A large Roy Morgan analysis has found sharp differences in support for One Nation among Australian voters born overseas, with South African-born and Italian-born electors recording substantially higher support for the party than voters born in many Asian countries.
The analysis, published on 15 September, combined interviews with 26,418 Australian electors conducted between January and July 2026. Roy Morgan reported that 24 per cent of all electors in the combined sample supported One Nation.
Among voters born in South Africa, support for One Nation was 36 per cent. Italian-born electors recorded 30 per cent, while voters grouped as being born in mainland Europe recorded 29 per cent.
Those figures were above the 25 per cent recorded among Australian-born electors and the 24 per cent among New Zealand-born electors. United Kingdom-born voters recorded 23 per cent, broadly in line with the national sample.
At the other end of the table, One Nation support was much lower among several Asian-born groups. Roy Morgan reported 5 per cent support among mainland China-born electors, 6 per cent among Nepal-born voters, 7 per cent among voters born in Vietnam, Sri Lanka and Pakistan, and 8 per cent among India-born electors.
The results complicate any simple claim that migrant voters behave as one political bloc. Country of birth was associated with large differences in party preference, and those differences extended beyond One Nation.
Labor support was particularly strong among several Asian-born groups in the Roy Morgan data. The company reported Labor support of 57 per cent among Vietnam-born electors, 52 per cent among Nepal-born voters and 49 per cent among Sri Lanka-born voters. Labor also recorded more than 40 per cent among voters born in mainland China, India, Malaysia and Pakistan.
The Coalition performed relatively strongly among some of the same groups. Roy Morgan reported Coalition support of 34 per cent among mainland China-born electors, 33 per cent among India-born voters and 30 per cent among Nepal-born electors.
The Greens recorded their highest figure among Pakistan-born electors in the published table, at 32 per cent. Roy Morgan suggested the result may be influenced in part by the profile of Greens senator Mehreen Faruqi, who was born in Pakistan, although that is an interpretation rather than a directly measured reason given by respondents.
The One Nation figures have drawn particular attention because of the party’s hardline position on migration. One Nation has called for much lower immigration and tougher action against visa overstayers, while presenting high migration as a major contributor to housing and infrastructure pressure.
At first glance, strong One Nation support among some overseas-born voters may appear counterintuitive. But voting behaviour is shaped by far more than a person’s migration history. Economic circumstances, age, religion, region, education, views about subsequent migration and experiences in a voter’s country of origin can all influence political preferences.
It is therefore important not to treat country of birth as a substitute for ethnicity, race or political identity. “South African-born”, for example, includes people from different racial and cultural backgrounds. “Mainland Europe” combines many countries with different political traditions. Even within a single country-of-birth group, individual voting preferences can vary widely.
The sample sizes also differ sharply between groups. The overall survey base was large, but some country-specific subgroups were relatively small. Roy Morgan listed 242 South African-born respondents, 82 Italy-born respondents, 96 Vietnam-born respondents, 49 Nepal-born respondents and 71 Pakistan-born respondents, compared with more than 21,000 Australian-born electors.
That matters because margins of error increase as sample sizes shrink. A figure based on several dozen respondents carries substantially more statistical uncertainty than a figure based on thousands. The published percentages are useful indicators of difference, but they should not be read as exact measurements of every community.
Roy Morgan itself notes that sampling error depends primarily on the number of interviews and that design effects from weighting and stratification should also be considered. The company’s general margin-of-error table shows how uncertainty declines as sample size increases.
The analysis covers interviews conducted over the first seven months of 2026 rather than a single week. That makes it different from a conventional voting-intention poll designed to capture the immediate political mood. Combining months of data increases the number of respondents available for smaller demographic groups but also averages across changes in party support over time.
That distinction is especially relevant in a volatile political year. One Nation’s national support has moved considerably in Roy Morgan’s weekly polling during 2026. A separate poll conducted from 7 to 13 September placed both Labor and One Nation on 25 per cent primary support, with the Coalition on 21.5 per cent and the Greens on 16 per cent.
The country-of-birth analysis therefore should not be used to claim that any group will vote at the published percentage in a future election. It describes the combined responses of people interviewed during a specified seven-month period.
Roy Morgan chief executive Michele Levine said the analysis showed that immigrant voting patterns differed substantially depending on country of birth and that the clearest contrast involved One Nation. She also noted that the results could matter strategically as parties prepare for upcoming state and federal elections.
Political parties increasingly use demographic data to understand where their support is growing or weakening. Country of birth can be one variable among many, alongside age, gender, income, occupation, region and housing status. But there are risks if parties or media organisations turn statistical groups into stereotypes.
A poll showing higher support for One Nation among South African-born voters does not establish why those respondents chose the party. It cannot demonstrate that race, migration policy or any single issue caused the result unless respondents were separately asked about their motivations.
Similarly, lower One Nation support among Asian-born voters should not be interpreted as uniform opposition to the party or as evidence that all Asian communities share the same political priorities. The Roy Morgan data itself shows meaningful differences between countries and between support for Labor, the Coalition and the Greens.
The most useful conclusion is therefore a limited one: Australian migrant voters are politically diverse, and their voting intentions vary substantially by country of birth.
That finding challenges broad political narratives from both supporters and opponents of One Nation. It shows that some immigrant groups are more receptive to the party than Australian-born voters, while others are far less receptive.
As Australia’s overseas-born population continues to grow, those differences will become increasingly important to campaign strategy. But the figures should be read with the same caution that applies to any subgroup polling: sample size, timing, methodology and uncertainty matter.
Roy Morgan’s analysis provides an unusually large dataset for examining those patterns. It does not provide a simple explanation for them, and it should not be used to reduce millions of Australians to racial or national categories. The numbers describe voting intention during a period of 2026; the reasons behind those choices remain more complex than any single table can show.
The wording used to describe the findings also matters. The Roy Morgan release refers to country or region of birth, not “white immigrants” as a formal survey category. Some secondary coverage has grouped European and South African respondents together and used racial language, but that is an editorial characterisation rather than the variable measured in the published data.
That distinction is important in a country where nationality, ancestry and race do not neatly overlap. South Africa is racially diverse, European-born Australians have many different backgrounds, and voters may identify culturally in ways that cannot be inferred from birthplace. Reporting the actual survey categories avoids assigning racial identities that respondents were not necessarily asked to provide.
The analysis is nevertheless politically significant because it undermines the assumption that support for stricter migration policies comes only from Australian-born voters. Some people who migrated to Australia themselves may favour lower future migration, just as others may strongly oppose it. That pattern is not unique to Australia and can emerge when voters distinguish between their own circumstances, the migration system they entered under and the policies they believe should apply now.
