The Quiet Foreclosure

The Pattern Already Named

On June 6, 1978, California voters passed Proposition 13 by an approximately sixty-five percent margin.¹ The proposition, drafted by Howard Jarvis and Paul Gann, did three things: it capped property tax rates at one percent of assessed value, capped annual assessment growth at two percent or the rate of inflation (whichever was lower), and required reassessment to current market value only at sale or substantial new construction. Subsequent academic and legislative analysis has cataloged its unintended effects — the lock-in of long-tenured owners, the regressive subsidies to high-value coastal homes, the structural starving of municipal services — at length.² What is harder to find in the contemporary literature is a clear statement of what the proposition was a response to. It was a response to elderly Californians being driven out of homes they owned outright by property tax bills they could not pay.

The mechanism was not new and not local. Through the early and mid-1970s, California housing prices climbed quickly. Assessors, working under standard market-value principles, raised assessments to follow. A homeowner who had purchased a Los Angeles bungalow in 1955 for fifteen thousand dollars and now lived in it on a Social Security check found that the same house, in 1976, was assessed at sixty thousand dollars, with the tax bill rising in proportion. The homeowner’s income had not risen in proportion. The asset had appreciated, but the appreciation could not be eaten or banked or used to pay the assessor; it was paper. The bill, by contrast, was real. A constituency of paid-off but illiquid homeowners discovered that the right of continued occupancy in a house they legally owned was conditional on continuing to pay an annually rising tax against an asset whose appreciation they could not access without selling. Selling, for many of them, meant moving out of communities they had lived in for thirty or forty years, into rentals they could not afford in the same neighborhoods at the new market rate. Selling was therefore not a relief mechanism. It was the failure mode the tax bill was forcing them into.

The constituency that supported Proposition 13 was not, in the main, ideological. It was demographic. It was retirees, widows on fixed pensions, working-class homeowners in neighborhoods that had appreciated past their wages. The Jarvis-Gann coalition assembled them by giving the simplest possible name to what was happening: “they are taking our homes.” Critics pointed out, accurately, that no one was literally taking the homes; the homeowners were free to sell at the new prices and pocket the gain. The critics were technically correct and politically irrelevant. To the homeowner, what was being taken was the practical content of ownership — the ability to live in a house one had bought, paid off, and intended to die in. The legal title remained. The right it conferred was being eroded.

Proposition 13 passed and became a template. Massachusetts adopted Proposition 2½ in 1980, capping municipal levy growth at two and a half percent annually.³ Michigan adopted Proposal A in 1994, capping individual-property assessment growth at the lesser of inflation or five percent and uncapping on transfer.⁴ Florida adopted the Save Our Homes amendment by referendum in 1992, effective 1995, capping homestead assessment growth at the lesser of three percent or CPI; portability was added by amendment in 2008.⁵ Texas, Oregon, New York City, and a dozen other jurisdictions adopted partial caps in subsequent decades. Each of these provisions originated in the same constituency Proposition 13 had named: long-tenured homeowners being priced out of homes they owned by escalating assessments.

The pattern of legislative response was, in retrospect, partial. Where the political coalition existed, the cap was enacted. Where it did not, the cap was not. The states most exposed to the underlying problem — the states with the highest property-tax burdens and the most volatile housing markets — were not always the states that produced the reform; New Jersey and Illinois, with the highest effective property tax rates in the country, never adopted a cap of any kind.⁶ National-level discussion treated the issue as resolved by the partial state-level fixes that had been enacted. The federal government, which had no jurisdiction over local property tax administration, had no policy on it. The political class had nothing further to say. The constituency dispersed.

It is now forty-seven years since Proposition 13. The constituency has not gone away. It has grown, and the squeeze it identified has acquired two new amplifiers that did not exist in 1978. The first is the climate-driven repricing of property insurance, which is doing to the insurance side of the carrying-cost equation what the 1970s appreciation cycle did to the property-tax side. The second is the wage compression now being induced by the AI transition described in the previous essay,⁷ which is removing the income against which any of these costs must be paid. The mechanism the Jarvis-Gann coalition named in 1978 — the conversion of paid-off ownership into involuntary forced sale through escalating non-mortgage carrying costs — is operating at larger scale than at any point since the 1930s, in a country whose policy response remains stuck where it stopped in the early 1990s. The remainder of this essay describes the shape of the contemporary squeeze, the mechanism by which it transfers wealth, the failures of both major parties to address it, and the policy response that the next decade will either produce or refuse.

The Two Bills

The carrying cost of an American home, in plain accounting, has four components: principal and interest on any mortgage, property taxes, hazard insurance (and where required, flood and earthquake), and maintenance.⁸ The first component is, for the great majority of conventional mortgages, fixed at origination. A homeowner who took out a thirty-year fixed-rate mortgage in 1995 paid the same nominal monthly principal-and-interest in 2024 as in 1996. The remaining three components, in contrast, float. They float against assessed market value, against replacement cost in a climate-priced insurance market, and against the trade-and-materials inflation that drives roof, HVAC, and structural maintenance. The fixed component diminishes as a share of total housing cost over the life of the mortgage. The floating components grow.

For homeowners who have paid off the mortgage entirely — a category that includes a substantial fraction of the over-65 American population⁹ — the principal-and-interest line is zero, and the floating components are the entirety of the carrying cost. This is the population the 1978 California revolt was about, and it is the population for whom the contemporary squeeze is most exposing. A retiree who owns a home outright is, in the conventional understanding, “set.” She has no rent and no mortgage; her housing cost is supposed to be a small residual line item against her Social Security and pension. The conventional understanding has not kept up with the math. In high-tax states, in climate-exposed states, and in markets that have appreciated significantly since the homeowner’s purchase, the residual line item has become a primary household expense — in some cases the largest single expense. The assessor and the underwriter, operating independently and in good faith, have rebuilt the carrying cost the paid-off mortgage was supposed to have eliminated.

