Inclusionary zoning: why affordable housing mandates fail

The instrument was called inclusionary zoning, and its premise was structurally simple: require or incentivize private developers to set aside a portion of newly built residential units for households earning below median income, thereby embedding affordable housing inside the market’s own production pipeline.

Half a century later, the premise remains widely accepted. The outcomes do not. Empirical studies have found that mandatory inclusionary zoning — particularly when it is poorly calibrated to local market conditions — can reduce overall housing supply, raise the cost of market-rate units, and produce affordable units in quantities far smaller than its political promise suggests. The policy was designed to solve an affordability crisis. In many markets, it has become a structural contributor to that crisis.

Mandatory inclusionary zoning operates as a tax on residential construction, extracting private capital to subsidize below-market units while reducing the total volume of construction the private sector is willing to deliver.

The Evolution of Inclusionary Zoning: From Fairfax to 800 Jurisdictions

Inclusionary zoning did not emerge from a vacuum. Fairfax County’s 1971 ordinance was a response to a localized affordability problem in a county undergoing rapid residential expansion, and it operated initially as a voluntary density-bonus mechanism: developers who agreed to provide below-market-rate units received permission to build at higher densities than baseline zoning allowed. The voluntary design carried an implicit economic logic. The density bonus could offset the developer’s lost margin, preserving the project’s financial viability while producing a public benefit.

That logic began to erode as IZ migrated. Across the 1980s and 1990s, municipalities from Montgomery County, Maryland, to cities throughout California adopted IZ programs, but with progressively more aggressive structures. Voluntary programs gave way to hybrid models, and hybrid models gave way to mandates. By the mid-2010s, when New York City launched its Mandatory Inclusionary Housing program in 2016, the voluntary phase of the policy’s history had effectively closed in many major metropolitan areas.

Today, more than 800 U.S. jurisdictions maintain some form of inclusionary zoning ordinance, and the typical set-aside requirement ranges from 10 to 20 percent of total units in qualifying developments. That range is not an economic constant. A 10 percent requirement in one market can have a very different effect from a 10 percent requirement in another, depending on land costs, construction expenses, rents, interest rates, expected absorption, and the amount of additional density a municipality permits in return.

The migration followed a familiar political gradient: as housing affordability deteriorated in major coastal markets, demand for visible, quantifiable affordability outputs grew louder than demand for calibrated, market-sensitive policy design. Elected officials wanted numbers they could announce. Developers wanted predictability. Housing advocates wanted enforceable obligations. The result was a policy instrument that satisfied all three constituencies rhetorically — and that, in many markets, satisfied none of them economically.

The distinction between these models matters because “inclusionary zoning” is not one intervention. It is a family of policies with different requirements, incentives, exemptions, affordability terms, and administrative rules. A voluntary density bonus, a mandatory set-aside, and an in-lieu fee are not interchangeable instruments. Treating them as one policy makes it easier to claim success or failure without examining the mechanism that produced the result.

The Supply-Side Paradox: How Mandates Function as a Development Tax

The structural mechanics of mandatory IZ are direct. When a municipality requires a developer to lease or sell 15 percent of a new building’s units at below-market rates, the developer forgoes revenue on those units. That lost revenue is not automatically absorbed by the public sector. It must be absorbed somewhere in the project’s financial structure — often by the remaining market-rate units, by the price paid for the land, by construction quality, by the expected return on capital, or by the project’s feasibility altogether.

If the remaining units can be sold or rented at higher prices without losing enough demand to undermine the project, the developer may recover part of the cost through the market-rate portion of the building. In that case, the mandate functions as a cost-loading mechanism: the subsidy for below-market units is embedded in the price structure of units sold or rented at market rates.

But the cost cannot always be passed through. Housing demand is not infinitely elastic, and market-rate tenants or buyers have alternatives. A developer cannot simply raise prices by whatever amount is required and assume that the project will still perform. The project has to clear several thresholds at once: construction costs must be covered, financing must be available, expected revenues must justify the risk, and the land must be worth more in development than in its current use.

This is why the supply effect is more important than the arithmetic of the set-aside alone. A parcel that pencils out as a six-story market-rate condominium building may not pencil out as a six-story building with a 15 percent affordability requirement. The same parcel may pencil out only as a five-story building, or as a building constructed with cheaper materials, or as a building that is postponed until market conditions improve. It may also be sold to another owner who reaches the same conclusion. In the most extreme case, it is not constructed at all.

That is the supply-side paradox at the core of mandatory IZ: the policy’s affordability output is purchased, in part, by reducing the total supply of housing the market would otherwise produce.

