The Architecture of Necessity

Fifty-nine point seven million Americans live under the same roof as two or more generations of family. Seventeen percent of all home purchases in 2024 were multigenerational — a record. And the $1.01 trillion in unpaid caregiving that holds these arrangements together is invisible to every measure of economic output that shapes policy. This is the story of why the most ancient form of human shelter is becoming the most urgent one.

I. The Return of the Oldest Arrangement

There is a version of this story that gets told as decline — as if the multigenerational household represents something going wrong, a retreat from the independence and prosperity the twentieth century promised. That framing has the history exactly backward.

The nuclear family — two married parents, their biological children, living independently — is not the ancestral form of human dwelling. It is a mid-century American anomaly, made possible by a specific convergence of postwar prosperity, suburban expansion, GI Bill homeownership, and a gendered economic bargain in which one income could sustain a household and one person, generally a woman, could manage the caregiving that family life requires. For most of human history, in most of the world, people lived in extended kinship networks. Grandparents raised grandchildren while parents worked. Adult siblings pooled resources. Elders were cared for by the people who knew them.

What is happening now is not a regression. It is a recalibration — forced in many cases by economics, enabled by remote work, and shaped by a caregiving crisis that the United States has steadily built toward for sixty years without ever deciding to address. The result is that 59.7 million Americans, representing 18% of the population, now live in households containing at least three generations. That figure was 7% in 1971. It has more than doubled in fifty years and shows no sign of reversing.

The forces driving this are structural, not sentimental. Housing has decoupled from wages. Institutional care has become financially catastrophic for most families. The aging of the Baby Boom generation has produced a caregiving demand that the formal care economy cannot absorb. And a generation of adult children — primarily Millennials — who delayed their own family formation for economic reasons now find themselves managing it simultaneously with elder care, inside the same compressed window of their late thirties and early forties.

Understanding the multigenerational home requires holding two things in tension simultaneously: it is, in aggregate, an arrangement that works — that produces meaningful benefits for children, parents, and elders — and it is also, in too many cases, an arrangement that extracts enormous cost from the people who make it function, most of them women, most of that cost invisible to the systems meant to support them.

II. The Numbers Describe a Country Reorganizing Itself

The scale of the multigenerational trend is not adequately captured by the headline figure of 59.7 million. What the data reveals, when examined carefully, is a phenomenon that is simultaneously older than it looks (the trend began well before COVID), broader than it appears (it crosses every demographic), and more structurally durable than its critics acknowledge.

A trend with deep roots

The modern multigenerational household trend traces to 1980, when the share of Americans in multigenerational homes bottomed out at 12% and began a long, uninterrupted climb. The 2008 financial crisis accelerated it: 5 million adults moved back in with parents between 2008 and 2011, the largest single-event spike before COVID. The 2020 pandemic produced a six-fold surge in doubled-up arrangements in a matter of months — by July 2020, 52% of adults aged 18–29 were living with a parent, the highest share since the Great Depression. But the pandemic didn't create the trend; it stress-tested an existing one and revealed that tens of millions of families were, at any given moment, one economic shock away from recombining.

What is notable about the post-pandemic data is that most of those who combined households during COVID did not immediately separate when it was over. A 2021 Pew Research survey found that 72% of people who had moved in with family during the pandemic reported the arrangement was working well enough to continue. A meaningful portion stayed.

The demographics of the multigenerational home

The pattern is not distributed evenly. Asian Americans live multigenerationally at a 26% rate, Hispanic Americans at 25%, and Black Americans at 24% — all roughly double the White rate of 13%. These gaps reflect real cultural differences: in many immigrant communities, multigenerational living is not an economic accommodation but an intergenerational expectation. The cultural capital embedded in those norms — the ease with which grandparents take on childcare, the absence of guilt in asking for help — represents a genuine advantage that policy thinking has largely failed to learn from.

But the White share of the multigenerational trend is growing fastest by volume. Of the total growth in multigenerational households between 2000 and 2021, 28% came from White households — not because the rate among White Americans is highest, but because they represent a large enough share of the population that even modest rate changes produce large absolute numbers. The multigenerational household is no longer a primarily immigrant or minority phenomenon. It is an American one.

