Poll briefing · Ace Strategies

Western Australia, June 2026

Labor's primary vote falls to 33 per cent and its two-party lead narrows four points to 53–47. The government's rating fell on four of the eight issues it was tested on, which is the reverse of what Queensland's did over the same months.

State: Western Australia · PremierNational
Fieldwork: 29 May – 11 June 2026
Sample: n=1,015 · eff. 616 · ±3.9%
Election: 10 March 2029

Labor primary

33%

Down 8.4 on its 2025 result, and its lowest of the cycle.

One Nation

18%

Up 14.0 since the election, and level with February.

Liberal

23%

Up 2 since February, and still 5.0 below 2025.

Labor is the only government in this batch whose two-party position has fallen

A change between the February and June waves carries about ±5.4 points at 95 per cent, from effective samples of 668 and 616. No single primary-vote movement clears it, and the two-party figure moved four points anyway.

Each bar is the change in a party's first preference from February to June 2026, in points. The pale band is the ±5.4 points a change of that size carries.

Others

Liberal

One Nation

Greens

−1

Nationals

−1

Labor

−3

The largest movement is three points and the interval is ±5.4, so this wave on its own is consistent with no change. What makes it worth reading is the direction of every bar: Labor and the Greens down, the Liberals, One Nation and the others up. Six small movements all pointing the same way are weaker evidence than one large one, but they are not nothing, and the two-party figure reflects their sum rather than any one of them.

Across the cycle the comparison with Queensland is the useful one. Both governments have shed primary vote: Labor here is down 8.4 points on its election result, the LNP there down 7.5. Queensland's two-party position improved by 4.2 points over that period and Western Australia's fell by 4.1. The difference is not in how the two electorates feel about their governments. It is that One Nation preferences return to the conservative side in both states, which helps a conservative government in Queensland and hurts a Labor one here, and that the return rate is 74 per cent there against 68.5 here. ACE-IU-2026-029 works the Queensland side of the same arithmetic.

Where the four two-party points went

Running DemosAU's published flow rates over the June primaries reproduces 53.6 against a published 53. The fall from February is carried by Labor's own three points and the others' two.

The preference flows DemosAU applied to the June 2026 Western Australian primary votes, and the two-party preferred totals they produce.
Source	Primary	To Labor	To Liberal
Labor	33	33.0	—
Liberal	23	—	23.0
One Nation	18	5.7	12.3
Greens	12	9.2	2.8
WA Nationals	3	0.7	2.3
Others	11	5.1	5.9
Total	100	53.6	46.4

The flow rates are DemosAU's and are unchanged from February: Nationals 77 per cent to the Liberals, Greens 76.5 per cent to Labor, One Nation 68.5 per cent to the Liberals, others 54 per cent to the Liberals. Each cell is rounded to one decimal place and the totals are struck on unrounded values. The published two-party figures move from 57 to 53 and the same flows over the published primaries reproduce them as 56.4 and 53.6, a move of 2.8; the difference is what rounding the primary votes to whole numbers costs at both ends.

The others category is the quiet mover. At 11 per cent it is the highest it has been in the cycle, above its 10.3 at the election and its 9 in February, and 54 per cent of it flows to the Liberals. Labor's three-point primary fall costs it three points directly; the two points moving into a category that splits against it cost about another fifth of a point. Those, with the Greens' one-point fall, account for the movement the two-party figure records.

The structural position is worth separating from the four-point move. Labor holds 53.6 per cent of the two-party vote on a primary of 33, which means roughly a fifth of its two-party support is borrowed from Greens preferences and would be returned the moment those voters stopped preferencing it. A government on 41.4 per cent of primary vote, as this one was at the election, does not depend on that. A government on 33 does.

The government's rating fell on four issues and rose on two

Asked whether the Cook government is doing a good job on each issue, respondents rated it worse than February on education, job creation, hospitals and the budget, and better on crime and cost of living. Housing and the economy did not move.

Each bar is the net rating on an issue, the good-job share less the bad-job share, in points. The note gives the three underlying shares and the movement since February.

Education and schools

−3

25% good job, 47% neutral, 28% bad job. The net is down 4 and has crossed from positive to negative.

Job creation

−6

23% good job, 48% neutral, 29% bad job. The net is down 7 and has also crossed below zero.

Crime and policing

−13

22% good job, 43% neutral, 35% bad job. The net is up 5, the largest improvement in the battery.

Budget management

−14

23% good job, 40% neutral, 37% bad job. The net is down 8, the largest fall in the battery.

The economy

−18

22% good job, 38% neutral, 40% bad job. The net is unchanged since February.

Hospitals and health system

−33

16% good job, 35% neutral, 49% bad job. The net is down 2.

Cost of living

−51

10% good job, 29% neutral, 61% bad job. The net is up 2 from a very low base.

Housing

−54

7% good job, 32% neutral, 61% bad job. The good-job share of 7 per cent is the lowest figure in any ratings battery in this batch.

Nothing in this battery is rated positively. In November 2025 education and job creation were both at a net of 1 positive, and they are now 3 and 6 negative, so the government has lost the only two issues it was ahead on. Housing's good-job share of 7 per cent is the lowest single figure in any ratings battery in this batch, and in a state where 34 per cent name housing as the biggest issue facing them.

