1863 Leadership · Issue Paper No. 15
The Missing Rung
Twenty-five million American adults under thirty-five live with their parents. Seven repairs, and a country that can build the thing people used to buy first.
Abstract
Sixty-five percent of American households own their home. Fifty-three percent of American adults do — because a household where a thirty-year-old lives in a parent's spare room counts as fully owning. A record 25.2 million adults under thirty-five live with their parents. In urban America the ownership rate is 49.8 percent. This paper argues that the shortfall is not mysterious and not mainly about interest rates: the country has made it systematically harder to build and sell a residence than to build and rent one. In Colorado, new construction now runs fourteen apartments for every condominium, because defect liability adds an estimated $30,000 per unit to the for-sale version of the identical building. Ten million mortgages priced below four percent have frozen their holders in place, preventing an estimated 1.7 million sales and raising prices seven percent on their own. Manufactured homes cost $121,300 against $487,300 — and appreciate at the same rate as site-built houses when titled as real property, which only 17 percent of them are. Seven repairs follow, none of which involves the government building or owning anything, and one of which is a rule against ever tilting the scale this way again.
Key findings
- Sixty-five percent of households own. Only 53 percent of adults own the home they live in — and a record 25.2 million adults under 35, nearly one in three, live with their parents.1
- Colorado now builds fourteen apartments for every condominium, against roughly 1.25 before 2009. Insurance runs 5.5 percent of hard costs for for-sale buildings against 1.1 to 1.65 for rental — an estimated $30,000 per unit.2
- Each percentage point by which market rates exceed a homeowner's fixed rate cuts their probability of selling by 18.1 percent. Lock-in prevented 1.7 million transactions and raised prices 7 percent.3
- A manufactured home averages $121,300 against $487,300 — and appreciated 203.7 percent from 2000 to 2024 against 200.2 for site-built, when titled as real property. Only 17 percent are.4
- Cutting mortgage rates from 7 to 5 percent lets the same monthly payment borrow 24 percent more. In a market that cannot add units, that is a 24 percent price increase, and the seller keeps it.5
Section 1The question before us
The homeownership rate is 65.3 percent, and it is measured per household. A house containing a married couple and their two adult children counts once, as owned. Measured instead as the share of American adults who own the place they sleep, the figure is 53 percent — and the gap is widest exactly where housing costs most. In Hawaii, 62 percent of units are owner-occupied and 43 percent of adults own their home.
The twelve points between those two numbers have a location. A record 25.2 million adults under thirty-five — nearly one in three — live with their parents. They are not renters and they are not owners. They are not in the statistic at all.1
Two out of three households, one out of two adults, and a generation in the bedroom it grew up in.
The distribution is worse than the average. Under thirty-five: 37.5 percent. Over sixty-five: above 79. Rural America: 73.9 percent. Urban America: 49.8. Americans aged 55 to 64 are the largest single group of homeowners at 23.4 percent of all owners; Americans aged 25 to 29 are 3.1 percent.1
This paper proposes nothing built or owned by government. It argues that the country has quietly made it harder to build and sell a dwelling than to build and rent the identical building, and that most of the remedy is undoing that.
1.2 Why ownership and not merely shelter
A reasonable person may ask why the tenure matters if the person is housed.
Median household net worth in 2022 was $396,200 for homeowners and $10,400 for renters — thirty-eight to one. Between 2019 and 2022 the owner's figure rose by more than a hundred thousand dollars while the renter's barely moved, and across thirty-three years homeowners' median wealth rose about $165,000 against roughly $5,800 for renters.9
The obvious objection is selection: people who can buy were already wealthier. So here is the figure that answers it. Among families in the bottom twenty percent of incomes, median net worth was $147,000 for homeowners and $3,400 for renters.9 Poor owners are not rich people. They are poor people who own something.
The mechanism is less glamorous than appreciation and more reliable. A mortgage is forced saving: every payment retires principal whether or not the owner had the discipline to save, and after thirty years the payment stops. Rent has no terminal point. That is why renters over sixty-five see net worth decline as they age while owners' equity continues to rise — the retiree who owns has a housing cost near zero and the retiree who rents has the same bill on a fixed income.9
This is also the tradition described in Issue Paper No. 11. On the first of January, 1863, the Homestead Act took effect — the same day Lincoln issued the Emancipation Proclamation. The Morrill Act had come six months before. The GI Bill followed. In each case the government handed a citizen an asset and then got out of the way. That is the American mechanism for distributing wealth, and it has never been a transfer payment.
1.3 And the tension we are not going to hide
This paper wants two things that pull against each other. More Americans holding an appreciating asset, and housing that more Americans can afford. If houses appreciate faster than incomes, every subsequent buyer pays more — which is the central contradiction in American housing policy and the reason the argument never resolves. The owner and the prospective owner want opposite things from the same building.
