How rental vacancy rates are measured, and why Queensland sources differ

In this article

Written by

Kooky

Builder of Shaka, the payment router that pays every agent their commission on closing date.

About Kooky and Shaka →

For the middle of 2026, three widely quoted sources gave three different vacancy rates for Brisbane. SQM Research put the city at 0.9 per cent for June. The Real Estate Institute of Queensland (REIQ) reported 1.0 per cent for the Brisbane local government area and 0.8 per cent for Greater Brisbane for the June quarter. Cotality, in its Rental Review for the same quarter, had Brisbane at 1.9 per cent.

None of these is a mistake. A vacancy rate is not a count taken from a register; it is an estimate, and each provider builds its own from different raw material. This guide explains what a vacancy rate is, how the three main sources used in Queensland arrive at theirs, why the results differ, what the "tight" and "healthy" labels rest on, and how to read a new release without being misled by a headline number.

SQM Research release of 13 August 2026; REIQ vacancy report published 30 July 2026; Cotality Rental Review published 9 July 2026.

What a vacancy rate is

The definition is simple. In its 2025 review of the measure, the Australian Housing and Urban Research Institute (AHURI) describes the rental vacancy rate as the number of vacant rental dwellings divided by the total number of rental dwellings, tenanted and vacant together. The REIQ gives the same formula in its own explainer: the number of vacant properties in an area divided by the total number of properties in that area.

Both halves of that fraction have to be estimated. Nobody holds a live list of every rental home in Queensland and whether it is occupied this week. So a provider has to decide what will stand in for "vacant", usually a home that is advertised for rent, and what will stand in for "all rental homes", usually an estimate anchored to the Census. Different choices at either step give different rates for the same city in the same month.

It also helps to be clear about what the measure is not. It is not the share of all dwellings that are empty, which would include holiday homes and houses awaiting sale. It covers the private rental market only.

The REIQ's quarterly report and its bands

The REIQ publishes a residential vacancy rate report each quarter. Its June quarter 2026 edition, released on 30 July, covered 50 regions, from the Brisbane local government area and its surrounding councils to regional centres and smaller markets such as Cook, Isaac and Goondiwindi. That geographic reach is its main strength for Queensland readers: it is the one regular source that gives a rate for Hervey Bay, Mount Isa or the Bay Islands as well as for the capital.

The institute also supplies the vocabulary most Queensland reporting uses. It sorts markets into three bands.

On how the numbers are gathered, the REIQ's release and its explainer do not go into detail. State real estate institutes have traditionally surveyed their member agencies, asking how many of the homes on each agency's rent roll are unlet. AHURI's review describes that survey approach and then notes that the REIQ, along with the institutes in South Australia and the ACT, has moved from conducting member surveys to sourcing data from external producers, naming SQM Research as an example.

That matters for interpretation. If the REIQ's regional figures now rest on listings data, they share the strengths and limits of the listings method described in the next section, and the closeness of the REIQ and SQM figures for Brisbane in mid-2026 is less surprising. The institute's own material does not state its current method, so this point rests on AHURI's account.

SQM Research: listings open for three weeks or more

SQM Research publishes a vacancy rate every month, for postcodes, regions and capital cities, and it states its method on its releases and its website.

The numerator comes from online rental advertisements. SQM says its vacancy count is based on all monitored and unique online listings for the calendar month, and that only listings advertised for three weeks or more are used. It removes advertisements with no address and removes duplicates where the same address appears on more than one site.

The denominator is the stock of rental homes. SQM says it starts from the number of established dwellings in each postcode as counted in the Census, multiplies by the share that are rented, and interpolates between Census years. For the 2016 and 2021 Censuses, its note adds, it used the number of renters as counted plus a proportion of the dwellings where the occupant's tenure was not stated or not applicable. That detail matters more than it looks: a Census never classifies every dwelling, and how the unclassified ones are shared out changes the size of the rental stock, and with it the rate.

The method rests on one assumption, which AHURI's review spells out: that monitoring the major listing websites captures more than 97 per cent of all real estate listings. SQM's own note acknowledges two known weaknesses of advertisements, false listings and homes that are taken up immediately, and says that checking addresses and counting over a full calendar month deal with them.