The mechanism by which property tax bills float is well-defined and rarely examined. American property tax administration, in nearly every jurisdiction outside the handful of cap states, operates on the principle of ad valorem assessment to current market value.¹⁰ An assessor’s office, working from comparable sales data and standardized appraisal methods, sets a market value for each property each cycle (annually in most jurisdictions, biennially or triennially in some). The tax rate, set by local authorities, is then applied. Two things happen. First, when neighborhoods appreciate, every owner’s tax bill grows in proportion to the appreciation, regardless of whether the owner intends to capture the appreciation by selling. Second, when neighborhoods gentrify around long-tenured owners — the classic case being a working-class neighborhood that becomes desirable as adjacent areas saturate — the long-tenured owners face the steepest assessment increases, because comparable-sales appraisal pulls their valuations toward the prices being paid by incoming buyers, who are by definition wealthier than they are. The doctrine that justifies all of this is one of horizontal equity: properties of equivalent market value should be taxed equivalently. The doctrine treats the homeowner as a market actor who can convert her home into cash at any time and is therefore appropriately taxed against its current cash value. The empirical homeowner, who lives in the house and cannot convert it without losing her home, is not the figure the doctrine has in mind.

The mechanism by which insurance bills float is more recent, more dramatic, and structurally similar. American homeowners insurance is written in the dominant HO-3 form against replacement cost: in the event of a covered loss, the insurer pays to rebuild the structure to its current condition.¹¹ Replacement cost is determined by current construction prices, current materials prices, and current labor rates. Through the 2010s, replacement cost rose modestly faster than headline inflation. After 2020, with pandemic-era supply-chain disruption, lumber spikes, labor shortages, and accelerating climate-driven catastrophe losses, replacement cost compounded much faster — and the insurance pricing required to cover it followed.¹² In the most exposed states, premium increases between 2019 and 2024 ranged from roughly 33 percent (national average) to over 60 percent (Florida and Louisiana).¹³ In some Florida counties, average annual homeowners premiums passed eight thousand dollars; in some California wildland-urban-interface zip codes, homeowners discovered that their carrier would not renew at any price.

The two bills compound. A homeowner facing a tax assessment that has tripled since 2010 and an insurance premium that has tripled since 2019 has, by 2025, often seen her annual non-mortgage carrying cost double or triple. If her mortgage was paid off, the doubling or tripling of carrying cost falls entirely against her income, which has not doubled or tripled. If she still has a mortgage, the floating components are added to the fixed P-and-I, and the total monthly housing cost grows without limit even though the mortgage was supposed to be the cap. In either case, the homeowner is absorbing market-value escalation she did not benefit from, against income that has not moved.

The doctrine treats this as the homeowner’s choice. She is, at any moment, free to sell — to convert her appreciated asset into cash, downsize to a smaller home in a cheaper market, and live the difference. The doctrine treats her continued residence in her appreciating, increasingly expensive house as a consumption decision. The empirical evidence is that a large fraction of homeowners in this position do not in fact freely choose; they are eventually forced to sell when the carrying-cost line crosses what their income can absorb, and they sell into markets where the cash from the sale does not buy a comparable home in a comparable community at any reasonable distance. They become renters, or move in with adult children, or relocate hundreds of miles to lower-cost markets where they have no community ties. The Census Bureau does not classify any of this as foreclosure. The bank was paid; the mortgage, where one existed, was satisfied at sale; the homeowner walked away with cash. The classification is technically correct. What it misses is that the homeowner did not, in any meaningful sense, choose.

The Assessor’s Doctrine

The argument so far has presented assessment-to-market-value as a neutral technical procedure that produces hardships for homeowners with stagnant incomes. It is in fact something more pointed. Christopher Berry, who runs the Center for Municipal Finance at the University of Chicago, has spent more than a decade documenting that American property tax assessment is not merely uncomfortable for low-income owners but systematically regressive against them: the lower a home’s actual market value, the higher, on average, its effective assessment ratio.¹⁴

Berry’s analysis, drawn from approximately twenty-six million arms-length residential sales across roughly two thousand six hundred U.S. counties between 2006 and 2016, found that in the typical American county the lowest-priced homes (bottom decile) were assessed at substantially higher fractions of true market value than the highest-priced homes (top decile). The headline pattern, replicated across most jurisdictions studied, was that effective assessment rates on bottom-decile homes ran in the range of fifty to one hundred percent higher than effective rates on top-decile homes.¹⁵ The mechanism is technical but not mysterious. Lower-priced homes are appraised through standardized procedures that smooth over the heterogeneity of cheap housing stock; higher-priced homes are evaluated with more idiosyncrasy and more attention to specific market evidence, which tends to capture downward adjustments more reliably than upward ones; appeal procedures, which are accessible to homeowners with the resources to engage them, are disproportionately used by higher-priced-home owners and disproportionately reduce their assessments. The cumulative effect is that, in most American jurisdictions, the property tax functions as a regressive tax on the assessed value, even where the nominal rate is uniform.

The most public test of this finding came in Cook County, Illinois, where the 2017-2018 Chicago Tribune investigative series “The Tax Divide” documented the assessment regressivity in the Berrios assessor’s office in painful detail.¹⁶ Residents of low-income, predominantly Black and Latino neighborhoods on the South and West sides of Chicago paid effective property tax rates roughly twice as high as residents of wealthy North Shore communities, controlling for the fair market value of their homes. The series cited Berry’s analysis directly; it contributed to assessor Joseph Berrios’s defeat in the 2018 Democratic primary by Fritz Kaegi, who ran on assessment reform. Six years later, the Cook County assessment system, despite Kaegi’s reforms, has narrowed but not eliminated the regressivity gap.¹⁷

A second body of work, by Carlos Avenancio-León (Indiana University) and Troup Howard (University of Utah), demonstrated in a 2022 Quarterly Journal of Economics article that assessment regressivity has a racial dimension that persists after controlling for property characteristics.¹⁸ Using a national dataset, they found that Black homeowners face property tax assessments approximately ten to thirteen percent higher, relative to true market value, than white homeowners owning equivalent homes in equivalent neighborhoods. The gap is not explained by property differences, by school district, or by jurisdiction. It is, in their analysis, the residual of assessment-administrative practices that systematically capture upward but not downward valuation drift in Black neighborhoods.