The housing market has registered this shift with unusual clarity. The National Association of Realtors reported that 17% of all home purchases in 2024 were multigenerational — a record high, and up from 11% in 2019. That figure represents roughly $400 billion in annual real estate transactions shaped by the need to house multiple generations under one roof. Major homebuilders have responded: Toll Brothers, KB Home, and Lennar now all offer dedicated multigenerational floor plans featuring separate entrances, secondary kitchens, and ADU configurations. What was once a custom-build request has become a standard product category.

Why people choose — or end up in — these arrangements

Pew Research's 2022 deep-dive into multigenerational living found that financial reasons were cited by 54% of residents as a primary factor. Caregiving — for an elder, a child, or a person with a disability — was cited by 42%. But the categories overlap: the family that combines households because Mom needs care is also the family that discovers the rent savings make other things possible. The caregiving decision and the financial decision frequently arrive together.

What the data does not fully capture is the texture of the choice. The Reyes family portrait that opens this page is composite, but its emotional arc is real: a sequence of individually manageable decisions — a health scare, a cost calculation, a garage conversion — that accumulates into a complete reorganization of how a household works and what it expects of itself.

III. The Caregiving Economy Nobody Accounts For

The most important number in understanding the multigenerational household is not 59.7 million. It is $1.01 trillion.

That is the value AARP placed, in its 2026 Valuing the Invaluable report, on the unpaid family caregiving provided by Americans to aging relatives — a figure that exceeds total Medicaid spending and represents a workforce of 63 million people contributing an average of 27 hours per week of labor that does not appear in GDP, is not covered by Social Security earnings records, and generates no employer match on a 401(k). When economists discuss the "labor market," they are describing an economy that has been offloading its most essential work onto unpaid family members, predominantly women, for the entirety of the modern era — and calling it a personal choice rather than a structural subsidy.

The 45% increase in caregiving Americans over the past decade reflects the aging of the Baby Boom. More than 4.1 million Americans are turning 65 each year — the peak of "Peak 65," a demographic phenomenon that will continue through 2027. By 2034, adults over 65 will outnumber children under 18 for the first time in American history. The number of Americans over 80 will nearly triple by 2050. The country is aging faster than any care infrastructure it has built.

The cost of the alternative

The financial case for multigenerational living becomes stark when set against institutional care costs. Nursing home care (private room) hit a national median of $108,000 per year in 2024, up 25% since 2019. Assisted living averages $64,000 annually. Home health aide costs run approximately $75,000 per year at full-time equivalent rates. Memory care facilities — which house roughly 1.2 million Americans with dementia — average $6,000 to $7,500 per month.

Against these figures, the calculus of the multigenerational home is transparent. A garage conversion costs $25,000 to $75,000. An ADU addition runs $100,000 to $200,000. Even at the high end, a family that avoids three years of memory care costs has saved $150,000 to $270,000 — net of the construction cost, net of the friction, net of the 27 hours per week of labor they are now providing themselves.

Only 3% of Americans carry long-term care insurance. The average person who shops for it at age 60 pays $3,000 to $5,000 per year for a policy that may cover $200 to $300 per day. The math discourages most buyers. The result is that an overwhelming majority of American families will face catastrophic care costs with no financial product to absorb them — and will make the multigenerational calculation not from choice but from necessity.

What the care infrastructure looks like from inside it

The Surgeon General's 2025 caregiving crisis report noted that family caregivers provide an average of 27 hours of care per week while simultaneously maintaining employment. For the 16 million Americans identified as sandwich generation caregivers — simultaneously supporting aging parents and raising children — the load is closer to 50 hours per week of combined caregiving: roughly 22 hours for aging relatives and 28 for children.

These are hours that don't show up in anyone's schedule as "work." They show up as early departures, late arrivals, phone calls taken in parking lots, vacations canceled, promotions declined, and retirement savings accounts that grow more slowly or not at all. A 2025 Allianz Retirement Study found that 60% of sandwich caregivers had reduced or stopped retirement contributions to manage caregiving costs, and 47% reported being unable to meet essential expenses. The estimated lifetime earnings and Social Security benefit loss for women caregivers is $320,000.

This is the architecture of the multigenerational home: a financial arrangement that appears to save money at the household level while redistributing a very specific set of costs — opportunity costs, health costs, retirement security costs — onto the individuals doing the caring.

IV. The Sandwich Generation's Impossible Position

No group occupies a more structurally difficult position in this transformation than Millennials, born between 1981 and 1996. They are the generation that delayed marriage and children by four to seven years compared to their parents, largely for economic reasons — student debt, unaffordable housing, precarious early career labor markets. They are now in their late thirties and early forties, finally moving through the family formation years they deferred, and doing so simultaneously with the demands of a parent generation beginning to require serious care.

