In 2025, for the first time since before the financial crisis, builders broke ground on more 2-to-4-unit structures than in any year since 2007, according to the National Association of Home Builders. It is a modest milestone by historical standards — a partial recovery from decades of policy-induced suppression — but it points to something real: the legislative tide on zoning is shifting at the state level, and the housing market is beginning to respond.
The reason 2-to-4-unit housing almost disappeared from the American production landscape was not market failure. Builders did not stop constructing duplexes and small apartment buildings because buyers stopped wanting them. They stopped because, lot by lot and city by city, local zoning codes made them illegal. The decades-long suppression of what housing economists call the "missing middle" — duplexes, triplexes, fourplexes, and small walk-up apartments — was a deliberate regulatory choice. So is its reversal.
Since 2019, more than a dozen states have enacted some form of zoning preemption: legislation that strips municipalities of their authority to prohibit certain housing types that the market would otherwise build. Oregon went first, and most ambitiously. Montana followed with a law that passed with near-unanimous Republican and Democratic support. Minnesota enacted duplex-as-of-right legislation. California removed local barriers to accessory dwelling units so effectively that ADU permits increased by more than 15,000 percent in a decade. The architecture of the housing shortage — built painstakingly over sixty years by local governments acting in the interests of existing homeowners — is now being dismantled at the state level. This is what supply-side reform actually looks like.
The Regulatory Tax on Every New Home
To understand why preemption matters, it helps to understand what the regulatory system it displaces actually costs. Economists Edward Glaeser and Joseph Gyourko have spent twenty years developing the analytical framework for measuring what they call the "zoning tax" — the gap between what housing would cost in a competitive, unregulated land market and what it actually costs given the constraints imposed by planning codes. Their most recent analysis, published in the Brookings Papers on Economic Activity (Spring 2025), finds that the gap between construction costs and market prices in constrained cities has widened to historic levels, with regulatory barriers accounting for the majority of the affordability gap in the highest-cost metropolitan areas.
Their foundational work — first published as a National Bureau of Economic Research working paper (WP 10124) and later in the Journal of Law and Economics — established the basic methodology: in a competitive market, the price of housing should approximate its marginal cost of production, including land. Where prices substantially exceed construction costs, regulation is functioning as a tax. The zoning tax they identify in high-cost cities ranges from tens of thousands to hundreds of thousands of dollars per unit — a toll levied not by any single agency, but by the cumulative effect of density limits, use restrictions, height caps, minimum lot sizes, parking mandates, and discretionary review processes that each individually appear benign.
The National Association of Home Builders and the National Multifamily Housing Council have estimated that regulatory costs now account for approximately $94,000 of the cost of every newly built home in the United States, a figure compiled from permit fees, compliance requirements, environmental reviews, and the soft costs imposed by the time delays that discretionary processes create. This is not a marginal inefficiency. It is the dominant driver of housing unaffordability in every constrained metro, and it is produced entirely by government action.
Oregon's Experiment: A Statewide Natural Laboratory
In 2019, Oregon enacted House Bill 2001, requiring all cities with populations above 25,000 to allow duplexes on every lot zoned for single-family use, and requiring larger cities to permit triplexes, fourplexes, and cottage clusters on similar terms. The bill effectively ended exclusionary single-family zoning as a legal instrument in Oregon's urban areas, shifting the default from "prohibited unless explicitly allowed" to "allowed unless there is a compelling reason to restrict."
Four years into the Oregon experiment, researchers at the National Bureau of Economic Research have begun measuring its effects. A working paper examining Oregon's subsequent housing reform legislation (Senate Bill 608), which streamlined infill approval and removed additional barriers to small-scale residential construction, found measurable increases in housing permit activity relative to what a counterfactual trajectory would have predicted. The results were not transformative in the way that advocates hoped — supply constraints are multi-layered, and removing a zoning prohibition does not instantly remove the financing, construction, and workforce constraints that also slow production — but they were unambiguously positive.
