ABS acknowledges concerns about bias in Census religion question
29 January 2023
The Australian Bureau of Statistics (ABS) has acknowledged concerns about potential bias in the wording of the Census question on religion.
In a statement on its website, the ABS has listed the problematic wording of the religion question among key concerns raised during a public consultation held as part of a review of the census questions.
The website notes that, from 193 submissions by individuals and organisations, feedback included: “Concerns about potential bias in the question wording because it assumes a person is affiliated with a religion.”
The campaign recommended, as its top priority, that the ABS re-word the current question – “What is the person’s religion?” – to remove the loaded bias that assumed each respondent had a religion.
The campaign said that, at a minimum, the ABS should reword the question and insert ‘if any’ at the end to make the question: “What religion does the person belong to, if any?” However, the preferred approach would be to use a two-part question: a) “Does the person have a religion?” b) “What is the person’s religion?”
In its response to the public consultation, the ABS said it would present proposed changes, including to the Australian Standard Classification of Religious Groups (ASCRG) and the Religious Affiliation Standard (RAS), by mid-2023.
Michael Dove, spokesperson for the Census21 – Not Religious? campaign, said it was important for the ABS to listen to the concerns raised about the religion question to ensure more accurate data in future.
“The response to the consultation has sent a clear message to the Australian Bureau of Statistics that it should fix the religion question. The biased nature of the questions is obvious, and it is continuing to produce inaccurate data,” he said.
“In assuming that all respondents have a religion, the current question leads to a result that greatly overstates the religious affiliation of Australians. While almost 40 per cent of Australians told the 2021 Census they were not religious, we believe the real figure would be over 50 per cent already.
“As governments rely on the Census data to inform policy-making and the allocation of public funds, it’s time for the ABS to fix the fundamental problems with the religion question.”