The 46% of Millennials who identify as sandwich caregivers in the 2025 Allianz study is not a coincidence of timing. It is the structural consequence of two overlapping demographic waves: Millennials' delayed family formation finally producing young children in their late-thirties households, and Baby Boomers aging into their late-seventies and eighties while their children are still in the thick of active parenthood.

What this looks like in practice

Gallup's 2025 State of the Global Workplace survey found that among full-time employed sandwich caregivers, average weekly work hours plus caregiving hours plus commute time totals 74 hours — leaving roughly 10 hours per day for sleep, meals, exercise, and everything else. This is not sustainable as a long-term condition. But it is a condition millions of families are in right now, and the policy response to it — episodic, inadequate, and mostly invisible — has not kept pace.

Millennial sandwich caregivers are also carrying the consequences of the multigenerational housing decisions they made during the pandemic. Forty-eight percent of Millennials lived in suburbs by 2020; 54% of homebuyers aged 31–40 bought in suburbs or secondary cities, many of them purchasing specifically because remote work made the geography viable. They structured mortgages, childcare, and family architecture around not commuting. When RTO mandates arrived, or when a parent's health declined and required proximity, the geometry of those decisions became suddenly costly to rearrange. The multigenerational home is frequently not a new decision but the revelation of a decision that was already embedded in earlier ones.

The retirement time bomb inside the sandwich

The financial consequences of sandwich caregiving on retirement security have not received adequate policy attention. The 60% of sandwich caregivers who have reduced or stopped retirement contributions are not making an irrational choice in the moment — they are managing a cash-flow crisis with the tools available to them. But the consequence is a retirement security gap that will compound for decades.

A person who earns $80,000 per year and suspends their 401(k) contribution for five years loses not just the $20,000 they would have contributed (at a 5% savings rate plus 3% employer match) but the thirty-year compounding on that base. At a 7% average annual return, five suspended years starting at age 38 represents roughly $200,000 in foregone retirement savings by age 65. For the woman who also exits the workforce for two years to manage intensive care — losing $160,000 in wages, employer benefits, and Social Security credits — the lifetime financial impact is structurally different from any financial event the system was designed to absorb.

V. What the Research Actually Shows About Life Under One Roof

The policy debate about multigenerational living tends to flatten it into either celebration (the family is taking care of its own) or pathology (poverty and precarity forcing unnatural arrangements). The academic literature is more precise, and more useful, than either framing.

Children in multigenerational homes

The most comprehensive review of evidence on child outcomes in multigenerational households — Sadruddin, Ponguta, and Zonderman's 2019 systematic analysis screening over 12,000 studies — produced a finding that cut against common assumptions: teenagers in multigenerational homes with single mothers had outcomes "at least as good and often better" than those in two-parent nuclear households. The mechanism appears to be the presence of an additional caring adult — typically a grandparent — who provides both material support (childcare, food security) and relational continuity (stability, mentorship, a second source of emotional attunement).

Collaborative grandparent-parent caregiving — where grandparents supplement but do not replace parental care — consistently produces the strongest child development outcomes across studies. The relationship is conditional: it holds when grandparents are in reasonable health, when the role is chosen rather than imposed, and when household resources are adequate. It does not hold under custodial grandparenting, where grandparents become primary caregivers due to parental absence, incapacity, or incarceration. In those configurations — which affect approximately 2.7 million grandparents in the U.S. — the grandparent health and wellbeing toll is significant, with 68% of studied cases showing negative health outcomes for the caregiving grandparent.

Grandparents in multigenerational homes

The research on grandparenting and wellbeing is more nuanced than the either/or framing suggests. A 2022 PMC systematic review found that grandparenting in moderate doses — a few hours of childcare per week, connection without overwhelming responsibility — is associated with improved cognitive function, reduced depression, and lower mortality risk in older adults. The presence of purpose, routine, and intergenerational connection appears to be genuinely protective.

The threshold matters. Full-time custodial care reverses these benefits sharply. The distinction between "beloved elder contributing to household life" and "de facto primary caregiver with no relief" is not subtle in lived experience but is frequently blurred in how families actually negotiate these arrangements under pressure.