What is most instructive about Oregon is not the magnitude of the supply response but its character. The housing that materialized after SB 2001 was not large-scale development requiring complex financing and long regulatory timelines. It was small-scale. Homeowners adding a backyard unit. Developers converting a single-family structure to a duplex. Small builders constructing a fourplex on an infill lot. This is exactly the kind of organic, dispersed, market-responsive production that Hayek's framework on distributed knowledge would predict: not the output of a planning program, but the accumulated decisions of thousands of property owners and small builders responding to price signals the market had been suppressing for decades.
The International Evidence: Auckland's Unitary Plan
The most carefully studied large-scale upzoning outside the United States is Auckland's 2016 Unitary Plan — the most ambitious residential deregulation in New Zealand's history. Auckland's Unitary Plan rezoned approximately three-quarters of the city's residential land, removing density restrictions that had previously confined new residential development to single-dwelling forms across most of Auckland's neighborhoods. Areas previously restricted to detached homes were opened to duplexes, terraced housing, and low-rise apartment buildings as-of-right, without discretionary review.
The supply response exceeded most economists' expectations. Auckland's residential building consent rate accelerated sharply following implementation, and the city's new housing construction outpaced comparable New Zealand cities that maintained existing restrictions. Independent analyses of Auckland's rent and price trajectory following the Unitary Plan have consistently found that rents and prices in Auckland grew more slowly than would have been expected given the underlying demand conditions, a pattern consistent with an effective supply response. The Auckland experiment has become the reference case in the academic literature on large-scale supply deregulation: it is the closest thing to a clean empirical test of what happens when a government withdraws its prohibition on density, and the answer is that the market builds more housing and prices respond.
The State Preemption Wave: Montana, Minnesota, and Beyond
Oregon's reform was explicitly positioned as a progressive housing policy. What followed has been something different: a genuinely bipartisan supply movement that has found support across the ideological spectrum precisely because it does not require new spending, new agencies, or new bureaucracy. It requires only that state governments assert their constitutional authority to supersede local exclusion.
Montana's housing reform — passed through the Montana Legislature in 2023 with near-unanimous support from both parties — represents the most striking political realignment. Montana's Republican-dominated legislature enacted a package of zoning preemption bills that mandate duplex and ADU allowances in cities with populations above 5,000, eliminate parking minimums near transit corridors, and impose by-right approval for housing projects that comply with existing bulk and dimensional standards. The bills passed because Republicans in Montana understand something that coastal progressives sometimes obscure: zoning restrictions are government interference in private property markets. The free-market argument for preemption is not a concession to the left; it is the core conservative position.
Minnesota has enacted complementary reforms through the Minnesota Legislature, including legislation that requires municipalities to permit duplexes as-of-right throughout all residentially zoned areas in cities above a population threshold. Minnesota's approach has been incremental — its reforms have not matched Oregon's breadth — but the direction is the same: states withdrawing the local government's authority to prohibit housing types that willing buyers and willing sellers would otherwise transact freely.
California's approach has been more granular but equally consequential. The state's Assembly Bill 68 and subsequent ADU legislation stripped municipalities of their authority to impose the barriers that had previously made accessory dwelling units uneconomical: prohibitive minimum unit sizes, owner-occupancy requirements, excessive setback standards, impact fees disproportionate to the unit's actual infrastructure burden. The result was a permit surge that has delivered more than 150,000 new units in California since 2020 — units that produced no subsidy expenditure, no public land, and no discretionary approval. They were built by private homeowners and small developers, on existing residential lots, in response to a market signal that regulation had been suppressing for decades.
Why Local Control Produces Exclusion, Not Efficiency
The standard objection to state preemption is that it overrides local knowledge. Local governments, the argument goes, understand their communities' needs better than Sacramento or Salem or Helena. This objection would be compelling if local zoning boards were neutral arbiters of community preferences. They are not.