The size of the effect depends on the market. In strong markets with rising rents and constrained land supply, developers can absorb mandates more readily because baseline project viability is robust. A project with unusually strong expected revenues may retain enough room to carry the cost of the set-aside. In weak markets, or in markets where construction margins are already thin, the same mandate can suppress project initiation entirely.

The variable most programs struggle to calibrate is not the nominal percentage of units reserved for affordability. It is the relationship between that percentage and the underlying economics of the project. The same requirement can be marginal in one neighborhood and decisive in another. A mandate that appears modest when expressed as a share of total units can become significant when combined with expensive land, high interest rates, restrictive height limits, or a long approval process.

Developers also respond before a project reaches the permitting stage. If a regulation makes future development less profitable, the price of developable land tends to adjust downward over time. That adjustment may eventually distribute part of the cost to landowners rather than builders. But land does not reprice instantly, and existing owners may prefer to wait rather than accept a lower value. In the interim, projects that would otherwise have moved forward can remain stalled.

This is the practical meaning of describing IZ as a development tax. The label does not mean that every dollar of the requirement appears as a dollar added to market-rate rents. It means that the mandate takes value from the development process, and that the burden is distributed through prices, land values, project design, returns, timing, and the number of projects that proceed.

Quantifying the Decline: Empirical Evidence on Housing Permit Contraction

The empirical literature on IZ’s supply effects is not uniform, but its central finding is consistent enough to matter. Powell and Stringham’s 2004 study of southern California municipalities found that jurisdictions adopting inclusionary housing policies experienced declines in residential housing permits ranging from 10 to 30 percent over the seven years following policy adoption. The variation correlated with the aggressiveness of local set-aside requirements and the underlying strength of the local construction market.

The finding is significant because permits are not merely an administrative measure. They are an early indicator of how much housing is likely to enter the pipeline. A decline in permits does not translate mechanically into an identical decline in completed homes: some permitted projects are never built, while others take years to finish. But fewer permits mean fewer projects have crossed the threshold from proposal to authorized construction. Over time, that difference compounds.

The Powell-Stringham finding has been reinforced by more recent work. Research by Noah Kouchekinia examining California’s post-2017 policy environment — following the state legislature’s “Palmer Fix” legislation, which restored municipal authority to mandate inclusionary requirements on rental developments after the 2009 Palmer v. City of Los Angeles ruling had restricted them — found that mandatory inclusionary zoning policies cut annual residential construction by approximately one-third.

That figure is not a theoretical projection. It is an observed supply contraction in one of the largest housing markets in the United States, measured against comparable markets without mandatory IZ. As with any empirical comparison, the result depends on the research design and the markets used as reference points. It should not be treated as a universal estimate for every jurisdiction. It does, however, establish the scale of the risk: a mandate intended to create affordable homes can also reduce the volume of new housing by a substantial amount.

The mechanisms behind these contractions are not mysterious. They are the predictable consequence of compressing developer margins below the threshold at which projects clear feasibility review. When a policy mandates a 15 or 20 percent set-aside at below-market rates, and when the developer cannot recover the lost margin through density bonuses, fee offsets, lower land costs, or sufficient market-rate rent growth, the project may not proceed.

The resulting supply loss is also difficult to see in the political accounting of an IZ program. The affordable units that are completed are visible. They can be counted, opened, and assigned to households. The market-rate units that were never proposed, or that were proposed but abandoned during the approval process, do not appear in the same tally. Neither does the housing that would have been built several years later if the first project had improved the economics of the site or demonstrated demand to lenders and competing developers.

That asymmetry creates a persistent bias in public evaluation. The program receives credit for the units it directly produces, while the opportunity cost remains diffuse. It is distributed across delayed construction, smaller projects, reduced density, and higher competition for existing homes. The policy may therefore look productive when judged building by building even as it is restrictive when judged across the entire housing market.

The Math of Market-Rate Displacement: Why Subsidies Often Fail to Scale

The most counterintuitive finding in the empirical literature concerns the relationship between set-aside requirements and the cost of the market-rate units that absorb the subsidy. Modeling of a 16 percent inclusionary zoning requirement in Los Angeles demonstrated that an additional unrestricted rent growth of just 0.8 percent per year would be sufficient to negate the entire economic value of the private subsidies provided by the IZ mandate.

In other words, if market-rate rents rise by less than one percent annually more than they otherwise would have, the cost imposed on market-rate renters exceeds the value of the below-market units the policy produces. The result does not mean that every tenant in an IZ building pays more, or that every affordable unit has no value. It means that the broader market response can offset the intended benefit when the lost supply is large enough.

