Analysis
What did tripling Vancouver's empty homes tax do? Not much.
Vancouver began charging owners to leave a home empty in 2017. In the first year, about 3,576 existing condominiums entered the rental pool beyond what new construction explains. The charge later rose to three times its original rate, and nothing comparable happened again.
The short version
- Vancouver counts its empty homes, which almost nowhere does. Owners declare every year and the city audits a sample.
- The count has fallen every year since the tax began, from 1.18% in 2017 to 0.49% in 2024, and not because properties moved into exemptions. Exempt properties fell too.
- In the tax's first year, rented condominiums rose from 28,144 to 34,064 while the surveyed stock grew 2,344, leaving about 3,576 existing homes that started being rented.
- No later year repeats it. In three of the next four periods, new construction outpaced the growth in renting entirely.
- The rate then tripled. Declared vacancy fell 11.6% a year under a 1% tax and 11.5% a year under a 3% tax.
- Four instruments arrived within about a year of each other. On the one that can be partly tested, Vancouver looks like its neighbours.
Companion video
What did tripling Vancouver's empty homes tax do? Not much.
Publishing shortly. The analysis below stands on its own in the meantime.
Recap: what Vancouver counts
If you have watched the companion video, this is a refresher and you can move to the findings. If you have arrived here cold, this is what you need before the analysis begins.
Most cities argue about empty homes without knowing how many they have. The argument runs on inference: a dark window, a building that never seems to fill, a name on a title search.
Vancouver has a number. Since 2017 the owner of every residential property in the city has had to file an annual declaration saying whether it was occupied, and the ones that were not have been charged a percentage of assessed value. That produces something unusual, which is a yearly count of empty homes established by law rather than estimated from a proxy, running across two changes in the amount charged.
So a narrower question becomes answerable than the one usually asked. Not whether homes are being left empty, but what happens to a home when leaving it empty starts to cost money.
Consider a toll appearing on a road that was free. Traffic reroutes, and it reroutes in the first week. Whether the toll later doubles matters less than whether it exists, because everyone who was going to change route already has. The question is whether housing behaves that way.
1. What the tax charges, and what it never touches
A count is only as wide as the definition behind it, so the definition comes first.
Vancouver's Empty Homes Tax took effect for the 2017 vacancy reference period. Owners declare, the city audits a sample, and a property declared or found vacant is charged a share of its assessed value:
- 2017 to 2019: 1%
- 2020: 1.25%
- 2021 to 2025: 3%
That is not the whole charge. British Columbia's Speculation and Vacancy Tax applies to the same properties. For the 2025 tax year it charged 0.5% for a Canadian citizen or permanent resident and 2% for a foreign owner or untaxed worldwide earner. Neither government credits the other: the by-law makes no reference to the province, to credits, to offsets or to deduction, and the province describes its tax as different from the municipal one.
So for the 2025 reference year an empty Vancouver home carried 3.5% of assessed value a year where the owner is a Canadian citizen or permanent resident, and 5% where the owner is a foreign owner or untaxed worldwide earner.
Both figures have since been overtaken on the provincial side. The Speculation and Vacancy Tax rose in 2026 to 1% for a citizen or permanent resident and 3% for a foreign owner, and rises again on 1 January 2027 to 1% and 4%. As of August 2026 the city has not published an empty-homes rate for 2026, so the 2026 combined figure cannot be stated without inventing half of it. The rate is a dial, and it is being turned again.
Against that sit eleven categories of exemption, and two of them carry most of the weight. Of 3,718 exemptions granted in the 2024 reference year, redevelopment or renovation accounts for 35% and property transfer for 41% (City of Vancouver 2025). Those two take 76% between them. The remaining nine share at most about 892 properties.
One of those nine is worth reading closely, because its wording is unusually open.
The developer clause
The by-law exempts a newly built home that has never been occupied, provided it is owned by a developer and is "actively offered to the public for sale throughout the vacancy reference period" (Vacancy Tax By-law No. 11674). Three features of that clause are visible in the text:
- No duration cap. Two other exemptions in the same by-law, for property under redevelopment and for an owner in care, are capped at two consecutive vacancy reference periods. This one is not.
