olds is about $74 billion a year at 90% participation (NIEER), and a child-care guarantee capping family costs at 7% of income was scored by CBO at roughly $27 billion a year averaged over its first decade, more at maturity. Together they would add $100–130 billion a year; they are listed as an extension, not counted in the totals.

The full bill

The Floor costs the federal budget about $2.2 trillion a year at full operation, 7.2% of 2025 GDP, and costs the country's real resources roughly nothing, within a range of about one percent of GDP either way. Those two sentences are both true, and the pamphlet needs both.

The first column is what the funding section has to raise: $2.2 trillion against fiscal 2025 outlays of $7.0 trillion and receipts of $5.2 trillion. The third column is the money that stops leaving households', employers' and states' accounts when the public programs take over: premiums, medical bills, excess rent, tuition, part of the grocery bill. The last column is the difference once the health system's own savings are counted, and it is the answer to the question of whether paying for the Floor breaks the economy. The Floor is a redistribution of about seven percent of GDP through the budget, not a seven-percent increase in what the country consumes. The real new spending is homes built, meals eaten and seats filled, about 0.5% of GDP, set against administrative and price savings in health care of about 0.7% of GDP on CBO's middle options.

Two things are priced but left out of the total. Long-term care adds about $500 billion to the federal line and about $230 billion to total spending when it arrives in year six, which is why it is phased rather than promised on day one. Early childhood (universal pre-K and a 7%-of-income child-care cap) adds $100–130 billion if the education pillar is extended downward.

On the OECD's measure, the Floor takes U.S. public social spending from 19.8% of GDP to about 27%, roughly Germany's 27.9%, and leaves total social spending, public and private combined, about where it already is: 33.2% of GDP, the highest in the OECD. The country is already paying for a floor. It is buying insurance company overhead and emergency rooms with the money.

Paying for it

The $2.2 trillion is raised in two tiers: $1.5 trillion of contributions that replace the premiums and bills they abolish, and $0.7 trillion from a progressive package aimed at capital, wealth and tax expenditures. The margin is thin and the package says so. Yields are annual averages for 2025–2034 from CBO's own options where CBO has scored the measure, scaled from other estimates where it has not, and every line names its source.

Tier A: contributions that replace what they abolish ($1.52 trillion).

Tier B: the progressive package ($0.68 trillion).

Tiers A and B raise about $2.20 trillion against a bill of $2.22 trillion. The package therefore names two reserves rather than pretending the margin is wide: a carbon fee at $25 a ton rising 5% a year ($92 billion a year, with the Floor itself as the rebate), and a half-point on the household contribution (about $70 billion). It also names the alternative that every European floor actually uses: a 5% value-added tax on a broad base raises $338 billion a year, which could replace the wealth tax and the deductions cap together with no constitutional question and more regressivity. Which of the two a country chooses is a political decision; the arithmetic closes either way. What it does not do is close on the top 1% alone. Everyone pays in, in proportion to income, and nearly everyone gets more back than they pay. That is what universal contribution looks like, and it is the same principle as universal benefit.

The household ledger. A family earning $80,000 with employer coverage now pays $6,850 toward a $26,993 family premium plus out-of-pocket bills that average $1,630 a person, call it $11,000 a year; their employer pays another $20,143 they never see. Under the Floor the family's contribution is 4% of the $40,000 above the threshold, $1,600, and its bills are zero. The employer's 7.5% on an $80,000 salary is $6,000, so about $14,000 of compensation is freed, and CBO's modeling converts that to taxable wages over time: roughly $10,900 after tax. The family is between $9,000 and $20,000 a year better off depending on how much of the employer's saving reaches its paycheck, before any rent or tuition it no longer pays. A two-earner household at $250,000 pays $8,400 in contribution and loses most of the employer pass-through, roughly a wash against today's premiums, and gives up some itemized deductions. A household at $1 million pays about $38,000 in contribution plus the capital gains and bracket changes, and comes out behind. Above $50 million the wealth tax is the bill. The Floor is paid for by the top tenth and by the health industry's overhead, and the median household is its largest beneficiary.

Deficits. The package is pay-as-you-go at maturity: the federal deficit was $1.78 trillion in fiscal 2025 with net interest above $1 trillion and debt at 100% of GDP, and the Floor is not designed to add to it. The one borrowing is the housing program's construction, $50–70 billion a year of bonds serviced from rents, which is investment in an asset the public keeps.