A historical extension of these findings appeared in 2024 with the publication of Andrew Kahrl’s The Black Tax: 150 Years of Theft, Exploitation, and Dispossession in America (University of Chicago Press).¹⁹ Kahrl, a historian at the University of Virginia, traces the use of property tax administration as a mechanism of Black land loss across the post-Reconstruction century: discriminatory assessment, the manipulation of tax-lien sales by local officials and private investors, the targeted use of tax foreclosure against Black landowners in the rural South and the urban North, and the cumulative effect on what economists now describe as the racial wealth gap. His core argument is that property tax administration, more than any other instrument of state policy, has been the long-running mechanism by which Black-owned property has been transferred to white owners and to municipal coffers across the modern American era. The mechanism Kahrl describes is older than the contemporary squeeze, but it is the same mechanism. The contemporary squeeze is what happens when that mechanism, whose disparate impact has long fallen hardest on Black homeowners, is applied at scale to a much broader population of owners under conditions of climate-driven cost escalation.

The implication is consequential and is the part of the assessment story that the policy literature has been slowest to integrate. Assessment-to-market-value, treated as a neutral technical procedure, in fact systematically taxes the asset more heavily the less of it the homeowner has. A homeowner whose primary asset is a home assessed at one hundred fifty thousand dollars pays a higher effective rate against that asset than a homeowner whose primary asset is a home assessed at one and a half million. The latter has more wealth, larger appeals resources, and assessment errors that systematically run in his favor. The former has the asset and little else, no resources to appeal, and assessment errors that systematically run against her. The American property tax, viewed across the distribution, is not a tax on real estate wealth. It is a tax on the home of homeowners who have no other significant wealth, and a much lighter tax on the multi-million-dollar assets of homeowners whose wealth extends well beyond the home. The policy debate continues to treat property tax as if it were an asset tax; the empirical structure is closer to a regressive consumption tax on the housing of the bottom and middle of the wealth distribution.

The Underwriter’s Withdrawal

The insurance side of the carrying-cost squeeze has, over the last five years, undergone a structural transformation that has not yet been politically processed. Until roughly 2017, American property and casualty insurers operated against a set of climate-loss assumptions that, while increasingly out of date, allowed pricing in catastrophe-exposed states to remain affordable for ordinary homeowners. The cushion came from cross-subsidy: low-risk regions subsidized high-risk regions through the diversification of large national insurers’ books; reinsurers absorbed the tail risk; state-level rate regulation kept individual policy prices stable across years even as underlying loss expectations crept upward. The cushion has now collapsed in the most exposed states and is fraying in the second-most-exposed tier.

The collapse in Florida is the most fully visible.²⁰ Through the 2010s, Florida’s residential insurance market was already in chronic stress: the state’s exposure to Atlantic hurricane risk, combined with a uniquely litigious assignment-of-benefits regime that drove up loss-adjustment expenses, produced repeated insurer insolvencies. After Hurricane Irma (2017), Hurricane Michael (2018), and a series of smaller named storms, the insolvencies accelerated. Between roughly 2020 and 2023, multiple Florida-domiciled insurers — including FedNat, Southern Fidelity, Lighthouse, UPC, Avatar, St. Johns, Weston — became insolvent and were liquidated by the Florida Office of Insurance Regulation. National carriers reduced or paused new-business writing in Florida; Farmers Insurance Group, in July 2023, announced the discontinuation of new and renewal policies for its branded lines in the state.²¹ Citizens Property Insurance Corporation, the state-created insurer of last resort, absorbed the residual demand. Its policy count grew from roughly four hundred twenty thousand in 2019 to over one and a half million by mid-2023, briefly making it Florida’s largest residential insurer.²² State legislative response — Senate Bill 2-D in 2022, Senate Bill 2-A in a December 2022 special session, House Bill 837 in 2023 — restructured the litigation regime, capitalized state reinsurance funds, and produced a partial depopulation of Citizens through 2024 as carriers re-entered.²³ What the legislative response did not do was reduce the underlying loss expectations that had driven the crisis in the first place. Florida homeowners pay, in 2025, the highest average homeowners premiums in the United States; in many counties, the average annual premium exceeds five thousand dollars and in some Gulf-coast and Miami-Dade zip codes exceeds eight thousand.²⁴

The collapse in California is more recent, more selective, and more revealing. California operates under Proposition 103 (1988), which requires prior approval of insurance rates by the state insurance commissioner and constrains insurers’ use of forward-looking catastrophe models in ratemaking.²⁵ For most of its history, this regulatory regime kept California insurance affordable in absolute terms. The trade-off was that insurers could not price wildfire risk against current expected losses; they had to price it against historical losses. Through the 2010s, as the wildfire loss series climbed sharply, the regulated rates fell further behind expected losses, and insurers absorbed the gap. By 2022, the gap had become unsustainable for several major carriers. State Farm General Insurance Company, the largest residential insurer in the state, announced in May 2023 that it would no longer accept new applications for homeowners and most personal-lines policies in California, citing wildfire exposure and construction-cost inflation against constrained ratemaking.²⁶ In March 2024, State Farm announced it would non-renew approximately seventy-two thousand California policies, including roughly thirty thousand homeowners policies.²⁷ Allstate had paused new-business writing in California in late 2022. Farmers capped its new homeowners writing at a percentage of monthly volume in mid-2023 before partially reversing in late 2024.

The California FAIR Plan, the state’s bare-bones insurer of last resort, was the absorber of last resort for the homeowners these decisions affected.²⁸ FAIR Plan policy counts, around one hundred forty thousand in 2018, exceeded four hundred fifty thousand by late 2024, with total exposure across covered properties exceeding half a trillion dollars.²⁹ Then, in January 2025, the Palisades and Eaton fires in Los Angeles County destroyed thousands of homes in some of the highest-value residential markets in the country. Insured losses were estimated by major modelers in the range of twenty-eight to forty-five billion dollars, depending on methodology.³⁰ The FAIR Plan, holding billions in concentrated exposure in the affected zones, levied a one billion dollar assessment on member insurers — the first assessment of its kind in over thirty years — to cover claims it could not pay from premium reserves.³¹ Insurance Commissioner Ricardo Lara, whose 2024 “Sustainable Insurance Strategy” reforms were intended to allow forward-looking catastrophe modeling in exchange for insurer commitments to write in distressed zones, found that the strategy’s first major test was a billion-dollar deficit in the program it had been designed to wind down.³²

