A survey commissioned before Australia's 2026 census reveals a stark finding: the way you ask Australians about religion dramatically shapes the answer. Essential Media tested two formats on 2,000-plus respondents each. The first used the current census approach — a list of religions with 'no religion' as an option. The second posed a direct yes-or-no question: do you have a religion? The difference was striking. The list-based format produced 43% selecting 'no religion'; the yes-or-no format produced 54% The Guardian.

Why this gap matters has everything to do with how census data works. The Australian Bureau of Statistics census, scheduled for 11 August 2026, is treated as the authoritative measure by governments, churches, and funding bodies that allocate resources for religious education, chaplaincy services, and multicultural programs. An 11-percentage-point swing across millions of people is not a statistical rounding error—it translates to roughly 2 million adults who would categorise themselves differently depending on question format alone. Michael Dove, spokesperson for the campaign that commissioned the poll, called the census the "gold standard" of data underpinning policy and funding decisions The Guardian.

The mechanics of question design explain the divergence. When respondents see a list with 'no religion' embedded among many options, they must actively locate and select it — a cognitive task that framing research shows can suppress rates of selection. When asked directly "Do you have a religion?" before any denominational menu, respondents engage a different mental process: they answer a binary first, then elaborate. Survey methodology has long documented this phenomenon. Leading with a yes-or-no choice tends to shift distributions relative to burying the negative option inside a longer list. Essential Media's findings replicate that pattern almost precisely.

The Essential Media poll also surfaced demographic texture. Under the yes-or-no format, religious affiliation was lower among 18-34 year-olds, at 34%, and higher among university-educated respondents, at 50%. Both patterns align with what secularisation research shows internationally: youth and education correlate with lower religious affiliation. The poll does not establish why — only that stated affiliation shifts across the two formats in predictable demographic groups.

One methodological caveat matters if you are treating this as a proxy for the actual 2026 census outcome. Essential Media surveyed adults only. The census counts children too, typically via an adult submitting answers on their behalf. Children's religious affiliation, often reflecting household default rather than personal choice, tends to follow different patterns than adult self-report. Any extrapolation from this poll to a full population census outcome under an alternative format should account for that structural difference.

The broader context here is that the ABS itself considered revising the religion question after the 2021 census, in which 43.9% of Australians nominated Christianity and 38.9% ticked 'no religion' The Guardian. The ABS subsequently published clarification reports on the religion question for the 2026 cycle ABS. That regulatory process has not gone uncontested. In June 2026, Catholic bishops wrote to Prime Minister Anthony Albanese expressing concerns about proposed changes to the religion identification question, signalling that religious institutions themselves recognise how sensitive question design is to their own reported membership, funding allocations, and public standing The Saturday Paper.

What distinguishes this dispute is that it is not fundamentally about belief. It is about instrument design — and both camps have material stakes in the outcome. The campaign behind this survey — Census: Not Religious? Mark No Religion — argues that the current format understates the no-religion cohort and points to other surveys on its website, censusnotreligious.org.au, showing higher non-affiliation shares than the 2021 census figure of 38.9%. Churches, chaplaincy funding, religious education allocations, and multicultural service provision all depend on census religion data. A lower affiliation count affects budgets and institutional visibility. A higher count shapes public perception of religious Australia's role in society.

The 2026 census will proceed using the ABS's existing list-based format. No format change of the kind tested in this poll has been adopted for this cycle. The Essential Media results function as an advocacy data point rather than a preview of the actual census outcome — a distinction worth holding as the 54% figure circulates in coming weeks. Whether the ABS revisits question design for future censuses will depend on how it weighs the methodological critique against institutional pushback from religious bodies already on record with government.