Housing Theory of Everything — Land Is the Master Variable

The core claim of the housing theory of everything is narrow enough to test and broad enough to matter: the binding constraint on housing in rich countries is not construction cost, capital or demography, but the legal right to build on land near where economic activity already happens. Restrict that right and prices rise to whatever incomes will bear. The rent that results does not go to builders. It goes to landowners, because the scarce input is the land — or more precisely, the entitlement attached to it. Every downstream symptom that dominates political debate in the developed world — young adults deferring children, workers unable to move to higher-wage cities, wealth concentrating in the hands of people who bought early — follows from that single upstream rationing decision ◈ Strong Evidence [1].

The strongest empirical support for treating land as the master variable comes from long-run price data. Knoll, Schularick and Steger assembled annual house price series for 14 advanced economies going back to 1870 and found something counterintuitive: real house prices were broadly flat from the late nineteenth century to the middle of the twentieth, then rose steeply and persistently in the second half of the twentieth century ✓ Established [2]. Construction technology did not get worse. What changed was the treatment of land. Their decomposition attributes approximately 80% of the post-war global house price boom to rising land prices rather than to replacement cost ✓ Established [2]. Housing became expensive because land became expensive, and land became expensive because the supply of buildable, permitted land stopped responding to demand.

This distinction is not academic. If housing costs were driven by construction inputs, the policy response would be industrial: subsidise materials, train more tradespeople, automate assembly. If they are driven by land values created by regulatory scarcity, the policy response is legal — change what may be built, and where. Countries have spent four decades running the first experiment and are only now, in a scattered and partial way, running the second.

The distributional consequence is visible in the OECD's own accounts. Across member states, one in three low-income tenant households now spends more than 40% of disposable income on rent — the threshold at which housing is formally classified as overburdening ✓ Established [16]. In Colombia, Chile, Costa Rica, Spain and the United States, more than half of low-income tenants cross it ✓ Established [16]. These are not marginal populations. In the United States the January 2024 point-in-time count recorded 771,480 people homeless on a single night, the highest figure since the methodology was standardised in 2007 and an 18% increase in a single year ✓ Established [17]. The 2025 count fell about 3% to 745,652 — the first annual decline since 2016 — but chronic and individual homelessness both hit record highs, and 28 states still recorded increases ✓ Established [17].

The second-order effect follows mechanically. When aggregate incomes rise in a place where housing supply cannot expand, the additional income is bid into rent. Households compete for a fixed stock, and the gain accrues to whoever owns the land. This is why productivity growth in high-wage metropolitan areas has translated so poorly into improvements in living standards for the people who moved there to capture it — and why the same growth has translated so well into balance-sheet gains for people who already owned property before the growth arrived ◈ Strong Evidence [1].

The phrase "housing theory of everything" was coined in September 2021 by John Myers, Sam Bowman and Ben Southwood in Works in Progress. Their argument was deliberately expansive: that shortages of housing where people want to live are implicated in falling fertility, weak productivity growth, widening wealth inequality, higher carbon emissions and worse public health ◈ Strong Evidence [1]. Much of that essay is a synthesis of existing literatures rather than new evidence, and its most ambitious claims are the least well identified. But the central mechanism — supply constraint capitalises into land values, which redistributes income upward and prevents mobility — is among the better-established results in urban economics.

This report tests the theory channel by channel. It examines the productivity and mobility evidence, including the most consequential retraction in the field; the fertility evidence, which is stronger than its critics allow and weaker than its advocates claim; the wealth channel, where the data are least ambiguous; and four natural experiments — Tokyo, Auckland, Minneapolis and England — in which land use rules were actually changed and the results measured. It then examines why the most popular political response, capping rents, reliably deepens the shortage it is meant to relieve.

The mechanism has three components, and all three must be present for scarcity to bind. The first is a rule that makes the default answer "no": a zoning envelope, height limit, density cap or minimum lot size that forbids the building that market demand would otherwise produce. The second is discretion — a process by which exceptions can be sought but not relied upon, which converts a legal question into a political one. The third is standing: a defined class of people, usually adjacent property owners, who can object and impose delay. Delay is the operative weapon. A project that is merely slowed by three years may be as dead as one refused outright, because land carrying costs and interest do not pause for consultation.