Two things follow, and we would rather state them than be caught between them.
The appreciating asset does not have to be the house. When the only one available to an ordinary family is the place they sleep, every household acquires a financial interest in housing being scarce — which is precisely the behavior Section 10.1 describes, and it explains the zoning hearing better than any account of snobbery. A country whose families also held the retirement accounts proposed in Issue Paper No. 11 would have less riding on the price of one building.
And the right target for housing is appreciation roughly in line with incomes, not faster. What built the postwar middle class was not nineteen-percent years. It was thirty years of amortization plus modest real appreciation. A house that outruns wages is a transfer from the young to the old, and this paper is written on behalf of the young.
1.4 And our own share of the failure
The people who most loudly defend property rights have, at the level of the town council, spent sixty years restricting what may be built on other people's property.
Minimum lot sizes, bans on multifamily construction, parking requirements, and discretionary review exist because homeowners showed up and voted for them. They are enforced by the same people who would describe themselves as opposed to government interference in markets, and they constitute the single largest body of government interference in any market in the United States.
Section 10.1 sets out the serious defense of that behavior, because there is one and it is not snobbery. But the contradiction should be named by us first: a movement cannot call zoning a property right while it is being used to tell a neighbor what they may not build.
Section 2Reform I: A rule against tilting the scale
No government at any level should adopt a rule, regulation or statute that makes building and selling a residence to a primary-residence buyer more costly than building and renting the same structure. Every existing rule should be tested against that standard and every proposed rule scored against it before adoption.
This is the structural reform and the others are examples of what it would catch.
Nothing here pretends government can equalize what markets do on their own. Selling to three hundred owners is genuinely a different business from renting to three hundred tenants, and no statute makes that difference disappear. What government can do is stop adding to it — which is what it has done, repeatedly, in ways nobody set out to do and nobody has since counted.
A rule that makes the for-sale version of a building more expensive than the rental version is a rule that manufactures renters, whatever it was written to accomplish.
We will say plainly that this is a thumb on the scale, and why we think it is permitted where picking a company is not. A free market requires broad participation the way a sport requires competitive teams — which is why leagues cap salaries and give the worst team the first pick, and why antitrust exists at all. Concentration is a threat to a market whether it comes from a government or a private party. A country where most of the land is held by a few families and none of them will sell has a market in name. Preferring dispersed ownership to concentrated ownership is a judgment about market structure. Preferring one firm to another is corruption. Those are different things and Issue Paper No. 7 draws the same line.
Section 3Reform II: Unbundle zoning
Zoning does three different jobs and they should be separated, because two of them are necessary and the third is the principal obstacle to American homeownership.
Life safety should be kept, with a route through it. Fire code is not zoning and conflating them discredits the argument. A warehouse converted to sleeping quarters without sprinklers, egress and fire separation kills people, and the answer is not to waive that. It is adaptive reuse pathways — alternative compliance routes that let an existing building be converted without meeting every standard written for new construction. That is what stands between a vacant office tower and the condominiums it could hold.
Genuine nuisance separation should be kept and kept narrow. A smelter beside a nursery school is the original justification, upheld in 1926 as codified nuisance law, and it remains sound. Most of what is now defended on nuisance grounds is not nuisance. Parking spillover is a parking meter problem.
Restriction on use and density should go. Minimum lot sizes, bans on duplexes and apartments, and floor-area limits protect nobody from harm. They protect an expectation about who the neighbors will be, and they are the mechanism by which a country with abundant land produced a housing shortage.
3.1 Form-based code and by-right approval
The replacement is not the absence of rules. It is rules about the shape of a building — height, setback, envelope, relationship to the street — and silence about who lives inside it or how many of them there are. That preserves every legitimate object of planning: industrial areas together, roads and sewers sized to demand, schools where the children are.
And the second half matters more than the first. If a proposal meets the code, it is permitted, without a discretionary hearing. Exclusion in America does not principally live in the zoning map. It lives in the hearing — in the eighteen months, the consultants, the appeals, and the risk that after all of it the answer is no. A developer prices that risk, and the price is passed to the buyer or the project is never started.
On the question of who does this: zoning is delegated state power, not inherent local power. States granted it and states can take it back without any federal involvement or constitutional question, and several have — Oregon, California, Washington, and Montana in 2023, which passed statewide preemption and called it a property rights reform. That is the clean path. Federal funding conditions are the second-best instrument where states will not act, and a federal mandate is neither available nor desirable.
Section 4Reform III: Make the condominium buildable again
This is the clearest single example of Reform I, and the numbers are stark.