The collapse in Louisiana followed a comparable shape after Hurricane Laura (2020) and Hurricane Ida (2021); at least a dozen Louisiana-licensed insurers became insolvent, and the Louisiana Citizens policy count quadrupled.³³ Comparable but smaller-scale stress is now visible in coastal Texas, in the Carolinas, in the Gulf parishes of Mississippi and Alabama. The pattern has spread inland: insurers are non-renewing policies in parts of Iowa and Minnesota over hail-and-wind exposure, in parts of Colorado and Idaho over wildfire exposure, in inland California, in parts of Montana and Oregon. The Senate Budget Committee’s December 2024 report, “Next to Fall: The Climate-Driven Insurance Crisis Is Here,” compiled non-renewal data from major carriers across all fifty states and found that the geographic footprint of insurance market stress now extends well beyond the conventional Atlantic-and-Gulf hurricane belt and the California wildfire zone.³⁴

The structural problem behind the state-level patterns is the absence of a federal catastrophe reinsurance backstop. The National Flood Insurance Program, which covers a narrow slice of catastrophe exposure (residential flood specifically, on a deeply discounted-from-actuarial basis for legacy policies), holds approximately four point seven million policies and carries roughly twenty billion dollars in outstanding Treasury debt, with statutory borrowing authority of just over thirty billion. It has been reauthorized in dozens of short-term extensions since 2017 with no structural reform.³⁵ No comparable federal backstop exists for hurricane wind, wildfire, severe convective storms, or earthquake. Multiple proposals — the Homeowners Defense Act of various Congresses, the catastrophe savings account proposals, the FEMA-administered backstop concepts — have been introduced and have not advanced.³⁶ The result is that the insurance side of the carrying-cost equation, in the catastrophe-exposed states, is held up by state-level mechanisms (Citizens, FAIR Plan, Louisiana Citizens, Texas Windstorm) that are structurally underfunded for the loss tail they are now absorbing. When those mechanisms fail in the next major event — and the Jan 2025 LA fires were not a major event by global catastrophe standards — the federal government will face a choice between an ad hoc bailout and a generation-defining failure of insurance access in the affected states. There is no policy framework prepared for either choice.

What the climate insurance crisis adds to the carrying-cost squeeze is not only the higher premiums. It is the discovery, by homeowners across the most exposed states, that homeowners insurance has structurally re-priced from a stable budget item into a volatile and partially-rationed expense. The premium rises, year over year, in ways that the homeowner cannot anticipate or budget against; the carrier may exit the market, leaving the homeowner with the FAIR Plan or Citizens or no coverage at all; the mortgage holder, who requires insurance as a condition of the loan, treats lapse in coverage as default. The combination is a slow-motion eviction mechanism that operates through the insurance side of the carrying-cost equation, rather than through the tax side. Both mechanisms produce the same outcome. They produce homeowners who can no longer afford to keep homes they own.

The Forced-Sale Pipeline

The previous two sections described the two upward pressures on the carrying cost. This section describes what happens when a homeowner cannot meet the upward pressure: the legal and institutional mechanism by which her home is converted, against her preference, into the property of someone else. The mechanism is not the foreclosure most Americans picture when they hear the word, which involves a mortgage holder accelerating a defaulted loan and selling the home at sheriff’s sale. The mechanism is older, simpler, and largely outside the headlines.

It begins with delinquent property taxes. In nearly every American state, when a homeowner falls behind on property taxes, the local tax authority places a lien on the property for the amount owed, plus statutory interest and fees. In approximately half the states, the local authority then sells the lien — that is, sells the right to collect the debt — to a private investor at a tax-lien auction.³⁷ The investor pays the back taxes to the locality, which is made whole; the investor now holds a senior claim against the property. The homeowner has a defined redemption period — six months, two years, three years, depending on the state — during which she can pay the investor the original tax debt plus all accumulated interest and fees, and the lien is extinguished. If she does not redeem, the investor can foreclose on the property.

The arithmetic of the redemption window is what makes the mechanism so productive of forced sales. Statutory interest rates on tax liens, set in many states decades ago when general interest rates were higher, can run from twelve percent to thirty-six percent annualized; investor-charged fees compound on top.³⁸ A homeowner who falls four hundred dollars behind on property taxes, and who cannot redeem within the statutory window, may face a redemption price of two thousand dollars within eighteen months and four thousand within thirty-six. If she still cannot redeem, she loses the property — often a property she owns outright, unencumbered by any mortgage, worth tens or hundreds of thousands of dollars more than the debt against it.

The mechanism’s most thoroughly documented failure mode is the Washington, D.C. tax lien crisis exposed by The Washington Post’s September 2013 investigative series “Left With Nothing,” reported by Michael Sallah, Debbie Cenziper, and Steven Rich.³⁹ The series found that the District of Columbia had sold tax liens to private investors who then, after the brief redemption period, foreclosed on the underlying homes — many of them owned outright, many of them owned by elderly residents who had inherited them, many of them lost over original tax debts of less than one thousand dollars. The series’s most cited case was Bennie Coleman, a seventy-six-year-old Marine Corps veteran with dementia who lost his fully paid-off Northeast D.C. home over an original tax debt of one hundred thirty-four dollars; after fees and interest from the lien purchaser, his redemption price had grown to over five thousand dollars, which he could not pay.⁴⁰ The Post identified roughly two hundred D.C. homeowners who lost homes through the program, many over tax debts under one thousand dollars. The District Council passed reforms in 2014 raising the minimum debt threshold for foreclosure and capping investor fees, but the underlying tax-lien-sale mechanism was preserved; D.C. continued to sell liens, on a more constrained basis, after the reform.

The mechanism’s largest documented mass operation occurred in Wayne County, Michigan, between roughly 2011 and 2016. After the 2008 housing collapse and the collapse of Detroit’s automotive employment base, large numbers of Detroit homeowners fell behind on property taxes — many of them, as Bernadette Atuahene’s research subsequently demonstrated, on properties that had been illegally over-assessed in violation of the Michigan Constitution’s requirement that assessed value not exceed fifty percent of true cash value.⁴¹ Wayne County tax foreclosures during the peak years exceeded approximately one hundred forty thousand properties, with the great majority concentrated in Detroit; the county became, by Atuahene’s account, the site of “the largest property tax foreclosure crisis in American history.”⁴² Many of the foreclosed properties were sold at auction at prices well below their nominal value — in some cases for the legal minimum of five hundred dollars — to investors and small landlords who then operated them as rentals or held them for resale. Subsequent litigation, including the Morningside Community Organization v. Sabree case filed by the ACLU of Michigan and the NAACP Legal Defense Fund in 2016, produced policy changes but not damages adequate to the scale of the loss.⁴³