Japan is the control group that demonstrates all three are choices. Under the City Planning Law, Japan's Ministry of Land, Infrastructure, Transport and Tourism sets 12 use zones nationally — seven residential, two commercial, three industrial — and applies the same categories from central Tokyo to the smallest rural municipality ✓ Established [21]. Two features distinguish the system. Zones are nested rather than exclusive: a use permitted in a more restrictive category is automatically permitted in the less restrictive ones, so the system accumulates permissions rather than segregating them. And because the categories are national, a local council cannot invent a bespoke overlay to block a building that complies. Compliance with the code is the permission. There is no discretionary hearing at which neighbours litigate the aesthetics of a proposal ✓ Established [21].

The political economy explains the persistence. Homeowners in supply-constrained markets hold a concentrated, highly liquid interest in continued constraint: the appreciation of a single house can exceed a decade of after-tax savings. The beneficiaries of new supply are diffuse, largely future residents, and disproportionately people who do not yet live in the jurisdiction and therefore cannot vote in it. Any system that routes land use decisions through local public hearings will systematically over-weight the first group. This is not a claim about bad faith. It is a claim about who shows up.

The result is a supply curve that is nearly vertical in precisely the places where demand is strongest. England illustrates the arithmetic. In 2024-25 England recorded 190,600 new build completions, which accounted for 91% of the net change in dwellings ✓ Established [23]; completions in the year to March 2024 had already fallen 6.5% to 198,600 ✓ Established [23]. The government's stated target is 1.5 million homes over the parliament, implying roughly 300,000 a year. The Office for Budget Responsibility, scoring the National Planning Policy Framework reforms, concluded they would raise housebuilding to its highest level in 40 years — and still projected only 1.2 to 1.3 million UK net additions to 2029-30 ✓ Established [15].

Canada and Australia present the same gap in different currencies. CMHC estimates that up to 4.8 million new homes must be built over the next decade simply to restore affordability to 2019 levels ✓ Established [18] — an earlier estimate had put the 2030 gap at 3.5 million units, roughly 60% of it concentrated in Ontario and British Columbia ✓ Established [18]. Australia's National Housing Accord target of 1.2 million homes is now not expected to be met until the end of 2030, with forecasts pointing to roughly 938,000 homes between mid-2024 and mid-2029, a shortfall of about 262,000 ◈ Strong Evidence [24].

What these numbers share is a structure. Each country has set a quantitative target, retained the discretionary permitting system that made the target necessary, and then expressed surprise at missing it. The target is a demand-side statement of intent. The constraint is a supply-side legal fact. Until the second changes, the first functions as a forecast of failure.

Begin with the mechanism, which is uncontroversial. Workers are more productive in some cities than others. If housing supply in high-productivity cities is capped, the price of living there rises until the wage premium is fully absorbed by rent. At that point the worker is indifferent between moving and staying — but the economy is not indifferent, because the output that would have been produced by that worker in the high-productivity city is never produced at all. Land use restriction therefore imposes a deadweight loss that appears nowhere in any city's own accounts. It is a national loss caused by local decisions.

Chang-Tai Hsieh and Enrico Moretti formalised this in a spatial equilibrium model covering 220 US metropolitan areas, published in the American Economic Journal: Macroeconomics in April 2019 ✓ Established [3]. Their headline finding was extraordinary: constraints on housing supply in high-productivity cities "lowered aggregate US growth by 36 percent from 1964 to 2009" ⚖ Contested [3]. A subsidiary estimate held that relaxing land use regulation in New York, San Francisco and San Jose to the level of the median American city would raise national GDP by 3.7% ⚖ Contested [3].

That figure became one of the most influential numbers in modern housing policy. It was cited in legislative testimony, in central bank speeches and in the founding documents of the YIMBY movement. It is also, on the current state of the evidence, wrong. In April 2026 the same journal published a comment by Brian Greaney documenting errors in the original code ✓ Established [4]. Attempting to replicate the result, Greaney found that the authors' own counterfactual would lower output rather than raise it, and that the model's findings depended on the arbitrary choice of population unit — a specification artefact rather than an economic result ✓ Established [4]. After proposing corrections that remove the unit dependence, he finds the liberalisation experiment does raise output, but by an amount "two orders of magnitude smaller than what they report" ✓ Established [4].

This matters beyond one paper. It is a test of whether the housing theory of everything is an evidence-driven position or a movement with a favoured statistic. The honest reading is that the direction of the effect is well supported and the size is unresolved, with credible estimates spanning at least two orders of magnitude. That is a wide band, and any policy argument that depends on the top of it is fragile.