Set this against Queensland's June wave, where the government's rating improved on five of seven issues while its vote fell. Here the ratings fall and the vote falls with them, which is the ordinary relationship. Queensland's is the pattern that needs explaining, not this one. What the two have in common is the bottom of the chart: both governments are rated worst on cost of living and housing, both by margins above 35 points, and in both states those are the two issues voters name most often.

One Nation takes 29 per cent of regional Western Australia and the Liberals take 14

DemosAU publishes no cohort sample sizes for this wave. If the three-way splits each hold about a third of the effective sample of 616, a gap between two of them carries roughly ±7.7 points.

June 2026 Western Australian first preferences by party within each demographic cohort, in per cent.
Cohort	ALP	LIB	NAT	ONP	GRN	OTH
Western Australia	33	23	3	18	12	11
Men	37	22	—	21	9	9
Women	29	24	5	15	14	13
Aged 18 to 34	37	10	—	10	28	13
Aged 35 to 54	35	23	—	18	11	12
Aged 55 and over	29	31	6	23	3	8
School education	34	21	4	22	11	8
TAFE education	32	21	3	17	13	14
University education	34	28	—	14	11	11
Own home outright	34	29	7	19	4	7
Mortgage holders	30	24	—	18	14	12
Renters	36	14	—	17	16	15
Income under $45k	33	21	5	16	14	11
Income $45k to $125k	33	24	—	19	10	12
Income over $125k	31	26	—	22	10	9
Inner metropolitan	38	27	—	13	9	13
Outer metropolitan	32	23	—	19	15	9
Regional and rural	23	14	13	29	9	12

These figures are read from the demographic charts on the report's pages 10 and 11. The report draws the WA Nationals segment in every row but prints its label only where the bar is wide enough; an en dash marks a cell whose figure the report does not state, and the residual implied by the other five columns is between 0 and 2 points in each case. Every row with a complete set sums to 100. Cohort figures rest on subsamples and carry wider margins than the ±3.9% statewide topline.

Regional and rural Western Australia is the only cohort where Labor does not lead. One Nation takes 29 per cent there against Labor's 23 and the Liberals' 14, and the WA Nationals take 13, so the conservative vote in the regions is split three ways with the smallest share going to the party that leads the opposition. That is a different problem from the one the Liberals have in the metropolitan cohorts, where they run 27 and 23 and are clearly second.

One Nation's distribution has sharpened since February, when it ran between 13 and 21 across every cohort. It now runs from 10 among under-35s to 29 in the regions, and it takes 23 per cent of the over-55s and 22 of those earning over $125,000. The Liberals' 10 per cent among under-35s is the lowest figure any major party records in this batch, and the Greens' 28 in the same cohort is more than twice their statewide 12. Among those voters the Greens run second to Labor's 37 and ahead of both the Liberals and One Nation, which are level on 10.

Data notes

Roger Cook led Basil Zempilas as preferred premier 44 to 33, with 23 per cent naming neither, a margin of eleven points against thirteen in both earlier waves. On personal ratings Cook recorded a net of 6 positive, unchanged since February, and Zempilas 3 negative, also unchanged. Rita Saffioti recorded 17 negative, down seven, Libby Mettam 6 negative, Rod Caddies of One Nation 12 negative and Brad Pettitt of the Greens 18 negative. Cook is the only one of the six in positive territory.

Asked whether Western Australia is heading in the right or the wrong direction, 36 per cent said right, 43 per cent wrong and 21 per cent did not know. The right-direction share is down three points since February and the state has moved from a net positive reading in November 2025 to a net of seven points negative.

Asked the biggest issue facing Western Australia in an open-ended question, respondents named housing at 34 per cent and cost of living and inflation at 33. Health and hospitals followed at 8, immigration at 5, the economy and crime at 4 each, government spending and debt at 3, then social cohesion, climate change, infrastructure and the current state government at 2 each, and energy, global economic uncertainty and taxation at 1 each. Miscellaneous answers took 9 and don't know 1. The categories were generated with the assistance of AI and may sum to more than 100.

Method

DemosAU/PremierNational Western Australia Poll. Fieldwork 29 May to 11 June 2026, internet panels, Western Australian residents 18+ eligible to vote, n=1,015, effective n=616, effective margin of error ±3.9%. Rim weighting on age by gender, education, personal income, residential tenure, language, region, 2025 federal vote and left–right self-identification. State voting intention was the first substantive question after demographics. Initial undecided voters were asked a leaner question and the remainder were excluded from the totals, which exclude 6 per cent of voters. Two-party preferred uses 2025 state election flows at Nationals 77–23 to the Liberals, Greens 76.5–23.5 to Labor, One Nation 68.5–31.5 to the Liberals and others 54–46 to the Liberals, under full preferential voting in which no ballot exhausts. Polling and methodology are the work of DemosAU, a member of the Australian Polling Council, in association with PremierNational. Ace Strategies had no role in the poll; analysis and presentation are Ace Strategies' alone. Subgroup samples carry wider margins than the ±3.9% topline.

Use

Polling is by DemosAU and the sponsor is PremierNational. Material from this briefing may be reproduced in client and internal documents provided it credits acestrategies.au, the pollster and the sponsor. Not for resale.

Contact

Ace Strategies · info@acestrategies.au · acestrategies.au

Filing

File: ACE-IU-2026-032
Series: IU-03 · Poll briefings
State: WA · Public release
Issued: October 2026

Every poll briefing The previous Western Australian wave The next Western Australian wave The wave as published