Across the eleven Colorado counties holding eighty percent of the state's population, condominium development between 2018 and 2022 ran 76 percent below its 2002–2008 level. Before 2009 the state built roughly one condo for every 1.25 apartments. Recently it has built fourteen apartments for every one condominium.2
The mechanism is construction defect liability and the insurance it requires. Coverage for a for-sale building runs about 5.5 percent of hard costs against 1.1 to 1.65 percent for a rental building — more than 233 percent higher. Testimony put the added cost at roughly $30,000 per unit, and publicly traded production builders left the condominium market almost entirely.2
A builder facing identical land, labor and materials builds the rental version, because the for-sale version might bankrupt them. And the unit that disappeared is precisely the affordable one: a Denver condominium ran about $425,000 against $625,000 for a single-family house.2
The rung of the ladder that people used to step on first is not expensive. It is illegal to finance.
The target is not eliminating liability. A buyer's protection against defects is the liability, and a condominium with no recourse is a worse product that fewer people should buy. The target is making exposure predictable, bounded and resolvable: a defined warranty term, a right to repair before suit, independent inspection at completion, and limits on converting a maintenance dispute into class litigation by board vote.
We note that Colorado passed a reform in 2025, that Texas did a version in 2015 and Nevada in 2009, and that Colorado's is voluntary and insurers have said publicly they are unsure it changes underwriting. Which is the argument for Reform I rather than for another single fix.
Section 5Reform IV: Let the mortgage move with the borrower
Mortgages should be portable to a subsequent primary residence, and the assumability that already exists on federally backed loans should be extended.
Nearly all fifty-two million active American mortgages carry fixed rates, and most are priced far below the market. The consequence is measurable and large: each percentage point by which market rates exceed a homeowner's rate reduces their probability of selling by 18.1 percent. In the fourth quarter of 2023 that produced a 57 percent reduction in home sales among fixed-rate borrowers.3
Cumulatively, lock-in prevented an estimated 1.7 million transactions and raised home prices by 7 percent on its own. The Federal Reserve attributes 44 percent of the entire decline in homeowner mobility to it.3
And the burden falls where this paper is aimed: the research finds the effect disproportionately harms first-time buyers and lower-income households.3 A family that cannot buy is bidding against a house that never came to market.
This is the rare housing reform that increases supply and costs the Treasury nothing. The loan already exists. Nobody is subsidized. What changes is that a household can take a job in another city without repricing its shelter from three percent to seven.
Of the fifty-two million outstanding mortgages, only about 23 percent — the FHA, VA and USDA loans — are assumable today.3
Section 6Reform V: Title the manufactured home as a house
A manufactured home on land its occupant owns should be titled as real property and financed with a mortgage, and federal law should preempt local rules that bar these homes or impose lot requirements on them that site-built homes do not face. Separately, residents on leased lots should have a route to buying the ground beneath them.
The cost difference is not marginal. An average manufactured home runs $121,300 against $487,300 for a single-family house — about $113 per square foot against $166 — with a monthly cost to own of $1,317 against $2,188.4
6.1 Three arrangements, and only one of them is ownership
The word "manufactured home" covers three legally distinct situations, and conflating them is how this subject is usually discussed badly.
Chattel on leased land. The resident owns the structure and rents the ground. The home carries a vehicle-style title and is financed at seven to ten percent over fifteen to twenty years. It cannot appreciate meaningfully, because the thing that appreciates is land and they have no interest in it. The ground rent rises, and a "mobile" home that has not moved in twenty years cannot practically be moved. This is the arrangement most manufactured-home residents have.
Chattel on owned land. The resident owns both, but the home remains personal property under a separate title. They capture land appreciation and still pay car-loan rates on the structure.
Real property on owned land. Home and land merge into a single real-estate title, financed by an ordinary mortgage at six and a half to seven and a half percent over thirty years, with secondary market access and standard closing.
Only the third is ownership. And it is the only one anybody should be quoting appreciation figures about.
Because the standard objection — that these homes depreciate — is false for the third case and true for the first. Manufactured homes titled as real property appreciated 203.7 percent from 2000 to 2024, slightly exceeding site-built homes at 200.2 percent.4 The depreciation everyone associates with them is a function of the title and the ground, not the building.
And only 17 percent of new manufactured homes are titled as real property.4
6.2 What titling reform is, and what it is not
We should be precise, because this is easily misread. Converting a manufactured home to real property does not give anyone land. It requires that they already have it. Nearly every state conditions the conversion on the occupant owning the parcel, along with a permanent foundation and removal of the axles and hitch.
The conversion is a legal procedure — retiring the vehicle title and recording an affidavit that the home is affixed to the land. It exists in most states, it varies enormously in cost and difficulty, many owners do not know it is available, and many lenders will not finance through it.
So the reform is narrow and worth doing: make the conversion routine, uniform and cheap wherever the occupant owns the ground, and ensure the resulting loan is an ordinary mortgage eligible for the secondary market.