The mobility data are more robust, and they point the same way without relying on a structural model. The annual US interstate migration rate has fallen from roughly 3% in the 1980s to about 1.5% today ✓ Established [20]. The Federal Reserve Bank of Richmond attributes the decline to three factors — demographic ageing, housing affordability, and changes in labour market dynamics ◈ Strong Evidence [20]. Americans used to move toward opportunity; increasingly they cannot afford to arrive.

The composition of moves has shifted alongside the level. Job-related relocation has fallen sharply as a share of all moves, consistent with a labour market in which the wage gain from moving to a high-productivity metro no longer covers the housing cost of living there ◈ Strong Evidence [20]. This is the mechanism working exactly as theory predicts — but it manifests as an absence, a move that never happens, which makes it invisible in most economic statistics and easy to under-weight politically.

There is a further channel that resists quantification. Cities are where new firms are founded, where specialised labour markets clear, and where the tacit knowledge that drives innovation circulates. If the marginal worker in a dense, high-productivity metropolitan area is priced out, the loss is not merely their forgone wage differential but their forgone participation in that agglomeration. The empirical literature on agglomeration externalities is old and reasonably consistent in finding that density raises productivity; what it cannot do is tell us the counterfactual size of a city that was never allowed to exist.

The defensible conclusion is narrower than the popular version but still consequential: housing constraint suppresses labour mobility, the suppression is measurable and large in the mobility data, and the aggregate output cost is real but currently unquantified within an order of magnitude. Advocates who cite 36% and sceptics who cite 0.02% are both quoting numbers whose confidence intervals do not support the certainty with which they are deployed.

The intuitive version of the argument is that children require space, space costs money, and when space becomes unaffordable people have fewer children. The empirical version is more specific and more interesting. Rising house prices affect fertility through two distinct channels that point in opposite directions for different households. For existing owners, a price rise is a wealth gain, and the literature has long found that home equity appreciation modestly raises fertility among owners. For renters and prospective buyers, a price rise is a cost increase, and it lowers fertility. The net population effect depends on the ratio of owners to renters among people of childbearing age — which is precisely what land use constraint has been changing.

van Wijk and Feijten, publishing in the European Journal of Population in 2025, decompose this directly. They find rising house prices are associated with lower aggregate fertility, and that the effect runs partly through the falling propensity of young adults to become homeowners at all, and partly through reduced childbearing among renters in expensive markets ◈ Strong Evidence [13]. The mechanism is compositional as much as behavioural: as constraint pushes more of the childbearing-age population into the renter category, the negative channel gains weight relative to the positive one, and the aggregate effect turns negative and grows.

The tenure collapse in the United Kingdom is the clearest illustration available anywhere. In 1995-96, 65% of 25-34 year-olds with middle incomes for their age owned their home. Twenty years later, that figure was 27% ✓ Established [19]. Reading the same shift by birth cohort makes it starker: at age 27, those born in the late 1980s had a homeownership rate of 25%, against 33% for those born five years earlier and 43% for those born ten years earlier ✓ Established [19]. The Institute for Fiscal Studies attributes the decline chiefly to house prices rising relative to incomes ◈ Strong Evidence [19]. Within a single generation, the modal 27-year-old in Britain moved from being a probable owner to an improbable one.

The honest counterweight is Japan. Japan has the most permissive land use system in the developed world, national zoning that removes the local veto, and large-city housing that is dramatically cheaper relative to income than London, Sydney, Toronto or San Francisco ✓ Established [21]. It also has one of the lowest fertility rates on earth. If cheap housing were sufficient for fertility recovery, Tokyo would be the counter-example that proves it. It is not.

This does not refute the housing channel, but it does bound it. The correct reading is that housing cost is one input into fertility among several — alongside female labour force participation, childcare availability, working hours, gender norms within households, and the opportunity cost of caregiving — and that removing the housing constraint removes one barrier without addressing the others. Advocates who present land use reform as a fertility policy are overselling; critics who present Japan as a refutation are ignoring the counterfactual, which is what Japanese fertility would be if Tokyo housing cost what London housing costs.

The more defensible framing is subtractive rather than additive. Expensive housing does not explain why fertility is falling everywhere, including in countries with cheap housing. It explains a specific increment of the decline in specific places, concentrated among renters in high-cost metropolitan areas — and it explains why the fertility gap between owners and renters has widened rather than narrowed as prices rose ◈ Strong Evidence [13].

