Last Updated on August 9, 2025 by Karl Thompson
The Office for National Statistics has downgraded its gender identity statistics from the 2021 UK Census. These statistics are no longer ‘official statistics’, they are now just ‘official statistics in development’.
According to the 2021 Census 262 000 people in England and Wales identified as transgender, equivalent to 0.5% of the population.
However, new analysis by Oxford Sociology professor Michael Biggs suggests these statistics may lack validity. His research was recently published in the journal British Sociological Association’s journal Sociology.
Why might the government’s gender identity statistics lack validity…?
The main reason is that some people with poor English may have misunderstood the following question:
“Is the gender you identify with the same as your sex registered at birth”?
Professor Biggs’ analysis suggests that people with poor English skills are more likely to have answered ‘no’ to this by mistake.
Unusual regional variations
Professor Biggs looked at how the number of reported trans people varied across regions in England and Wales.
Many areas in the top 20 are not well-known trans-friendly areas, but many are well-known as having high proportions of people with poor English.
For example, Newham, Brent, Tower Hamlets, it does seem unusual that these have higher proportions of trans identifying people in them compared to Brighton and Hove, for example.
Professor Biggs thus went on to analyze the relationship between those identifying as transgender and their English language proficiency . His statistical analysis is summarized below:
- 0.4% of the population who spoke English as their main language identified as transgender
- 1.5% with English not as their main language identified as transgender
- More than 2% of people who didn’t speak English well identified as transgender.
Professor Biggs also looked at the type of trans identity in relation to proficiency in English.
Higher proportions of people with poor English skills identified as being a biological male and identifying as a trans male, or a biological female and identifying as a trans female, or as unspecified gender*.
Professor Biggs takes this as being evidence of people being ‘confused’ about their gender identity. However this is just his interpretation. It is possible to be biologically male and identify as a trans male, for example.
Poor reliability compared to other sources
Professor Biggs further points out that the statistics in the Census on transgender identity are different to other sources.
He cites data from from a petition put together by Reform the Gender Recognition Act, 2021 which had 118 000 signatories, as well as the number of referrals to the Gender Identity Development Services.
According to the former data source 0.75% of people from Brighton and Hove identified as being transgender, compared to only 0.16% of people in Newham. The former is broadly in line with the Census findings, the later is grossly different.
So it seems reasonable to suggest that lack of English is distorting the results on reported transgender identity in the Census
Relevance to A-level sociology
This is an excellent debate that raises serious issues about value freedom. It is important to keep in mind that Professor Biggs has been accused of being transphobic.
However Professor Biggs himself argues that the ONS have failed to be objective themselves. He thinks they have bowed to pressure from LBGTQ pressure groups and delayed downgrading their statistics.
You will need to read through Professor Biggs’ research yourself and decide whether you think this is truly objective analysis!
The ONS certainly seems to think his criticism of their statistics has some merit, as they have now downgraded them.
Sources
Biggs, M (2024) Gender Identity in the 2021 Census of England and Wales: How a Flawed Question Created Spurious Data, published in Sociology.