- No price test. The phrase "actively offered to the public for sale" appears once in the by-law's twenty-three pages and is never defined. Nothing requires the asking price to be reasonable, or a sale to be sought within any period.
- No anti-avoidance provision. The by-law carries no general clause directed at arrangements whose main purpose is obtaining an exemption.
The city's public summary of this exemption says it applies to a developer holding five or more residential properties. The by-law says two or more, and the by-law governs by its own terms.
The city does not publish how many properties claim it. The province publishes the equivalent figure for its own tax, which carries a similarly worded exemption for unsold new inventory: in Vancouver, 964 such exemptions were claimed in the 2024 tax year, against 1,011 in Burnaby and 660 in Surrey (Province of British Columbia 2025).
Two gaps in the by-law reading
- The 964 is a provincial figure, not a municipal one. The two exemptions are written by different governments in different statutes and need not cover the same properties. The number bounds the scale of developer-held new inventory in Vancouver; it does not measure the municipal clause.
- Exemption counts are published as shares, not counts. The 892 residual is a ceiling across nine categories rather than a figure for any one of them.
2. The count fell, and it was not reclassification
With the boundaries of the tax established, the series inside them can be read.
The declared vacancy rate has fallen in every year of the tax, from 1.18% for the 2017 reference period to 0.49% for 2024. In 2024 that is 979 properties out of 201,739 required to declare (City of Vancouver 2025).
Figure 1. Declared vacancy rate by reference period, against the empty homes tax rate in force. Source: City of Vancouver 2025.
A falling count invites an obvious objection. Properties might not have become occupied; they might have moved into an exemption, in which case nothing changed except paperwork. The city publishes the exempt count as well, which lets the objection be tested.
Table 1. Vacant and exempt properties, Vancouver, by vacancy reference period. Sources: City of Vancouver 2024; City of Vancouver 2025. The two editions publish identical values for the years they share.
| Reference period | Vacant | Exempt | Combined |
|---|---|---|---|
| 2021 | 1,398 | 4,440 | 5,838 |
| 2022 | 1,156 | 3,756 | 4,912 |
| 2023 | 1,073 | 3,930 | 5,003 |
| 2024 | 979 | 3,718 | 4,697 |
Both columns fell. Between 2021 and 2024 vacant properties fell 30% and exempt properties 16.3%, so the combined total fell 19.5% against a denominator that rises every year as more properties come into scope.
What the declaration counts, and how well
- It is self-reported, and audited on a sample rather than in full.
- Audit non-compliance has itself been falling, from 4.8% in 2020 to 3.0% in 2024, so part of any improvement in the series is improvement in the measurement. Audits for the 2024 reference year can be started until the end of 2026, so both the 2024 count and that 3.0% are still open to revision.
- The rate series is published as a chart, not a table. Values were recovered by mapping each plotted point back to its year, then checked two ways: the rate ladder reproduces the one stated in an earlier report's body text, and 195,012 properties at the published 0.90% rate for 2020 gives 1,755, the published count for that year.
3. The first year
A falling count says the number of empty homes went down. It does not say those homes went anywhere in particular. For that, a second source is needed, and the tax's own reports cannot supply it.
Canada Mortgage and Housing Corporation surveys Vancouver's condominium apartments each October and reports two counts: how many are being rented, and how large the surveyed stock is in total (CMHC 2019).
In that year the rented count rose by 5,920. Over the same twelve months the surveyed stock grew by 2,344, as new buildings finished and entered the count for the first time. At most 2,344 of the additional rentals can therefore be newly built.
The other 3,576 were already standing and already counted. They were not being rented before, and then they were.
What the residual does not cover
- Condominium apartments only. Purpose-built rental buildings, laneway houses, secondary suites and rented houses are outside this survey.
- It is a subtraction, not a count. Nobody counted 3,576 homes changing use. The figure is what remains after removing the growth new construction can account for, and it inherits the sampling error of both columns.
- It is not comparable to the tax's own counts. The city's reports footnote that Empty Homes Tax counts and this rental universe are different populations, collected differently and at different moments. They are never plotted together here.
4. Every year since
One large move in one year is consistent with a response to the tax. It is also consistent with several other things, and the way to tell them apart is to keep watching.
