Housing: supply vs subsidy

What the evidence says

On the supply side, land-use regulation is the most contested policy lever in these sources. Several studies associate the withdrawal of buildable land — caps on development, restrictive zoning, urban growth boundaries, permit-processing delays, and coastal development restrictions — with higher housing prices: towns where more than 20% of vacant land was regulated had prices about 6% higher, and California's coastal-zone development restrictions were found to have forced prices higher. But a comparable body of work finds no measured price or growth effect from the same category of regulation: a review spanning many land-use and growth-control studies found little or no price effect, a survey of 443 California jurisdictions found little association between growth controls and actual local growth, and a direct comparison of controlled versus uncontrolled cities found that only 3 of 7 controlled cities grew more slowly, without appreciably higher prices. The literature reviewed here explicitly declines to collapse this into one causal estimate, citing variation in both the regulation studied and the precision of the methods used to measure it.

Separate from land-use rules, one source frames housing-supply elasticity itself as a determinant of price behavior, predicting that countries with less elastic housing supply should experience greater house price volatility — stated as a theoretical expectation rather than a directly measured cross-country result in this source.

Differences in mortgage-finance institutions shape how housing demand converts into purchases, independent of construction supply. Loan-to-value ratios available to first-time buyers range from about 40% in Italy to 95% in the UK, and mortgage debt as a share of GDP ranges from 6% in Greece and 7% in Italy up to 65% in Denmark, variation attributed to differences in legal foreclosure speed, seller taxes, and mortgage-market institutions. Deregulated systems such as those in the UK, Spain, Finland, and Sweden ration credit through interest rates rather than direct quantity limits, while systems that preserve a regulatory monopoly on mortgage-bond issuance decrease competitive pressure and consolidate the market. Residential investment itself represents around one-fifth of gross domestic capital formation.

On the subsidy side, the clearest and most consistent effects in these sources are on homelessness recurrence rather than on prices. Among people who had already exited homelessness, holding a permanent rental subsidy — a Housing Choice Voucher, public housing, or permanent supportive housing (PSH) — decreased the hazard of returning to homelessness by 57% (adjusted hazard ratio 0.43, 95% CI 0.29–0.64). Housing-loss incidence followed the same pattern: participants in PSH had a housing-loss rate of 0.12 events per 100 person-years, and those in independent apartments or rooms 0.17, against an overall rate of 24.5 per 100 person-years across the full sample. Separately, a meta-analysis of randomized controlled trials found Housing First programs increase days in stable housing and decrease emergency-department visits and hospitalizations, without a measured worsening of substance use disorders.

These subsidy effects are large for the people who receive them, but coverage is limited, and that gap sits next to a documented link between market rent and homelessness. Only one in three eligible adults aged 65 and older receive a Housing Choice Voucher. Consistent with that gap, most people who exit homelessness do so without a permanent subsidy: of 327 people, 70% (n=226) moved into unsubsidized rented or owned housing or in with family or friends, versus 21% who received a permanent subsidy. On the price side, a $100 increase in median rent was associated with a 9% increase in the estimated homelessness rate. HUD-VASH, which bundles subsidized permanent housing with supportive services for veterans, is described as having contributed substantively to a 50% reduction in veteran homelessness over the past decade, though the source flags that this attribution has not been isolated from other concurrent policy and economic changes.

Where the evidence splits

Studies disagree on whether land-use and growth-control regulation raises housing prices or slows growth. Towns with more than 20% of vacant land regulated showed prices about 6% higher, and California coastal-zone development restrictions were found to have forced prices higher, and a general claim about withdrawing buildable land also links regulation to higher prices. Against this, a broad review of land-use and growth-control studies found little or no price effect, a 443-jurisdiction California survey found little association between growth controls and actual growth, and a controlled-versus-uncontrolled city comparison found only 3 of 7 controlled cities grew more slowly with no appreciable price difference. The sources themselves offer an explanation for part of this split: the null-result review notes that local regulation is "sometimes... symbolic, ineffectual, or only weakly enforced," and the general claim adds that "variations in both observed regulation and methodological precision frustrate sweeping generalizations" — meaning the disagreement is attributed to how regulation is measured and enforced, not fully reconciled to a single estimate.

Thresholds and dose-response

Towns in which more than 20% of vacant land was regulated had housing prices about 6% higher than towns below that threshold. Separately, a $100 increase in median rent was associated with a 9% increase in the estimated homelessness rate in high-cost housing markets.

Open empirical questions

• What is the causal magnitude of land-use regulation on prices once regulation type and enforcement intensity are controlled for — the sources here state this is not established?
• Does supply elasticity's predicted link to price volatility hold up under direct empirical cross-country testing, rather than the theoretical framing given here?
• What is the marginal effect of expanding voucher/subsidy coverage rate (beyond the current one-in-three eligibility rate) on homelessness rates, as distinct from the effect of already holding a subsidy?
• How much of the 50% reduction in veteran homelessness is attributable to HUD-VASH specifically versus other concurrent policy or economic factors?