Macroeconomics: work, wages, prices, growth

The evidence says the Floor would cut hours worked by a few percent, raise take-home pay by about a tenth, leave total U.S. social spending roughly where it already is, and move GDP somewhere between 1% below and 2% above its path, with the sign decided almost entirely by how the health pillar is financed. Collapse is not in the evidence. A financing mistake is.

The country already spends this money. On the public measure the U.S. is a low social spender: 19.8% of GDP against an OECD average of 21.2% and France's 30.6%. Count private spending net of taxes, as the OECD also does, and the U.S. is the highest social spender in the OECD at 33.2% of GDP, above France (32.0%) and ten points above the OECD average (22.6%). The Floor is mostly a change of channel: about 7% of GDP of private health premiums and out-of-pocket bills moves onto the public ledger, and the three non-health pillars add about 1.4% of GDP of new claims on real resources. Economies run at that level everywhere in Western Europe. The risk is the transition, not the destination.

Work. Every guaranteed-income experiment finds people work somewhat less, and none finds them stopping. In the 1970s negative-income-tax trials husbands cut hours about 7%, wives and single mothers about 17%. OpenResearch's $1,000-a-month study found participation about 4 points lower and 1.3 fewer hours a week, with the time going to caregiving, school and job search. Alaska's dividend produced no employment effect and more part-time work; Finland's experiment found employment slightly higher and mental strain lower; Denver's found no between-group differences at 24 months. CBO's modeling of single payer alone puts hours per person 3.6% to 4.7% lower in 2030, with a larger labor force over time as people live longer and after-tax wages about 11.5% higher because employer premiums become taxable pay. In-kind guarantees have smaller income effects than cash (a grocery credit cannot buy leisure), so the working estimate here is hours 2% to 4% lower, concentrated among caregivers, students, the sick and people over 60, which is who the pamphlet means. The labor-market effect that matters more is on the other side of the table: when nobody has to take any job on any terms, low-wage employers compete on pay and conditions, and the price of restaurant meals, retail and care rises once by a few percent. That is the hidden subsidy on the pamphlet's first page being repaid.

Growth. CBO's five single-payer systems, financed by trimming other government purchases so that debt is unchanged, leave 2030 GDP between 0.3% lower and 1.8% higher than its path: administrative savings of $533–562 billion (1.8% of GDP), employer health costs worth 7% of compensation converted to wages, better health. Financed instead by raising income or payroll tax rates, CBO finds GDP 1% to 10% lower, the 10% belonging to its most expensive design (long-term care included) paid for with higher marginal rates. The whole economic case for the Floor's financing lives in that gap. Today's premium is a flat per-head tax on jobs, $26,993 a family plan in 2025 whether the worker earns $30,000 or $300,000. Replacing it with proportional contributions, holding marginal rates on middle earners near where they are, and raising the rest from capital, wealth and tax expenditures that CBO's labor channel barely touches is what the revenue package below is built to do. The working estimate is 2030 GDP within 1% below to 1.5% above its path.

Prices. The starting point is not benign: consumer prices rose 3.7% in the year to September 2026 (core 2.6%, shelter 3.2%) and the Fed held at 3¾–4% after a September increase, so the phasing has to respect the cycle. In-kind delivery narrows the channel. The food credit adds perhaps $75 billion of grocery demand once substitution is counted, against a system carrying a 29% surplus. Single payer sets prices and is disinflationary by construction. The rent guarantee is the real inflation risk, since a subsidy on a fixed stock goes to landlords; the answers are the 7.6% rental vacancy rate it starts from, the building program that is funded in year one and delivers from year three, and the four-year ramp. The bottleneck is the trades: an industry already short 349,000 workers is why the education pillar's apprenticeship fund exists.

The window. Goldman Sachs puts 6% to 7% of U.S. workers in jobs AI can displace, with unemployment half a point above trend during the transition and productivity 15% higher at full adoption; Stanford's 2025 data already show a 13% relative employment decline for 22- to 25-year-olds in the most exposed occupations. A Floor whose benefits switch on the month income falls is an automatic stabilizer at national scale: in a downturn, spending on food and rent flows to the households that spend every dollar, and nobody loses coverage with a job. That is the macroeconomic version of the pamphlet's fourth page. The productivity dividend goes the same direction: the end of job lock, 68,000 fewer deaths a year on the Lancet model, graduates who pay more tax, children who stay in one school.