Economists Rebecca Diamond, Timothy McQuade, and Franklin Qian published a landmark study in the American Economic Review (2019) documenting the mechanics of how politically organized homeowners use discretionary land-use processes to restrict supply in ways that directly serve their financial interests. Using San Francisco's rent control and development restrictions as a case study, they found that housing restrictions consistently reflect the preferences of existing property owners — who benefit from scarcity — rather than the needs of prospective residents who are excluded by it. The "local knowledge" that zoning boards actually exercise is knowledge of how to perpetuate the conditions that make their members' assets appreciate.
Hayek's insight about the knowledge problem cuts in an unexpected direction here. The knowledge that is relevant to housing production is not held by planning boards — it is held by builders, buyers, renters, and property owners who understand the local demand for specific unit types. Local zoning processes do not aggregate that dispersed knowledge; they override it with the organized preferences of incumbent homeowners. State preemption does not centralize knowledge. It removes the veto on the market's own knowledge aggregation mechanism.
The Mercatus Center's analysis of light-touch density and states' role in zoning reform makes this point with particular clarity: the most efficient level of zoning control is not local. Local zoning creates a collective action problem in which each jurisdiction that restricts supply imposes costs on the regional housing market as a whole, while capturing the benefits of scarcity for its own homeowners. The regional externality can only be corrected at a level of government that internalizes it — which means the state.
The Friedman Test: Does This Cost Money?
Milton Friedman had a useful heuristic for evaluating government programs: follow the money. Every subsidy creates a constituency and an administrative apparatus that outlast the problem they were created to address. The housing subsidy complex — LIHTC, Section 8, CDBG, home buyer assistance programs, state housing trust funds — has grown continuously for fifty years while the housing shortage it was designed to address has deepened. This is not a coincidence.
State zoning preemption fails the Friedman follow-the-money test in the most desirable way possible: there is no money to follow. Preemption legislation does not create a new agency, a new entitlement, or a new administrative process. It eliminates an existing prohibition. The fiscal cost is zero. The supply effect is real. Thomas Sowell would recognize the structure: the housing shortage is not a problem that requires new government programs to solve. It is a problem that was created by government programs. The solution is to stop doing the thing that caused the problem.
This is why the state preemption wave is significant beyond its immediate housing effects. It represents a rare instance of a government correctly diagnosing that its own prior intervention was the cause of the problem it is now trying to fix, and responding by removing the intervention rather than layering a new one on top of it. The political economy of that diagnosis is fragile — the homeowner interests that benefit from scarcity do not disappear when states pass preemption laws, and many preemption measures have faced legal challenge from municipalities asserting home rule — but the direction is correct.
What the Data Now Requires
The Census Bureau's New Residential Construction data for 2025 shows the partial recovery in missing middle starts that the NAHB highlighted: the best year for 2-to-4-unit structures since 2007. Context matters: 2007 was itself a depressed base, reflecting the unwinding of the mid-2000s boom. The historical peak for small multifamily construction occurred in the 1960s and early 1970s, before the full spread of exclusionary zoning through American suburbs. Recovery to those levels would require multiple decades of sustained permit growth at rates the current trajectory does not yet support.
What preemption has delivered so far is proof of concept: when the prohibition is removed, the market begins to build. The pace of recovery depends on how completely the prohibition is removed and how aggressively states enforce preemption against municipalities that attempt to reinstate restrictions through other regulatory mechanisms — design review, discretionary approval, excessive impact fees. The Oregon and Montana experiences suggest that enforcement is the hard part: states can write preemption law, but municipalities have tools to nullify its effects if the state is not prepared to monitor compliance and impose consequences for evasion.
The free-market case for supply reform is sometimes framed as a position in tension with affordability concerns. The evidence from Oregon, Montana, California, and Auckland resolves that tension. The most effective housing affordability intervention available to any level of government — measured by units produced per dollar of public expenditure, measured by speed, measured by the breadth of the income distribution served — is removing the government's authority to prohibit the housing that the market would otherwise build.
The missing middle was not missing because the market failed to produce it. It was missing because the government made it illegal. The states that have acted on that diagnosis are now watching the market fill the gap. The states that have not acted are still asking why housing is unaffordable.