Zoning is the other half of it. Forty-nine percent of manufactured homes sit outside a metropolitan area, against 22 percent of single-family detached housing.4 A great many are on leased lots because an ordinary residential lot is closed to them by rule. Permit the home where a site-built house is permitted and the occupant will simply buy the lot, which solves the title problem without any conversion procedure at all.
And the Consumer Financial Protection Bureau found Black, Hispanic, American Indian and elderly borrowers disproportionately in chattel loans even controlling for whether they owned the land — which means part of this is a financing market failing people who qualify for better terms.4
6.3 The leased-lot resident, whom titling reform does not reach
Some 43,000 communities hold roughly 4.3 million homesites where the resident owns a structure on ground they rent, and ground rents rose 7.3 percent in a single recent year — the largest increase in two decades.4
Nothing in Section 6.2 helps that person. They cannot convert a title they are not eligible to convert, and no financing reform improves a position in which they own the depreciating half of the asset.
The answer for them is acquiring the ground, not improving the lease. Residents can and do buy their communities collectively through cooperatives, and the mechanism that makes it possible is a right of first refusal — notice when the owner intends to sell, and a window in which the residents may match. Some states grant it. Most do not. Extending it, and removing regulatory obstacles to cooperative formation and financing, is the reform.
We considered and rejected the alternative, which is to make long-term ground leases mortgage-eligible so that leased-lot residents can at least borrow on better terms. It is a half-measure, and this organization has already declined a ninety-nine-year leasehold model in Issue Paper No. 11 for the same reason: a very long tenancy is not ownership, and calling it ownership is how people end up holding an asset that expires.
So the position stands. A house on rented ground does not count as ownership, should not be counted in any ownership statistic, and should receive no benefit conditioned on ownership in this paper. The remedy for those households is the path to the land set out above — not a better lease described as a home.
Section 7Reform VI: Stop subsidizing conversions to rent only
Where a jurisdiction offers tax benefits to convert commercial buildings to housing, those benefits should be available to for-sale and rental conversions alike — and where an affordability requirement attaches, limited-equity cooperative and deed-restricted ownership should be permitted to satisfy it.
This is Reform I violated in statutes passed within the last two years, in the fastest-growing category of American housing production, in exactly the markets where ownership has fallen to 49.8 percent.
7.1 What is being built, and who will own it
The office-to-residential pipeline went from 23,100 units in 2022 to 70,700 in 2025, tripling in three years and now accounting for 42 to 47 percent of all adaptive reuse. New York leads with 8,310 units, Washington follows with 6,533, Los Angeles with 4,388.10
Every one of those figures is counted in apartments.
And converting to ownership is worth more than converting to rental. The New York City Comptroller found that office buildings converted to residential condominium trade at $591 per square foot against $276 for rental conversions — more than twice.10 It is still not what gets built.
Because the incentive is written for rental. New York State's Section 467-m, enacted in 2024, provides property tax exemptions of twenty-five to thirty-five years for commercial conversions — and requires that the building "be operated as rental housing," with all units rentals. The deepest benefits go to Manhattan below 96th Street. Washington's program offers twenty-year abatements.10
7.2 The argument for rental, which is better than we expected
We went looking for a bad motive and found a real one, so we will state it at full strength.
The statute requires that 25 percent of units be affordable, and its operative language is that a unit must qualify "upon initial rental and upon each subsequent rental following a vacancy" — affordable in perpetuity.10
That is administrable in rental and leaks in ownership. A landlord under a regulatory agreement re-qualifies each new tenant indefinitely. A condominium sold at a discount to an income-qualified buyer is resold at market price a few years later and the subsidy departs with the first owner. Preventing that requires deed restrictions or shared-equity structures that are hard to enforce across decades. The legislative history is explicit: the incentive was designed to produce affordable units "instead of just building condominium units."10
And there is a second argument we accept. Rental serves lower incomes. A household with no down payment and thin credit can rent and cannot buy, and if the objective is housing the most people fastest, rental does it.
7.3 Three answers, and New York already built the first one
The limited-equity cooperative solves the perpetuity problem. The resident owns, accumulates equity, and resells at a capped price that preserves affordability for the next buyer. New York invented the form and built tens of thousands of units with it under Mitchell-Lama. A commenter on the 467-m rulemaking asked exactly this — why limit conversions to rentals when the co-op model has proven to be the only sustainable form of affordability.10 The tool exists, the state that made it is the state now declining to use it.
And the affordability rationale governs one unit in four. The other 75 percent are market-rate apartments receiving a multi-decade tax exemption with no affordability obligation whatever. There is no perpetuity argument for those units being rentals. They are rentals because the statute says so.
A converted tower with five hundred apartments has one owner. As condominiums it has five hundred.
