The Office for National Statistics has corrected the census after it mistakenly inflated Britain’s “pansexual” population by more than half.
But this was incorrect and it has since been changed to 48,000 people.
Officials blamed the blunder on “a coding mistake”, but critics said it brings the reliability of the entire census into question.
The gaffe centres on the “other sexual orientation” box which invited 165,000 open-ended responses from people who self-defined.
Initially, the ONS website reported that of those who selected “other sexual orientation, the most common write-in responses included: pansexual (112,000, 0.23 per cent), asexual (28,000, 0.06 per cent) [and] queer (15,000, 0.03 per cent). Another 10,000 (0.02 per cent) wrote in a different sexual orientation”.
This has now been corrected to read: “Pansexual (48,000, 0.10 per cent), asexual (28,000, 0.06 per cent), queer (15,000, 0.03 per cent). Another 75,000 (0.15 per cent) wrote in a different sexual orientation, or ticked the ‘other sexual orientation’ box but did not write anything in.”
Michael Biggs, associate professor of sociology at the University of Oxford, said: “This error raises further questions about the competence of the Office for National Statistics and the reliability of the 2021 census.
“Overall, though, the estimates of the lesbian, gay, and bisexual populations appear valid. The problems are much worse when it comes to estimates of the transgender population.
“These come from a very complicated census question which confused many people, especially those who lacked English fluency. The ONS has yet to fully acknowledge the flaws in their transgender data.”
Prof Alice Sullivan, a census expert at University College London, said: “This illustrates the pitfalls of using open-text responses, especially when concepts are not necessarily clearly defined.
“Perhaps the ONS might have just stuck with the protected characteristic because that is, on the whole, what you would think people need to know.
“The way they have developed this question may have given too much weight to activists and not enough towards the actual purpose of the question.”
Overhaul unemployment rate
The latest census blunder comes after Whitehall sources claimed the ONS “hugely overestimated” the number of transgender people in the UK. An official inquiry found that the 260,000 people identified as transgender may need “additional probing” because of “uncertainty”.
In October, the Bank of England also faced a knife-edge call on interest rates after the ONS was forced to overhaul its survey of workers to estimate the unemployment rate, which underpins their decision.
An ONS spokesman said: “The error identified does not change the statistics for the overall LGB+ population. The only data affected are the ‘pansexual’ and ‘all other sexual orientations’ categories of our detailed sexual orientation classification. These are both part of the LGB+ population.
“We’ll produce around 5 billion statistics from Census 2021. This error, affecting responses from 0.1 per cent of the population, was caused by a coding mistake in the early stages of census processing. Because there were no previous statistics on pansexuals, we had no cause to think the original statistics were incorrect.”