The three-week filter is the key design choice. A home advertised on a Monday and let by the following weekend was never really spare capacity; it was changing hands. Counting only listings that have been open for 21 days or more leaves such homes out, and the effect is a lower figure than a method that counts every listing. AHURI's review adds that SQM's coverage leaves out short-term rentals, build-to-rent buildings and purpose-built student accommodation.

For July 2026, SQM counted 3,057 vacancies in Brisbane, a rate of 0.9 per cent, against 40,771 and 1.3 per cent nationally.

Cotality: a shorter filter and a different stock estimate

Cotality, formerly CoreLogic, reports vacancy rates in its quarterly Rental Review and in its monthly Home Value Index. Those publications give the result without a method note, so the description here comes from AHURI's review of the providers.

According to AHURI, Cotality counts rental properties that have been advertised for at least two weeks, measured at the end of the month, using unique addresses drawn from several listing sources. Its denominator is a rolling eight-year estimate of rental stock, built by matching sales and rental listings for individual properties and benchmarked against the Census.

Two differences from SQM stand out. The filter is 14 days, not 21, so more listings qualify as vacant. And the stock estimate is built from the firm's own property records over time, not directly from Census counts by postcode. Either difference alone would move the result. Together they help explain why Cotality's Brisbane figure for the June quarter, 1.9 per cent, was about double SQM's 0.9 per cent.

A worked example shows the size of the effect. Suppose a city has 300,000 rental homes, and at the end of a month 5,400 of them have been advertised for at least 14 days, of which 2,700 have been advertised for at least 21 days. Counted with the shorter filter, the rate is 5,400 divided by 300,000, or 1.8 per cent. Counted with the longer one, it is 2,700 divided by 300,000, or 0.9 per cent. Now keep the 2,700 but change the stock estimate to 270,000 homes: the rate becomes 1.0 per cent. These are illustrative figures, not market data, and they are not either firm's counts.

The gap is not confined to Brisbane. In its index for August 2026, Cotality put the national vacancy rate at 1.9 per cent. SQM's national figure for July was 1.3 per cent.

Three methods side by side

Two other national providers appear in Queensland coverage from time to time, and AHURI describes them as well. PropTrack, part of REA Group, counts properties listed and available for rent on realestate.com.au on the last day of each month, with no minimum time on the market, against a stock estimate built from matched sales and rental records. Domain counts properties advertised on its own site for more than three weeks against a Census-based stock estimate. Each is tied to one listing site, which is a further source of difference.

Why the same market gives different numbers

AHURI's review groups the reasons into a handful of causes, and they can be put in plain terms.

The first is the definition of vacant. A filter of two weeks, three weeks or no filter at all captures different things. The review observes that a two-week advertised period is shorter than the notice a tenant must give to leave in some states. In Queensland, a tenant leaving a periodic tenancy gives at least 14 days' notice, so a home can be advertised for a fortnight while the outgoing tenant is still living in it and paying rent.

The second is the stock estimate. A rate is a small number divided by a large one, and the large one is itself modelled. Two providers with identical vacancy counts would still publish different rates if their estimates of total rental homes differed.

The third is coverage. A listings method assumes that most available rentals are advertised on the sites being monitored. Homes let by word of mouth, through a waiting list or directly by an owner never appear.

The fourth applies to survey methods. AHURI notes that survey-based rates depend on which agencies respond, and that sample sizes are not publicly reported. Over the period from 2005 to 2024, it found that the survey-based rates published by the real estate institutes ran consistently above SQM's listings-based rates, most of all in smaller markets, with long-run gaps of 1.8 percentage points for Darwin, 1.2 for Hobart and 1.0 for Perth.

The fifth is timing. Some figures are a snapshot at month end, some cover a calendar month and some, like the REIQ's, describe a quarter.

The 3 per cent benchmark and what "healthy" rests on

Commentary often treats a vacancy rate of about 3 per cent as the point where supply and demand are in balance. The REIQ's healthy band of 2.6 to 3.5 per cent sits around that figure.

