Thank you for the opportunity to provide information to the Department of Housing and Urban Development (HUD) pertaining to the White House Council on Eliminating Regulatory Barriers to Affordable Housing. The Mercatus Center at George Mason University is dedicated to bridging the gap between academic ideas and real-world problems and to advancing knowledge about the likely consequences of proposed regulation for private markets. Accordingly, this comment represents the views of no particular party or interest group.
The president and Secretary Carson have directed attention to the problem of expensive housing in cities across the United States and correctly tied it to artificial scarcities induced by overregulation. Since land use regulation is under the authority of state governments and is largely administered by local governments, federal authorities have relatively little scope for corrective action.
In this comment, we first note two avenues for immediate federal action to alleviate artificial unaffordability. Then, in response to query IV(6)(a), we provide an annotated list of research resources on regulation, emphasizing those published by the Mercatus Center. Finally, we address query IV(6)(b) directly.
We recommend two avenues of federal executive action to lighten the regulatory burden and ensure that federal housing funds are being effectively used.
First, the Department of Labor should switch to modern, accurate survey methods to estimate “prevailing wages” that influence the cost of construction and the efficacy of federal housing subsidies. The Davis-Bacon and Related Acts require the Department of Labor to publish “prevailing wages” for each region. Many states require that contractors using Low-Income Housing Tax Credits (LIHTC) or other affordable housing subsidies pay “prevailing wages,” as calculated by the Department of Labor. The calculations, however, are riddled with errors: they rely on unrepresentative samples, are often out of date, cover regions with differing market wages, and vary widely between markets. Economist James Sherk has shown that the Bureau of Labor Statistics could vastly improve the accuracy of prevailing wage calculations.
Inaccurately calculated prevailing wages redirect labor subsidies that Congress intended for housing affordability, but only in regions where the errors in the prevailing wage calculation are on the high side. (When prevailing wages are inaccurately low, the invisible hand of the market protects workers from being underpaid.)
Second, HUD should propose and finalize an Affirmatively Furthering Fair Housing (AFFH) rule that disentangles HUD from exclusionary local policies. HUD does not (and should not) have the power to overturn local policies, even very bad ones. But it can and should limit the ways that funds are spent in jurisdictions where HUD’s intended beneficiaries—usually low- and moderate-income persons—are excluded by the paper wall of zoning. In a previous public interest comment we laid out one way to implement this principle; HUD ought to consider that and other approaches and find an administratively workable way to evaluate whether grantees are in fact affirmatively furthering fair housing.
Finally, Congress and the administration deserve credit for taking steps, large and small, that have chipped away at the privileged status of single-family homes in federal policy. The Tax Cut and Jobs Act of 2017 limited the value of the mortgage interest deduction, a tax incentive that nudged Americans toward high-leverage home purchases and favored high-cost, strictly regulated places. By increasing the standard deduction, the bill also sharply decreased the number of taxpayers who itemize in order to take the mortgage interest deduction. Secretary Carson has been a leader on housing policy reform by promoting supply-side reforms, including praising the elimination of single-family zoning in Minneapolis.
Research on Land Use Regulations
The following papers, from the Mercatus Center and elsewhere, provide relevant findings about the relationship between land use regulations and housing affordability, along with problems caused by constraints on new housing in the locations where people want to live.
Measuring the Cost of Zoning
A 2003 study by Edward Glaeser, Joseph Gyourko, and Raven Saks uses the difference between construction costs and house prices to estimate the “zoning tax.” As of then, Boston, Los Angeles, Newport News, New York, Oakland, Salt Lake City, San Jose, and Washington, DC, all had zoning taxes that accounted for more than 10 percent of housing costs. In San Francisco, the zoning tax accounted for more than 50 percent of housing costs. The zoning tax across the United States has certainly increased since Glaeser, Gyourko, and Saks made these estimates. HUD could replicate this study’s methodology to get an updated estimate of the portion of housing costs that’s driven by zoning.
