Future present, Future Perfected

Before you read another word here, I'd rather you read A Room of One's Own Devices first — a short story I wrote about a version of myself five years from now, a week out of wrist surgery, propped on the couch with my dog, figuring out how technology and a little community infrastructure let me recover at home instead of paying someone else to manage my decline. It's fiction. It's also the thing I kept picturing while I read the piece that prompted it.

That piece is a Washington Post investigation (if no subscription, here) into what growing old actually costs American families, and once I read it, I couldn't stop doing the math on my own life.

The number that stuck with me

The Post's analysis, drawn from a federally funded survey that has followed thousands of Americans from their 50s until death since 1992, found that the typical American spends around $19,000 out of pocket on care in the last decade of life. That's the median. One in six people spend more than $50,000. One in twenty spend more than $100,000. And the share of people who end up with nothing left after paying for their own care rose from about 6 percent among those who died in the late 2000s to nearly 11 percent among those who died in the early 2020s.

Those numbers actually undersell it, because they don't fully capture the cost of the roof over your head — the room-and-board charge at an assisted-living facility, which is usually the biggest line item of all. A private room in assisted living now runs a bit over $74,000 a year nationally. A nursing home, if you need one, is closer to $115,000 a year for a semi-private room and $130,000 for a private one. Stay seven years with a dementia diagnosis, and the bill approaches a million dollars.

I wanted to know how we got here, so I went looking for more than the Post had room for.

How the cost curve actually bent

I found a chart in my head before I found one on paper, so I built the paper one. Genworth and its successor CareScout have run the same private-pay cost survey since 2004, and older HHS actuarial data lets you push the line back further:

The median annual cost of a nursing home room went from around $8,600 in 1977 to roughly $115,000 in 2025 — and that's in nominal dollars, no inflation adjustment needed to make the point. From 2004 to 2024 alone, nursing home costs rose 138 percent, nearly double the 71 percent rise in general inflation over the same stretch. Zoom out further and it's just as stark: total U.S. spending on nursing home care was about $1 billion in 1960. By 2024 it was roughly $220 billion. Some of that is simply more people living longer. But the per-person cost of a single year of care has outpaced ordinary inflation for decades, and the industry itself points to the same two culprits the Post found: a shrinking supply of caregivers driving up labor costs, and a wave of baby boomers large enough to strain every part of the system at once.

A researcher I hadn't come across before the Post piece, Jessica Forden at the New School, put a sharper point on it than I expected from an academic paper: the idea of a $68 trillion "great wealth transfer" from boomers to their kids is mostly a story about the wealthy. For everyone else, the money gets spent on the way out the door, and the inheritance conversation quietly becomes a Medicaid-eligibility conversation instead.

The insurance question I've been avoiding

Here's my own confession, since I'm asking you to think about this alongside me: I don't have long-term care insurance, I've been meaning to for years. Researching this piece made clear that someday has a shrinking window.

Two different insurance questions get tangled together, so it's worth separating them.

The first is Medigap versus Medicare Advantage. Original Medicare doesn't cap what you pay out of pocket, so most people layer on either a Medigap policy (a fixed monthly premium that fills in the gaps — deductibles, coinsurance, the stuff Medicare doesn't cover) or a Medicare Advantage plan (a private insurer bundles your coverage, often for a low or zero premium, in exchange for a narrower network and more variable extra benefits like dental or vision). Medicare Advantage has grown fast — more than half of all Medicare beneficiaries are now enrolled in one — but 2026 has been a rough year for the category: several large insurers have pulled out of markets or cut benefits in response to lower federal reimbursements, and Milliman found the average plan's over-the-counter allowance dropped double digits. That's pushed a wave of people back toward Medigap, whose enrollment is projected to keep climbing through the early 2030s. Neither option touches long-term custodial care, though — that's the second, separate problem.

The second is actual long-term care insurance, and this is the part of the industry that's quietly collapsed. At its peak in the early 2000s, more than a hundred companies sold standalone LTC policies. Today it's down to roughly fifteen, and most of the household names that used to dominate the space — John Hancock, MetLife, Prudential — stopped writing new policies years ago after badly underpricing the original generation of them. What's replaced standalone LTC insurance is mostly "hybrid" policies: a life insurance or annuity product with a long-term-care rider attached, so the money isn't "wasted" if you never need care, but the upfront cost is steep — often a six-figure lump sum — and it's really only a fit for people with a few hundred thousand to a couple million in liquid assets. Below that, as Forden's research suggests, you're likely planning around Medicaid spend-down. Above a few million, self-insuring often makes more sense than paying an insurer's markup at all. It's the broad middle — which is most of us — that the market has essentially stopped serving well.

I don't have a tidy resolution to offer here, because I haven't made the decision myself yet. But I'd rather write the sentence "I need to actually get quotes this year" in public, where you can hold me to it, than keep filing it under someday.

Why I wrote the story I wrote

None of this is an argument that technology solves the underlying problem. The Post's numbers are structural — fewer children per family, more geographic distance between generations, a caregiver shortage that isn't going away, a for-profit senior living industry whose prices are rising nearly twice as fast as everything else. No sensor mesh fixes that.

What the story imagines is something narrower and, I think, more honest: that a fair amount of what currently gets billed at $3,000 to $5,000 a month — the supervision, the safety net, the routine monitoring that justifies moving someone into a facility or hiring someone into their home — can be handled by a combination of better-designed homes, wearable and ambient sensors, remote physical therapy, and, maybe most importantly, neighbors who've organized themselves to share the load. My own co-housing community has spent the last couple of years building exactly that kind of shared infrastructure, not because we're utopians, but because we all did the math on assisted living and didn't like the answer.

That's not a substitute for insurance, or for the country figuring out long-term care financing at a policy level, which frankly it needs to. But it's a lever an individual family can actually pull today, and it buys something the industry's pricing model doesn't sell: staying yourself, in your own house, next to the people who love you, for longer than the actuarial tables assume.

If you read the story first, I'd love to hear which part of it felt plausible and which part felt like wishful thinking. And if you've already been through this with a parent, or you're staring down your own version of it — I'd especially love to hear from you.

This is exactly the kind of story I want more of us telling out loud.