There is a house in Akron that Patricia Osei has been trying to make safe for three years. The house belongs to her mother, Abena, who is 81 and has moderate Parkinson's disease and the early memory changes that her neurologist calls "mild cognitive impairment" and that Patricia calls "she left the stove on twice last month." Abena has lived in the house for 34 years. She raised Patricia and her two brothers there. She knows where everything is. She does not want to leave.
Patricia does not want her to leave either. She has read enough about nursing home costs — the median for a semi-private room in Ohio is $97,000 a year — to understand that the house is not just her mother's preference. It is also the only financially viable option. So Patricia has been making it work. She installed grab bars in both bathrooms herself, after watching four YouTube videos. She paid $2,200 out of pocket for a walk-in shower conversion when her mother started falling getting in and out of the tub. She bought a medical alert pendant ($35 a month), a medication dispenser with alarms ($180 upfront, $29 a month), and a ring camera for the front door so she can see when her mother leaves and whether she comes back. She drives to the house four days a week, 22 miles each way, to help with meals, laundry, and the medication management that the dispenser handles imperfectly.
She has applied twice for Ohio's PASSPORT Medicaid waiver, which funds home and community-based services for older adults who would otherwise qualify for nursing home care. Both times, Abena was assessed as not quite meeting the functional threshold. The assessor came to the house, asked Abena a set of questions, watched her walk to the kitchen and back, and determined that she could perform her activities of daily living with minimal assistance. The assessment did not ask Patricia how many hours a week she was providing that minimal assistance, or what would happen to Abena's functional status if Patricia stopped.
Patricia is 54 years old. She works full time as a hospital administrator. She has a husband and a 17-year-old at home. She has not had a vacation in two years. She is, in the language of eldercare policy, a family caregiver providing informal support that allows an older adult to age in place. She is also, in the language nobody uses in policy documents, the reason the system works.
The policy consensus and how it formed
Aging-in-place — the idea that older adults should be able to remain in their own homes and communities as they age, rather than moving to institutional care — is not a controversial policy goal. It is the stated preference of roughly 90 percent of Americans over 50, according to AARP surveys conducted over the past two decades. It is endorsed by the Administration for Community Living, the Centers for Medicare and Medicaid Services, the National Council on Aging, and virtually every state aging services agency in the country. It is the organizing principle of the Older Americans Act, which has been reauthorized repeatedly since 1965 with broad bipartisan support.
The policy consensus formed for good reasons. Nursing homes and assisted living facilities are expensive — expensive enough to exhaust most families' savings within two to three years, at which point Medicaid picks up the cost. Institutional care, when it is not excellent, can be isolating, infantilizing, and dangerous; the COVID-19 pandemic, which killed more than 200,000 nursing home residents in the United States, made that danger visible in a way that decades of advocacy had not. And the evidence on outcomes is reasonably clear: older adults who remain in familiar environments, near their social networks, with some control over their daily routines, tend to do better on measures of cognitive function, depression, and quality of life than those who move to institutional settings involuntarily.
So the policy goal is right. The problem is not the goal. The problem is the gap between the goal and the infrastructure required to achieve it — and the way that gap has been quietly filled, for decades, by families who were never asked to fill it and are rarely acknowledged for doing so.
What aging-in-place actually requires
The phrase "aging in place" sounds passive, like something that simply happens when an older adult stays put. It is not passive. It is a continuous, active, resource-intensive process that requires at minimum four things: a home that is physically safe and accessible, transportation to medical appointments and daily necessities, personal care assistance for activities of daily living, and care coordination across the multiple providers and systems that a person with multiple chronic conditions touches.
Each of those four things has a cost. The question is who pays it.
Home modification. The average American home was not built for an 80-year-old with balance problems. Stairs, narrow doorways, bathtub-shower combinations, and kitchen layouts designed for a standing adult all become hazards as mobility declines. The AARP Public Policy Institute estimates that a basic home modification package — grab bars, handheld showerhead, lever door handles, improved lighting, removal of trip hazards — costs between $3,000 and $8,000. A more comprehensive modification, including a roll-in shower, widened doorways, a first-floor bedroom conversion, or a ramp, runs $10,000 to $30,000 or more. A full home elevator, which is sometimes the only option for a two-story house with a resident who can no longer climb stairs, starts at $20,000.
Medicare does not cover home modifications. Medicaid covers them in some states, through Home and Community-Based Services waivers, but coverage is inconsistent, waitlists are long, and the scope of covered modifications varies widely. The Department of Veterans Affairs covers modifications for eligible veterans. For everyone else, the cost is out of pocket, or it is not done.
Transportation. An older adult who can no longer drive safely — which describes a significant and growing share of people over 75 — needs someone to take them to appointments, to the pharmacy, to the grocery store, to the social activities that prevent the isolation that accelerates cognitive decline. Medicaid covers non-emergency medical transportation for eligible beneficiaries, but the coverage is for medical appointments only, the scheduling is often unreliable, and the service is not available in many rural areas. Ride-share services exist but require a smartphone, a credit card, and the cognitive capacity to use an app — none of which can be assumed. In most cases, the transportation is provided by a family member, at the family member's expense in time and fuel, on the family member's schedule.