In May 2023, the United States Supreme Court issued a unanimous decision in Tyler v. Hennepin County⁴⁴ that partially closed one of the more egregious features of the tax-lien mechanism. The case concerned Geraldine Tyler, a ninety-four-year-old Minneapolis woman who owed approximately fifteen thousand dollars in property taxes, interest, and fees on a condominium. Hennepin County seized the condominium, sold it for forty thousand dollars, applied the proceeds to the tax debt, and kept the entire twenty-five thousand dollar surplus, as Minnesota law at the time permitted. The Court, in an opinion by Chief Justice Roberts, held that the county’s retention of the equity surplus beyond the actual debt constituted a taking under the Fifth Amendment without just compensation. The decision affected approximately a dozen states whose tax-sale statutes had permitted government retention of equity surplus; most have since amended their statutes or are litigating. What the decision did not do was alter the underlying tax-lien-sale mechanism itself, the redemption-window arithmetic that produces the forced sales, or the disproportionate concentration of these losses among the elderly and the bottom of the wealth distribution. Tyler v. Hennepin County closed one egregious surcharge on top of the mechanism. The mechanism continues.

Two features of the contemporary forced-sale pipeline deserve particular emphasis because they connect this section’s mechanism to the broader argument of the essay. The first is the institutional buyer at the auction. The second is the absence of public attention.

The institutional buyer at the tax foreclosure auction is, in many of the most affected metro areas, the same class of buyer that purchased distressed assets through the 2007-2012 mortgage foreclosure cycle and the post-Hurricane catastrophe sales of the 2010s. Invitation Homes, Progress Residential / Pretium Partners, American Homes 4 Rent, Tricon Residential, Amherst Residential, and a longer tail of regional and local operators have institutionalized the acquisition of single-family residential at scale.⁴⁵ As of late 2023, the largest of these operators each held portfolios in the range of forty thousand to ninety thousand homes; the combined institutional portfolio in the United States included on the order of three to five hundred thousand single-family rental properties.⁴⁶ Aggregate institutional ownership represents perhaps three to five percent of the national single-family rental market, but that aggregate hides a much higher concentration in specific Sun Belt metros; in metro Atlanta, institutional investors owned roughly eleven percent of single-family rentals by 2018, and in some neighborhood-scale tracts the share has exceeded thirty percent.⁴⁷ These operators have reliable access to capital at rates unavailable to retail buyers, professional bidding operations at auction, and the legal and financial infrastructure to absorb tax-foreclosure inventory at scale. They are positioned to be the residual buyer of properties that paid-off elderly homeowners cannot continue to carry.

The absence of public attention is the more politically significant feature. Mortgage foreclosure produces visible signs: the bank-owned listing in the local MLS, the sheriff’s sale in the courthouse newspaper, the Zillow REO category. Tax foreclosure does not produce any of these. The properties are sold in administrative auctions; the sale prices are public records but are not aggregated into the housing-market data series that economists and the press follow. Insurance-driven displacements — the elderly homeowner who could no longer afford the new premium and sold to a cash buyer — do not produce any administrative record at all; they appear in the Census Bureau’s data only as ordinary sales. The aggregate effect — the conversion of paid-off owner-occupiers into renters or downsizers via carrying-cost squeeze — is happening below the level at which conventional housing market reporting can see it. The displaced homeowner does not appear in the foreclosure rate; she appears, if anywhere, in the year-over-year decline in the homeownership rate, which the Census Bureau reports to a fraction of a percent without explaining the mechanism.

The U.S. homeownership rate, after peaking at sixty-nine point two percent in the second quarter of 2004, fell to a contemporary low of sixty-two point nine percent in 2016 and has since recovered to roughly sixty-five point seven percent.⁴⁸ The recovery is uneven. The Black homeownership rate, which had reached approximately forty-nine percent at the 2004 peak, was approximately forty-four percent by 2024 — a five-point gap that has not closed across two decades.⁴⁹ The headline figure does not distinguish between first-time-buyer access (the YIMBY supply problem) and incumbent-owner displacement (the carrying-cost problem). Both are operating. Both are contributing to a slow downward drift in the share of American households that own the home they live in, in an economy in which homeownership remains the principal vehicle of intergenerational wealth transmission for households below the top decile.

The Policy That Won’t Come

A reader who has followed the argument so far may reasonably expect a turn toward the policy responses available to address the carrying-cost squeeze. Several are available, each with historical precedent. The problem is that the political coalition required to enact any of them has not assembled in the United States in the four decades since Massachusetts’s Proposition 2½, and the structural reasons it has not assembled apply to both major parties.

The Democratic position on housing is shaped by three commitments that interact in ways that make a coherent response to the carrying-cost squeeze nearly impossible to formulate. The first commitment is to property tax revenue as the foundation of K-12 public school funding. American public schools are funded predominantly through local property taxes, a structural arrangement that has been the subject of equity-litigation in many states (notably San Antonio v. Rodriguez in 1973 and a long sequence of state supreme court decisions thereafter) but has never been fundamentally restructured.⁵⁰ The Democratic coalition includes teachers’ unions, school administrators, and a broad professional-class constituency for whom property tax revenue and school quality are linked in a way that makes any cap on assessment growth read as a threat to public education. The result is that the state-level cap movements of the late twentieth century — Proposition 13, Proposition 2½, Save Our Homes, Proposal A — were either Republican-led, bipartisan-with-Republican-leadership, or referendum-driven against the active opposition of Democratic state-level officials. The Democratic Party has not, in the contemporary period, advanced an assessment-cap proposal at any level of government.

The second Democratic commitment is to tenant-protection and rent-stabilization frameworks as the primary tool of housing affordability. The post-2010 housing-policy debate within the Democratic coalition has been dominated by the tenant-organizing left, which has prioritized rent control extensions, tenant-protection statutes, eviction-defense funding, and the expansion of Section 8 vouchers and the Low-Income Housing Tax Credit. None of these instruments addresses the carrying-cost squeeze, which is a problem of owners — the population least visible in the contemporary tenant-protection framework. Section 8 does not help an elderly homeowner whose property tax bill has tripled. The Low-Income Housing Tax Credit does not produce a single new owner-occupant. Rent control, where applied, does not reach the owner-occupied market at all. The framework’s silence on owner-occupiers is structural; the political base that drives the framework is composed of renters, and the constituency of paid-off homeowners being carried out by escalating taxes and insurance is not part of it.

