Housing Unaffordability: How We Got There and What to Do About It


Capital will always go where it’s welcome and stay where it’s well treated. Capital is not just money. It’s also talent and ideas. They, too, will go where they’re welcome and stay where they are well treated.

—Walt Wriston1

From the end of World War II until 1970, owner-occupied housing was broadly affordable across the entire country. The standard measure for measuring affordability—the price-to-income ratio—was at about 2.8 in 1950, 2.5 in 1960, 2.6 in 1970, 3.4 in 1980, and 4.2 in 2020.2 This meant that, to a large extent, factors other than housing, such as climate, amenities, and job and economic opportunities, drove migration, which builders were in a position to respond to. However, as shown in Table 1, a number of metros on the coasts now have much higher ratios today, evidence that supply has not kept up with demand.

This ought not to be the case. The United States, unlike many other developed countries, has an abundance of habitable land—even in many coastal areas—and high levels of internal mobility. However, a variety of past and present policy mistakes—chief among them restrictive land-use policies—have driven up both the cost of land and new construction, as reflected in higher home prices. Fourteen of the 15 severely unaffordable major metros in the United States as defined by Demographia are located on the East and West Coasts, where land-use constraints are especially strong. The one exception is Denver, which implemented strong land-use constraints.

A robust literature on the relationship between land-use restrictions and housing supply provides evidence that local land-use policies limit housing

Table 1. 2020 Housing Affordability in Select Housing Markets

Source: Urban Reform Institute, Demographia United States Housing Affordability 2021 Edition: Data from 2020 3rd Quarter, 2021, https://urbanreforminstitute.org/wp-content/uploads/2021/11/Demographia-United-States-Housing-Affordability-2021-Edition.pdf

construction and, in turn, reduce housing affordability.3 Figure 1 depicts the relationship between median house prices across metropolitan areas and one measure of local regulatory constraints on housing construction in metro areas, the Wharton Residential Land Use Regulation Index. This measure of housing-supply constraints and prices has a positive correlation, with the San Francisco metro, an outlier on prices, having the highest score on the index.

Figure 1. Wharton Residential Land Use Regulatory Index and Metro-Area House Prices

Note: Only larger metros are labeled in this figure.
Source: Index values are from Joseph Gyourko, Jonathan Hartley, and Jacob Krimmel, “The Local Residential Land Use Regulatory Environment Across U.S. Housing Markets: Evidence from a New Wharton Index” (working paper, National Bureau of Economic Research, Cambridge, MA, December 2019). Median house-price data are from Zillow, “Zillow Home Value Index,” accessed February 24, 2020, https://www.zillow.com/research/data.

Consider this comparison. Across America, roughly 5.3 million hotel rooms—which like residential units are real property and need to be financed long-term—are available at an astonishingly varied set of price points based on age, location, and amenities.4 New rooms are easily built in most areas based on supply and demand. It is routine to rent hotel rooms for under $50 per night to over $200 per night in the same city, or even in the same location. Basically, anyone who wants a hotel room and can pay for it can rent one without subsidies. The hotel business is regulated, but no one prohibits customers with incomes above a certain level from renting lower-end hotel rooms, and there is no political pressure for subsidies for low-income hotel guests.

Contrast this with the housing market, in which supply is constrained by zoning that fails to recognize a location’s highest and best use, procedures and approvals that add time and expense, and urban-growth boundaries that limit available land and where demand pressure is provided by federal financing guarantees, subsidies, and accommodative monetary policies. The result is a government-run system that is unresponsive to market conditions and leads to housing unaffordability.

Land-use policies have demographic and economic consequences. Yet for many elected officials, pricing out lower- and middle-income Americans was seemingly not a major concern, as long as their cities could attract the “best and brightest” by agglomerating firms and people near one another.5 But this model has now come under threat due to the work from home (WFH) revolution, which is allowing talent and ideas to move to “where they’re welcome and stay where they are well treated.”