The Journal of Child Psychology and Psychiatry's 2024 meta-analysis by Wang and colleagues found that grandparental care was associated with better child mental health outcomes when it was structured as supplementary care, and worse outcomes — for both child and grandparent — when it was total care driven by parental absence. The policy implication is that the multigenerational home works best as a network, not a substitution.

The satisfaction paradox

Survey data on multigenerational household satisfaction is uniformly high — 98% of residents in the Generations United survey describe their arrangement as "functioning successfully." This figure is real but requires interpretation. "Functioning successfully" is not the same as "optimal." Sociologist Vern Bengtson's Intergenerational Solidarity Model, developed over four decades of research on family dynamics, identifies solidarity and conflict not as opposites but as co-existing dimensions of intergenerational relationships. Families that report high solidarity also report high conflict at similar rates. The question is not whether multigenerational living produces friction — it does — but whether the benefits are understood as worth the friction by the people experiencing both.

The more honest framing comes from Jennifer Glass's landmark 2016 study in the American Journal of Sociology, which found that the U.S. had the largest wellbeing penalty for parenthood of 22 developed nations studied — and that the penalty was "entirely explained by the presence or absence of social policies allowing parents to combine paid work with family obligations." Countries with strong family support infrastructure did not show the same penalty. In countries where parents are systematically supported — paid leave, subsidized childcare, caregiver tax relief — the arrangement of family life does not consume the people doing it. The finding does not apply only to parenthood. The same logic applies to multigenerational caregiving: the cost is a policy variable, not a fixed feature of human nature.

VI. The Structural Forces That Will Deepen This Trend

The multigenerational household is not a pandemic anomaly or a recession artifact. It is the leading edge of a structural transformation that will intensify over the next fifteen to twenty years regardless of economic conditions. Four forces are driving the intensification.

Peak 65 and the care demand curve

The Baby Boom generation — 73 million people born between 1946 and 1964 — is aging at a rate the care system was not designed to absorb. More than 4.1 million Americans are turning 65 each year through 2027. By 2030, all Boomers will be 65 or older. By 2034, adults over 65 will outnumber children under 18 in the United States for the first time in recorded history.

The math of dementia alone is staggering. Alzheimer's and related dementias currently affect approximately 7 million Americans; that figure is projected to reach 13 million by 2050. The average duration of dementia care from diagnosis to death runs eight to ten years. Memory care facilities cost $6,000 to $7,500 per month. Medicaid covers long-term care only after a person has spent down virtually all assets. Medicare covers short-term skilled nursing and nothing resembling long-term custodial care. The gap between what families need and what institutional systems provide will be filled — by families, in their homes, at their own cost.

The elder orphan horizon

Perhaps the most underreported demographic story of the next generation is the projected growth of what geriatric researchers Maria Carney and Carol Mintz first identified in a 2016 PMC paper as "elder orphans" — older adults who age without a spouse, children, or close family contacts. Carney and Mintz estimated that 22.6% of adults over 65 were already at high risk of elder orphan status. Demographers Rachel Margolis and Ashton Verdery, modeling fertility and mortality trends forward, project that by 2060, approximately 21 million Americans over 50 will be without a living partner or biological children.

This figure is the demographic consequence of the family-formation shifts documented in the companion research brief. The generation that is currently delaying or forgoing children — today's adults in their twenties and thirties — will be in their seventies and eighties in the 2060s. They will not have children to call on when they need care. The multigenerational household, as a caregiving solution, depends on having multiple generations. A generation that declines to form the next one leaves a structural gap that the formal care system — already strained beyond capacity — will be expected to fill.

Remote work as architectural enabler

The stabilization of remote and hybrid work at roughly 22–23% of the workforce has had a profound and underappreciated effect on multigenerational household viability. Before 2020, the multigenerational home required at least one household member to either sacrifice labor market participation or commute from a geography determined by family need rather than employment. The sandwich generation caregiver who needed to be near an aging parent in suburban Ohio could not hold a San Francisco software job.

Remote work dissolved that constraint. The 35–44 age cohort — the precise cohort most likely to be in sandwich caregiving situations — is now the most remote-work-concentrated age group in the labor market, with 27.4% working fully or primarily remotely. Advanced-degree holders work remotely at a 42.8% rate. These are exactly the workers whose income makes multigenerational household economics function. The confluence is not coincidental: the families who most need the arrangement are now, more often than before 2020, able to maintain it without sacrificing employment.