That is the concentration point, and it is worth being careful about where it applies. Nationally the fear of corporate landlords is overstated — small investors holding one to five properties own 87 percent of investor-owned single-family homes, and the largest corporate owners have been net sellers for six consecutive quarters.7 But a converted office building is not a scattered portfolio. It is a single asset with a single deed, and the public is buying a generation of single ownership in the most valuable district in America with a thirty-five year exemption.
One further consequence deserves naming, because it is the design and not an accident. A unit required to remain affordable to every successive tenant forever is a unit that will never transfer wealth to anyone who lives in it. Permanent affordability is permanent non-accumulation. That may still be the right policy for some share of the stock — a person needs a roof before they need an asset — but it should be chosen deliberately, and it should not be the default for an entire category of housing production.
Section 8Reform VII: Pay for units that do not exist yet
Where public money supports ownership, it should attach to the creation of a new unit sold to an owner-occupant, and where it attaches to a buyer it should be capped against a benchmark so that the buyer keeps what they save by buying less.
The reason is arithmetic and this organization has now made it three times. At seven percent, a $2,000 monthly payment supports a loan of about $300,600. At five percent it supports $372,600. At four, $418,900. A two-point subsidy does not make housing 24 percent cheaper. It lets every buyer bid 24 percent more, and where supply cannot respond the seller collects it.5
We ran this experiment in 2021 with mortgage rates near three percent — the cheapest money in American history — and home prices rose roughly nineteen percent in a year while the ownership rate barely moved.
The cost compounds as well. A two-point buydown is about $8,000 a year per household on a $400,000 loan. At three million qualifying purchases annually, that reaches $120 billion a year within five years and $240 billion within ten, because each cohort's obligation runs thirty years while new cohorts keep arriving.5
Two structures survive that objection. A credit to the builder, conditioned on sale to an owner-occupant, pays for a unit that did not exist rather than bidding for one that did. And a benchmark-capped benefit — set against the median price for a unit of that size in that county, with the buyer keeping the difference if they buy below it — rewards buying less rather than more.
That second structure is the inverse of what we do now. The mortgage interest deduction pays you more for borrowing more. It is a subsidy scaled to consumption, it bids up the stock, and it is on the list of preferences Issue Paper No. 8 proposes to eliminate. Capping or flattening it saves money, stops rewarding larger loans, and requires no new program.
We considered proposing expanded federal loan guarantees and rejected it. Fannie Mae and Freddie Mac acquired more than $650 billion of single-family mortgages in 2024 alone, guaranteeing them for an average fee of 65 basis points; the Federal Housing Administration insures at 3.5 percent down and more than eight in ten of its purchase borrowers are first-time buyers; the Department of Veterans Affairs guarantees at zero down.6 There is no shortage of lenders willing to lend against a federally guaranteed note. Adding capacity to a supply-constrained market raises prices, and the 2008 lesson is that a guarantee separates the party making the loan from the party bearing the loss.
Section 9Reform VIII: Stop the code favoring the landlord over the resident
A person who buys a house to live in and a person who buys the identical house to rent out should face the same tax treatment.
They do not. The investor depreciates the structure against income, deducts interest as a business expense, deducts maintenance, and may defer capital gains indefinitely by exchanging into another property. The resident gets none of that — and since 2017, most residents do not itemize at all, so the mortgage deduction reaches a minority of them.
We want to be careful with the popular version of this argument, because it is largely wrong. Investors bought roughly 30 to 33 percent of single-family homes in 2025, but small investors owning one to five properties hold 87 percent of investor-owned single-family homes. Institutional purchases ran about one-fifth those of small investors, and the four largest corporate landlords have been net sellers for six consecutive quarters.7
So this is not a paper about Wall Street buying America. It is a paper about a tax code that treats the same building differently depending on whether the buyer intends to sleep in it — and that asymmetry is a government action tilting toward rental, which is what Reform I prohibits.
Section 10 · The strongest case against
10.1 The homevoter is not being irrational
The serious defense of zoning is William Fischel's, and any reader who dismisses it will lose arguments to people who have thought about this harder than they have.
For most American households the home is by far the largest financial asset, and unlike shares it cannot be diversified — nobody owns a tenth of a house in ten neighborhoods. And unlike fire or theft, adverse neighborhood change cannot be insured against. There is no policy that pays out when the school declines.8
So the homeowner does the only thing available and votes that house at the hearing. Zoning is a substitute for an insurance market that does not exist, and the people defending it are protecting an undiversified, uninsurable position in the way the system left open to them.
Two observations follow. Home equity assurance — tried in Oak Park and Syracuse and never scaled — addresses the actual problem more directly than a veto over what neighbors may build. And the irony is worth naming: the more concentrated a family's wealth is in one house, the more fiercely it will defend the rules that make housing scarce. A country whose households held diversified assets, including the accounts proposed in Issue Paper No. 11, would have less at stake in the hearing.
