Dear ABC,
If you're looking for opinions on the Census Data, here's mine.
I'm a former data analyst (crime), but I also designed a data matching algorithm still used today by Victoria Police because of its effectiveness, so I have some professional skin in the game.
I am unable to defend the 2026 prioritisation of Census questions.
How does "We urgently need nationally exhaustive information about whether someone identifies as straight, gay, bisexual or another sexual orientation"...
While we didn't obtain nationally exhaustive information about whether households can access a GP,
whether they have unmet childcare needs,
whether working people commute three days rather than five,
whether households are involuntarily overcrowded because they cannot find housing.
Or these types of questions, which would surely be hard to argue against from a prioritisation perspective of best value for data capture, and best opportunity for Australians to provide a holistic picture of their lives;
Citizenship / visa or residency status (because Country of birth doesn't tell government whether someone is a citizen, permanent resident, international student, temporary worker, etc. Those populations can impose quite different infrastructure/service demands).
Year a non-Australian-born person obtained permanent residency/citizenship (because there is much better longitudinal information about settlement and integration than birthplace alone)
Primary mode of transport normally used (Census asks about journey-to-work circumstances, but hybrid/WFH has made a single Census-day travel snapshot much less informative than it once was).
Housing capacity/occupancy pressure (Bedrooms alone is crude. Ask whether people are sharing bedrooms, using converted garages/studies, or accommodating people who would otherwise live elsewhere. That's enormously relevant during a housing shortage).
Reason adults remain in/shared household (Understanding drivers of Financial necessity, caring, cultural preference, disability, housing unavailability etc...would distinguish genuine housing stress from chosen multigenerational living & inform other pressure points th Government can focus on).
Internet quality/access (As infrastructure in its own right, particularly with WFH, telehealth and online government services).
Household energy characteristics (Solar, battery, EV charging, gas/electric heating etc... would be extraordinarily useful for electricity-grid infrastructure planning)
Many of these variables have extremely obvious relationships with hospitals, roads, rail, schools, housing, childcare, workforce participation and infrastructure investment.
The ABS also removed “number of children ever born” from the 2026 Census (it will return in 2031).... while introducing gender and sexual orientation.
From a strategic-planning perspective, fertility and family formation are among the most consequential long-term variables imaginable. They affect future school enrolments, workforce size, dependency ratios, healthcare demand and population ageing.
This missed opportunity to capture all those above mentioned data sets to inform in policy/planning areas isn't just a shame, it's borderline failure of professional judgement.
We removed access to data that could inform those things, and replaced it with gender questions.
Who managed to convince a ABS strategic review board
“We don't have enough justification to collect fertility information every Census, but sexual orientation warrants universal national collection every five years”??
I'd love to see the cost-benefit/consequences methodology behind that decision !
The reality is the Census can only ask so many questions, and we should prioritise the ones that provide the greatest benefit for planning Australia's future, not pander to the "feelings" of a tiny majority of people that got warm and fuzzy because a national census survey "affirmed" their gender.
Thank you, if you read this.