Experts indicate that the United States requires an additional 2 million to 20 million new housing units to bridge the deficit, highlighting the formidable challenge of meeting the nation's housing needs. January 26, 2026, Adams Morgan neighborhood, Washington, D.C. Quantifying America's housing demand hinges on a series of assumptions regarding housing costs, population numbers, and living space. The United States is grappling with a severe housing shortage. Moody's estimates that resolving this issue would necessitate the construction of over 2 million additional homes. However, analysts at Goldman Sachs peg the deficit at 3 million units; Zillow's estimate exceeds 4 million; the Brookings Institution forecasts 5 million; while McKinsey claims the shortfall reaches 8 million units. Concurrently, Congressional Republicans insist the gap is closer to 20 million units. Some economists even argue that there is no housing shortage in the United States at all. These vastly divergent projections reflect the immense challenge of quantifying U.S. housing demand—a puzzle centered on a series of assumptions about housing affordability, household size, and living space. Housing affordability is a critical political issue and is becoming increasingly out of reach for many Americans. Consequently, determining national housing needs is not merely an academic exercise but is crucial for formulating policies to address the problem. Vacancy Rates and "Missing Households" U.S. Census Bureau data shows there are 146 million housing units nationwide. Among these, 8.1 million are categorized as "shared households," where people live with non-relatives. Zillow's housing gap estimate assumes these individuals would prefer to have their own homes. The real estate website also states there are 3.4 million vacant homes available for rent or sale across the country. Therefore, Zillow's economists subtract the number of available homes from the number of shared households, concluding that the U.S. needs an extra 4.7 million new housing units. Multiple analyses focus on two core questions: What constitutes a reasonable housing vacancy rate? And how many consumers are delaying independent living due to cost pressures? Counterintuitively, a healthy real estate market requires a certain level of vacancy. For instance, vacant properties might be between tenants/buyers, undergoing renovation, or indicate owners rotating between multiple residences. According to the National Association of Home Builders (NAHB), there are over 6 million housing units nationwide (roughly 1 in 20) that serve as secondary residences. Experts disagree on the definition of a healthy vacancy rate, with estimates ranging from 3% to 13%. Following the 2008 housing crisis, which caused a collapse in construction, vacancy rates plummeted to near 20-year lows: the homeowner vacancy rate fell below 1%, and the rental vacancy rate was just 5%, levels from which the market has not fully recovered.

The ideal number of homes might seem simply definable as: one home per household + a certain percentage of vacant units. The problem, however, is that we cannot accurately count the number of households. When housing costs become prohibitively high, adult children often live with their parents for longer periods. A Pew Research Center report indicates that in 2023, 18% of adults aged 25 to 34 lived with their parents, compared to just 8% in the 1970s. For many economists, the formula for the housing gap should be: current number of households + a reasonable number of vacant homes + the number of new households that would naturally form if housing inventory were sufficient and prices decreased. Yet, even using the same framework, different researchers arrive at starkly different conclusions about the housing shortfall. Moody's Analytics and PolicyMap suggest that 800,000 housing units are needed to restore the U.S. housing market to the equilibrium seen between 1985 and 2000. Adding 1.2 million units for "pent-up household formation" (households that haven't yet formed) leads to the conclusion that the U.S. needs an additional 2 million homes. The Brookings Institution's calculation aims to restore the vacancy rate to over 12% seen in 2006 (near its historical peak). Using a complex statistical model, the institution parsed out how much of the decline in household formation rates since then was due to home prices, versus other factors like difficulty finding jobs for young people or later marriage. They ultimately concluded that the U.S. needs an extra 4.9 million housing units. Similar analyses include a Freddie Mac report, which estimated a gap of 3.7 million units. Goldman Sachs analysts employed a "vacancy rate + pent-up demand" approach, combined with a mathematical model calculating the number of homes needed to restore housing affordability relative to income to 1990s levels. Both methods yielded results between 3 and 4 million units. McKinsey added together new household formation, vacant housing