10.2 We cannot promise a surge
Estimates of the American housing shortage run from 1.5 million to 6.8 million units. Nothing proposed here closes that, and the largest lever — land use — is exercised by thousands of local governments this paper cannot reach.
What these reforms do is remove specific, identifiable obstacles that make ownership units harder to build and harder to trade than rental units. That is a real and measurable objective. It is not the same as a number.
10.3 Portable mortgages have costs we are glossing
Portability complicates securitization, which is the reason it is not standard here and is standard in Canada. And an assumable low-rate loan is itself a valuable asset that a seller can price into the house, so some of the benefit capitalizes — less than a rate subsidy, because it is finite and attached to one existing loan, but not zero.
10.4 The buyer's protection is the liability
Section 4 proposes making construction defect exposure predictable and bounded. A reader who believes we have simply found a polite way to reduce what a homebuyer can recover from a builder who did shoddy work is entitled to that suspicion, and the history of such reforms gives them reason for it. Every limitation on recovery is a transfer of risk from the builder to the family living in the building.
10.5 The remedy for leased-lot residents is slower than the alternative
Section 6.3 proposes that residents on leased land acquire the ground collectively, supported by a right of first refusal when the owner sells. That is the right remedy and it is not a fast one: it requires a state to grant the right in the first place, which most have not, and it moves community by community.
The faster alternative is to make long-term ground leases mortgage-eligible, so leased-lot residents could at least borrow on ordinary terms next year rather than waiting for a cooperative conversion that may never come. We rejected it in Section 6.3, and a reader who weighs speed more heavily than we do is making a serious argument.
Our reason is that financing a leasehold more cheaply entrenches the leasehold, and that the appreciation evidence this paper relies on does not hold for homes on ground the occupant does not own. But that is a judgment about which remedy to pursue, not a claim that the other one is worthless.
10.6 The stewardship claim is softer than it sounds
It is often said that people care for what they own and neglect what they rent, and the research broadly supports it — owners maintain more, invest more, and participate more in local affairs.
But a substantial share of the measured effect runs through expected duration rather than ownership as such. Owners move less, and people who expect to stay invest in where they are; a long-term tenant behaves more like an owner than a short-term one. And the counterexample is real: a professional landlord with a maintenance budget may keep a building better than a low-income owner who cannot afford the roof. Deferred maintenance in owner-occupied housing is a documented problem, not a hypothetical one.
We therefore put stewardship third among our reasons rather than first. The wealth argument in Section 1.2 does not depend on it.
10.7 Tenure neutrality is a thumb on the scale, and we know it
Section 2 argues that preferring dispersed ownership is a structural judgment rather than favoritism. A reader who believes government should be indifferent between tenures, and that a household's choice to rent is as worthy as its choice to buy, is making a coherent argument that this paper does not accept and cannot refute from first principles.
We rest on the empirical claim — that ownership is how American households have historically accumulated — and on the observation that we are proposing to remove a tilt rather than create one. But a tilt removed in one direction is a tilt applied in the other, and we will not pretend otherwise.
10.8 What we concede, and what we do not
We concede that the homevoter argument is serious and that our answer to it is partial. We concede that we cannot promise a number. We concede that portable mortgages complicate securitization and partly capitalize. We concede that bounding defect liability shifts risk onto homebuyers. We concede that our manufactured housing reform reaches only households who own or can acquire the land, and that the route we propose for the rest is slower than the one we declined. And we concede that tenure neutrality is a preference, not a neutrality.
We do not concede that fourteen apartments per condominium is a market outcome. It is the product of a liability rule, an insurance market, a zoning code and a tax code, each adopted for its own reasons, none of them intended to decide who in America gets to own anything — and together deciding exactly that.
Section 11What we are not claiming
We are not claiming renters are failing at anything. Millions rent by preference, and mobility has real value that ownership reduces.
We are not claiming Wall Street caused this. Section 8 reports that small landlords hold 87 percent of investor-owned houses and the largest corporate owners have been selling.
We are not claiming cheaper credit is the answer. Section 7 shows it is the opposite, and 2021 is the proof.
And we are not claiming any of this is fast. Land use is decided in thousands of hearings this paper cannot attend, and the reforms that move fastest — portable mortgages and manufactured home titling — are the least discussed.
Section 12The argument you can carry
The paper compressed to what a person can remember and repeat.
And the one that indicts the argument's own side. The people who most loudly defend property rights have spent sixty years, at the level of the town council, restricting what may be built on other people's property — and minimum lot sizes and multifamily bans are the largest body of government interference in any American market. A movement cannot call zoning a property right while it is being used to tell a neighbor what they may not build.
Section 13Conclusion
There is an American mechanism for spreading wealth and it has never been a transfer payment. It was a hundred and sixty acres to anyone who would work them, a land-grant college in every state, and a mortgage for every veteran who came home. In each case the government handed a citizen an asset and then left them alone with it.