Distribution. Who pays is in the next two sections, but the shape is simple. Premiums become wages, so after-tax pay rises for everyone with a job. The top tenth pays more in tax than it saves in premiums. The health industry's administrative overhead, 1.8% of GDP, is the one group that loses outright, and the Floor itself is what catches its workers.

Transition: eight years, revenue first

The order is the policy. The revenue package is enacted before any guarantee switches on, the cheap and visible guarantees come in the first year, health takes four years, housing takes the longest because homes have to be built, and long-term care waits for the health system to be running.

Transition timeline (eight years, four pillars, revenue first): revenue package enacted 2027; Floor card, free school meals and free community college live in 2027; Housing Development Authority chartered 2027, Housing First surge 2027–2030, rent guarantee entitlement 2030, 200,000 public homes a year from 2029; health: drug prices set and uninsured enrolled 2027, Medicare for under-26s and over-55s 2028, everyone covered 2029, zero cost sharing with dental and vision 2030, long-term care from 2032; full operation January 2035.

Year one is deliberately front-loaded with the three guarantees that are cheap, fast and felt: the Floor card with the food credit, free school meals, and free community college, together about $170 billion, funded by the first revenue measures to take effect (the corporate rate, the top brackets, the deductions cap, the IRS). The Housing Development Authority is chartered the same year so that its first homes land in year three, before the rent guarantee becomes an entitlement in year four; a subsidy that arrives before supply goes to landlords. Health runs on its own four-year track and brings its own contributions with it: the employer and household contributions begin the year premiums end, so no household pays twice. Each pillar carries its revenue line into law in the same bill, and no pillar's benefits start before its money does. That is the discipline that keeps the package from becoming the inflation it is accused of.

Risks, sensitivities, and where this could be wrong

Six dials move the bill by more than everything else combined, and one design flaw has to be engineered out before the package is enacted. The table gives the swings; the paragraphs say what to do about them.

The stacking problem. This is the flaw. A family of four between the poverty line and 250% of it faces the food credit tapering at about 26 cents on the dollar, a rent guarantee that takes 30 cents of each new dollar as contribution, the 4% health contribution, and ordinary income and payroll taxes. Stacked, those can take 70 cents or more of a raise, on top of a system where families with children just above poverty already face a median marginal rate of 51%. The fix is a single Floor taper: one schedule, administered on the tax return, that caps the combined withdrawal of all three guarantees at about 25 cents on the dollar and stretches the tapers higher as needed. It costs roughly $40–60 billion a year more than the uncoordinated version and it is not optional; it is the difference between a floor and a trap.

Provider rates. Hospitals paid near Medicare rates lose money on patients who were privately insured, and rural and safety-net hospitals close first. The package picks a middle rate and a five-year stabilization fund. If the political settlement lands near today's private rates instead, the federal health line rises toward $1.9 trillion and the country's total saving shrinks toward zero, which is CBO's Option 4 and still a universal system.

Rent capture. A subsidy that arrives before supply goes to landlords. The package starts from a 7.6% rental vacancy rate, charters the building authority in year one, makes the rent guarantee an entitlement only in year four, and ties the subsidy to a modest-rent standard rather than to whatever the landlord asks. The sensitivity is still real: the guarantee's cost scales with the rents it is measured against.

Construction capacity. Two hundred thousand extra homes a year into an industry short 349,000 workers means wage pressure in the trades and competition with private building. Materials cost 40% more than in 2020 and tariffs add about $10,900 a home. The ramp is four years, the apprenticeship fund points at the trades, and the construction is bond-financed so that a cost overrun lengthens a repayment schedule rather than cutting a guarantee.

Revenue is scored conventionally. CBO's option estimates do not include macroeconomic feedback. Penn Wharton's dynamic analysis of the wealth tax cut its yield by about 15%; something similar could apply to the capital gains lines. The reserves exist for that reason, and the package should not be enacted without them.

Health costs grow faster than the economy. CMS projects health spending rising 5.4% a year against 4.1% for GDP through 2034, to 20.6% of GDP. A single payer's advantage is that it can set a budget; if it does not, the health contributions have to rise a tenth or two of a point every year to keep pace. Whether Congress holds a global budget against providers is the political test of this pillar.