The third Democratic commitment is to the YIMBY (“Yes In My Back Yard”) supply-side framework, which holds that the central problem of American housing is insufficient construction, driven by exclusionary zoning, and that the central solution is federal preemption of local zoning restrictions to allow multifamily construction by right.⁵¹ This framework, broadly correct on the supply problem, also has nothing to say about the carrying-cost squeeze. Adding new units to the housing supply does not lower the assessment on a paid-off bungalow in a neighborhood that has appreciated; it does not lower the insurance premium on a property in a wildfire-exposed zone; it does not address the regressive incidence of assessment-to-market-value across the wealth distribution. New supply is necessary. It is not a substitute for an answer to the question of what the existing owner-occupier owes the assessor and the underwriter against income that has not grown.

The Republican position fails for opposite-but-symmetrical reasons. The Republican commitment to “local control” treats municipal property tax administration as a sacred preserve of local government, into which neither federal nor state policy should intrude. This commitment, defensible on federalist grounds, makes any federal-leverage proposal — for example, conditioning HUD or transit funding on assessment-cap adoption — politically inaccessible to the Republican coalition. It also makes the Republican commitment to “deregulation” peculiarly absent in the area where deregulation would most directly benefit the constituency the Republican coalition claims to represent, namely the small homeowner. Republicans have argued for the deregulation of nearly everything — energy, finance, healthcare, environmental review — but have made no parallel argument for federal preemption of municipal assessment practices, despite municipal assessment being the single largest tax most American homeowners pay.

The Republican silence on institutional ownership of single-family residential is the second feature of the Republican position. The institutional acquisition of paid-off and foreclosed homes by Blackstone-derived and similar operators is, in financial terms, a paradigmatic Wall Street operation. The Republican coalition’s relationship to Wall Street is constitutive; the institutional SFR sector is not a constituency the Republican coalition is positioned to constrain.⁵² Senator Merkley’s 2022 Stop Wall Street Landlords Act and Senator Smith’s 2023 End Hedge Fund Control of American Homes Act, both Democratic-introduced, did not advance, in part because they had no Republican cosponsors of consequence.⁵³

The third feature of the Republican position is the 2017 Tax Cuts and Jobs Act’s cap on the state-and-local tax deduction. The SALT cap, as it has come to be called, limited the federal deduction for combined state and local taxes (including property tax) to ten thousand dollars per filer.⁵⁴ The cap was, in dollar terms, the single largest revenue raiser in TCJA, scored at approximately six hundred sixty-eight billion dollars over ten years.⁵⁵ Its incidence fell heavily on homeowners in high-property-tax states — New Jersey, Illinois, Connecticut, New York, California, Massachusetts — many of whom found that the cap effectively raised their federal taxable income by ten to thirty thousand dollars, depending on the magnitude of their property tax bill above the cap. The cap was passed by a Republican-only vote and was widely understood at the time to be designed to fall on Democratic-state taxpayers. Its actual incidence, after the dust settled, fell on owner-occupiers in high-tax states across the political spectrum. The Republican coalition has not, since 2017, proposed any structural revision to the cap; the Trump administration’s 2024 campaign rhetoric included some movement toward raising or repealing the cap, but the legislative posture has been ambiguous. The result is that, on the carrying-cost squeeze, the Republican Party has presided over a tax change that materially worsened the squeeze for tens of millions of middle-class homeowners and has neither defended the change nor reversed it.

The off-the-shelf federal program that comes closest to a response — Trade Adjustment Assistance, established in 1974 to help workers displaced by trade liberalization — is not an instrument with any application to the housing carrying-cost problem. Federal disaster assistance through FEMA covers some categories of catastrophic loss but is structurally back-end (it pays out after a declared disaster, not in advance against carrying-cost escalation), and the long-running debate over expanding the National Flood Insurance Program into a broader federal catastrophe backstop has produced no legislation in two decades. The closest historical precedent for a federal program addressing the carrying-cost squeeze is the Home Owners’ Loan Corporation, the New Deal-era agency that purchased and refinanced approximately one million mortgages between 1933 and 1936 to prevent foreclosure during the Depression.⁵⁶ HOLC operated for three years and was wound down. It has no contemporary analog. The political conditions that produced it — bank failures, mass urban unrest, an active electoral mandate for a New Deal — are not the political conditions of 2025.

The realistic policy response to the carrying-cost squeeze, when it eventually assembles, will require components that do not currently sit anywhere on the active legislative agenda of either party. Capping owner-occupied primary-residence assessment growth, the lever the late-twentieth-century state revolts identified, will require either federal preemption of municipal property tax administration (which the Republican coalition cannot accept on local-control grounds) or a wave of state-level constitutional referenda that the Democratic coalition cannot lead. Building the federal catastrophe reinsurance backstop the climate insurance crisis requires will require either federal underwriting of state catastrophe pools (which the Republican coalition cannot accept on federal-spending grounds) or a national risk-pricing program that absorbs the cross-subsidies the Democratic coalition cannot bring itself to make explicit. Reforming the tax-lien-sale mechanism — closing it down completely as several states have already done after Tyler, capping the interest rates and fees the lien purchasers can charge, requiring full preservation of homeowner equity in any forced sale — will require either state-level legislation against the active opposition of the lien-investor industry or a federal floor against the active opposition of the local-finance lobby. Constraining institutional ownership of single-family residential, through ownership caps, punitive taxation above thresholds, or restrictions on auction participation, will require a politics neither party currently inhabits. None of these moves is available within the current configuration. All of them will become available, on the historical record, only when the alternative becomes intolerable for reasons that are not principally ethical. The next section describes what that alternative looks like.

The Cascade That Closes

The previous essay in this series described the AI transition as producing a wage cascade: the compression of knowledge-work wages forcing displaced workers downward through the labor distribution, with terminal incidence falling hardest on workers at the bottom who have nowhere further to fall.⁵⁷ The cascade described in this essay operates against a different variable but moves in the same direction. The carrying-cost cascade compresses ownership: rising assessments and insurance force out the homeowners least able to absorb them, who in many cases sell at the floor of the market rather than at the peak, transferring their accumulated equity to the institutional class that is positioned to acquire it. The two cascades collide in the next decade in a way that neither cascade, considered alone, makes legible.