ADU permitting data tracks the same story. Accessory dwelling unit permits — the garage conversions, backyard cottages, and in-law suites that are the physical infrastructure of multigenerational living — have risen dramatically in California, Texas, Florida, and North Carolina since 2020. The multigenerational home is not just a living arrangement; it is a building type that the housing industry has begun to produce at scale.

The immigration variable

The United States' multigenerational trend is not demographically self-sustaining. The native-born fertility rate — 1.60 total births per woman in 2024 — falls well below the 2.1 replacement threshold. Without immigration, the working-age population would have begun declining in 2012. In 2023–2024, net international migration of 2.8 million people accounted for 84% of U.S. population growth.

Immigrant households live multigenerationally at roughly twice the rate of native-born households across income levels. The cultural infrastructure of multigenerational living — the normative expectation that parents will live with adult children, the absence of stigma in asking for family help — is often more intact in immigrant communities than in native-born ones. As immigration drives a larger share of American demographic growth, its cultural patterns will shape what American households look like. The multigenerational home is already the majority arrangement in many immigrant communities; it may become the majority arrangement in America more broadly as those communities grow in relative size.

VII. The Policy Gap and What Fills It

The United States does not have a multigenerational household policy. It has a collection of programs designed for other purposes that interact, often poorly, with the reality of how millions of families actually live and care for each other.

Social Security does not credit unpaid caregiving years. A person who exits the workforce for five years to care for an aging parent receives lower lifetime benefits than if they had continued working — not because they contributed less to society, but because the contribution they made was invisible to the earnings record. Medicare does not cover custodial long-term care. Medicaid covers it only after asset spend-down that wipes out multigenerational wealth building. The Dependent Care FSA caps at $5,000 annually — a figure established in 1986 that has never been indexed to inflation and now covers roughly two weeks of full-time childcare in high-cost states.

The Dependent Care Tax Credit, theoretically available for elder care as well as childcare, phases out above $43,000 in household income — precisely the income range of the Millennial sandwich caregivers who most need it. The Family and Medical Leave Act provides up to twelve weeks of unpaid leave for elder caregiving; unpaid leave is inaccessible to most caregivers without substantial savings. Eleven states have implemented paid family leave programs with varying levels of adequacy; thirty-nine have not.

What works elsewhere

The global policy laboratory offers sharp lessons that the American debate largely ignores.

France sustains a total fertility rate of 1.80 — the highest in the European Union — through a comprehensive family support system spending nearly 4% of GDP on family-related programs. Its crèche system provides heavily subsidized infant and toddler care. Its parental leave structure provides near-full salary replacement. Its approach treats the combination of paid work and family life as a social problem with a social solution, not a personal management challenge.

The Nordic countries offer a different lesson. Sweden, Denmark, Norway, and Finland have long had the most generous family policy infrastructure in the developed world — generous parental leave, heavily subsidized childcare, strong work-life balance norms. Yet since 2008, Nordic fertility has declined steeply. Finland has fallen below the EU average. The Institute for Family Studies has characterized this as "The New Nordic Paradox": even generous welfare states may not fully compensate parents for the total cost and opportunity cost of child-rearing in high-productivity, high-expectation labor markets.

The most important lesson from South Korea — which has spent over $270 billion on pronatalist policies since 2006 and watched its fertility rate fall from 0.98 to 0.72, the lowest recorded anywhere — is that financial incentives directed at the decision to have children are nearly powerless against structural conditions that make having children genuinely incompatible with the lives people are already living. Money helps at the margin. It does not overcome housing costs five times median income, childcare costs equivalent to a second rent, and a care system that expects families to perform its most essential functions for free.

What the evidence most consistently supports is not a pronatalist policy or an anti-natalist one, but what demographer Lyman Stone calls a "fertility gap" strategy: identifying the difference between the number of children people say they want and the number they have, and removing the structural obstacles to that gap closing. Americans consistently report wanting 2.5 children on average while having 1.6. The gap is not a values story. It is a policy one.

VIII. Technology as Partial Answer and Partial Risk

The eldercare technology sector has grown rapidly in response to the care demand crisis. The global eldercare robot market, valued at $3.14 billion in 2025, is growing at 12.5% annually. AI companion products — Israel's ElliQ, South Korea's Hyodol, Japan's PARO robotic seal, and a growing array of smartphone-based companions — are being piloted in assisted living facilities and, increasingly, in private homes.

