For durable housing supply, broad upzoning wins; foreign-buyer taxes cut prices fastest but faintest

Asked (summary):

Which interventions have the strongest measured evidence for reducing housing prices or increasing housing supply — upzoning and zoning reform, new transit lines, foreign-buyer taxes, vacancy taxes, or demand-side subsidies? Cite specific studies with numbers. And which Canadian open datasets — house price indexes, building permits, census housing data, transit station opening dates, or tax records — could compare these interventions head to head?

The evidence set holds 30 study and policy-evaluation rows — six per intervention — spanning Auckland, Zurich, Vancouver, Toronto, France, Chile, England, Taipei and more, plus an inventory of 27 Canadian open datasets across six data categories. Effects below are each study's own estimate in its own units and horizon; they are not directly comparable.

What 30 evaluations measured, by intervention and outcome

causal / quasi-experimental model-based or structural program monitoring (no counterfactual) secondary source / non-causal

Each dot is one study's estimate for that outcome; a study appears in every outcome column it reports. Interventions are ranked top to bottom by strength of evidence for the asked goals. Hover or tap a dot for the study, design and numbers. Dots show direction and design quality, not comparable effect sizes.

What the dots cannot say alone

Auckland's citywide upzoning is the standout supply result: permits roughly doubled within five years — about 52,200 above the synthetic counterfactual after seven — and three-bedroom rents sat 26–33% below counterfactual after six years (tandfonline.com); a broad eight-metro US study found only a 0.8% supply gain in 3–9 years and no rent reduction, so design and bindingness matter (urban.org).

Foreign-buyer taxes show the sharpest short-run price estimates — about −5% in Vancouver over 15 months and −7% to −9% in Toronto over 8 months (econpapers.repec.org) and −6% in high-foreign-buyer BC neighbourhoods as the foreign share of detached sales fell from 13.2% to 1.7% (econpapers.repec.org) — but these are short-run gaps against counterfactuals, not lasting price-level cuts; the same authors' New Zealand estimate is −2.5% and insignificant (jonathanhartley.net).

Vacancy taxes move empty homes into use rather than lowering prices: France's matched difference-in-differences on exhaustive administrative records finds a 13% drop in vacancy over 1997–2001, mostly conversion to primary residences, with no detected construction effect (econpapers.repec.org); Vancouver's 54% fall in declared vacant homes is program monitoring without a counterfactual (progress.org).

Transit and demand subsidies mostly capitalize into prices: Vancouver's SkyTrain expansion raised the housing-cost differential by $5.07 a month in newly connected neighbourhoods and produced about $0.50 billion of appreciation (bankofcanada.ca), while Chile's 2011 homeownership subsidy raised prices 8.87% against an average supply elasticity of 0.543 (jifuenzalida.github.io). Taipei shows transit can add supply — 2.26% more units a year near stations — when land-use rules permit (cre.mit.edu).

Canadian open data: coverage and gaps for a head-to-head design

No single harmonized Canadian dataset supports a national head-to-head comparison. Transit opening dates are fragmented by agency; tax data are aggregate reports, not open property-level microdata; Ontario warns NRST revenue cannot proxy foreign transactions; Vancouver's Empty Homes Tax reporting is city-specific; CHSP excludes Quebec and carries suppressions; census outcomes are mostly five-year snapshots. A credible design needs municipality or neighbourhood panels, harmonized boundaries, coded policy dates and rates, and event-study or synthetic-control methods with many untreated places.

All 30 studies

Intervention	Study and place	Evidence	Measured effects	Main caveat	Source

Method: 30 study and policy-evaluation rows (six per intervention) and 27 Canadian open-dataset rows, compiled in a scan on 2026-10-09. The study rows span 22 independent hosts and 29 distinct URLs (no URL backs more than 6.7% of rows); the dataset rows span 7 hosts and 23 URLs (max 11.1% per URL), with Statistics Canada appropriately the dominant official publisher at 16 rows. Each effect is the study's own estimate in its own outcome, geography and horizon — effect sizes are not directly comparable. Long effect, caveat and coverage texts are shortened for space; full detail sits at each source link.

This report was generated automatically by Keenable SELECT at a user's request, from publicly available web sources linked herein. Keenable does not review, verify, or endorse its contents and makes no representation as to accuracy, completeness, or timeliness; AI-based extraction may contain errors. Nothing in this report is investment, legal, financial, or other professional advice. All trademarks and referenced content remain the property of their respective owners; no affiliation or endorsement is implied. To report an error, rights concern, or request removal: legal@keenable.ai.