States. The maintenance-of-effort payment and local delivery assume states participate. The package has a direct federal channel for the ones that refuse, as a dozen refused Medicaid expansion; it costs more to run and it should be in the law from the start rather than improvised.

Durability. Finland's homelessness rose 20% in 2025 after funding was cut; a guarantee is only as universal as its appropriation. The structural answers are the ones the pamphlet already gives: universality, mandatory rather than annual spending, dedicated contributions, and benefits indexed to the cost of the thing guaranteed rather than to a frozen dollar amount.

Measurement. USDA ended the food-security survey in 2025 and the homelessness count depends on volunteers one night a year. A guarantee that cannot be measured cannot be defended; the package budgets about $50 million a year for its own statistics.

What no number here captures. The pamphlet's deepest claim is that removing the survival question changes what people do with their minds. The experiments measure hours and well-being, and well-being rises everywhere; the larger effect, if it exists, is outside the evidence either way. This package does not count it as a benefit, and it does not need to in order to balance.

Text for page 2 of the Floor

The two placeholders in the "full bill" box, in the pamphlet's own register. The short version fits the box as drawn; the long version is a fifth page if the pamphlet grows one.

Short. Total annual cost: about $2.2 trillion in new public spending, 7% of GDP, of which $1.8 trillion replaces premiums, medical bills and state Medicaid payments that end. Net new cost to the country: close to zero. Primary funding: contributions that replace premiums (7.5% of payroll from employers, 4% of income above $40,000 from households, the states' Medicaid share), plus taxes on capital, wealth and tax breaks: a 28% corporate rate, higher top brackets, gains taxed at death, deductions capped, a wealth tax above $50 million, a rebuilt IRS.

Long. The Floor costs the federal budget about $2.2 trillion a year once it is fully running, seven percent of the economy. Most of that money is not new. About $1.8 trillion of it is the health care bill the country already pays through premiums, deductibles, copays and state Medicaid budgets, routed through one public payer instead of a thousand private ones, at a saving of roughly $200 billion a year in overhead and prices. The genuinely new spending is $420 billion: $180 billion for housing (an end to homelessness, a rent guarantee at 30% of income, and 200,000 public homes a year), $150 billion for food (a card every resident holds, free school meals), and $90 billion for tuition-free public college and training. It is paid for in two tiers. Employers contribute 7.5% of payroll and households 4% of income above $40,000, in place of the premiums both now pay; nearly every household comes out ahead, the median family by $9,000 to $20,000 a year. The rest comes from the top: a corporate rate of 28%, higher top brackets, capital gains taxed at death, a cap on itemized deductions, a wealth tax above $50 million, and an IRS rebuilt to collect the $600 billion a year that goes unpaid. Everyone pays in, in proportion to what they earn, and everyone is covered. That is the same principle twice.

Numbers on the pamphlet that have moved since it was drafted. The homelessness count: HUD's January 2025 figure is 745,652, the first decline since 2016, so "a record" now belongs to the chronic count (155,750) rather than the total. Vacancies: 3.87 million homes vacant and for rent in the third quarter of 2026, more than five per person homeless, with 3.47 million seasonal units on the same survey. Food waste: ReFED's 2026 report puts it at 29% of the supply, 70 million tons, inside USDA's 30–40%. Health spending: $15,474 a person and 18.0% of GDP on the CMS measure for 2024, $14,885 on the OECD's; the pamphlet mixes the two, which is fine if the sources line says so. Everything else on pages 2 and 4 checks out against the pages opened for this document.

Sources

Pages opened for the figures above, as of October 28, 2026. Where a figure is marked derived in the text, the arithmetic is this document's, applied to the cited source.

