International evidence and the risks of reframing the sex question in the census

At the time of writing, the three UK census authorities (National Records of Scotland , Office for National Statistics, and the Northern Ireland Research and Statistics Authority) are planning to provide guidance to accompany the sex question in the next census, which will advise respondents that they can answer based on their self-defined gender identity.
Proposed guidance: ONS and NRS

Proposed guidance: NISRA

The framing of the long-standing sex question in this way, in effect conflating sex and gender identity in the same question, has prompted criticism from leading quantitative social scientists, who have repeatedly raised concerns with the census authorities, in particular, highlighting the risks to data reliability at a sub-group level.
In September 2020, as part of its ongoing accreditation assessment of the census the Office for Statistics Regulation (OSR) wrote to the ONS, questioning their assumption that the impact of the guidance on data quality would be negligible.
‘The assessment team thinks it essential for ONS to consider the concerns raised by users during its further testing and research on the guidance on the sex question, and consider the impact of data quality on the analysis of small sub-groups of the population’.
For the ONS, this task is likely to be challenging, in part because the size and nature of the UK transgender population is unknown. While it is reasonable to expect some variation in terms of sex and age, with higher proportions among younger age-groups, one of the key difficulties is estimating what these proportions might look like.
Population data are however available in some other jurisdictions. For example, a recent large-scale study of adolescents in China, based on data collected in 18 secondary schools, reported that of the 5,590 female respondents, around a quarter did not identify with their sex at birth: 15.4% identified as boys, 2.0% as non-binary, and 8.5% as ‘questioning’. By contrast, among the 6,518 male respondents, only 3.2% identified as girls, 2.1% as non-binary, and 4.9% as ‘questioning’ (Wang et al. 2020).
A 2018 large-scale Swedish study, based on a population-representative sample of 50,157 Stockholm County residents, reported that 2.0% of males and 3.5% of females wanted to live or be treated as someone of a different sex. This proportion increased to 6.3% among the 22 to 29-year old age-group (for both sexes) (Åhs etl al, 2018)
A New Zealand study of 8,166 adolescents, based on a national, cross-sectional, population-based survey of school students, reported a much lower prevalence, with 1.2% of participants self-identifying as transgender, and 2.5% as ‘not sure’. Within the transgender group of respondents, 46% of respondents reported their sex as male, and 54% as female (Clark et al. 2014).
In a study on gender identities and gender dysphoria, based on data from a sexual health study among the general Dutch population, researchers reported that 4.6 % of males and 3.2 % of females reported an ‘ambivalent’ gender identity, and 1.1 % of males and 0.8 % of females identified strongly with the opposite sex (Kuyper and Wijsen, 2013). An earlier study conducted in 2008 among 760 Dutch 11–18 years-olds reported that 5.0% of boys and 8.2 % of girls indicated that they “a bit or sometimes” wished to be of the opposite sex (reported in Kuyper and Wilson, 2013).
Lastly, an older study of 5,010 college students conducted in Taiwan (Lai et al. 2009) found that wanting (very much) to be the opposite sex was more prevalent in female respondents (7.3%), compared to male respondents (1.9%).
Taking an overview of these studies, two observations can be drawn. Firstly, for the most part, among young people the proportion of females reporting as transgender is consistently higher than that of males. This patterning is also evident in referrals to the Gender Identity Development Services (GIDS) in the UK, where the number of girls outstrips that of boys. It is also reflected in the distribution of applications to change legal sex in Belgium, following the introduction of gender self-identification in 2018. As we previously reported, in 2018/19 transmen aged 16 to 24 years accounted for nearly a third of all legal sex change registrations (30%). Among those 16 to 24-year olds registering a change in legal sex, the proportion of transmen was more than double that of transwomen, at 65% and 27% respectively.
Second, and equally striking is the variation in prevalence, both between different countries, and over time, with higher proportions reported in more recent studies. Both these factors make it difficult to estimate what similar figures might look like in the UK.
Against this varied international backdrop, without reliable UK data, and in what appears to be a period of rapid societal change in relation to self-declared gender identity, the proposed conflation of sex and gender identity in the next census should be ringing loud warning bells in the offices of the UK census authorities.