Nothing in this paper builds anything, owns anything, or leases anything for ninety-nine years. It observes that the country has made it more expensive to build and sell a dwelling than to build and rent the identical building, and proposes to stop.
Fourteen apartments for every condominium is not a preference anyone expressed. It is a liability rule interacting with an insurance market interacting with a zoning code interacting with a tax code — four decisions made separately, by people who were not thinking about who would own a home in 2026, which together answered the question.
The generation now living in childhood bedrooms did not fail a test. Twenty-five million of them are waiting for a rung that the country stopped building, and there is no mystery about why it stopped.
Notes
- U.S. Census Bureau, Housing Vacancy Survey, third quarter 2025: homeownership rate of 65.3 percent, 133.1 million total households, 37.5 percent among householders under 35, and the 2004 peak of 69.2 percent; see eyeonhousing.org. On the distinction between the owner-occupancy rate and the share of adults who own the home they live in — 53 percent nationally, with a 19-point gap in Hawaii — see the Federal Reserve Bank of Minneapolis analysis reported at newsnationnow.com, which is also the source for the Realtor.com finding that a record 25.2 million adults under 35 lived with their parents in 2025. On rates by area type — rural 73.9 percent, suburban 72.6, urban 49.8 — and by state, USAFacts, usafacts.org. On the composition of the owner population by age, including 23.4 percent aged 55 to 64 against 3.1 percent aged 25 to 29, and the figure of above 79 percent for those 65 and over, see the compilation at insuranceopedia.com; that source reports a different national rate using a different base and its figures should be checked against Census releases before publication.
- Common Sense Institute, "The Decline of Condo Development in Colorado" and related analyses, reporting condominium development across eleven Front Range counties 76 percent below its 2002–2008 level, the shift from roughly 1.25 apartments per condominium before 2009 to approximately 14 recently, insurance at 5.5 percent of hard costs for for-sale construction against 1.1 to 1.65 percent for rental, and the approximately $30,000 per unit figure from 2025 legislative testimony. On Colorado House Bill 25-1272 (2025), effective January 2026, and on insurer scepticism that it will change underwriting, see contemporaneous Colorado reporting. The Common Sense Institute advocates for construction defect reform; the underlying permit data is public and any published version should cite it directly. Texas enacted a version of these reforms in 2015 and Nevada narrowed its defect definition after 2008.
- Ross M. Batzer, Jonah R. Coste, William M. Doerner and Michael J. Seiler, "The Lock-In Effect of Rising Mortgage Rates," Federal Housing Finance Agency Working Paper 24-03, fhfa.gov and fhfa.gov. Source for the 18.1 percent reduction in sale probability per percentage point of rate gap, the 57 percent reduction in fixed-rate home sales in the fourth quarter of 2023, the estimate of 1.33 million prevented sales through 2023 rising to approximately 1.72 million through the second quarter of 2024, the 7.0 percent price effect, and the finding that the burden falls disproportionately on first-time buyers and lower-income households. On the 44 percent share of the mobility decline and the independent 8 percent price estimate, Board of Governors of the Federal Reserve System, "Locked In: Mobility, Market Tightness, and House Prices," federalreserve.gov. On the roughly 52 million outstanding mortgages of which about 23 percent are federally backed and assumable, and on the securitization complications of portability, Bipartisan Policy Center, bipartisanpolicy.org.
- On the average sales price of a manufactured home against a site-built home, cost per square foot, and monthly cost to own, and on the finding that manufactured homes titled as real property appreciated 203.7 percent from the first quarter of 2000 to the fourth quarter of 2024 against 200.2 percent for site-built homes, see the Manufactured Housing Institute's industry data; the appreciation series derives from FHFA house price index data and should be verified against the underlying index before publication, as the comparison is sensitive to the sample of homes included. On the share of new manufactured homes titled as real property, chattel loan pricing and terms, and the finding that Black, Hispanic, American Indian and elderly borrowers are disproportionately in chattel loans even controlling for land ownership, Consumer Financial Protection Bureau, "Manufactured Housing Finance: New Insights from the Home Mortgage Disclosure Act Data," consumerfinance.gov. On the 49 percent of manufactured homes outside metropolitan areas, the 43,000 communities and 4.3 million homesites, and ground rent increases, see Manufactured Housing Institute and Lincoln Institute of Land Policy materials. The Manufactured Housing Institute is the industry trade association and readers should weigh its framing accordingly.
