Watchdog launches probe amid concern ONS figures on the transgender population have been ‘substantially’ overestimated
- ONS figures said there are 262,000 transgender people in England and Wales
The statistics regulator yesterday launched a formal probe into fears that the transgender population has been ‘substantially’ overestimated.
The Office for Statistics Regulation (OSR) said it was reviewing findings from the Office for National Statistics (ONS) suggesting that there are 262,000 transgender people in England and Wales.
In a letter to Jen Woolford, the ONS’s director of population statistics, OSR boss Ed Humpherson said it was launching the review after ‘concerns being raised with us regarding the correlation of gender identity with other characteristics’.
It will consider whether the ONS breached the Code of Practice for Statistics.
Academics believe the 262,000 figure is wrong because the question used to record gender identity in the 2021 census may have confused respondents whose main language was not English.
The Office for Statistics Regulation (OSR) said it was reviewing findings from the Office for National Statistics (ONS) over fears that the transgender population has been ‘substantially’ overestimated
OSR boss Ed Humpherson (pictured) said the review was launched after ‘concerns being raised with us regarding the correlation of gender identity with other characteristics’
They say that this likely explains why the ONS found that one in 67 Muslims said they were transgender, and why the London boroughs of Newham and Brent, where many speak English as a second language, recorded the highest proportion of trans people.
MPs and campaigners believe that the apparent over-estimation may have led to an ‘exaggerated’ focus on transgender issues.
Professor Carl Heneghan, director of Oxford University’s Centre for Evidence-Based Medicine, said: ‘I think the information in this category is invalid so it’s welcome that they’re going to review it.
‘I think this review will say that, actually, there was a substantial problem with the question when it came to people whose main language was not English.
‘I think the problem here is that they clearly haven’t road-tested the question well enough with people whose first language was not English. They need to set out a strategy so we don’t see this problem again. The error is so great that I think it will be deemed invalid.’
He added: ‘It’s an important lesson for the ONS to learn because this information is essential to how we plan future policies and services.’
He said an ‘overhaul’ of ONS processes, such as getting data peer-reviewed, was needed.
Professor Alice Sullivan, head of research at the UCL Social Research Institute, said: ‘I strongly welcome this intervention from the Office for Statistics Regulation.
Professor Carl Heneghan (pictured), director of Oxford University’s Centre for Evidence-Based Medicine, welcomed the review
MPs and campaigners believe that the apparent over-estimation may have led to an ‘exaggerated’ focus on transgender issues
‘The Code of Practice for official statistics states that, ‘Quality means that statistics fit their intended uses, are based on appropriate data and methods, and are not materially misleading.’
‘The anomalies in the 2021 census data call into question whether the data produced on gender identity meet this standard.’
The ONS published figures in January showing that 262,000 people, or 0.5 per cent of the population over 16, declared themselves as transgender.
It was based on answers to the 2021 census, which explored gender identity for the first time.
Of the 262,000, 118,000 did not provide further detail, while 48,000, or 0.1 per cent of the population aged 16 and over, identified as a trans man.
A further 48,000 identified as a trans woman, around 30,000 as non-binary and 18,000 people wrote a different identity.
Those who speak English poorly were five times more likely to be trans.
Adults whose first language was not English accounted for 10 per cent of the population but made up 29 per cent of the trans total.
Academics believe the apparently flawed figures likely stemmed from the question asked.
Rather than keeping it simple with a question such as ‘Are you transgender?’, respondents were asked: ‘Is the gender you identify with the same as your sex registered at birth?’
Tory MP Sir John Hayes said: ‘Statistically, this appears to be an open and shut case that reveals a glitch.
Tory MP Sir John Hayes (pictured) suggested it was ‘an open and shut case that reveals a glitch’
‘The problem is, when you look at the whole trans debate you might assume that the trans community is enormous. But in fact it’s a tiny number of people.’
Fellow Conservative Paul Bristow said: ‘It certainly raises a few eyebrows, having a larger trans community in Brent or Newham than in places like Brighton.
‘That’s why we need to ensure every care is taken to ensure that these surveys are as thoughtful as possible to communities that might not have English as a first language.
‘Official statistics are incredibly important because they help government and agencies design services.
‘If they are proven to be wholly inaccurate, this could mean policies being created and public money being spent in ways which do not necessarily reflect local communities and wider society.’
An ONS spokesman said: ‘We talked to a number of groups and engaged in wide-scale testing before the questionnaire was finalised for Census 2021.
‘This included running a census rehearsal, in which over 100,000 households took part, in diverse areas around England and Wales including areas with high numbers of second language English speakers.’
Source: Read Full Article