Personal care. Activities of daily living — bathing, dressing, grooming, toileting, transferring from bed to chair — are the core of what home health aides and personal care attendants provide. The median hourly rate for a home health aide in the United States is $33, according to the 2024 Genworth Cost of Care Survey. For someone who needs four hours of assistance a day, that is $48,000 a year. For someone who needs eight hours, it is $96,000. Medicare covers skilled nursing and therapy services at home for a limited period following a hospitalization or skilled nursing facility stay; it does not cover ongoing personal care. Medicaid covers personal care for eligible beneficiaries, but eligibility thresholds are strict, waitlists for HCBS waivers in many states run to years, and the workforce shortage in home care means that even families with funding often cannot find workers.
When paid care is unavailable or unaffordable, the care is provided by family. The economic value of that informal care — estimated by the National Alliance for Caregiving and AARP at $600 billion annually in the United States — is the largest single component of the long-term care system. It is also the component that appears in no federal budget, no state appropriation, and no insurance product.
Care coordination. A person with Parkinson's disease, hypertension, and mild cognitive impairment sees a neurologist, a cardiologist, a primary care physician, and possibly a geriatric psychiatrist, a physical therapist, and a speech therapist. Each of those providers generates orders, prescriptions, referrals, and follow-up requirements. Coordinating across them — making sure the neurologist knows what the cardiologist prescribed, making sure the physical therapy exercises are being done, making sure the follow-up MRI that was ordered three months ago actually gets scheduled — is a job. It is a job that, in a well-resourced system, would be done by a care manager or a social worker. In most American households, it is done by a family member, in the margins of their other work and life.
The cost-shift in plain numbers
The federal government's preference for home and community-based care over institutional care is not only a quality-of-life argument. It is a budget argument. A Medicaid-funded nursing home stay costs the federal and state governments a combined average of roughly $80,000 to $100,000 per year. A Medicaid-funded HCBS package — home health aide, adult day services, care management — costs roughly $30,000 to $50,000 per year for a comparable level of need. The difference is real and it is one reason that every state Medicaid agency has been working, with federal encouragement, to shift its long-term care spending from institutional to home-based settings.
What the budget comparison does not include is the cost of the informal care that makes the HCBS package work. The home health aide who comes for four hours a day is not providing 24-hour coverage. The other 20 hours are covered by someone — a spouse, an adult child, a sibling — who is not counted in the cost comparison because their labor has no price tag attached to it.
When researchers do attach a price tag — using replacement-cost methodology, which asks what it would cost to hire someone to do the same work — the numbers change the comparison substantially. A 2023 analysis published in Health Affairs estimated that for every dollar of Medicaid HCBS spending, informal caregivers contribute between $2 and $4 in equivalent unpaid labor. The policy savings from home-based care are real. They are also partially an accounting artifact that disappears when you count all the labor.
This is not an argument against home-based care. It is an argument for being honest about who is paying for it.
The families who hold the system up
Robert Tanaka is 61 and lives in Sacramento. His wife, Keiko, had a stroke four years ago at 58. She has left-side weakness, aphasia, and depression. She is home. Robert retired early — he had been a civil engineer — to care for her full time. He did not choose early retirement in any meaningful sense of the word "choose"; the math of hiring full-time care versus his salary made the decision for him. A full-time home health aide in Sacramento costs roughly $75,000 a year. Robert was earning $110,000. After taxes, the net gain from continuing to work was not enough to justify what it would have cost Keiko in continuity of care.
Robert now provides what he estimates is 60 hours a week of direct care: transfers, bathing, dressing, meals, medication management, speech therapy exercises, transportation to the three specialist appointments per month, and the emotional labor of being the primary human presence in the life of a 62-year-old woman who is grieving her own capacities. He has not calculated the dollar value of his labor. When asked, he says it doesn't matter because there is no alternative. That is the logic of a person who has internalized a cost-shift so completely that he no longer experiences it as a cost-shift. It is also the logic that the system depends on.
Diane Moreau is 48 and lives in rural Louisiana, 40 minutes from the nearest city. Her father, Claude, is 79 and has COPD, Type 2 diabetes, and the early stages of vascular dementia. Diane's mother died six years ago. Diane is the only one of Claude's three children who lives nearby; her brothers are in Texas and Georgia. She works as a school librarian, 10 months a year, and uses her summers to do the home modifications, the care coordination, and the stockpiling of supplies that keep Claude's situation stable during the school year. She has applied for Louisiana's Community Choices waiver twice. The waitlist, as of her last inquiry, had 3,200 people on it. The state has been working to reduce the waitlist for several years. It has not gotten shorter.
Diane drives Claude to his pulmonologist in Baton Rouge — 80 miles each way — four times a year. She manages his insulin, his inhalers, and the blood thinner he started last year. She has installed a hospital bed in his living room, a bedside commode, and a portable oxygen concentrator. She paid for the bed and the commode out of pocket; the concentrator is covered by Medicare. She estimates she spends $400 to $600 a month of her own money on Claude's care, not counting her time. Her salary as a school librarian is $44,000 a year.