The collision works as follows. A homeowner in his late fifties, employed in a knowledge-work occupation that the AI transition is in the process of compressing — a paralegal in Cleveland, a mid-tier financial analyst in Atlanta, a junior engineering manager in the Sun Belt — owns a home he purchased fifteen years ago and on which he has been paying down a thirty-year fixed-rate mortgage. His income, in 2026, is at the level he had reached by 2022 and has been flat or compressing since. His mortgage P-and-I is fixed and is, in inflation-adjusted terms, a smaller line item every year. His property tax assessment, in a market that has appreciated forty percent since 2020, is now substantially higher than the assessment he was paying in 2020, and continues to grow each cycle. His homeowners insurance premium, written by a national carrier whose cost of reinsurance has tripled since 2019, has doubled since 2020. The total monthly housing cost, in 2026, is materially higher than it was in 2020, against an income that is materially lower. The cushion that allowed him to absorb a small annual escalation in carrying costs against an annual raise no longer exists. The carrying cost continues to grow. The income does not.

The same homeowner, in the standard 2007-vintage analysis, would respond to this stress by tapping home equity through a HELOC or cash-out refinance. The 2007 mechanism is not available in 2026: rates are higher, lender appetite for cash-out against deteriorating borrower income is constrained, and the homeowner whose income has compressed does not present as a creditworthy borrower regardless of the equity in his home. The other standard response — selling and downsizing — is constrained by the same mechanism that constrained it for the elderly Californian in 1976. The cash from the sale does not buy a comparable home in a comparable community at the new market prices; the move requires either a different community at lower cost-of-living, which is a major life disruption mid-career, or a smaller home in the same community, the supply of which is sharply limited by the very zoning and supply constraints the YIMBY framework correctly identifies. The third option — renting — converts a homeowner with locked-in housing cost (the fixed P-and-I) into a renter exposed to annual rent escalation in a market that has been growing rents faster than wages for two decades.

The most likely actual outcome, in this configuration, is that the homeowner falls behind on a single line — typically property taxes, because property taxes are billed annually rather than monthly and a single missed payment can produce a year-long arrears before the lien is sold. The lien is sold. The redemption period runs. The homeowner, whose income has continued to compress, cannot redeem at the inflated redemption price. The home is foreclosed and sold at auction, often to one of the institutional operators described above, often at well below replacement cost. The homeowner’s accumulated equity — which may have been the bulk of his lifetime wealth — is partially or wholly wiped out.⁵⁸ He is added to the rental population. The institutional buyer adds the property to its portfolio. The wealth has transferred. The Census Bureau records a small downward tick in the homeownership rate. The homeowner, who never declared default on a mortgage, is not counted as a foreclosure victim in the federal data series. He is invisible in the statistical aggregate that political commentary follows.

This pattern, projected across the full population of mid-career and late-career homeowners experiencing AI-driven income compression in markets experiencing tax-and-insurance escalation, defines the next phase of the wealth transfer the previous essay identified. The 2007-2012 transfer moved housing equity from the bottom and middle of the distribution to the top, predominantly through mortgage default. The 2026-2032 transfer will move housing equity from the bottom, middle, and upper-middle of the distribution to the top, predominantly through the carrying-cost mechanism described in this essay. The mortgage holder, in the second transfer, may be paid off entirely; the home may be unencumbered; the foreclosure may not occur. The transfer happens anyway, through the assessor and the underwriter rather than through the bank.

The transfer’s terminal incidence — the question of who, ultimately, ends up holding the accumulated housing equity that the transfer has shaken loose — points back to the same institutional class that emerged from the 2007-2012 transfer with portfolios of single-family residential the country had never previously possessed. Invitation Homes, Progress Residential, American Homes 4 Rent, Tricon, Amherst, and the longer tail of regional operators are positioned to absorb the inventory the carrying-cost cascade is generating. Their access to capital is structurally cheaper than the access of the displaced homeowners. Their ability to absorb the carrying costs themselves — the same property taxes and insurance premiums that displaced the prior owner — is materially higher, because they hold the asset against rental income rather than against compressed wage income, and because they can pass much of the carrying cost through to the new tenant in the form of rent. The rental price of the converted single-family home, after the institutional acquisition, is typically higher than the homeowner’s prior carrying cost; the prior owner, now renting in the same neighborhood from the new landlord, pays more per month for the same square footage, with no equity accumulation, in a unit she previously owned outright.

The inheritance question follows. American household wealth in the bottom and middle deciles is overwhelmingly held in primary residence equity.⁵⁹ The mechanism by which a working-class or middle-class household transmits wealth across generations is, in the dominant case, the inheritance of a paid-off home. If the carrying-cost cascade liquidates that equity in the homeowner’s last decade — through forced sale to cover taxes and insurance, through tax foreclosure, through the slow draining of savings to meet escalating carrying costs — the inheritance does not occur. The asset has been monetized to pay for the right to continue living in it, and the cash from the monetization has been spent on continued shelter. The next generation does not inherit a house. They inherit, at best, the small residual cash from the prior generation’s last decade of housing carrying-cost expenditures, after taxes, after Medicare gaps, after long-term care. The asset that was supposed to transfer has dissipated.

The aggregate effect, across the population the carrying-cost cascade is most affecting, is the cancellation of intergenerational wealth transmission for the bottom and middle of the distribution over the next thirty years. The wealth held in the bottom and middle of the distribution today, which was largely built by the post-war expansion of homeownership — the GI Bill, the FHA, the suburban build-out, the long appreciation of single-family residential — was supposed to be the inheritance of the bottom-and-middle children and grandchildren of the post-war owners. It will, on the trajectory described in this essay, instead be transferred to institutional holders during the original owners’ last decade, in exchange for the right to continue living in the houses the owners thought they had bought.

The descendants will continue to live in those houses. They will live in them as renters.

What’s at Stake

The American political economy has confronted the shape of its housing arrangements three times before, and each time produced a settlement distinctive to its moment.

The Homestead Acts of the mid-nineteenth century redistributed federal land into private ownership at scale, producing the largest expansion of household property ownership in any country up to that date.⁶⁰ The settlement produced was, in retrospect, the foundation of the American agrarian middle class. It also explicitly dispossessed the indigenous nations whose territories the federal government distributed; that dispossession was constitutive of the settlement and is reckoned with separately in American historical memory. The pattern at the household level, however, was a deliberate use of state power to transfer ownership of a specific asset class — agricultural land — into the hands of ordinary households on terms designed to anchor the political economy of the country in broad asset ownership.

