Every quarter the department publishes a workbook of Star Ratings for every residential aged care home in the country. We read the May edition in July and said then that the interesting variation was not in the overall score, where seven in ten homes hold exactly 4 stars, but underneath it in Staffing and Quality Measures. The August edition has landed, and it is a good illustration of why a quarterly dataset has to be compared carefully rather than simply re-totalled.
The two answers
Take every home with a published Staffing rating in each extract and count the ones at 1 or 2 stars. In May that was 575 homes of 2,180 rated, or 26.4 per cent. In August it is 717 of 2,390, or 30.0 per cent. On that reading, staffing deteriorated sharply in a single quarter.
Now restrict the comparison to homes that appear in both extracts and carry a Staffing rating in both, which is 1,895 of them. Among that group, 1 and 2-star homes fell from 496 to 480, or 26.2 to 25.3 per cent. Within it, 412 homes improved their Staffing rating and 361 declined. On that reading, staffing improved slightly.
Staffing rating, 1 or 2 stars, two ways of counting
| Cohort | May 2026 | August 2026 |
|---|---|---|
| All homes with a published rating | 575 of 2,180 (26.4%) | 717 of 2,390 (30.0%) |
| Only homes rated in both quarters | 496 of 1,895 (26.2%) | 480 of 1,895 (25.3%) |
Where the difference comes from
The May extract listed 2,596 services and published an overall rating for 2,180 of them, leaving 416 with no overall score. August lists 2,600 services and rates 2,390, leaving 210. Those homes did not arrive from nowhere. They are largely the same services, moving from unrated to rated as their assessments complete.
So we split the August figures by whether a home already had a Staffing rating in May.
August 2026 Staffing ratings, split by whether the home was rated in May
| Group | Homes | At 1 or 2 stars | Mean rating |
|---|---|---|---|
| Rated for Staffing in both quarters | 1,895 | 480 (25.3%) | 3.32 |
| Gained a Staffing rating since May | 235 | 187 (79.6%) | 2.19 |
That is the whole story of the sector-wide move. A group of 235 homes entered the published staffing statistics this quarter, and nearly four in five of them are at the bottom two rungs, against one in four among homes already being counted. Their mean Staffing rating is 2.19 against 3.32.
The same pattern holds on the overall score, less starkly. The 230 homes that gained an overall rating average 3.37 stars, against 3.88 for the 1,801 rated in both quarters. Of those 230, four reached 5 stars.
What did change among the homes themselves
Comparing like with like, the quarter was mildly positive. Among 2,181 services matched across both extracts, 231 improved their overall rating and 174 declined. Quality Measures 1 and 2-star homes fell from 26.1 to 24.3 per cent within the matched group. On care minutes, restricted to the 1,914 homes with complete data in both extracts, the share missing at least one of their two mandated targets edged down from 26.4 to 25.6 per cent, with 181 homes stopping failing and 165 starting.
One home moved by two or more stars: Westmont Homestead at Baranduda in Victoria, from 1 star to 3. It is the same home we named in July as the sector’s largest faller, having gone from 4 stars to 1 between February and May. Australia now has no 1-star aged care home, and none at 2 stars either.
Our view, labelled as such
The honest headline from this quarter is not that staffing got worse, because among homes being measured consistently it did not. It is that 235 homes were absent from the published staffing statistics until now, and they are dramatically worse than the homes that were in them. A family comparing homes in May could not see a staffing star for those services at all. That is a real gap in a tool built specifically so families can compare, and it is closing, which is good.
It also means the sector-wide averages of the last few years have been flattering, not because anyone fudged them but because the worst-staffed homes were disproportionately the ones without a published rating. If the remaining 210 unrated services resemble the 235 just added, the national staffing picture has further to fall before it is complete. That is a prediction, and we will check it against the next extract.
The general point is worth holding on to for any quarterly dataset. Re-totalling a new release and comparing it to the last one measures two things at once: what changed, and who is being counted. When the counted population is still growing, those can point in opposite directions, and this quarter they did.