- Author's calculation. A level-payment 30-year loan supporting a $2,000 monthly principal-and-interest payment amounts to approximately $300,615 at 7 percent, $333,583 at 6 percent, $372,563 at 5 percent and $418,922 at 4 percent. The annual interest subsidy figures assume a $400,000 loan balance, and the aggregate projections assume three million qualifying purchases annually with obligations persisting. These are illustrative and ignore prepayment, refinancing and sale, all of which would reduce the accumulated obligation; they are intended to show order of magnitude rather than to serve as a score. On 2021 house price appreciation against prevailing mortgage rates, cite Federal Housing Finance Agency and Freddie Mac series directly before publication.
- Federal Housing Finance Agency, "Fannie Mae and Freddie Mac Single-Family Guarantee Fees in 2024," fhfa.gov, for acquisitions exceeding $650 billion and an average guarantee fee of 65 basis points. On FHA borrower composition, U.S. Department of Housing and Urban Development, "Financial Status of the FHA Mutual Mortgage Insurance Fund FY 2024," hud.gov, reporting that more than eight in ten FHA purchase borrowers over four recent calendar years were first-time buyers.
- National Association of Realtors research on investor purchases, reporting investors at roughly 30 to 33 percent of single-family purchases in 2025, small investors holding one to five properties accounting for 87 percent of investor-owned single-family homes, institutional purchase volumes at approximately one-fifth those of small investors, and comparable offer prices between institutional and non-institutional buyers. On the four largest single-family landlords being net sellers across six consecutive quarters, see contemporaneous industry reporting. These figures move quickly and should be updated before publication.
- William A. Fischel, The Homevoter Hypothesis: How Home Values Influence Local Government Taxation, School Finance, and Land-Use Policies (Harvard University Press, 2001), hup.harvard.edu, and Zoning Rules! The Economics of Land Use Regulation (Lincoln Institute of Land Policy, 2015). Source for the argument that an owner-occupied home is an undiversifiable asset subject to uninsurable neighborhood risk, and that local land-use politics is best understood through that lens. We cite Fischel against our own proposal because it is the strongest available defense of the behavior this paper seeks to change. On home equity assurance programs in Oak Park, Illinois and Syracuse, New York, document from primary sources before publication.
- Board of Governors of the Federal Reserve System, Survey of Consumer Finances 2022, reporting median household net worth of approximately $396,200 for homeowners against $10,400 for renters, and 2019 figures of $295,480 and $7,270. On the thirty-three-year comparison — homeowners' median wealth rising about $165,000 against roughly $5,800 for renters, and the median wealth gap reaching almost $390,000 in 2022 — Urban Institute, "The Wealth Gap between Homeowners and Renters Has Reached a Historic High," urban.org. On the bottom-quintile comparison of $147,000 against $3,400, First American, blog.firstam.com, analysing the same survey; First American is a title insurer and benefits from transaction volume, and the underlying figures are the Federal Reserve's. On renters' net worth declining with age, see the Pennsylvania Association of Realtors' summary of the same data, parealtors.org. These are cross-sectional comparisons and do not by themselves establish that ownership causes wealth; the bottom-quintile figure substantially narrows but does not eliminate the selection concern, and any published version should engage the literature on selection directly.
- On the office-to-residential pipeline growing from 23,100 units in 2022 to 70,700 in 2025, the share of adaptive reuse, and the leading metropolitan areas, RentCafe's annual adaptive reuse report, rentcafe.com; RentCafe is a rental listing platform and counts units in apartments. On conversion economics — condominium conversions trading at approximately $591 per gross square foot against $276 for rental conversions — Office of the New York City Comptroller, "Office-to-Residential Conversions in NYC: Economics and Fiscal Estimates," comptroller.nyc.gov, which notes the condominium sample is small. On the statutory rental requirement, the 25 percent affordability threshold, the perpetual affordability language and the twenty-five to thirty-five year exemption terms, New York Real Property Tax Law § 467-m and the implementing guidance at nyc.gov. On the legislative intent to encourage affordable units "instead of just building condominium units," see the practitioner summary at rosenbergestis.com. On the limited-equity cooperative objection raised during rulemaking, rules.cityofnewyork.us. Washington, D.C.'s Housing in Downtown abatement should be documented from primary sources before publication.
A note on the author
Issue papers are published under the name of 1863 Leadership rather than an individual byline.
A note on sources
The two most important citations here are a Federal Housing Finance Agency working paper on mortgage lock-in and William Fischel's work on why homeowners behave as they do — the second of which is the strongest case against this paper's central reform, and we have given it a full section rather than a sentence. Several figures come from trade associations and advocacy organizations, and the notes say which: the Manufactured Housing Institute, the Common Sense Institute, the National Association of Realtors. Note 5 is our own arithmetic with its assumptions stated, and notes 2, 4, 5 and 7 flag figures requiring verification against primary data before publication.
Recommended citation
1863 Leadership. "The Missing Rung." Issue Paper No. 15.
September 2026. 1863leadership.org
Corrections: None to date. Errors of fact are corrected on this page within one business day of notice, with a dated note describing the change.