Health. CMS, NHE Fact Sheet (2024 data) · CMS, National Health Expenditure Projections 2025–34 · AMA, Policy Research Perspectives on 2024 NHE (Medicaid federal/state split) · CBO, How CBO Analyzes the Costs of Proposals for Single-Payer Health Care Systems (Working Paper 2020-08) · CBO, Economic Effects of Five Illustrative Single-Payer Health Care Systems (2022) · CBO, Federal Subsidies for Health Insurance, 2026 to 2036 · CBO, Baseline Projections: Medicare (Feb 2026) · CBO, Baseline Projections: Medicaid (Feb 2026) · CBO letter on expiring premium tax credits (June 2025) · KFF, 2025 Employer Health Benefits Survey · KFF, Medicaid spending after the 2025 reconciliation law · Himmelstein, Campbell and Woolhandler, Annals of Internal Medicine 2020 · PERI, Economic Analysis of Medicare for All (2018) · LAWCHA, Jobs and Medicare for All · Galvani et al., The Lancet 2020, via PNHP · OECD, Health at a Glance 2025 · NCHS, Mortality in the United States, 2024 · Sanders, Options to Finance Medicare for All

Housing. HUD, 2025 AHAR Part 1 release · HUD, 2022 Annual Estimates of Sheltered Homelessness · NLIHC, The Gap 2026 · NLIHC, Out of Reach 2025 · NLIHC, FY26 appropriations analysis · Harvard JCHS, The State of the Nation's Housing 2026 · CRS, Federal Rental Assistance and Affordability (2026) · CRS, Estimates of a Housing Shortage (2025) · CRS, The Mortgage Interest Deduction (2026) · Urban Institute, full funding and use of housing vouchers (2023) · Census, Housing Vacancy Survey Q3 2026 · Census, New Residential Construction, December 2025 · Census, housing costs 2020–2024 · NAEH, cost of Housing First for all shelter households (2025) · Community Guide economic review of permanent supportive housing (CDC Stacks) · RAND, Housing as Health Care (2025) · RAND, Why it is so expensive to build affordable homes (2025) · USICH, Houston's 62% drop · Varke, Homeless people in Finland 2025 · Homes Act (Smith and Ocasio-Cortez, 2024) · City of Vienna, housing policy · Novogradac on JCT tax expenditures 2025–29 · Treasury, Tax Expenditures, fiscal 2025 · BLS, Employment Situation, September 2026 · ABC, construction workforce shortage 2026 · NAHB, how tariffs affect home building

Food. USDA FNS, SNAP participation and costs · USDA, Thrifty Food Plan, September 2026 · USDA FNS, federal cost of school food programs · USDA ERS, Household Food Security in the United States in 2024 · USDA, termination of the food security survey (2025) · CRS, SNAP in P.L. 119-21 · CBPP, SNAP participation losses (2026) · FRAC, Healthy School Meals for All · ReFED, Progress on the Plate: 2026 U.S. Food Waste Report · KFF, distribution of the population by poverty level (2024) · Census, Vintage 2025 population estimates

Education. SHEEO, State Higher Education Finance FY2025 · Georgetown CEW, The Dollars and Sense of Free College · ED, FY2026 Congressional Justification: Student Financial Assistance · CRS, Workforce Pell Grants (2026) · Federal Register, PY2026 WIOA allotments · DOL, apprenticeship fact sheet (2026) · NIEER, funding pre-K for all · CBO, economic effects of subsidized child care and universal preschool (2021) · FSA, student loan portfolio update (Sept 2026)

Revenue and macroeconomics. BEA, GDP annual (FRED) · BEA, wages and salaries (FRED) · CBO, Monthly Budget Review: Summary for Fiscal Year 2025 · CBO, Options for Reducing the Deficit: 2025 to 2034, with the options cited: individual rates, corporate rate, VAT, financial transactions, assets at death, greenhouse gases, itemized deductions, AGI surtax · Saez and Zucman, letter on the Ultra-Millionaire Tax Act (2021) · Penn Wharton Budget Model, Warren wealth tax (2019) · Glogower, The Constitutional Limits to the Taxing Power, Fordham Law Review (2024) · Treasury, IRS funding revenue estimate (2024) · IRS, tax gap projections for tax year 2022 · OECD social expenditure, public, via Our World in Data · OECD net total social expenditure, via DBnomics · Vivalt et al., NBER Working Paper 32719 (OpenResearch) · Burtless, income maintenance experiments (Boston Fed) · Jones and Marinescu, Alaska Permanent Fund (AEJ: Policy 2022) · Finland, basic income experiment results · Denver Basic Income Project, 24-month results · ASPE, Marginal Tax Rate Series · BLS, CPI September 2026 · Federal Reserve, October 28, 2026 statement · Goldman Sachs, How will AI affect the global workforce (2025) · Brynjolfsson, Chandar and Chen, Canaries in the Coal Mine (2025)