What the infrastructure gap looks like from inside it
The HCBS workforce shortage is not a background condition. It is the central operational reality of aging-in-place for most American families. There are not enough home health aides, personal care attendants, and home care workers to meet current demand, let alone the demand that will arrive as the baby boom cohort moves through its eighties. The Bureau of Labor Statistics projects that home health and personal care aide positions will be among the fastest-growing occupations in the country through 2033 — which means the demand is growing faster than the supply, which means the gap is widening.
The workforce shortage is a wage problem. Median pay for home health aides is $15 to $17 an hour in most states — above minimum wage, but not enough to attract and retain workers in a labor market where Amazon warehouse jobs and fast-food management positions pay comparably and offer more predictable hours. The Medicaid reimbursement rates that fund most home care have not kept pace with inflation or with the wages required to build a stable workforce. States that have raised rates have seen workforce improvements; states that have not have seen agencies close, waitlists grow, and families absorb the difference.
The home modification gap is a financing problem. The programs that exist — Area Agency on Aging grants, CDBG-funded local programs, state weatherization and accessibility programs — are real but small, geographically uneven, and means-tested in ways that exclude the middle-income families who are too wealthy to qualify and too cash-constrained to self-fund. A family with a home worth $200,000 and $30,000 in retirement savings does not qualify for most means-tested modification assistance, but also cannot easily spend $20,000 on a bathroom renovation without depleting the savings that are supposed to cover emergencies. The result is that modifications happen late, after a fall, after a hospitalization, after the moment when they would have prevented the crisis.
The care coordination gap is a recognition problem. The work of coordinating care across multiple providers, managing medications, tracking symptoms, communicating with insurance companies, and making the hundreds of small decisions that keep a complex medical situation stable is clinical-grade work. It is performed, in most American households, by people with no clinical training, no institutional support, and no compensation. The healthcare system benefits from this work — it reduces hospitalizations, reduces emergency department visits, improves medication adherence — but does not pay for it, does not train the people doing it, and does not formally acknowledge that it is happening.
What an honest accounting would require
The policy goal of aging-in-place is worth keeping. The dishonesty is in pretending the goal is being achieved without naming the cost and who is bearing it.
An honest accounting would start with the caregiver assessment — a formal, structured evaluation of what the family caregiver is doing, how many hours, at what personal cost, and what would happen to the care recipient if the caregiver's capacity changed. Caregiver assessments are recommended by the National Academies of Sciences, Engineering, and Medicine. They are required by some state Medicaid programs. They are not universally required, not universally reimbursed, and not universally connected to any intervention that reduces the burden they identify. Making them mandatory and consequential — meaning that the assessment results actually change what services are offered and funded — would be a start.
An honest accounting would also require Medicaid HCBS waitlists to be treated as the policy failures they are, not as administrative facts of life. A person on a three-year waitlist for home care is not aging in place by choice. They are aging in place by default, held there by a family caregiver who is absorbing the cost of a system that has not built the capacity it promised. The Medicaid HCBS funding that has been authorized in recent years — including the investments made through the American Rescue Plan's HCBS enhancement — is real and meaningful. It is also not enough to close the gap, and it requires sustained political will to maintain.
An honest accounting would require a home modification financing mechanism that reaches middle-income families — not just the very poor and the very wealthy. A federal loan program, structured like the existing home improvement loan programs but specifically designed for accessibility modifications, with income-based repayment terms, would make the difference between a family that can afford a ramp and a family that cannot. The cost to the federal government would be a fraction of the cost of the hospitalizations that unmodified homes produce.
And an honest accounting would require naming the informal caregiver as a participant in the long-term care system, not as a private individual making private choices. Paid family leave that covers caregiving — not just parental leave — is one piece of this. Caregiver tax credits are another. Social Security credit for caregiving years, which reduce lifetime earnings and therefore lifetime benefits for the people who provide the most care, is a third. None of these are new ideas. All of them have been proposed, debated, and not enacted at the federal level, in part because the labor they would compensate has never been made visible enough to generate the political pressure required.
Patricia Osei is still driving to Akron four days a week. Abena is still in the house she has lived in for 34 years. The grab bars are holding. The medication dispenser is working most of the time. The walk-in shower has not produced another fall. The system, by every measure the system uses to count things, is working.
What the system is not counting is Patricia. It is not counting the 22 miles each way, the four days a week, the two years without a vacation, the $2,200 bathroom renovation, the $64 a month in monitoring subscriptions, the hours on hold with the PASSPORT program, the weight of knowing that the whole arrangement depends on her continuing to show up. It is not counting what happens if she gets sick, or her husband gets sick, or her 17-year-old needs more of her than she currently has to give.
The house is holding Abena. Patricia is holding the house. The policy is holding Patricia responsible for all of it, without saying so, without paying her, and without building the infrastructure that would make her job possible to sustain.
That is the arrangement. It is time to call it what it is.