Land Use Regulation and Housing Affordability
Land use is regulated in a variety of ways in American cities, all of which raise the cost of housing to some degree. Common regulations include such diverse restrictions as zoning, parking requirements, and historical preservation areas. These regulations are generally not subjected to cost-benefit analyses before they are implemented. One paper by Sanford Ikeda and Emily Washington collects research that measures the economic effects of these regulations.
Housing Supply Variation across Regions in the United States
The growth in supply regulations on housing through the 20th and 21st centuries is well known to urban economists and housing experts, but the effects of any given regulatory environment are difficult to quantify. Regulations are made in a great variety of ways across many small jurisdictions and can be difficult to compare across localities. Using census tract data, a paper by Salim Furth identifies which census tracts had housing growth above or below the average in the 2010s, allowing for comparisons of the restrictiveness of regulatory environments.
The Value of the Right to Build
In housing-constrained markets, the right to build a home increases land’s value precipitously. One method for measuring the value of the right to build a home is to compare the value of land that sits under a house to land in a home’s yard. One study using this approach finds that, in the greater Boston area, land that sits under a home is worth 20 times more than comparable land that doesn’t.
Minimum-lot-size regulations are one of the rules that can most clearly limit density and thus housing supply in a city. To limit supply, however, the regulations must be binding; if no one were to build on a smaller lot even in the absence of a regulation, then the regulation does not limit supply. Evidence from suburban Texas shows that even in one of the least regulated parts of the country, minimum lot sizes are binding. In these localities, lot sizes cluster at the minimum and exemptions are often granted, indicating that the minimum lot sizes are binding, restricting supply and density in these suburbs.
Inclusionary zoning programs require developers to build some number of below-market units to be allocated to low-income people in new developments. These regulations are growing in popularity as a proposed solution to housing affordability problems in expensive cities. However, it is likely that inclusionary zoning requirements act as a tax on housing, and they provide a limited number of subsidized housing units to low-income people only by raising prices for people who cannot secure one of the subsidized units. Evidence from the Baltimore-Washington region indicates that jurisdictions with inclusionary zoning have higher housing prices than what they could expect without the program.
Ranking Land Use Regulations in California Localities
Land use regulations vary widely across jurisdictions, even in California, perhaps the most regulated state. In any given jurisdiction, land use restrictions appear as both formal regulations, such as minimum lot sizes, and informal restrictions, such as slow processing of permits. One paper by Furth and Olivia Gonzalez attempts to construct measures of intensity of formal and informal restrictions in California. It shows that informal restrictions are difficult to compare across districts. It also shows that despite there being a reasonable range of intensity of formal regulations, that measure of intensity explains relatively little of the variation in supply limitations from regulation. This finding suggests that regulatory changes may have to be large to effectively increase supply.
State Preemption in Land Use Policy
Land use regulations restrict people’s ability to move into cities. The regulations are often locally popular in part because they raise the value of homes for owners already in cities. Land use regulations are thus a way for municipal voters to seek rents at a cost to people who would like to move in. Because people who are unable to move into a city cannot vote in it, there is no democratic recourse at the local level. However, state governments can preempt municipal governments and limit land use regulations as a democratic response to municipal restrictions.
New Market-Rate Construction and Housing Affordability
Opponents of new housing construction often argue that because new market-rate construction generally commands high rents or sale prices, it doesn’t help the cause of housing affordability. They point out that new construction can drive up costs when it replaces older, less expensive homes or when it leads to improved neighborhood amenities. This argument stands in contrast to the theory of filtering, which posits that when new homes are built, relatively high-income residents move into them, freeing up their older homes for residents with lower incomes, and so on. Evidence from two recent working papers supports the filtering theory; new construction that is not affordable to low- or moderate-income residents frees up other homes that are.
Along with density restrictions such as single-family zoning and minimum-lot-size requirements, parking requirements are one of the most significant land use rules that drive up housing costs. Meeting parking requirements with surface lots means that valuable space has to be dedicated to parking rather than housing, which makes areas with expensive land even more costly to live in. Meeting parking requirements with either above-ground or below-ground parking garages costs tens of thousands of dollars per spot.