In the future, cities will increasingly have to compete with each other for human and financial capital. As people can increasingly locate where they prefer, cities and regions will be judged based on the services they provide and housing costs that result from their policy decisions. This healthy competition should hold policymakers more accountable and lower housing costs.

An increasing number of states and cities are already taking steps in this direction, and those that do may experience the benefits.

History of Zoning

Up until zoning became widespread in the 1920s, detached single-family houses were commonly intermingled with accessory dwelling units (ADUs), attached single-family houses, and duplexes, triplexes, and fourplexes.

The push for zoned single-family districts across the country that began in 1921 was the first in a series of expanding land-use constraints. These have limited both new housing supply and the conversion of existing supply to other more intensive uses. In 1916, New York City became the first US municipality to adopt a zoning ordinance. Commencing in 1921, the federal government, in an effort spearheaded by the US Department of Commerce, became the driving force behind the widespread adoption of zoning by municipalities and the near-universal move to adopt one-unit detached zoning districts.

The Commerce Department, with its Standard State Zoning Enabling Act, promoted the use of geographically separated zoning districts consisting of either one-unit, single-family detached houses or all other structure types (including two to four units or townhouses, which were common in single-family detached zones) as a tool to keep racial and ethnic groups separate. While the stated goals of zoning were written in high-minded prose advocating American values including thrift and independence, the true purposes were thinly veiled efforts to isolate immigrants from southern and central Europe and to keep African Americans segregated from whites who were welcome to higher-cost, one-unit, single-family detached housing.

After being established in 1934, the Federal Housing Administration (FHA) took over from the Commerce Department and went on to play a pivotal role in the use of zoning and housing finance to segregate residential development and neighborhoods. The lasting impact of the federal government’s role through the actions of the Commerce Department and FHA is clear: The vast majority of residential land in major American cities is zoned exclusively for single-unit homes.

The Legacy of the Planning Regime

The zoning laws implemented under the model of the Commerce Department’s Enabling Act limited property and development rights by creating single-family detached zones to the exclusion of multifamily and commercial development. Owners generally still retained the right to build what was legally permitted within a zone, called by-right approval. The push for single-family zoning promoted continuing residential racial segregation but did not create broad unaffordability until, starting in the 1950s, state and local policymakers increasingly adopted laws and regulations that added discretionary approval processes. These granted local and state government bodies a greater ability to stop, slow down, or demand expensive changes to development proposals.

Policymakers in San Francisco and other California localities were early adopters in the curtailment of by-right approval and the birth of discretionary review. In 1954, the San Francisco city attorney determined that “the city had ‘supreme control’ to issue building permits and could use its own discretion to decide whether projects were compliant.”6 Following this decision, city policymakers adopted a policy of holding discretionary-review hearings for many proposed developments, providing a platform for anyone with the resources to attend these hearings to delay or prevent building permits.

Shortly after the “not in my backyard” (NIMBY) movement took hold, the environmental movement imported urban-growth boundaries from the United Kingdom, where the Town and Country Planning Act 1947 had established development-prohibited greenbelts around large conglomerations.7 The goal of urban containment was to increase land costs with the aim of encouraging infill development and redevelopment inside the urban area.8 Ultimately, urban-growth boundaries (also known as “urban containment”) resulted in a constrained housing market, especially if combined with local zoning restrictions that prevent increases of supply within the urban core.9

Zoning and attendant land-use regulations had changed from a means to regulate development types and costs to a means to limit growth of any sort, with California leading the way by enacting growth-control regulations beginning in the 1960s.10 California’s enactments include the California Land Conservation Act of 1965, the California Environmental Quality Act of 1970, and the California Coastal Commission and California Coastal Act (1972 and 1976, respectively).11

The move to discretionary approval was not limited to California. Critics of suburbia after World War II played a key role in the evolution of zoning into a growth-control regime.12 Therefore, it is no coincidence that “virtually all of the markets with severely unaffordable housing have urban containment policies that severely restrict building on the periphery.”13

