At the most fundamental level, housing prices — like prices for most goods — reflect the balance between supply and demand. When the number of households seeking housing exceeds the number of units available, prices rise. When supply expands faster than household formation, prices moderate. The United States has, for an extended period, been producing housing at a pace substantially below what household formation and demand replacement require. Understanding why supply falls short, and why it responds slowly even when prices send strong economic signals, is essential to understanding the persistent nature of the affordability crisis.
The Size of the Gap
Estimates of the housing deficit in the United States vary by methodology, but converge on a large number. Freddie Mac, using household formation data and housing unit production data, estimated in 2021 that the United States was short by approximately 3.8 million units relative to demand. The National Association of Realtors estimated a deficit of 5.5 million units using a different methodology. Up for Growth, a housing policy research organization, found a shortage of more than 3 million units based on vacancy rate and formation analysis.
The variation in these estimates reflects different assumptions about what vacancy rate constitutes a “healthy” market, how migration and demographic change affect demand projections, and how to account for housing quality and geographic distribution rather than only counting units. But the direction of the finding is consistent across methodologies: the United States has substantially underbuilt relative to demand, particularly over the decade following the 2008 financial crisis, when new construction fell sharply and did not recover to prior rates for many years.
Single-family construction fell precipitously after 2008. Annual single-family housing starts dropped from above 1.7 million in 2005 to below 500,000 in 2009, and did not return to the 1 million level until 2019. During that decade, millions of new households formed — young adults who aged into the housing market, immigrants who established households, existing households that split — without a corresponding supply response. Multifamily construction recovered more quickly, as apartment demand surged in the aftermath of the financial crisis, but multifamily starts never fully made up for the lost single-family production.
Why Housing Supply Responds Slowly
The supply of housing does not respond to price signals quickly, even when those signals are strong. This is a structural feature of the housing development process, not a temporary market failure, and understanding its mechanics clarifies why rents can rise significantly before new supply arrives to moderate them.
The pipeline from a developer’s initial decision to build to the delivery of completed units is long. It begins with land acquisition and site control, which may take months or years in competitive markets. It continues through entitlement — the process of obtaining the legal right to build what is proposed, which requires navigating zoning regulations, environmental review, public hearings, and permit approvals. In many jurisdictions, this process takes one to three years for straightforward projects and longer for complex ones. It then moves to financing — securing construction loans and equity commitments from investors — which in a high-interest-rate environment may require renegotiating the project’s economics. Construction itself takes 12 to 24 months for a typical apartment building, and longer for larger or more complex projects.
The total timeline from decision to delivery is commonly three to five years in well-functioning jurisdictions, and longer where regulatory barriers are higher. This means that when rents rise sharply in 2021, the supply response to those rising rents will not appear in the market until 2024 or later — and in the intervening period, households seeking housing compete for a stock that has not yet expanded. The lag is inherent to the production process.
Permitting Delays and Regulatory Barriers
Local permitting processes are a significant source of delay and added cost. In some jurisdictions, the path from building permit application to permit issuance is measured in weeks. In others — particularly in California and some other high-cost markets — it is measured in years. A 2022 study by the Terner Center for Housing Innovation at UC Berkeley found that permitting timelines in California cities were among the longest in the nation, with significant variation even across cities in the same county.
Delays arise from multiple sources: understaffed permitting offices that cannot process applications quickly; mandatory environmental review under laws like CEQA that require extensive documentation and create opportunities for litigation; public hearing requirements that give opponents access to delay mechanisms; and design review processes that require multiple rounds of revision before approval. Each month of delay adds costs — interest on land acquisition financing, staff time, consultant fees — that ultimately flow into the price of the finished unit.
Beyond outright delays, the regulatory environment shapes what kinds of projects get built. Minimum parking requirements, setback rules, height limits, and density caps constrain unit counts. Floor-area-ratio limits cap the total buildable space on a site. Historic preservation designations remove developable land from the supply. Impact fees — charges that cities levy on new development to fund infrastructure — have grown substantially in many jurisdictions and can add $30,000 to $80,000 per unit in some California cities, a cost that is passed through to buyers or renters.
The Construction Labor Shortage
The supply side of housing production faces a second significant constraint beyond regulation: a shortage of construction workers. The 2008 financial crisis devastated the construction industry. Employment in residential construction fell by approximately 40 percent between 2006 and 2010. Many of the workers who left — including a significant share of immigrant workers who returned to their home countries — did not return when demand recovered. The construction workforce aged, and the pipeline of new entrants narrowed as vocational training programs contracted and fewer young people pursued trades careers.
By the early 2020s, construction industry surveys consistently identified labor availability as among the top constraints on building activity. The National Association of Home Builders regularly reported that the majority of its members faced shortages of carpenters, electricians, plumbers, and other skilled tradespeople. The labor shortage pushed up wages in the construction sector — a development with obvious benefits for workers — but also added directly to construction costs.
The immigration policy environment has intersected with the labor shortage. Research estimates that immigrant workers account for a substantial share of the construction workforce — approximately 24 percent nationally, and higher in certain markets and trade categories. Policy changes and enforcement environments that affect immigrant workers have the potential to affect construction labor supply, though the precise magnitude of these effects is difficult to quantify.
What Underbuilding Looks Like Locally and Nationally
National statistics on housing production mask enormous geographic variation. Some markets have built aggressively in response to demand. Phoenix and Houston, with relatively permissive land use environments and large supplies of developable peripheral land, produced new housing at rates that moderated price growth relative to coastal peers. Austin built aggressively enough in the early 2020s that by 2024, rents had actually fallen from their 2022 peaks — a direct outcome of supply catching up with demand in a permissive regulatory environment.
In contrast, cities in California, the Northeast, and other highly regulated environments have chronically underbuilt. San Jose — the heart of Silicon Valley — permitted fewer than 3,000 new units per year in most years of the 2010s against a backdrop of massive employment growth and household formation demand. Greater Boston consistently permitted housing at a rate well below the level that the Metropolitan Area Planning Council projected was necessary to meet regional demand. Greater New York City similarly fell short of its housing production targets across multiple decades.
The concentration of well-paying jobs in cities that have restricted housing supply creates a compounding problem: workers who could be most productive in those cities cannot afford to live there, and either accept long commutes, take jobs elsewhere, or do not seek the opportunities available. Economists have estimated that housing supply restrictions in the most productive American cities impose significant costs on national economic output by preventing workers from accessing high-productivity locations — a finding documented in research by economists Chang-Tai Hsieh and Enrico Moretti, published in the Quarterly Journal of Economics, which estimated that zoning restrictions in just New York, San Jose, and San Francisco reduced aggregate U.S. GDP growth by more than two percent per year.
The supply shortage is not a problem that will resolve itself quickly even if regulatory barriers are reduced substantially. The pipeline of new housing takes years to fill once it begins. The construction workforce takes years to expand. The implications for affordability are that the present gap will likely persist, to varying degrees, for many years — making the combination of supply expansion and near-term demand-side assistance a more complete response than either approach alone.
This article was researched and drafted with AI assistance under human review. See our full AI and editorial practices.