Land Use Regulations and Homelessness
In recent years, the rate of homelessness has been increasing, in spite of federal programs to reduce homelessness and eliminate it among certain subpopulations. The states with the highest rates of homelessness—California, New York, and Oregon—are also home to localities where housing is difficult to build, rents are high, and vacancy rates are low. Research indicates that housing markets where housing is expensive for everyone have high rates of homelessness. HUD defines households as cost burdened when they spend more than 30 percent of their income on housing, and evidence indicates that, in fact, communities experience higher rates of homelessness when median rents exceed 32 percent of median income.
The Economy-Wide Effects of Land Use Regulations
Land use regulations that drive up the cost of housing cause severe consequences for renters and are even associated with homelessness. But their effects don’t stop there; local policy that prevents housing from being built in the places where people want to live reduces opportunities for economic mobility and economic growth. Evidence shows that the rise of constraints on new housing supply in high-cost cities has reduced both income convergence and growth in gross domestic product. Macroeconomists’ estimates vary as to how much spatial misallocation affects aggregate output. For a critique of economists Chang-Tai Hsieh and Enrico Moretti’s approach (cited in footnote 16), see a 2018 paper by Glaeser and Gyourko. See also a paper by Kyle Herkenhoff, Lee Ohanian, and Edward Prescott. These authors estimate that land use regulations reduce US GDP by 1.5 percent.
HUD should be guided by economic theory in adopting performance measures and should remain flexible and open to changing those measures if the initial choices prove to be incorrect. Local regulatory reforms cannot achieve every policy goal of current or future HUD leadership. Grounding evaluations of policies and reforms in correct conceptual models of housing markets will keep HUD’s focus on outcomes within localities’ control.
Several concepts guide the choice of evaluation metrics:
- Housing markets are regional, so an individual locality cannot be held accountable for the general price level.
- In high-demand markets, net housing production is the best indicator of regulatory freedom. In comparing cities, the baseline density or availability of vacant land is an important conditioning variable, since net housing growth can be much higher when land is available.
- Low-demand markets are much more difficult to evaluate: the rate of construction is low regardless of regulation. And when regulation is nonbinding, we do not expect it to be a significant determinant of rent.
- Households balance a variety of factors in choosing where to live, so a family’s rent-to-income ratio is expected to be higher if it chooses to live in a place with better amenities (such as job proximity or well-run schools). Furthermore, local rent-to-income ratios say just as much about incomes as they do about rents. Thus, when we see that Baltimore City has twice the rate of severe housing cost burden as suburban Carroll County, Maryland, we should not conclude that Carroll is more “affordable” in a meaningful way. We are wary about using localized rent-to-income ratios as a measure of housing affordability.
- Cost-additive regulations have different effects from density restrictions. Environmental requirements, regulatory delays, parking requirements, and design requirements can increase the cost of new building, but at a high-enough market price, construction will occur. Density restrictions, however, can only be overcome through politics.
- Cost-additive regulations are the principal barrier to affordability in places where land is abundant; density restrictions are the principal barrier where land is scarce and where prices are already high. HUD should promote the loosening of both kinds of restrictions, but the emphasis ought to differ according to the type of market.
- HUD should be concerned with the availability of low-cost housing typologies. A place that is only building 5,000-square-foot mansions is unlikely to be affordable even if the per-square-foot prices are reasonable.
- In many jurisdictions, the affordable housing stock consists of mainly older buildings, and there’s nothing wrong with that as long as these buildings meet current health and safety standards.
Conceptual rigor can only be implemented as far as accurate measurement allows. Some key variables are difficult to measure:
- Quantities: Housing production can best be measured by net growth in the number of residential addresses in the HUD Aggregated USPS Data on Address Vacancies. The more commonly used Building Permit Survey has extensive missing data and systematic underreporting. It also does not distinguish luxury rebuilds from net new units. However, Building Permit Survey data is widely used and is better than nothing.