AHURI's review tested the idea against two decades of data and found limited support for it. Its conclusion is that the 3 per cent benchmark commonly accepted as representing balance does not appear to fit the empirical evidence it examined. The long-run averages it calculated for 2005 to 2024 are mostly lower.

For Brisbane, both series average a little over 2 per cent across twenty years. That does not make 2 per cent a new target, and AHURI does not propose one. Its recommendation is that policy makers watch how far a market's rate has moved from its own long-term level, not whether it has crossed a single threshold.

Read with that in mind, the REIQ's bands are a convention, and a useful one, since they are applied the same way in each report and let one quarter be compared with the next. A Brisbane reading near 1 per cent is tight on the institute's scale and also about half the city's own long-run average on AHURI's figures. The two framings agree about the present even though they would disagree about where "healthy" begins.

How other countries count, and a rate built from bonds

AHURI's review looked abroad and found Australia doing two things differently. In most countries, it reports, the vacancy rate is produced by the government, often the national statistics agency, and not by private companies. And most countries include unoccupied dwellings of all tenures in the rate, where Australia's measure looks only at vacant rental properties. The review gives the Housing Vacancy Survey of the United States Census Bureau and the work of the Canada Mortgage and Housing Corporation as examples of government-produced figures.

Those differences mean an Australian vacancy rate cannot be set beside an overseas one without care. A rate that counts every empty dwelling, whatever its tenure, measures something broader than homes waiting for a tenant.

The review then tested whether a public source could do the job in Australia. It took the records of Victoria's Residential Tenancies Bond Authority and measured, for each rental property, the number of days between the refund of one bond and the lodgement of the next. A home between two bonds is, for that period, a home without a tenant. AHURI reports that vacancy rates derived this way track well against those of the major commercial producers, and cover more regional areas than the commercial series do.

The same data let the researchers look at how markets behave over time. For 2011 to 2024, AHURI found evidence of an equilibrium vacancy rate in both Melbourne and regional Victoria, meaning a level the market returns to as supply and demand adjust. Regional Victoria was slower to return to it after shocks such as the pandemic.

The review attaches a caution: rental bond data are not collected or managed consistently across the states and territories. It presents no analysis of Queensland's bond records. Queensland does have a single bond authority, the RTA, which holds the bonds for the state's tenancies, but no vacancy rate is published from its records.

What a vacancy rate leaves out

A vacancy rate describes availability. It says nothing directly about price, and it does not describe the households looking.

For price, Queensland has a public source. The Residential Tenancies Authority publishes median rents every quarter, calculated from new rental bond lodgements, by location and dwelling type. The RTA's schedule puts June quarter data out in July and September quarter data in October. Because they come from bonds lodged for new tenancies, these medians describe agreements that were actually entered into, not advertisements.

AHURI's review also warns against reading a rising vacancy rate as proof that pressure has lifted, since a higher rate can mask problems such as overcrowding or homelessness among households that have given up looking for a home of their own. Among its recommendations are that governments consider producing the rate themselves, as an independent and potentially cheaper source, with rental bond data explored as a basis, and that vacancy be examined for smaller areas than the administrative boundaries the published series use.

Five questions to ask of any vacancy figure

Applied to mid-2026, those questions turn three conflicting numbers into a consistent account. SQM's Brisbane rate of 0.9 per cent in July was the same as a year earlier. The REIQ's Brisbane rate of 1.0 per cent for the June quarter was unchanged on the quarter before, while 27 of the 50 regions it tracks recorded a rise. Cotality's 1.9 per cent sits on a higher scale because of how it is built. Each source, read against its own history, describes a capital with little spare rental stock and a state where some regional markets have begun to loosen.

No provider counts every empty rental home. Each one estimates, and the estimate carries the method that made it.

Kooky, from Shaka

Kooky edits Queensland Estate and builds Shaka, the payment router he made for Queensland property professionals. One payment comes in, and every agent, agency and party in the deal receives their signed share on closing date.

About Kooky and Shaka →