This finding exposes a feedback loop that much of the political argument over IZ does not address. The policy is designed to embed affordable units inside new construction. But by reducing the total volume of new construction, the policy also constrains the supply of market-rate units that would have relieved upward pressure on rents in the broader market.

The below-market units are produced; the market-rate units that would have absorbed demand are not. Households that might have moved into those new market-rate units continue competing for older apartments. Some remain in units that no longer fit their needs. Others bid against lower-income households for existing housing. The affordability benefit inside the regulated building can therefore coexist with affordability losses elsewhere.

When a 0.8 percent annual rent increase can offset the entire economic value of private subsidies built into a 16 percent set-aside mandate, the structural calculus of affordability shifts decisively against the policy’s stated goals.

The arithmetic of displacement is easier to understand if the housing market is treated as a connected stock rather than a collection of isolated buildings. A new development does not serve only the households who move into it. It also creates vacancies in the homes those households leave, which can then be occupied by other households. This chain is not perfectly orderly, and new units do not all filter through the market in the same way. But additional supply can reduce pressure beyond the boundaries of the original project.

The reverse is also true. When a project is reduced, delayed, or canceled, the market loses more than the specific units in that project. The households that would have moved into the new building remain active participants in the existing market. They compete for older units, smaller units, units farther from employment centers, or homes in adjacent neighborhoods. The pressure is not eliminated; it is redirected.

These cascading effects extend well beyond the buildings subject to the mandate — to older buildings, secondary markets, and neighborhoods adjacent to the constrained primary market. The policy’s affordability output is real. Its affordability shadow is rarely measured with the same precision.

The effects are particularly severe in markets with already-low vacancy rates and constrained construction pipelines. In these markets, even modest reductions in permit issuance can translate into higher rent growth and greater household displacement. The markets most desperate for affordability solutions are, structurally, the markets in which mandatory IZ is most likely to impose the largest hidden costs.

This is also why the question “Does inclusionary zoning increase housing prices?” has no single answer detached from implementation. The mandate can raise the required price of market-rate units within a project, reduce the value of development land, or cause a project to be redesigned. Across the wider market, the effect depends on how much construction is lost and how responsive rents are to the resulting shortage. The relevant question is not whether the policy creates a cost. It does. The relevant question is who bears that cost and whether the affordable units created outweigh the housing that is no longer supplied.

Legal and Economic Friction: The Battle Over Nexus Studies and Property Rights

The economic tensions embedded in mandatory IZ have generated corresponding legal friction. Property-rights organizations — most visibly the Pacific Legal Foundation — have mounted constitutional challenges against inclusionary zoning ordinances, frequently invoking the Fifth Amendment’s Takings Clause.

The legal theory advanced by these challengers is that a municipality may be imposing an exaction on private development: requiring a developer to surrender part of a project’s economic value in order to obtain approval. In that setting, challengers often seek evidence connecting the required contribution to a burden associated with the proposed development. That evidence is commonly discussed through the language of nexus and proportionality.

The important point is narrower than the broad claims often made about IZ litigation. The supplied record does not establish that courts have generally required nexus studies for every inclusionary zoning mandate, nor does it establish a uniform pattern in which mandates lacking such studies are struck down while mandates supported by them survive. Litigation outcomes depend on the ordinance, the legal theory presented, the jurisdiction, the record developed by the parties, and the way a court characterizes the obligation.

What can be said with confidence is that property-rights challengers often invoke the Takings Clause and seek nexus evidence. They argue that a municipality should not be able to impose a particular affordability obligation without explaining how the obligation relates to the development’s alleged effects and why the chosen structure is justified. That argument places pressure on cities to make the economic logic of the mandate visible rather than treating the set-aside percentage as a purely political decision.

A nexus inquiry is not the same as a simple requirement that a municipality prove the exact amount of affordable housing created by a development. Nor is it accurately described as a demand that the set-aside be no larger than the entire public burden allegedly imposed on private developers. The legal and economic questions are more specific: what connection exists between the development and the burden identified by the municipality, and how does the required exaction relate to that burden under the governing constitutional framework?

Those questions are difficult because housing markets generate broad, cumulative effects. A single building may add demand for infrastructure, alter the local mix of households, or contribute to an existing affordability problem, but the effects of one project are not always easy to isolate. A city may be responding to a market-wide shortage, while a property owner may challenge the decision to place the cost of addressing that shortage on one development.