Table 2. Vancouver condominium apartments: rented homes, growth in the surveyed stock, and the residual between them. Sources: CMHC 2019; CMHC 2022. October 2020 and 2023 levels are carried forward from published changes.
| Oct to Oct | Rented homes | Change | Stock grew by | Beyond new supply | Tax rate |
|---|---|---|---|---|---|
| 2018 to 19 | 34,064 | +5,920 | +2,344 | +3,576 | 1% |
| 2019 to 20 | 36,519 | +2,455 | +3,497 | −1,042 | 1% |
| 2020 to 21 | 37,134 | +615 | +2,849 | −2,234 | 1.25 to 3% |
| 2021 to 22 | 38,765 | +1,631 | +4,054 | −2,423 | 3% |
| 2022 to 23 | 40,217 | +1,452 | +952 | +500 | 3% |
Rented homes rise every single year, which is the column most people expect to be the story and is not. What changes is how much of that rise the surveyed stock already accounts for.
Figure 2. Rental growth beyond what new supply explains, Vancouver condominium apartments. Positive means existing homes entered the rental pool faster than the surveyed stock grew. Sources: CMHC 2019; CMHC 2022.
In three of the five periods the final column is negative. That does not mean homes left the rental pool. It means new construction outpaced the growth in renting, so no existing home needs to have changed use to explain the year. Only the first period shows a large positive residual.
What the series cannot settle
- Stock exhaustion was tested and does not fit. If the homes easiest to convert went first, the proportional decline should get shallower as the remaining stock gets harder to move. Across the seven year-on-year changes it does not: the three earliest average 8.5% a year and the four latest average 13.9%. That is weak evidence against selection dominating, on seven noisy observations.
- The pandemic sits inside this window and moved rental demand for reasons unrelated to a vacancy tax.
- Two levels are chained. October 2018, 2019, 2021 and 2022 are read directly. October 2020 and 2023 are carried forward from published changes, and CMHC restated October 2019 by 22 units between editions, so the levels are precise to about that order.
5. Then they tripled it
So the movement is concentrated in a single year at the start. That leaves the question the series exists to answer, because the charge did not stay where it started.
The 2021 reference period raised it from 1.25% to 3%. If holders respond to how expensive it is to leave a home empty, this is where the response should be largest.
It is where the response is smallest. The 2020 to 2021 period produced the smallest rise in rented homes in the series, 615, and a residual of −2,234 once the stock's own growth is removed.
The rate of decline tells the same story.
Figure 3. Average annual proportional decline in the declared vacancy rate, within each period of constant tax rate. Source: City of Vancouver 2025.
Declared vacancy fell 11.6% a year while the tax stood at 1%, and 11.5% a year while it stood at 3%. Those are the same number.
How the comparison is built, and what it leaves out
The measure is the year-over-year percentage change in the declared vacancy rate, averaged within each period of constant tax rate. Percentage rather than percentage-point, because a percentage-point drop from a high base is not comparable to one from a low base.
Two years are excluded, and they are the two extremes of the series. A year in which the rate moved cannot measure behaviour under a steady rate.
Table 3. Every year of the declared vacancy series, showing which enter the steady-rate comparison. Source: City of Vancouver 2025.
| Year | Rate moved | Proportional decline | In the comparison |
|---|---|---|---|
| 2018 | no | −8.5% | yes, at 1% |
| 2019 | no | −14.8% | yes, at 1% |
| 2020 | 1% to 1.25% | −2.2% | no |
| 2021 | 1.25% to 3% | −21.1% | no |
| 2022 | no | −18.3% | yes, at 3% |
| 2023 | no | −6.9% | yes, at 3% |
| 2024 | no | −9.3% | yes, at 3% |
What the comparison cannot show
- The steepest single year in the series is the year of the increase, at 21.1%. Taken alone it would suggest the opposite conclusion.
- The following year, under an unchanged rate, is nearly as steep at 18.3%. A move that repeats when nothing changes cannot be attributed to the change, which is why the proportional comparison carries the weight and the single year does not.
- Two periods, no control group. Five observations either side of one step is suggestive, not decisive.
