The Caregiver's Wage

The problem

In March 2026 the AARP Public Policy Institute published its twentieth-anniversary accounting of family caregiving in the United States. The findings, in three numbers:

59 million Americans caring for an adult. 49.5 billion hours a year. $1.01 trillion in value, if the same work were purchased in the market.

That labor is equivalent to 23.8 million full-time workers — about 17 percent of the entire American full-time workforce, employed invisibly, without wages, benefits, or a payroll line. It exceeds what all private businesses in America spent on health care in 2024 ($967 billion) and it exceeds the entire Medicaid program ($932 billion). Twenty years ago, the first report in this series estimated the figure at $350 billion. It has nearly tripled.

Here is the part that makes it a policy failure rather than a sentiment. American long-term care is financed on the assumption that this labor is free and infinitely elastic. Medicaid will pay a stranger to do the work. It will pay an institution tens of thousands of dollars a year to do the work. In most of the country it will not pay the daughter who quit her job to do the work, and it will not count the years she spent doing it toward her own retirement. She will reach 67 with a hole in her earnings record, no employer pension, and a diminished Social Security benefit — a poverty risk manufactured by the same government that saved money on her.

The system does not merely fail to pay caregivers. It is funded by not paying them, and it converts today's caregivers into tomorrow's beneficiaries.

The proposal

Pay family caregivers a wage for certified hours of heavy care, credit those years to their Social Security record, and guarantee them respite.

The precedent is not theoretical and not foreign. The Department of Veterans Affairs has paid monthly stipends to family caregivers of seriously injured veterans since 2011 under the Program of Comprehensive Assistance for Family Caregivers. More than half the states already let Medicaid beneficiaries self-direct their care budgets and hire a family member. America already does this. It does it narrowly, inconsistently, and by accident of geography.

How it would work

1. A functional-need standard, not a diagnosis. Eligibility attaches to the recipient: a person who requires assistance with a defined number of activities of daily living — essentially the standard that already qualifies someone for institutional care. If the nation would pay a facility to do this, it will pay a family member to do it.

1. A wage for certified hours, capped. The caregiver is paid an hourly wage at the local home-health-aide rate for certified hours, capped at 30 hours per week. Capped, because this is a wage for work, not an income-replacement program, and because an uncapped entitlement is how good ideas die in committee.

1. Social Security caregiving credits. Years of certified caregiving generate earnings credits, so a caregiver's retirement is not confiscated to fund somebody else's. This costs comparatively little and repairs the most indefensible feature of the current arrangement.

1. A respite guarantee. A statutory minimum of paid substitute care per year. Caregiver collapse is the single most common route to institutionalization, which is both the human catastrophe and the expensive outcome.

1. Portable, not employer-linked. The wage rides with the caregiver under the Portfolio Act's benefit rails (№ 013), so caring for a parent for two years is not a career-ending gap.

The numbers

We will not pretend this is cheap, and we will not use the $1.01 trillion figure to imply savings that do not exist.

Target the wage at caregivers of people meeting the institutional level-of-care standard — roughly the population Medicaid home and community-based services already recognizes, on the order of 3 million caregiving relationships. At 25 certified hours a week and a $16 hourly rate, the direct cost is approximately $62 billion a year. Social Security credits add several billion more in deferred liability. Respite adds a few billion.

Call it $70 billion a year. That is a real number and a large one, and anyone proposing it owes the public a funding source rather than a hand-wave.

Against it, genuine offsets: the median annual cost of a semi-private nursing-home room now runs well over $100,000. Every 100,000 institutional placements deferred saves roughly $10 billion, and the effect of paid family care on delaying institutionalization is one of the better-supported findings in long-term-care research. Add reduced caregiver workforce exit — the labor-force participation gains are real but modest — and a plausible offset is $15 to $25 billion. Which leaves a net cost around $45 to $55 billion a year, unfunded.

Our answer is that this is what the Covenant is for, and that the honest place to fund it is general revenue, not the Fund. The American Permanent Fund pays an equal dividend to everyone; it is not a fiscal patch for programs that lack votes. Raiding it for a good cause is how permanent funds stop being permanent. If Congress will not fund the Caregiver's Wage, the correct outcome is that it does not happen — not that we quietly attach it to the endowment.

The honest objections

"A trillion dollars is an imputed number. Families do this out of love, and pricing it is a category error." The trillion is a replacement-cost estimate and should be read as one — it is what the market would charge, not what anyone would pay. But the counterfactual is not imaginary: when the caregiver stops, someone bills Medicaid at the real rate the next week. Love is doing the work. It is not doing the financing, and confusing the two is what lets a budget depend on it.

"Paying for care that is currently free will pull in enormous latent demand." This is the objection that determines whether the proposal is affordable, and it is correct. The woodwork effect is well documented in home and community-based services expansions: people who were coping unassisted appear the moment a benefit exists. Our cost estimate above assumes a substantially larger claiming population than today's HCBS caseload for exactly this reason, and it could still be low. The functional-need standard and the 30-hour cap are the controls. If the take-up exceeds projections, the honest options are to tighten the standard or raise the money — not to discover the overrun after enactment and cut the wage.

"This will pay women to leave the labor force and deepen the very inequality it addresses." The most serious objection in this entry. Roughly two-thirds of family caregivers are women, and a wage at a home-aide rate is a poor substitute for a career. Three partial answers: the respite guarantee and the 30-hour cap are designed to keep part-time employment feasible; the Social Security credits directly repair the retirement penalty that current policy imposes; and the relevant comparison is not "career versus caregiving" but "unpaid caregiving versus paid caregiving," because the caregiving is happening either way. We concede this does not fully answer it. A policy that makes an unjust allocation of labor more survivable has not made it just.

"Paying relatives invites fraud that paying agencies does not." Self-directed Medicaid programs and the VA caregiver program have both had oversight failures, and the VA's has been the subject of critical inspector-general findings. Certification of the recipient's functional need by an independent assessor, periodic re-certification, and hour verification are the minimum. We would rather build the audit into the statute than defend a scandal later.

"Institutional care exists because some conditions genuinely require it." Yes, and nothing here discourages institutional placement or pays families to attempt care beyond their competence. The wage applies to certified hours of a defined scope. The goal is to end the arbitrage where the state pays a stranger and not a daughter — not to declare that families should manage dementia alone.

Sources

• AARP Public Policy Institute, Valuing the Invaluable: 2026 Update — 59 million caregivers, 49.5 billion hours, $1.01 trillion, 23.8 million full-time-equivalent workers (aarp.org/pri)
• AARP, comparison to 2024 private business health spending ($967B) and Medicaid ($932B); 2006 baseline estimate of $350B
• U.S. Department of Veterans Affairs, Program of Comprehensive Assistance for Family Caregivers (va.gov/family-member-benefits)
• KFF, Medicaid home and community-based services participants, spending, and self-direction by state (kff.org)
• Genworth Cost of Care Survey, median nursing facility and home health aide costs
• Proposals № 004 (CitizenCare), № 013 (The Portfolio Act)