As seen in Figure 2, until the 1970s, affordability levels were in lockstep across virtually all metro areas. Starting in the 1970s, San Diego, San Francisco, Los Angeles, San Jose, and other California metros started to experience a sharp uptick in their price-to-income ratios, soon followed by other coastal metros such as New York City, Boston, Seattle, and Portland. Compounding these misguided policies was a belief by urban planners that they could ultimately deliver more livable communities.14 Later on, they created policies to subsidize the purchase of homes by poorer families. This ended badly during the financial crisis: These policies contributed to some 10 million or more foreclosures and distressed sales, which disproportionally affected low-income and minority neighborhoods.15

Figure 2. Historical Price-to-Income Ratios Grouped by Price-to-Income Ratio Today: 1950–2020

Note: “Severely unaffordable” is defined as having a median price-to-income ratio of 5.1 or higher.
Source: Urban Reform Institute, Demographia United States Housing Affordability 2021 Edition: Data from 2020 3rd Quarter, 2021, https://urbanreforminstitute.org/wp-content/uploads/2021/11/Demographia-United-States-Housing-Affordability-2021-Edition.pdf.

The Crux of Housing Unaffordability

Despite the federal government’s efforts to promote homeownership, the homeownership rate in the fourth quarter of 2020 was 65.8 percent, only marginally higher than the rate of 63.0 percent in the fourth quarter of 1964.16 Today, the twin legacy of racially motivated zoning combined with discretionary approval and poorly designed affordable-housing policies continue to make the housing market unaffordable for many lower-income and middle-class families.17

Even after the crisis and the recovery of the housing market, the price- to-income ratio grew and in many places affected the homeownership rate. In some metros, the implication is that many lower-income and middle- class families are being priced out of ownership entirely. (See Figure 3.)

Housing supply constraints and affordability problems are most severe in some of the country’s most productive regions. By limiting the number of people who can move into these regions, local restrictions on housing construction reduce wages, economic output, and innovation. (See Figure 4.)

Figure 3. 2020 Median Price-to-Median Household Income and Homeownership Rate: 40 Largest Core-Based Statistical Areas

Source: Data from Joint Center for Housing Studies, Census Bureau, and AEI Housing Center.

Figure 4. Distribution of Census Tracts by Income Level and Year (Largest  100 Metros)

Source: Data from Census Bureau, Federal Financial Institutions Examination Council, and AEI Housing Center

The WFH Great Rebalancing and Its Implications for Cities

Unaffordable metros and states all too often have become synonymous with anti-growth, high-tax, and high-regulation regimes; high crime rates; and low rates of homeownership. As a consequence, many have been losing population since before the pandemic. Nowhere is this as pronounced as in California, which has lost 3.5 million people based on cumulative net foreign and domestic migration as tracked by the IRS from 1990 to 2019.18 (See Figure 5.) This trend almost certainly accelerated in 2020 and 2021.

These same policies have also sparked a corporate hegira. Hewlett Packard, Oracle, and Tesla have recently decamped for Texas. When asked about the move, Elon Musk acknowledged some of the challenges of staying in the Bay Area: “It’s tough for people to afford houses, and a lot of people have to come in from far away.”19 As noted at the head of this chapter: “[Talent and ideas] will go where they’re welcome and stay where they are well treated.” The question will be how long and how much of a brain drain and taxpayer shift these areas can sustain.

In contrast, more-affordable states and metros have gained brain power and higher-income taxpayers. These areas experienced tailwinds from relatively lower home prices, tax policy, job growth, new construction, and attractive climates. States in mostly the South and Southwest have far outgrown others over the past three decades, with Florida, Nevada, Texas, and Utah consistently coming out ahead.

The WFH revolution, which has unshackled many higher-income and more-mobile workers, has turbocharged the more gradual process of reshuffling, thereby pushing back against the economics provided from agglomerating people and firms.

