Was the sex question in the 2011 census ‘self-identified’?

1.   Introduction

Less than six weeks before the date of the 2021 census in England and Wales, the Office for National Statistics (ONS) published the guidance which will accompany the compulsory sex question. The guidance is of critical importance because it defines what is meant by ‘sex’ in the census. The finalised guidance is shown below (emphasis added).

Your answer is key to understanding trends in the population. It also helps your local community by allowing charities, organisations, and local and central government to understand what services people might need.

This information will be used for equality monitoring between groups of people of different sexes in your local area. Your answer also helps public bodies to identify discrimination or social exclusion based on sex, and work to stop it from happening.

The sex question has been asked since 1801.

This question is vital for understanding population growth and equality monitoring. Please select either “Female” or “Male”.

If you are considering how to answer, use the sex recorded on one of your legal documents such as a birth certificate, Gender Recognition Certificate, or passport.

If you are aged 16 years or over, there is a later voluntary question on gender identity. This asks if the gender you identify with is different from your sex registered at birth. If it is different, you can then record your gender identity.’

The guidance provided by ONS means that the definition of sex in the next census will conflate three separate demographic categories. These are biological sex, legal sex (for those who have acquired a Gender Recognition Certificate and who may or may not have changed the sex marker on their birth certificate), and self-declared gender identity or gender reassignment (for those who do not have a GRC, but have changed the sex marker on their passport or other unspecified legal documents). Due to the voluntary nature of the separate question on gender identity, it will not be possible to disentangle these categories.

To justify its approach, the ONS has published a methodological paper that outlines how it reached its final decision. Among its rationale for the current guidance, ONS cite continuity with previous datasets, principally the 2011 census, which ONS interpret as having gathered data based on ‘self-identified’ sex on the basis that similar guidance was provided that year. ONS also interpret the sex data in the 2001 census (where guidance was available on request), as ‘self-identified’, and all censuses prior to that.

[Sex] is interpreted as self-identified sex both in 2011 and before, although in 2001 guidance to that effect was only available on request, and prior to that no guidance was provided’.


Indeed, ONS characterise all surveys that ask if a person is male or female, without accompanying guidance, as capturing ‘living/presenting’ sex (p.14)

This blog suggests that this interpretation is not borne out by the historical context, and is indicative of the confusion that has been sown around what is an ordinary and easily understood concept.

2.   How did we get here?

The introduction of ‘self-identification’ guidance in the census can be traced to the late 1990s, and a series of policy decisions that took place without proper consultation or scrutiny.

In 1998, a campaigner from the trans advocacy group Press for Change wrote to the ONS concerning completion of the census form in the forthcoming census. ONS responded in the following terms, stating that the individual could reply using the sex that they ‘believed to be correct’:

Should there be the traditional question on the individual’s Sex in the 2001 Census, it would be reasonable for you to respond by ticking either the ’Male’ or ’Female’ box whichever you believe to be correct, irrespective of the details recorded on your birth certificate’.

(Appendix 1)

Drawing on this exchange, in February 1999 Press for Change announced on their website that trans people could ‘now give an answer which is both truthful and legal’

For most people, the ten-yearly census is just another form to fill in … but for thousands of trans people around the country, it raises yet again the question of how to legally and truthfully answer the question about sex. Because UK law does not fully recognise trans people in their true gender, the possibility arose that once again, it would be impossible to provide a truthful answer which also matched legal requirements. There is a legal obligation to complete the census form accurately, so trans people faced an unpleasant dilemma.

However, this letter from the Office of National Statistics (in response to an enquiry from Press For Change campaigner Paula Thomas) makes it clear that trans people can now give an answer which is both truthful and legal:

Trans women should record their sex as female if they believe that to be correct, even though their birth certificates will still say ’male’

Trans men should record their sex as male if they believe that to be correct, even though their birth certificates will still say ’female’.’