There is a distributional corollary that deserves more attention than it receives. Because the fertility effect operates through tenure, and tenure is increasingly inherited — determined by whether a young adult's parents can supply a deposit — housing constraint converts fertility itself into a partially hereditary outcome. The children of owners become owners and have children earlier; the children of renters remain renters longer and have fewer. That is a mechanism by which land use law reaches into demography across generations, and it is invisible in any aggregate national fertility statistic.

The arithmetic of capitalisation is unforgiving. Suppose a city's productivity rises, raising wages by 10%. If housing supply can expand, some of that gain is spent on more or better housing and the rest is retained as higher real income. If housing supply cannot expand, households bid against each other for an unchanged stock, and the wage gain flows through to rents and then to land prices. The worker is no better off. The landowner is 10% richer without having done anything. Nothing was built, nothing was invented, and a real productivity improvement was converted into a pure transfer.

At $393.3 trillion — approximately four times global GDP — real estate is the largest asset class on earth, exceeding the combined value of global equities, debt and gold ✓ Established [14]. Residential property accounts for $287 trillion of that total ✓ Established [14]. When an asset class of this scale appreciates faster than wages for four decades, the distributional consequence is not a side effect of the economy. It is the economy's principal distributional mechanism.

The Knoll, Schularick and Steger series makes the historical break visible. Real house prices across 14 advanced economies were broadly flat for eighty years, from 1870 to roughly 1950 ✓ Established [2]. Over that period, capital gains on housing were not a meaningful source of wealth accumulation for anyone; housing was a consumption good that depreciated. The post-1950 divergence, driven approximately 80% by land prices, created an entirely new asset — the appreciating owner-occupied home — and with it a class of households whose balance sheets grow through no economic activity of their own ✓ Established [2].

The intergenerational consequence compounds. Where entry to ownership depends on a deposit that exceeds several years of median savings, the binding variable becomes parental wealth rather than the buyer's own income or productivity. Two people with identical earnings and identical savings rates arrive at radically different tenure outcomes depending on whether their families own appreciated property. This is the reintroduction of inherited position into economies that spent the twentieth century dismantling it, achieved not through inheritance law but through land use law.

The measurement problem is that this transfer appears almost nowhere as a transfer. Rising house prices are recorded as household wealth creation in national accounts, and reported in the press as good economic news. But an unbuilt house that becomes more expensive has created no wealth in any aggregate sense — it has revalued a claim. For every household whose balance sheet improved, there is a future household whose lifetime housing cost rose by the same amount. Aggregate housing wealth is, to a first approximation, a measure of how much the young owe the old.

The rental market carries the same transfer in flow rather than stock form. Across the OECD, one in three low-income tenant households now spends over 40% of disposable income on rent, and in the United States, Spain, Chile, Colombia and Costa Rica more than half do ✓ Established [16]. That expenditure is not investment. It buys no equity, accumulates no claim, and its principal economic function is to service the land value that regulatory scarcity created.

This is why the wealth channel is the strongest leg of the housing theory of everything. It does not depend on a contested structural model, as the productivity channel does, or on disentangling housing from a dozen confounders, as the fertility channel does. It follows directly from the price data, and the price data have been assembled consistently across 14 countries and 140 years ✓ Established [2].

Auckland is the cleanest test available anywhere in the world. The 2016 Auckland Unitary Plan upzoned roughly three-quarters of the city's residential land for medium and high density in a single instrument, on a schedule set by central government rather than by neighbourhood negotiation ✓ Established [5]. Because the change was large, abrupt, and applied to one city within a country of otherwise comparable cities, it supports synthetic control estimation — constructing a counterfactual Auckland from a weighted combination of non-upzoned New Zealand cities and measuring the divergence.

Greenaway-McGrevy and So, publishing in Economic Inquiry in July 2026, applied exactly that design to Ministry of Housing and Urban Development rental bond data covering 2017 to 2024. Their finding: "Eight years after the reform, the synthetic control under our preferred empirical specification implies that rents decreased by 23% due to the reform" ◈ Strong Evidence [5]. The preferred specification is the conservative one — it excludes Rotorua, the largest-weighted donor city, following leave-one-out robustness testing ◈ Strong Evidence [5]. Earlier specifications at the six-year horizon produced a larger estimate of 28%.

The dissent is real and should be stated plainly. A body of critical work, including an independent review of the Auckland literature published by Motu, questions how much of the observed construction increase can be causally attributed to the Unitary Plan rather than to concurrent interest rate movements, migration cycles and infrastructure investment ⚖ Contested [6]. The synthetic control method is only as good as its donor pool, and New Zealand has few cities. This is a legitimate methodological objection, not motivated reasoning. It does not overturn the finding, but it should moderate confidence in the point estimate.

