posted on 2026-08-11 by Kait.
the 2026 Australian census questions are bad #
on statistical and ethical grounds
the questions
as someone who isn’t trans, can you please help me understand what is wrong with the current questions (of iirc “Sex assigned at birth” followed by “What gender do you identify as”), im just genuinely curious and trying to better understand it
put simply, there are other ways to get the information which are both easier and vastly more tactful, and the ABS should do better.
if the two questions were only used to infer gender identity then it would be frustrating but it would still be tolerable. the ABS will use sex at birth for other purposes though, to present data categorised by sex. this is despite the fact that it does not serve the goal of the census - to get useful data from which to make informed decisions. sex at birth is not an accurate indicator of… anything. not about how you live your life, not about which services you might use, and not about how others perceive you.
so, asking and recording the sex at birth data is less than worthless, and only causes distress (and likely messy data) and it invites misuse of the data
separately, the linked article makes more arguments against the two question method when used by organisations in general (outside of census). it’s a good article and its thesis is that organisations should ask themself: do you have a legitimate need to know if someone is trans? if so, is this a good way of asking and storing that information? the article says that for most places that are uncritically using the two question method, they fail on both counts. the census fails on the second.
official sources #
A question directly asking trans status or including transgender or trans and gender diverse as a response option in a gender question is discouraged. These practices give unreliable statistical results and are unlikely to offer a meaningful population count.