- A third tax covers the whole of the 3% arm and none of the 1% arm. The federal Underused Housing Tax, an annual 1% charge on vacant or underused housing, took effect on 1 January 2022 and was removed for 2025 onward. Its active years are 2022, 2023 and 2024, which are exactly the three years averaged on the right of this comparison. Whichever way that pushes, the two arms are not measuring the same tax environment.
6. Three measures, one shape
Those are four separate observations: a falling count, one large conversion, four quiet years, and a rate change that did not register. Together they describe one pattern, and the pattern is easier to argue with than any of the pieces.
Table 4. Three measures of the response, compared across the tax appearing and the rate rising. Sources: City of Vancouver 2025; CMHC 2019; CMHC 2022.
| Measure | When the tax arrived | When the rate tripled |
|---|---|---|
| Existing homes entering the rental pool | +3,576 in one year | −2,234 |
| Rise in rented homes | +5,920 | +615, the smallest in the series |
| Annual decline in declared vacancy | 11.6% at 1% | 11.5% at 3% |
The observable response clusters at the moment a cost appeared, and does not track how large that cost became.
What this cannot establish
An owner facing a new annual charge has four options: rent it, sell it, occupy it, or pay. Whichever is chosen is chosen at the point the arithmetic changes, and owners who choose to pay have by construction decided the charge is smaller than their reason for holding. Tripling a number that was already outweighed need not reach them.
That account fits the three measures. So do others. No data here observes an owner deciding anything; what is observed is a count of homes and a count of rentals.
7. What else could have done this
Which leaves the largest question in the piece, and the one the evidence answers least well.
Four instruments arrived within about a year of each other:
- Vancouver's Empty Homes Tax, from the 2017 vacancy reference period
- The provincial Speculation and Vacancy Tax, from 2018
- The federal B-20 mortgage stress test, tightening qualification from 2018
- A widened and raised foreign buyers tax
A fifth arrived later. The federal Underused Housing Tax charged 1% a year on vacant or underused housing from 1 January 2022, and was removed for 2025 onward. It does not bear on the first-year move, which precedes it by four years, and it covers the whole of the period used to measure behaviour under a 3% municipal rate.
Any of them could produce a one-time move in the same direction at the same time.
One can be partly tested. Vancouver carries two vacancy taxes, its own and the province's. Every other municipality in the provincial tax area carries only the province's. If the municipal tax does work of its own, Vancouver should look different on the provincial measure.
Table 5. Residential properties not exempt from the provincial Speculation and Vacancy Tax, 2024 tax year. Vancouver is the only municipality here that also levies a municipal empty homes tax. Source: Province of British Columbia 2025.
| Municipality | Non-exempt properties | Total | Share |
|---|---|---|---|
| Surrey | 482 | 159,661 | 0.30% |
| Vancouver | 628 | 200,360 | 0.31% |
| Abbotsford | 144 | 44,995 | 0.32% |
| Burnaby | 341 | 89,615 | 0.38% |
| Richmond | 585 | 79,208 | 0.74% |
| Kelowna | 561 | 53,001 | 1.06% |
Vancouver sits between Surrey and Abbotsford, and below three of the six.
Why this is not the experiment
- There is no baseline. The province publishes per-municipality counts one year at a time, with no municipal figures from before 2018. A single cross-section cannot separate the municipal tax adding nothing from Vancouver having started higher and converged.
- The two instruments measure different things. Non-exempt under the provincial tax includes foreign owners and untaxed worldwide earners who may be living in the home. Declared vacancy under the municipal tax is about occupancy alone. No figure in the table above belongs beside the 0.49%.
- Four municipalities are missing. The province suppresses cells covering fewer than five entities.
What survives is narrow. Whatever moved these homes moved them once, and making the charge three times larger did not move them again.
8. What to do with this
Nothing here lowers the price of a home.
What it offers is a distinction worth carrying into the next argument you hear. "Did it work" and "would more of it work better" are different questions with different answers, and the second is the one most debates about this tax are actually having. Vancouver ran that experiment without meaning to: the charge tripled, and the series did not bend. Anyone arguing from either direction now has a number to argue with.
It is also a reason to be careful with a chart that starts at an intervention. Declared vacancy has fallen every year since 2017, which looks decisive until you notice that three other instruments moved at the same moment and that the pool of homes able to respond gets smaller every year. Easy to show and hard to attribute are not the same property, and most charts you will be shown are the first kind.