Added 18 August 2026: the Staffing star is a pass or fail on care minutes, with three grades of pass
This piece treated the Staffing rating and the care-minutes test as two measures pointing the same way. They are closer than that, and the extract makes it checkable. Cross-tabulating every August home that has complete care-minutes data produces a table with no exceptions in it: all 774 homes rated 1 or 2 stars for Staffing miss at least one legislated target, and all 1,721 rated 3, 4 or 5 stars meet both. Not mostly. Every one.
That is the algorithm working as documented. The department’s
Star
Ratings Provider Manual says it plainly: The redesigned Staffing rating algorithm
means that aged care homes need to meet both of their legislated care minute targets to
receive a Staffing rating of 3 or more stars.
The manual dates that redesign to the
May 2026 update, which is the quarter this story compares against, so the two extracts here
are on the same basis and the comparison holds.
What that means for anyone reading these ratings. One and two stars for Staffing carry a single meaning: the home missed a legislated care-minutes target. Three, four and five all mean it met both, and separate only on how far above. So the staffing count and the care-minutes count in this story are not independent corroboration; they are largely one finding seen twice, and we should have said so first time.
The compression that this masthead keeps returning to is visible here too. Of the 774 homes on 1 or 2 stars for Staffing, none carries an overall rating below 3: 420 are 3-star overall, 297 are 4-star and 57 are unrated. Of the 171 homes on the lowest possible staffing score, 19 hold four stars overall. A family reading only the headline number would not know.
How we worked this out
Every figure comes from two workbooks published by the Department of Health, Disability and Ageing: the August 2026 Star Ratings quarterly data extract, downloaded on 16 August 2026, and the May 2026 extract we analysed in July. Both are kept on file. The counts are ours, computed from the workbooks with scripts kept alongside the data so the same method runs each quarter.
The join, and its known weakness. The extract carries no persistent service identifier, so services are matched across quarters on the combination of Service Name, Provider Name and Service Suburb. A home that is renamed, or whose provider changes, drops out of the matched set and would be counted as newly rated. That is the main way these figures could be wrong, and it is why we report both the whole-cohort and the matched-cohort numbers rather than choosing one.
What we mean by rated. A home counts as having a rating where the workbook records a whole number in that column. Homes with a blank are treated as unrated rather than as zero.
What we have not done. We have not established why any individual home was unrated in May, and the department does not publish a reason per service, so we do not assert that any particular home was withholding information or under assessment. We have not contacted the department, and we would publish a response. We have not linked any of these homes to enforcement action. The prediction that the remaining 210 unrated services will resemble the 235 just added is ours and is untested.
Sources
- Department of Health, Disability and Ageing, Star Ratings Provider Manual (PDF, read 18 August 2026): the Staffing rating algorithm, that homes must meet both legislated care minute targets to receive a Staffing rating of 3 or more stars, and that the redesign is reflected in published ratings from the May 2026 quarterly update.
- Star Ratings quarterly data extract, August 2026, Department of Health, Disability and Ageing (workbook downloaded and analysed 16 August 2026): every August figure here, including the 2,600 service rows, the 2,390 with a published overall rating, the Staffing and Quality Measures distributions, and the care-minutes targets and actuals.
- Star Ratings quarterly data extract, May 2026, Department of Health, Disability and Ageing (data as at 5 May 2026, released 26 May 2026): the comparison quarter, and the source of every May figure.
- Star Ratings quarterly data extracts, Department of Health, Disability and Ageing: the collection page confirming the quarterly cadence and the August 2026 release.
- The Long View, Seven in ten aged care homes carry the same star rating: our reading of the May extract, including the finding that the variation sits in Staffing and Quality Measures, and the identification of Westmont Homestead as the quarter’s largest faller.
Spotted an error? The correction form is on our tips page; we check every correction against the cited sources and log the outcome here.