The post-war housing settlement, constructed over roughly the period from 1934 (the Federal Housing Administration) through 1968 (the Fair Housing Act), did the same thing for urban and suburban single-family residential. The federal government, through the FHA, the GI Bill, the secondary mortgage market institutions (Fannie Mae and later Freddie Mac), the long expansion of the mortgage interest deduction, and the explicit subsidy of suburban infrastructure, produced the contemporary American homeowning middle class. The settlement was racially constrained — redlining excluded Black households from the federally-supported mortgage product through the early 1960s, and the consequences of that exclusion structure the contemporary racial wealth gap — but the structural commitment was to broad household ownership of the principal residential asset class. That settlement structured the American economy from approximately 1950 to approximately 2000.

The third settlement, the financialization era beginning in the 1980s, partially dissolved the second. The mortgage interest deduction was preserved but capped (and capped further by TCJA in 2017); secondary-market institutions were privatized in form and bailed out in fact; the regulatory framework that had constrained mortgage origination and securitization was loosened, then collapsed, then partially rebuilt. The settlement’s defining failure was the 2008-2012 housing collapse, which moved an unprecedented volume of single-family residential equity from the bottom and middle of the distribution to a newly assembled institutional class. That settlement is the one currently in operation. It is the one this essay has described as failing.

A fourth settlement is now required, and it is structurally different from any of the three that preceded it. The Homestead Acts created ownership where none had existed. The post-war settlement subsidized ownership at scale. The financialization settlement loosened the institutional structures that had supported broad ownership and produced, as its principal output, a transfer of ownership from the broad middle to a narrow institutional top. The settlement now required is none of the above. It is a settlement that protects existing owner-occupiers against the carrying-cost mechanisms — assessment, insurance, tax-lien sale — that are converting their ownership into involuntary tenancy, and that constrains the institutional class that is the residual buyer of the property the conversion produces. The components are identifiable and have been listed in the policy section of this essay. The political coalition required to enact them does not currently exist.

The fourth settlement, when and if it assembles, will have to do four things simultaneously. It will have to cap assessment growth on owner-occupied primary residences against a measure that does not float with market value — the acquisition-based mechanism that California voters identified in 1978 and the Massachusetts, Michigan, and Florida voters subsequently confirmed, designed for owner-occupiers only and uncapped on transfer to prevent the regressive subsidies that Proposition 13 produced over time. It will have to construct a federal catastrophe reinsurance backstop that absorbs the climate-driven repricing of property insurance in the most exposed states, paid for by the carriers and reinsurers whose books the backstop relieves rather than by the public, with mitigation requirements for the property classes most exposed. It will have to reform the tax-lien-sale mechanism nationally, capping the interest rates and fees the lien purchasers can charge, requiring full preservation of homeowner equity beyond the actual debt — Tyler v. Hennepin established the constitutional minimum but stopped well short of the full reform — and providing federal funding for redemption-window legal aid in the jurisdictions where tax foreclosure operates at scale. It will have to constrain institutional single-family ownership through the mechanism The Intelligent Party has proposed — punitive taxation above an ownership threshold sufficient to force divestment of large portfolios back into the owner-occupier market, and direct prohibition of institutional bidding at tax-foreclosure auctions.⁶¹ None of these is, on its own, a solution. The four together are a restoration of the second settlement’s commitment to broad household ownership, retrofitted for the climate-and-AI conditions of the twenty-first century rather than the post-war conditions of the mid-twentieth.

The Vienna model of public housing, paired alongside the four-component reform of the existing private market, supplies the pressure release the private market cannot supply on its own.⁶² Vienna’s gemeindebau and limited-profit cooperative sector house roughly sixty percent of the city’s residents in cost-rent apartments built and operated by the municipality and by regulated non-profit operators. The system is funded by a dedicated payroll-tax mechanism that the city has maintained across nearly a century, and it produces a market in which the private rental sector competes against a public alternative whose rents are calculated to cost rather than to market. The American social-housing tradition — from the Public Works Administration of the 1930s through the deeply underfunded HUD public-housing system of the present day — has never been allowed to operate at scale comparable to Vienna’s. Constructing such a sector in the contemporary United States would require a federal commitment to social housing on a scale not attempted since the 1930s; it would also produce, within a decade, a market in which the carrying-cost squeeze on the private homeowner was bounded above by the rent of an alternative that did not depend on the assessor or the underwriter for its viability.

None of this is politically available in 2026. All of it is structurally identifiable. The essay about the housing collapse that someone writes in 2040 will describe either the construction of a fourth settlement of the kind sketched above or its refusal. The shape of the settlement is, narrowly, in the hands of the people who will live under it. The constituency that would benefit from the settlement is the constituency the 1978 California revolt named: long-tenured homeowners being driven out of paid-off homes by escalating carrying costs they cannot control. That constituency, in 2026, includes most American homeowners over fifty in the most exposed states, and a steadily growing share of homeowners under fifty in the second tier. Its political assembly has not yet occurred. The mechanical conditions for its assembly, on the trajectory this essay has described, will be in place within the next decade.

The argument this essay has made is colder than the policy preferences it concludes with. The mechanical claim — that the carrying-cost squeeze is converting American owner-occupiers into involuntary forced sellers, transferring their accumulated wealth to a newly-assembled institutional class, and structurally cancelling intergenerational wealth transmission for the bottom and middle of the distribution — does not require the policy response. The mechanical claim is true whether or not the policy is built. What the policy decides is who pays for the transition. In the absence of the policy, the people who pay are the ones with the least capacity to absorb the carrying-cost escalation: the elderly, the displaced, the wage-compressed, the heirs of the asset that was supposed to transfer to them. They have been paying since at least the early 1970s; the bill is now past due, and the constituency the bill is being collected from has not yet learned to organize against the collection.

The previous essay in this series ended with Paul Starr’s borrowed sentence about the “hard-won realism” that would have to win out again. The realism that essay called for was a return to a regulatory framework that had been built and dismantled. The realism this essay calls for is the construction of a framework that has not been built, against political coalitions that have not yet assembled, in defense of a constituency that has not yet learned to name itself. The realism is harder. The need is the same.

---

Notes