Minneapolis ran a smaller version of the same experiment. The 2040 Plan, adopted in 2018 and implemented from 2020, ended exclusive single-family zoning citywide, removed parking minimums and granted administrative approval for apartments near transit and commercial corridors. Pew's analysis found that from 2017 to 2022 Minneapolis grew its housing stock by 12% while rents rose 1%; over the same period the rest of Minnesota added 4% to its stock while rents rose 14% ✓ Established [7]. Nearly 21,000 units were permitted, of which 87% were in buildings of 20 or more units and just 1% in two-to-four-unit buildings ✓ Established [7].

That composition detail matters more than the headline, and it cuts against the popular narrative. The abolition of single-family zoning — the reform that generated the national coverage — produced almost none of the new housing. What produced it was the administrative approval pathway and the height minimums along transit corridors ✓ Established [7]. Reform that removes the discretionary veto delivers supply; reform that merely relegalises duplexes in principle, while leaving discretion and process intact, largely does not.

The Federal Reserve Bank of Minneapolis has published the necessary caution: some of the divergence between Minneapolis rents and the rest of Minnesota reflects softening demand rather than new supply alone ⚖ Contested [8]. Any single-city comparison over a period that includes a pandemic, a remote-work shock and an interest rate cycle is vulnerable to this critique. The supply effect is well supported; its precise share of the total is not.

Tokyo is the long-run control rather than a discrete experiment, because Japan never adopted the discretionary model. Nationally set use zones, nested permissions and code-based approval mean the local veto that dominates Anglosphere planning simply does not exist as a legal instrument ✓ Established [21]. The consequence is that Tokyo has absorbed decades of population growth without the price trajectory seen in comparably sized global cities. It is the strongest available evidence that the Anglosphere's housing costs are institutional rather than inevitable.

England is the counterfactual in the other direction — a jurisdiction that has set the most ambitious numerical target in the developed world while retaining the discretionary system that made the target necessary. Completions were 190,600 in 2024-25 against an implied requirement of roughly 300,000 a year ✓ Established [23]. Even the OBR's optimistic scoring of the planning reforms, which it judged sufficient to produce the highest housebuilding in 40 years, projects only 1.2 to 1.3 million UK net additions to 2029-30 ✓ Established [15]. The reforms are real and their direction is right. They are also, on the government's own watchdog's numbers, insufficient.

The pattern across all four is consistent. Where reform removed discretion and made permission a matter of code compliance — Auckland's blanket upzoning, Minneapolis's administrative approval pathway, Japan's national zones — supply responded and prices moderated relative to counterfactuals. Where reform adjusted the parameters of a discretionary system without removing the discretion, as in England, output improved marginally and targets were missed. The operative variable is not how much density is nominally permitted. It is who has the power to say no, and how long they can take to say it.

The appeal is obvious and the diagnosis is half right. If rents are the problem, cap rents. And for tenants who hold a controlled tenancy, the policy works exactly as intended — their rent falls, their tenure stabilises, and their displacement risk drops. The dispute concerns what happens to everyone else, and to the stock itself, over the following decade.

Berlin ran the most dramatic version. The Mietendeckel, in force from February 2020 until the Constitutional Court struck it down in March 2021, froze and rolled back rents across the city. Advertised rents inside the cap fell by roughly 11% ✓ Established [10]. The ifo Institute found the supply of rental listings shrank by up to 60% ✓ Established [10] — landlords withdrew units, converted them, sold them to owner-occupiers, or simply stopped advertising. The shortage did not fall; it was reallocated to whoever already held a lease. And it spilled outward: rents in the commuter belt rose, with Potsdam up 12% as displaced demand searched further out ✓ Established [10].

Saint Paul provides the North American counterpart, and it is instructive because the policy was adopted by referendum in a city with a directly comparable neighbour. In the six months after the November 2021 ordinance passed, Saint Paul permitted 352 housing units, against 2,180 in the same period a year earlier — a fall of 84% ✓ Established [9]. Across the river, Minneapolis was simultaneously running the opposite policy and recording 12% stock growth against 1% rent growth ✓ Established [7]. The Federal Reserve Bank of Minneapolis's 2026 review of Saint Paul's experience describes mixed results, noting that subsequent amendments exempting new construction partially restored development activity ◈ Strong Evidence [9].

