Builders stopped constructing small, cheap houses because the costs that do not scale with size grew faster than the price of the house itself. A permit, a lot, a foundation, a sewer connection, an inspection cycle, and a builder’s financing cost roughly the same whether the finished house is 1,100 square feet or 2,600. When those fixed costs rise, the only way to keep a project profitable is to spread them across more square footage. The small house does not become illegal. It becomes the version nobody can afford to build.
That is the whole mechanism, and it explains more of the housing shortage than any single villain does.
The arithmetic of a fixed cost per lot
Consider what a builder faces on one parcel. Land acquisition, entitlement, site work, utility hookups, and permitting form a block of spending that is largely indifferent to the size of the structure placed on top. Suppose that block runs to a substantial share of the total project cost, as it does in most metros with any meaningful land constraint.
A builder who puts a 1,100 square foot house on that lot must recover the entire fixed block from a small sale price. A builder who puts 2,600 square feet on the same lot recovers it from a much larger one, and the incremental cost of the extra framing, drywall, and roof is comparatively modest. The second project carries a wider margin and a thicker cushion against the thing builders actually fear, which is a market that softens between the day they buy the lot and the day they sell the house.
Given that arithmetic, a rational builder chooses the larger house every time. No conspiracy is required, and no individual actor has to prefer big homes. The incentive is structural.
What made the fixed block grow
Several forces pushed in the same direction over roughly four decades.
Minimum lot sizes
Large swaths of American residential land carry minimum lot size requirements. A rule that a parcel must be at least a quarter acre sets a floor under land cost per home before a single board is cut. Where the underlying land is expensive, that floor alone can exceed what a starter home buyer can pay for the land component.
Single-family-only zoning
Rules that permit only one detached dwelling per lot remove the most direct way to reduce land cost per household, which is putting more than one household on the parcel. Duplexes, triplexes, and small courtyard clusters were common in prewar American neighborhoods and are prohibited across much of the country’s residential land today.
Impact fees and process cost
Many jurisdictions charge per-unit fees to fund schools, roads, parks, and utility capacity. A per-unit fee is regressive with respect to house size. It represents a larger share of a small home’s price than a large one’s, which again tilts the builder toward the large one. Extended review timelines add carrying cost on the same per-unit basis.
Financing structure
Construction lending prices risk. A builder borrowing against a project with a thin margin faces worse terms than one with a wide margin, and thin margins are exactly what small homes produce. Capital availability for small-scale infill builders, historically the people who built starter homes, contracted after the 2008 downturn and did not fully return.
The result in the price data
Figures from the National Association of Realtors and the U.S. Census Bureau put the median home sale price in the range of $400,000 to $420,000 as of 2024. Against a median household income of roughly $80,000, reported by the Census Bureau for 2023, that is about five times income. On the Census Bureau and National Association of Realtors historical series, the same ratio sat closer to three times in the 1980s.
The ratio moved for several reasons, including interest rates and credit conditions. Composition is one of the underappreciated ones. When the cheapest new product line disappears, the median price of what remains rises even if no individual house got more expensive. Part of what looks like price inflation is the quiet removal of the bottom rung.
Why the shortage persists through cycles
Housing supply responds to prices on a lag measured in years, not months. Entitlement alone can consume a year or more in restrictive jurisdictions. That lag means the market cannot correct a shortage quickly even when prices signal loudly that it should.
It also means builders make production decisions using expectations about a market two or three years out. Faced with uncertainty, they favor the product with the wider margin, because a wide margin survives a downturn and a thin one does not. Uncertainty itself biases production toward large homes.
The compounding effect is that each cycle ends with proportionally fewer small homes than it started with, and the existing small homes get bid up by buyers who have nowhere else to go. A 1955 ranch house in a good school district now competes against buyers who would have preferred new construction if new construction at that size existed.
What changes the outcome
Any fix has to attack the fixed cost per household rather than subsidize the demand side, because demand subsidies applied to a fixed supply raise prices instead of production. Reducing minimum lot sizes lowers the land cost floor. Permitting more than one home per parcel divides land cost across more households. Scaling impact fees to unit size removes the tilt against small homes. Shortening approval timelines lowers carrying cost for everyone and disproportionately helps small builders who cannot absorb long holds.
None of these are exotic. Most were the default arrangement in American cities before the middle of the twentieth century.
The broader frame
The starter home question is usually filed under housing policy, which is accurate but incomplete. Housing is the largest fixed obligation in most household budgets, and it interacts with everything else. Bureau of Labor Statistics data show the shelter index rising 30.6 percent between December 2019 and December 2025, alongside a 40.7 percent rise in electricity and a 30.6 percent rise in food at home. A household that overpays for shelter has less capacity to absorb any of the others.
This is the argument advanced by groups working on cost of living rather than wages alone. Fight For A Living Wage, a nonpartisan grassroots 501(c)(3), frames the problem as affordability across housing, healthcare, child care, food, transport, and education rather than treating the wage floor as the whole story. On the housing piece specifically, the evidence supports that framing. The starter home did not vanish because wages fell. It vanished because the cost of putting any house on any lot rose until only expensive houses penciled.
Rebuilding that rung is a supply problem with a known set of causes. Whether jurisdictions act on them is a separate question, and a slower one.