needs, housing to address homelessness, and the need to replace overcrowded housing (multiple people per bedroom), arriving at a deficit of 8.2 million units. Envisioning a Market Free from Regulatory Constraints A 2022 Congressional report took a fundamentally different approach. Most analyses attempt to replicate housing market conditions from two or three decades ago. But Republicans on the Joint Economic Committee argued that the correct deficit figure should equal the number of homes developers would build in a market free from regulatory constraints—that is, without permits or zoning rules that prohibit constructing housing the market demands. The Republican estimate is based on the logic that land value should constitute 20% of housing costs. A higher proportion suggests the market is artificially constrained—when land is harder to develop, its price rises. To align home prices in all U.S. counties with this ratio, they concluded the housing gap is a massive 20 million units. According to their calculations, North Dakota and West Virginia have almost no housing shortage, while California is short 4.5 million units. While comprehensively eliminating zoning and building restrictions across hundreds of U.S. jurisdictions may be impractical, they predict that any substantial relaxation would lower prices. For example, they believe building an additional 2.7 million homes would be sufficient to significantly reduce prices, making homeownership economically feasible for nearly 5 million more consumers. Single-family homes under construction in Menifee, California, 2024. "If we relaxed all housing supply-related regulations across all U.S. markets, this is the number of homes we would get... I believe this is the correct way to think about how many homes America should have," said Kevin Corinth, who helped author the report as Senate staff and now works at the libertarian think tank Alset EHome International Inc. (AEI). "If you truly want to bring prices down to levels affordable for the average person, you must build far more homes than current estimates suggest." Per Capita Housing Construction Spending Housing analyst Kevin Erdmann recently made a startling calculation. After adjusting for inflation, spending on housing construction as a share of personal consumption expenditures has been steadily declining, falling 23% since 1990. He stated that if this share had remained at its 1990 level, the U.S. would have an additional 40 million homes. "Virtually all professional estimates of the housing shortage are wildly too low," wrote Erdmann, author of two books on housing markets, on his Substack platform. He argues that the slowdown in housing construction spending means people are forced to live in homes smaller than they need. However, he does not explicitly state that the U.S. lacks 40 million homes. Instead, based on aggressive assumptions about missing households and necessary vacancy rates, he suggests the U.S. needs 15 to 20 million additional units. Perhaps There Is No Overall Shortfall At All Urban planning professors Kirk McClure and Alex Schwartz studied 900 U.S. metropolitan areas and found that since 2000, only 19 regions saw population growth outpace housing construction. They argue that developers built an excess of homes before the 2008 recession, creating a buffer that accommodated the subsequent years of underbuilding. "It's true, there is a shortage of housing at the low-income end, but there is no overall shortage," McClure said. He advocates that government assistance to help low-income families rent or purchase existing housing is far less costly than building new units. "The best housing policy right now is raising the minimum wage. Increase the hourly wage to $20, and people's lives improve immediately—we cannot solve this problem simply by building homes." This perspective transcends partisan lines, with both the highest and lowest estimates of the housing gap coming from right-leaning camps. Economists at the libertarian Cato Institute argue that housing construction has kept pace with population growth. They claim people simply want to live in large homes in expensive, densely populated areas, which does not equate to a housing shortage. "A shortage means people have nowhere to live. That is not our current situation," said Norbert Michel, a writer at the Cato Institute, in an interview. Ultimately, the controversy stems not only from the choice of mathematical models but also from differing interpretations of what constitutes a housing shortage itself. "If I have difficulty finding my ideal apartment in Washington, D.C., I can still move to Maryland," Michel said. "A situation where I have absolutely no choice, no place to live at all, constitutes a true shortage. And the data does not support that." Erdmann holds a different view: "There are 28-year-olds living with their parents who could live independently if suitable housing were available. If that doesn't qualify as a shortage, I don't know what does."