The San Francisco evidence establishes the long-run mechanism with the strongest identification. Diamond, McQuade and Qian exploited the 1994 expansion of rent control to small multi-family buildings and found that affected landlords reduced the supply of rental housing by 15%, and that controlled buildings were 8% more likely to be converted to condominiums ✓ Established [11]. Crucially, the authors found the policy achieved its intended effect for incumbent tenants while reducing the citywide rental stock — raising rents for everyone not protected by it. The policy redistributed within the renter population rather than from landlords to renters.

None of this makes tenant protection illegitimate. Security of tenure, notice requirements, deposit regulation and protection against retaliatory eviction address genuine power asymmetries and have no comparable supply penalty, because they do not attack the price. The specific failure mode belongs to price ceilings, which suppress the signal that would otherwise call forth supply while leaving the underlying scarcity untouched.

There is a deeper point here about the housing theory of everything. Rent control is what a political system reaches for when it has diagnosed the symptom and cannot touch the cause — because the cause is the land use regime that its own median voter owns a stake in. The popularity of price control is therefore evidence for the theory rather than against it: it is what constrained supply looks like when it becomes politically unbearable but remains politically untouchable.

Assessed channel by channel, the theory's claims have very different evidentiary standing, and treating them as a single package does the argument a disservice. The wealth channel is the strongest: it follows directly from 140 years of price data across 14 countries, requires no structural model, and is confirmed by the tenure collapse visible in national statistics ✓ Established [2] ✓ Established [19]. The mobility channel is next: the decline in US interstate migration from roughly 3% to 1.5% is a measured fact, and housing affordability is one of three leading explanations offered by the Richmond Fed ✓ Established [20].

The fertility channel is real but bounded. The tenure mechanism identified by van Wijk and Feijten is well specified and consistent with the cohort data ◈ Strong Evidence [13], but Japan demonstrates that cheap housing is not sufficient for fertility recovery ✓ Established [21]. The aggregate GDP channel is the weakest, and it is the one most often quoted. Hsieh and Moretti's 36% figure ⚖ Contested [3] does not survive Greaney's replication ✓ Established [4], and the corrected effect is two orders of magnitude smaller ✓ Established [4].

The most useful finding to emerge from these experiments is not about density at all. It is about discretion. Minneapolis's headline reform — ending single-family-only zoning — produced 1% of the new units; 87% came in buildings of 20 or more units built under the administrative approval pathway ✓ Established [7]. Freemark's Chicago study found upzonings that raised permitted density without changing the approval process raised land values and produced no additional housing over five years ◈ Strong Evidence [12]. Permission on paper is not permission in practice. What binds is the process, and the process is where incumbent power lives.

Three policy implications follow with reasonable confidence. First, by-right approval matters more than nominal density limits: a jurisdiction that permits six storeys as of right will out-build one that permits twelve subject to discretionary review. Second, reform must be broad enough to prevent displacement of objection — parcel-by-parcel upzoning, as in Chicago, invites capitalisation into land without construction ◈ Strong Evidence [12]. Third, price controls and supply reform are not complements. Saint Paul and Minneapolis demonstrate the two policies working against each other within a single metropolitan labour market ✓ Established [9] ✓ Established [7].

The 2025-26 legislative cycle suggests the diagnosis is spreading faster than the evidence base. California's SB 79, signed on 10 October 2025, permits up to seven storeys within a quarter-mile of major transit stops across eight counties and takes principal effect on 1 July 2026 ✓ Established [22]. New Zealand introduced the Planning Bill and Natural Environment Bill on 9 December 2025 to replace the 1991 Resource Management Act entirely. England's reforms were scored by its own fiscal watchdog as producing the highest housebuilding in 40 years — and still falling short ✓ Established [15].

What remains unresolved is the aggregate magnitude. If the corrected estimates are closer to Greaney's than to Hsieh and Moretti's, land use reform is a large distributional policy and a modest growth policy ✓ Established [4] — which changes the political argument considerably, because distributional policies create identifiable losers and growth policies do not. The evidence supports the second more strongly than the first.

The honest summary is this. Land use regulation is the master variable for housing costs, and housing costs are a master variable for wealth distribution, household formation and internal migration. It is not the master variable for aggregate growth, fertility or inequality as such — those have multiple independent causes, and treating housing as sufficient for all of them invites the disappointment that follows every over-promised reform. The theory of everything is better understood as a theory of one thing that touches many: who is permitted to build, where, and who gets to object.