(Appendix 1)

In 2008 the ONS commissioned an independent consultancy, Diversity Solutions, to undertake an Equality Impact Assessment Screening for the development of the ‘sex/gender’ question in the 2011 census. As part of the screening, the assessors consulted with one individual: the same Press for Change campaigner who previously contacted the ONS in 1998, whose view is shown below:

For this equality impact assessment screening process, consultation has been conducted with Paula Thomas, a trans community representative who was asked to give her expert view. Ms. Thomas also made a relevant enquiry of ONS related to the 2001 Census. Her current view is as follows:

“Where no Gender Recognition Certificate has been issued: Clear guidance is required here. For example, it should be made clear that someone who believes herself to be female, despite having a male birth certificate and no Gender Recognition Certificate, can tick the female box without fear of prosecution under the Perjury Act.”’


The screening report stated that ONS should consult with trans community groups such as the Gender Trust and Press for Change, and that it was ‘essential’ to consult with Mermaids, but made no wider recommendation.

Crucially, the report erroneously concluded that that ONS had a statutory duty to provide self-identification guidance for the sex question:

To comply with their statutory duty to promote gender equality, ONS must issue clear guidance to trans people, including those who do not have a Gender Recognition Certificate so that they, or others completing the census form on their behalf, are enabled to give accurate answers to the question.


The screening assessment was signed off by the census Deputy Director in late 2008, with the following agreed actions:

  • Provide guidance to trans people, including those who do not have a Gender Recognition Certificate so that they, or others completing the census questionnaire on their behalf, are enabled to give accurate answers to the question asking ‘what is your sex?’
  • Consult on this guidance with trans community group

Both ONS and the National Records of Scotland subsequently introduced self-identification guidance in the 2011 census. Consistent with recommendation in the screening report, there is no evidence of any wider consultation beyond the trans community, nor additional layers of scrutiny.

3.   Can the sex question in the 2011 census be described as ‘self-identified’?

In making the case for a similar approach in the 2021 census, the ONS routinely depict the sex question in the 2011 census as having collected data on ‘self-identified sex’, and argue that there should be consistency between the two datasets.

There is, however, no evidence to suggest any meaningful public awareness of the change made in 2011. To the best of our knowledge, ONS made no announcement as to its introduction of the guidance. The guidance was available online but not linked to the online questionnaire. Moreover, most people completed a paper questionnaire in 2011, with only 16% of returns in England and Wales submitted online (House of Commons Library 2013: 28). The guidance was also available via the telephone helpline, although as the ONS paper notes, this was not widely used. This means that for both online and paper questionnaires, respondents had to purposively seek out the guidance. In Scotland’s 2011 census, where similar guidance was also made available, it appears that even Stonewall, the UK’s largest LGBT advocacy group, was unaware that such guidance existed (CTEEA Official Report 12 September 2019 col. 6).

4.   Self-reported or self-identified?

The ONS interpret all previous census sex data as relating to respondents’ ‘self-identified’ sex. We think this approach is conceptually mistaken, and that it is historically implausible to characterise all census data, as far back as 1841, as based on people’s ‘self-identified’ sex.

ONS confuse the common practice of asking people to report a characteristic without providing evidence (‘self-report’), and actively guiding people to consider a characteristic as having no definition beyond the personal and subjective (‘self-identified’).

In practice, biological sex is easily recognised and understood. In numerous species, including homo sapiens, it simply means the state of being male or female for reproductive purposes. For almost everyone, sex is observed accurately at birth from outward appearance.

In recent years however, an extraordinarily unhelpful degree of confusion has been sown around the ordinary concept of sex, as exemplified in the five ‘concepts’ proposed by the ONS. There is, however, no evidence to suggest that ideas about sex, as defined in terms of subjective self-identity, extend to previous generations as far back as 1841, in terms of common public understanding.

5.   Discontinuity in the 2021 census

Contrary to the view of ONS, we think that the most significant break in continuity is likely to occur in 2021, as a result of the proposed guidance. This is because, in contrast to 2011:

  • It is anticipated that around 70% of returns in 2021 will be made online.
  • The guidance will be directly linked to the online questionnaire and therefore readily accessible
  • As a result of media coverage, academic engagement and engagement by advocacy groups, there is much greater public awareness of the census guidance, compared to the 2011 census.
  • There is persistent international evidence of sharp growth in the numbers of people identifying as transgender, particularly among younger age groups.