And it is a smaller scoreboard than the argument assumes. Vancouver counts its empty homes, which is rare and worth valuing, and in 2024 the count was roughly one property in two hundred. That number is worth knowing precisely because it is small.
Which returns to the toll. Traffic rerouted in the first week, and the price after that mattered less than anyone expected. The useful question was never how much the toll should cost. It was how much of the traffic was ever going to take that road.
The next edition asks what else was tried, and what followed.
Appendix A. Methods and scope
Empty Homes Tax data. City of Vancouver Empty Homes Tax annual reports, 2024 and 2025 editions. Where the two overlap, on the 2022 and 2023 reference periods, their published values are identical.
Dates. A vacancy reference period is defined in the by-law as the calendar year before the tax year it applies to. Comparisons between a by-law date and an annual report date carry that one-year offset.
Rental data. Canada Mortgage and Housing Corporation rental market survey, Vancouver, condominium apartment universe, Table 4.3.1, editions 2019 and 2022. Both the rented count and the surveyed stock come from the same survey, which is what makes the subtraction in section 3 meaningful.
The proportional decline. Year-over-year percentage change in the declared vacancy rate, averaged within each period of constant tax rate.
Provincial data. Speculation and Vacancy Tax data by municipality, 2024 tax year, Table 3 for properties by exemption status and Table 2 for exemptions claimed.
By-law text. Vancouver Vacancy Tax By-law No. 11674, sections 1.2, 3.3, 3.9, 3.10, 4.12 and 5.3. The by-law and the city's public summary disagree on the number of properties that makes an owner a developer for the exemption in section 1: the summary says five or more, the by-law says two or more.
What is not here. Assessment values, the split between land and building, and price-to-income series were pulled during this research and are not used. They answer the next edition's question rather than this one's.
Appendix B. References
Canada Mortgage and Housing Corporation. 2019. Rental Market Report data tables, Vancouver. Table 4.3.1. Ottawa: CMHC. Accessed 2026-08-06. https://assets.cmhc-schl.gc.ca/sites/cmhc/data-research/data-tables/rental-market-data/rmr-data-tables/2019/rmr-vancouver-2019-en.xlsx
Canada Mortgage and Housing Corporation. 2022. Rental Market Report data tables, Vancouver. Table 4.3.1. Ottawa: CMHC. Accessed 2026-08-06. https://assets.cmhc-schl.gc.ca/sites/cmhc/data-research/data-tables/rental-market-data/rmr-data-tables/2022/rmr-vancouver-2022-en.xlsx
City of Vancouver. 2024. Empty Homes Tax Annual Report, covering the 2023 vacancy reference period. Indicator #2. Accessed 2026-08-06. https://vancouver.ca/files/cov/2024-empty-homes-tax-annual-report.pdf
City of Vancouver. 2025. Empty Homes Tax Annual Report, covering the 2024 vacancy reference period. Indicators #1 to #3, and the vacancy rate chart on page 4. Accessed 2026-08-06. https://vancouver.ca/files/cov/2025-empty-homes-tax-annual-report.pdf
City of Vancouver. Vacancy Tax By-law No. 11674. Consolidated. Sections 1.2, 3.3, 3.9, 3.10, 4.12 and 5.3. Accessed 2026-08-07. No web archive copy exists: the host rejects automated clients, including the Internet Archive's own fetcher, so this entry rests on the by-law number and section, which is the stable legal citation, rather than on the URL resolving. https://bylaws.vancouver.ca/11674c.PDF
Province of British Columbia. 2025. Speculation and Vacancy Tax: Data by Municipality, 2024 tax year. Tables 2 and 3. Accessed 2026-08-07. The province replaces this file annually, so it carries one year at a time. https://news.gov.bc.ca/files/SVT_DatabyMunicipality.pdf
Cite this article
Appendix C. Changelog
- 2026-08-07 First publication.
Appendix D. Corrections
Corrections are made in place, dated, and listed here. If you believe something in
this analysis is wrong, write to vancouver-empty-homes@animatti.ca with the
section and the figure in question. Substantive corrections are logged below with
the date and what changed. Typographical fixes are not logged.
No corrections have been made since publication.