We’ve Built 500,000 Apartments a Year Before. Here’s How.
America is short somewhere between 2 and 7.4 million homes, depending on how you count, and half of all renters now spend more than 30% of their income on housing. Yet a new report shows annual apartment construction has held near 350,000 homes for 40 years. That’s through strong job markets and recessions.
In Raising the Housing Investment Level: The History and Future of Multifamily Investment Policy, Center for Public Enterprise Executive Director Paul Williams argues that the absence of federal construction financing policy helps explain the stall. He points to two postwar periods when lending and tax programs pushed annual home starts well above 500,000. Those tools, he contends, remain available but largely dormant today.
Key Takeaways:
- Pipeline stagnation. An estimated 750,000 apartments have cleared zoning and land-use approvals but remain unbuilt. That’s six times the number of homes with permits but haven’t yet broken ground.
- Historical precedent. The last two times annual apartment starts exceeded 500,000 homes coincided with deliberate federal financing policy, and both ended when that policy was repealed.
- Financial bottlenecks. The Federal Housing Finance Agency’s (FHA) flagship construction loan program finances just 4% of new apartments despite offering better terms than private lenders, because its approval process takes three to five times longer.
Williams draws on U.S. Census Bureau housing start data from 1959 to the present, cross-references shortage estimates from Freddie Mac and the Urban Institute, and consults Federal Reserve lending surveys from 2023 and 2024 to document the current pullback. To estimate unbuilt projects — homes with legal permission to build, but no construction permits yet — the study draws on metro-level planning data from San Francisco, Boston, and New York, Yardi Matrix’s 4.27-million-unit private database, and a model that estimates how many projects are likely stuck in the entitlement stage. No federal agency tracks this stage, so the resulting figure should be understood as a reasonable estimate rather than a precise count.
This is what he found:
- Approved projects are sitting idle. An estimated 750,000 multifamily homes nationwide have cleared zoning and land-use approvals but haven’t yet applied for construction permits. That figure is about six times the 117,000 homes with permits but not yet under construction, and slightly larger than the entire 667,000-homes under-construction pipeline. San Francisco alone accounts for 52,000 such stalled homes, Boston reports 23,000 more, and New York City holds roughly 100,000 in its broader active development pipeline. Many of these projects have completed architectural drawings and engineering work. What they lack is financing on terms that make them viable.
- Federal policy built the last two booms. The 500,000-home annual threshold has been sustainably reached only twice in the postwar era — both times driven by specific federal financing programs, and both times lost when Congress reversed them. The first came after HUD introduced Section 236 interest-rate subsidies in 1968, which dropped effective mortgage rates to as low as 1% and spurred the construction of roughly 3 million apartments between 1970 and 1973. The second unraveled when the Tax Reform Act of 1986 repealed the depreciation rules fueling the early-1980s building boom. Multifamily starts collapsed from nearly 600,000 to under 140,000 by 1991. They have only briefly and temporarily exceeded 500,000 since. Williams argues the pattern is hard to ignore: sustained construction followed deliberate federal support, and dried up without it.
- The government’s own tools are broken. FHA’s 221(d)(4) program offers fixed-rate construction loans covering up to 90% of project costs with no personal liability if the project fails — what Williams calls the best construction financing product in the country on paper. Yet its approvals take 270–360 days, compared with 60–90 days for conventional lenders. The culprit is process: no codified single-underwriter model, inadequate IT systems, and layers of environmental and compliance paperwork. Banks have meanwhile tightened construction lending standards since 2022, pushing developers toward private debt funds charging rates several hundred basis points higher. The result, the report argues, is that developers are rationally avoiding the government program meant to help them.
Williams proposes five federal policy levers; the three requiring the least new spending are highlighted here. Codifying a single-underwriter model inside FHA’s loan guidelines — alongside environmental review and compliance reforms — could meaningfully reduce approval timelines from their current 270–360 days. A one-time $2 billion federal match for state Housing Finance Agency loan funds, which recycle capital back into new projects as loans are repaid, could support an estimated 30,000 additional apartments annually at no ongoing cost. Directing the Federal Housing Finance Agency to update its risk scorecards — while Congress separately clarifies Fannie Mae and Freddie Mac’s authority to offer construction-to-permanent loans — would unlock a significant new source of construction capital at limit cost to taxpayers
The housing shortage is serious but not intractable — Williams points to two postwar periods when federal financing policy drove sustained apartment construction well above today’s levels, and a substantial pipeline of approved projects is already waiting for the financing conditions to move them forward.