The evidence on AI companionship for older adults is early and mixed. Positive findings include reduced self-reported loneliness, improved medication adherence, and cognitive stimulation effects that appear to correlate with maintained function over time. Negative findings include concerns about dependency, the substitution of technology for human contact rather than supplementation of it, and the particular risk that families use the availability of AI companions to feel less obligated to provide human presence.

Remote monitoring technology — fall sensors, medication dispensers, home-based vital sign tracking — has stronger evidence for genuine practical value in multigenerational households where adult children need situational awareness about a parent's safety without being physically present. The combination of sensors that alert to falls with video connection that enables human response represents a genuine improvement in the quality of multigenerational household management.

The more speculative frontier involves humanoid robots. Boston Dynamics, Figure, and Apptronik are all developing general-purpose humanoid platforms that could theoretically perform domestic care tasks — helping an elder transfer from bed to chair, preparing meals, managing medications. None of these are commercially available for home elder care as of early 2026, and the cost, reliability, and ethical questions involved in deploying robots as primary caregivers remain substantial. The care workforce shortage projected at 13.5 million across OECD nations by 2040 is real and growing; whether technology can close a gap of that magnitude without sacrificing the relational quality that makes care effective is among the most important design questions of the next decade.

IX. Chosen Family and the Emerging Alternatives

The multigenerational household is the most common adaptive response to the structural pressures this report has described, but it is not the only one. A set of alternative arrangements is developing that may represent early signals of how American households will reorganize over the next generation.

Cohousing — intentional communities of typically eight to thirty households sharing common spaces while maintaining private dwellings — has 134 established communities in the United States as of 2025, with 176 in formation. The model, imported from Denmark in the 1980s by architect Kathryn McCamant, has remained a niche choice; its growth has been slow relative to the scale of the housing and care challenges it addresses. But its demographic composition is shifting: where early cohousing communities were predominantly young families with children, new formations increasingly include aging cohorts planning for elder care mutual aid.

Chosen family networks — the relational structures documented in LGBTQ+ communities by anthropologist Kath Weston in her 1991 ethnography Families We Choose — have expanded well beyond their origins. For adults whose biological kinship networks are strained, distant, or absent, chosen family offers a parallel infrastructure of support and obligation. As family sizes shrink and geographic mobility continues to scatter biological kin, chosen family becomes not a cultural identity but a practical necessity for a growing share of the population.

The structural challenge facing both cohousing and chosen family as caregiving solutions is the same: they depend on proximity, commitment, and the slow accumulation of relational trust that human care requires. They are not scalable in the way that policy is. They are not deployable in the way that technology is. They are exactly as difficult to build and maintain as any human relationship — which is to say, very.

X. What the Multigenerational Home Reveals About What We Value

The deepest argument embedded in this data is not demographic. It is philosophical.

The United States has organized its economy and its policy on the premise that care — of children, of the elderly, of the disabled, of the ill — is a private family matter. The consequences of that organizational choice are not incidental; they are the subject of this report. A trillion dollars of unpaid labor. Sixty percent of sandwich caregivers foregoing retirement savings. Women forfeiting $320,000 in lifetime earnings and Social Security benefits. An estimated 21 million Americans projected to age without family support by 2060.

These outcomes are not the result of families failing to care for each other. The 63 million Americans providing unpaid family care are evidence of the opposite — of people absorbing extraordinary personal cost to meet obligations that no formal system will cover. The outcomes are the result of a society that has decided, through policy and through the repeated choice not to make policy, that this labor is not worth paying for.

The multigenerational home is, in its best versions, something genuinely valuable: an arrangement in which multiple generations share space, resources, and responsibility in ways that benefit all of them. Children grow up knowing their grandparents. Elders remain embedded in living family life rather than warehoused in institutional care. Adults in the middle generation receive support as well as provide it. These benefits are real and documented.

What they require to be sustainable — rather than extractive — is the same thing that the French parenting research revealed about parenthood: that the arrangement work or fail is determined not by human nature or family willingness but by the presence or absence of structural support. The multigenerational home that functions as a chosen, resourced, negotiated network of mutual care is among the most human-scale solutions to the problems of aging, isolation, and economic precarity that modern life has produced. The multigenerational home that functions as the last resort of a family that ran out of other options is the same physical arrangement with a very different emotional interior.

The policy question is not whether Americans should live multigenerationally. Fifty-nine point seven million of them already do. The question is whether the society they live in will be organized to support what they are doing — or whether it will continue to rely on their unpaid labor while accounting for it as nothing.