Just like online shopping, WFH is likely here to stay. Post-pandemic levels will be much higher than pre-pandemic and will likely increase over time. A recent study has estimated that perhaps 20–25 percent of the workforce will continue to WFH after the pandemic.20 That translates to an additional 20 to 25 million workers with the option to move. In May 2021, about a quarter of all media, communications, software, and IT services job postings listed remote work, up from about 5 percent a year earlier.21

Higher-income workers, who have greater opportunities to WFH, can profit from arbitrage opportunities offered by vastly different home prices

Figure 5. Cumulative Net International and Domestic Migration for California Based on IRS Data

Note: These data are based on net changes in personal exemptions claimed due to international and domestic migrations, which approximate the net number of individuals migrating to or from California. California in-state births and deaths are excluded by this calculation.
Source: Internal Revenue Service, “SOI Tax Stats—Migration Data,” Individual Master File, Statistics of Income, https://www.irs.gov/statistics/soi-tax-stats-migration-data.

across metros and regions. Nearly a quarter of US households live in metros with an average median home price–to–median household income ratio of 6.9, while the rest of the country has a ratio of 3.3.22

Consider the extent of the arbitrage opportunity for the median homeowner in San Jose. In April 2021, the median sale price of a San Jose home was $1,353,000, over 10 times the 2019 median household income. This compares to the median sale price of a Phoenix home of $388,000, or below six times the median income. If an owner sold in San Jose and purchased a home in Phoenix, he or she could save around $1 million.23

Those areas that offer lower prices and a competitive environment stand to benefit the most. A recent AEI survey of over 5,000 people found that individuals preferred smaller localities due to a desire for greater influence at the local level.24

While the WFH revolution will leave many higher-income workers better off, it also has its downsides. Combined with a large millennial generation entering its prime home-buying years, this additional demand could quickly overwhelm certain still relatively affordable markets, even

Table 2. States with Highest Growth and Negative Growth in 1990–2000, 2000–10, and 2010–20

Note: * State growth is measured by percentage growth over the decade; this excludes the District of Columbia.
Source: Data from US Census Bureau and Brookings Institution.

in places that have high new-construction shares. For example, prices for an identical home in Austin are appreciating at around 20 percent over the past year despite a new construction share of sales that has consistently hovered around 30 percent.25

The differential impact of house-price inflation on lower-income households is crucial. AEI Housing Center research has demonstrated how supply regulations and demand boosters from looser underwriting or lower interest rates have inflated the cost of lower-priced houses more than others. This research shows that rapid price increases crowd out low-income potential home buyers in housing markets.26 The longer this trend continues, the more lower-income but also middle-class families will be priced out of the market—similar to what happened in many severely unaffordable metros on the coasts.

So even considering the arbitrage opportunities noted above, what used to be pretty affordable metros before the pandemic and WFH are increasingly becoming less affordable—particularly for local low- or moderate-income households unable to benefit from such arbitrage. In other words, they are becoming, on a less dramatic scale, similar to certain California metros, which does not bode well since we have already described the negative outcomes in unaffordable metros: lower homeownership rate, slower economic growth, income segregation, and hollowing out of the middle class—or increasing wealth inequality.

The Future of Cities and How to Bring Back the Middle Class

To solve today’s affordability challenges, policymakers need to focus more on increasing supply and less on increasing demand. While it would be best to remove all distortions, including urban-growth boundaries, immediately, public policy is the art of the possible.

The low-hanging fruit lies in taking advantage of the growing momentum building around adding to supply by focusing on what we have dubbed light-touch density (LTD). LTD would allow for by-right housing including detached single-family houses with ADUs, small-lot single-family houses, attached single-family houses, and duplexes, triplexes, and fourplexes in areas currently zoned for one-unit homes. This has been enacted in such places as Minnesota, California,27 and Oregon. Many other jurisdictions are debating similar changes.28

