Roy Morgan's mid-September analysis is the first large-sample look at One Nation support by country of birth. It is based on 26,418 electors interviewed across January–July 2026, when One Nation's national primary was running at 24 per cent. Australian-born electors sat almost exactly on that average (25 per cent). The variation is among people born overseas.

Highest One Nation support:

South Africa: 36 per cent

Italy: 30 per cent

Mainland Europe as a whole: 29 per cent

Around the national average:

Australia: 25 per cent

New Zealand: 24 per cent

United Kingdom: 23 per cent

Well below average:

Middle East: 17 per cent

Philippines: 16 per cent

Central and South America: 12 per cent

Africa excluding South Africa: 11 per cent

Malaysia: 10 per cent

Asia overall / US and Canada: 9 per cent

India: 8 per cent

Vietnam, Sri Lanka, Pakistan: 7 per cent

Nepal: 6 per cent

Mainland China: 5 per cent.

Reporting also noted relatively strong (unspecified) support among electors born in Greece and Israel. Sub-samples for smaller source countries are modest: Roy Morgan cited ranges from 49 Nepal-born respondents to 242 South Africa-born, so treat the tails as directional, not precise.

That pattern is the electoral fact. Below is how a campaign that wants to use it would treat it.

What the numbers actually say:

One Nation's vote among migrants is not uniformly low. It is high among some European and southern African source countries and low among most East Asian, South Asian and North American source countries. Australian-born, UK-born and NZ-born voters look like the national mean. The party's problem is not "migrants" as a bloc. It is particular diasporas that currently give it single-digit primaries, plus the fact that those diasporas are large and growing in a number of outer-suburban seats.

South African-born support at 36 per cent is the standout. That community is concentrated, English-speaking, and already over-indexed in several growth corridors. Italian-born and broader mainland-European-born support sitting 5–6 points above the national figure is also usable. Those groups are not the same as "the migrant vote." Treating them as interchangeable with China-born or India-born electors is how a party wastes effort.

Practical targeting:

1.Fish where the fish are. Prioritise South African, Italian and other mainland-European networks for membership, booth teams, candidate pipelines and community events. These are the overseas-born groups already voting One Nation at or above the rate of the Australian-born. A party that talks only to "the Australian-born base" is leaving its strongest migrant supporters unorganised.

2.Do not confuse low support with zero support. Middle East-born (17 per cent) and Philippines-born (16 per cent) are below average but not in the 5–8 per cent band. That is a different problem from China-born (5 per cent) or India-born (8 per cent). Message testing should be separate. A housing-and-infrastructure argument may travel further than a culture argument in some of those communities; the reverse may be true in others. The data do not tell you which, only that the starting points differ.

3.Candidate selection should match the map, not a diversity quota. If the goal is converting the Roy Morgan pattern into seats, candidates who look and sound like the high-support migrant groups (and like the Australian-born majority) will do more work than symbolic selections aimed at groups currently at 5–8 per cent. Token candidacies in low-support diasporas rarely move those numbers and can confuse the high-support ones.

4.Geography first. South African-born and European-born concentrations are not evenly spread. Campaign infrastructure should follow those postcodes and the Australian-born suburbs that already poll near 25 per cent, rather than national media buys aimed at "migrant Australia" as if it were one market.

5.Keep the issue frame tight. One Nation's recent primary strength has been tied to migration volume, housing capacity and cultural cohesion. The birthplace split is consistent with that: groups whose political memory includes high-crime transitions, farm seizures, or rapid demographic change (South Africa) or who arrived in earlier, more assimilative waves (Italy, much of mainland Europe) are more open to that frame. Groups whose current partisan home is Labor or the Greens are not going to be talked across with the same lines. Trying to split the difference usually blurs the offer that is already working with the 24–36 per cent bands.

What not to do:

Do not read the table as a licence to write off entire communities, and do not read it as proof that "migrants love One Nation." Most of the large recent source countries do not. Do not treat country of birth as a stand-in for the voting behaviour of Australian-born children of those groups, the survey does not measure that. Do not over-interpret tiny cells. And do not let a 36 per cent figure in one community become an excuse to ignore the Australian-born 25 per cent, which is still the bulk of the electorate.

The strategic implication:

One Nation is already at parity with the Australian-born average. Its upside among overseas-born electors is concentrated, not universal. A campaign that organises South African and European-born supporters, holds its Australian-born/UK/NZ floor, and stops pretending the 5–8 per cent groups are a near-term conversion target is aligned with the evidence. A campaign that treats "the migrant vote" as a single hostile or single friendly bloc is not.

That is the advice the Roy Morgan table supports as I read it.

https://www.roymorgan.com/findings/10341-voting-intentions-by-country-of-birth-september-2026