- Prices: HUD’s own Small Area Fair Market Rents data are probably the best choice for HUD analyses, although they are subject to the same questions of measurement accuracy as any other metric. In areas where few homes are rented (i.e., most homes are owned), home prices may be a better way to measure cost. Rent and price data should strive to compare apples to apples, either through hedonic regression or by limiting the sample to comparable units.
One particularly promising metric is new home prices (adjusted for size and quality if possible), since new homes are more likely than old homes to have comparable quality and style, and they reflect current regulations. Although brand-new homes are rarely low-cost housing, they are a key indicator of the cost of adding new supply in the local market.
- Regulations: Measuring land use regulations is notoriously difficult and subjective. The evidence we have suggests that jurisdictions are not reliable self-reporters. Given the available data, we do not recommend that HUD use any published metric or index, although any efforts to add to the available data on zoning are welcome.
The nature of land use regulation is that there are many ways to accomplish the same thing. Rather than trying to keep track of all the barriers, HUD could instead set targets for permissiveness, allowing localities to achieve these through whatever reforms are locally relevant. A set of graded benchmarks for regulatory freedom could include items such as these examples:
- At least 2 percent (low pass) / 5 percent (high pass) / 10 percent (exceed) of land in a jurisdiction has no zoning, environmental, historical, aesthetic, or other regulatory barriers that would prevent the construction of multifamily housing by right, at a density of at least 24 units per acre. (Count land containing existing multifamily development only if the zoning allows for a density increment of 24 units per acre more than currently exists).
- No zoning, environmental, historical, aesthetic, or other regulatory barriers, nor fees in excess of $10,000, prevent the creation of an accessory dwelling unit in at least 20 percent (low pass)
/ 50 percent (high pass) / 80 percent (exceed) of the single-family homes in a jurisdiction.
For more concepts that reflect a free and open market for housing, see State Policy Network’s How to Build Affordable, Thriving Neighborhoods: A State and Local Zoning Reform Toolkit, which we coauthored.
Setting several of these benchmarks would allow HUD to identify areas of success and areas for improvement when interacting with local governments. Failure to get the lowest passing grade across several benchmarks would indicate to HUD that a local government does not share HUD’s goals for fair, accessible housing. In order to comply with its statutory requirements, HUD should apply extra scrutiny to any grants made to such jurisdictions to guarantee that those grants are spent in accordance with statutory requirements despite the jurisdictions’ rejection of HUD’s goals.
Another approach is to use a “market test” to identify jurisdictions that have high and rising rents and are failing to add substantial housing supply. This test holds jurisdictions accountable for outcomes rather than processes. Jurisdictions that fail the market test would be subjected to extra scrutiny in grant applications.
A jurisdiction could pass the market test by showing that any of the following four conditions are true:
- Rent is below the US median.
- Rent is below the median in its own metropolitan area.
- Real rent declined over the past five years.
- The jurisdiction issued net building permits for new housing units equal to at least 5 percent of its housing stock over the past five years.
The market test could be used alongside or in place of the graded benchmarks mentioned earlier. The strength of the market test is that it relies on actual outcomes; its weakness is that it identifies only poor performers, not exemplars.
Scarcity and high prices in US housing markets are caused by state and local policies that constrain new housing construction. Because the federal government has an appropriately small role in local land use policy, substantial reform to local land use regulation will be driven by state and local decisions.
Nonetheless, some tools to encourage reform are available to the White House Council on Eliminating Regulatory Barriers to Affordable Housing. Within HUD’s grant programs, the most natural opportunity to encourage local land use reform is to apply extra scrutiny to Community Development Block Grants in those localities that stand in the way of new housing, particularly relatively low-cost multifamily housing and accessory dwelling units. HUD has an opportunity to institute this rule by implementing standards for land use regulation for all of its grantees and standards for housing construction for high-cost grantees.
HUD has requested ideas for quantifying the cost of regulations that affect home prices. Replicating the methodology that Glaeser, Gyourko, and Saks use to estimate the zoning tax in their 2003 paper is one way that HUD could take steps toward this goal. This method would not be a way to determine the total cost of the regulatory burden in housing, but the zoning tax is an important component of that total.