However, the devil is in the details of these changes, and NIMBY opposition has been particularly adept at thwarting or slowing these enactments. For example, for California’s ADU law, which was first passed in 1982, it took the state legislature repeated attempts to overcome local laws to preempt ongoing local obstacles to ADUs.29 Only a 2016 law, passed 34 years after the state’s initial law failed to make ADU construction feasible, required localities to reduce many of the barriers standing in the way of ADU construction, including a mandate that localities permit ADUs by right rather than through a discretionary review process.30

The widespread adoption of zoning and other land-use restrictions across the country has corresponded with a declining share of LTD as a portion of the total housing stock. Thus, LTD is one solution to the housing shortage caused by onerous land-use restrictions. It allows incremental increases to density to help alleviate the supply shortage of newly constructed homes. These reforms have proven to be a path toward increased infill-housing construction in some locations. They represent a commonsense return to the housing typology that existed up to the 1940s, when diverse types of residential and commercial development were interspersed and the topology (unit types) and tenure (owner and renter) of the varied residential stock provided opportunities for people of different income levels to live in the same neighborhoods. This variability also promoted intergenerational family living.

LTD is not a total solution to supply shortfalls and unaffordability, but if LTD construction increased enough to return to its 1940 share of the one- to four-unit housing stock all else held constant, LTD construction could be expected to contribute eight million additional housing units to the total stock over the next 20 years.

Relative to one-unit zoning, LTD is a market-driven approach as it allows a broader base of landowners to naturally realize their land’s highest and best use. The highest and best use of a property is the reasonable and probable use that will support the highest present value, and it must be legally permissible, physically possible, financially feasible, and maximally productive.31 In parts of the country where demand for housing is high or increasing, allowing LTD will result in it being the highest and best use of land.

By-right LTD zoning would help reestablish the balance between the interests of homeowners who wish to limit change and exert control over neighboring properties versus current and future property owners. By widely increasing development rights across what might be tens of millions of properties, the impact on any one neighborhood will be reduced relative to reforms applied to small areas.

A further tweak that could garner even more traction is focusing LTD around areas of commercial activity, which we have dubbed walkable- oriented development (WOD). A focus on WOD areas would bring housing closer to jobs, particularly service jobs, thereby reducing transportation

Figure 6. Share of One-to-Four Housing Units by Type: 1940 vs. 2019

Note: 1940 is the earliest date for which there are consistent definitions for comparison purposes. Source: US Census Bureau, 1940 Decennial Census; and 2019 American Community Survey

and housing costs while freeing up time for other activities such as recreation and childcare. Nationally, the AEI Housing Center estimates that a focus on WOD could add two million private homes over 10 years to our housing stock while taking advantage of existing developed land and infrastructure at little incremental taxpayer cost.32

Moderately increasing density by right would also reverse the trend to either single-family detached and attached units or multifamily development of 50-plus units, which likely has contributed to growing income segregation. As seen in Figure 6, the share of attached single-family houses and two-, three-, and four-family homes out of the total housing stock has shrunk over time, down from 26.5 percent in 1940 (the earliest year for which data comparable to 2019 are available) to 18.3 percent in 2019.

The hollowing out of LTD—townhouses and two-, three-, and four-family homes—decreases opportunities for potential homeowners to start out on the lower rungs of the housing ladder. The reintroduction of these mixed-use neighborhoods would increase opportunities for lower-income and middle-class families to move into areas with more economic opportunity.

Encouraging case studies show that LTD and WOD can work. Palisades Park, located across the George Washington Bridge from Manhattan, first adopted a zoning ordinance in 1939, which implemented combined one- and two-family zoning districts for nearly the entire jurisdiction with no zones that exclusively permit single-unit detached homes. A direct result of this by-right zoning flexibility is that Palisades Park has added 51 percent to its housing stock since 1969, while neighboring Leonia, which is zoning almost exclusively for single-unit structures, has added only 24 percent over the same period.33

When supply-enhancing policies are combined with commonsense demand-side policies, they represent a path forward to build resilient communities around resilient borrowers, which encourages sustainable homeownerships for all across broad swaths of the country, including many of today’s cities that have essentially lost these groups.