The result is a collision between two different ways of seeing the same policy. Municipal officials see a housing system in which development contributes to cumulative demand and public costs. Property-rights challengers see a specific private project being required to finance a public objective. Nexus evidence is one attempt to connect those levels of analysis. It does not eliminate the underlying dispute, but it forces the dispute into a more concrete form.

Property-rights challengers often invoke the Takings Clause and seek nexus evidence; that is a demand for justification, not proof of a uniform court rule for every inclusionary zoning mandate.

The legal friction therefore mirrors the economic friction. If a city sets a requirement without considering whether the project can absorb it, the mandate may function as a generalized charge on construction. If the city calibrates the obligation to local market conditions and provides meaningful offsets, the same policy can operate differently. The legal argument does not by itself determine the economic outcome, but it highlights the importance of explaining how the requirement was chosen and what burden it is intended to address.

The Calibration Gap

The empirical record on inclusionary zoning is not a record of uniform failure. Voluntary programs with calibrated density bonuses have produced affordable units in markets where they were paired with strong baseline conditions. Hybrid programs that combine modest set-aside requirements with generous density bonuses and predictable in-lieu fee structures can perform better than aggressive mandatory programs. The policy instrument itself is not inherently broken; the policy’s typical implementation is.

What the empirical record does show is that mandatory inclusionary zoning can reduce housing supply, raise market-rate costs, and produce affordable units in quantities that fall short of the policy’s stated goals when requirements exceed what project economics can support. The 10 to 30 percent permit declines documented in southern California, the approximately 33 percent construction reduction observed in the California analysis after 2017, and the 0.8 percent rent-growth threshold in the Los Angeles model are not interchangeable findings. They come from different studies and different settings. Together, however, they identify the same vulnerability: the affordability benefit of an IZ mandate can be overwhelmed by its effect on the volume and price of new housing.

A serious evaluation of an inclusionary zoning program therefore has to track more than the number of affordable units delivered. It has to ask:

• How many projects were proposed before and after the mandate?
• Did the typical project become smaller, less dense, or slower to approve?
• Were density bonuses large enough to offset the required subsidy?
• Did in-lieu fees produce more units through a housing fund than an on-site requirement would have produced?
• How did the policy affect market-rate rents, land prices, and the existing housing stock?
• Were the affordability requirements adjusted when construction costs, interest rates, or local rents changed?

These are not technical details added after the political decision. They are the decision. A mandate that looks moderate on paper may be severe in a market where land and construction costs leave little room for cross-subsidy. A higher set-aside may be viable in a market with substantial density capacity and strong demand. The percentage alone cannot tell us which outcome to expect.

The policy’s defenders argue that affordability cannot wait for perfect calibration, and that the cost of imperfect action is lower than the cost of inaction. The argument is not unreasonable at the level of political ethics. But it is weaker at the level of systems design. A policy that produces below-market units by suppressing a larger volume of market-rate construction may redistribute housing benefits without increasing the total amount of housing available to households under pressure.

That does not make the below-market units meaningless. A household that receives a regulated apartment experiences a real benefit, and that benefit should not be erased by an abstract market calculation. The question is whether the program can expand that benefit without worsening conditions for households that remain outside the regulated portion of the building. If it cannot, the program may help a visible group while intensifying the scarcity faced by a much larger one.

The choice is not between caring about affordability and caring about development. It is between different ways of paying for affordability. A municipality can place the cost on new private construction, fund it through general public revenue, sell additional development capacity, use public land, or combine several tools. Each approach has distributional consequences. What mandatory IZ does is make private development carry a significant part of the cost while presenting the resulting units as if they were free of trade-offs.

That is the central analytical failure. The policy counts the subsidy but discounts the supply response.

We do not need to abandon the goal of embedded affordability to acknowledge that the current mandatory instrument, as deployed across more than 800 U.S. jurisdictions, can function less as a solution to the affordability crisis than as a tax whose incidence falls partly on the households it purports to help. A voluntary program with a credible density bonus is not economically equivalent to a rigid mandate. A modest requirement in a high-margin market is not equivalent to the same requirement in a marginal project. And an ordinance supported by an explicit economic record is not automatically equivalent to one adopted without a clear account of its expected effects.

The durable lesson is not that every affordable housing mandate fails. It is that mandates cannot be evaluated only by the units they reserve. Their full impact includes the units they make more expensive, the projects they make smaller, the permits they prevent, and the households forced to compete for housing that already exists.

Until those costs are measured alongside the promised affordability output, inclusionary zoning will continue to look more successful in political announcements than in housing markets. The policy can produce affordable homes. But without careful calibration, it can also produce a less affordable market around them — and leave the households outside the set-aside carrying the bill.