What’s Wrong with Sex and Gender Identity in the 2021 Census for England and Wales – Part I

Giovanna Gilleri is a PhD Researcher in International Human Rights Law at the European University Institute (EUI), Florence, Italy.

Photo credit: Mačka

The relation between sex and gender identity is intricate. The definition and the interpretation of these concepts are at the centre of a long-lasting debate in legal, psychological, sociological, and philosophical circles. Indeed, sex and gender identity are key components for the identification of the challenges faced by different individuals throughout their lives, and the human rights breaches they suffer. The meanings attributed to sex and gender (identity) can have significant practical consequences. A paradigmatic negative example of this is the 2021 census for England and Wales. Here the question on sex is mandatory, while that on gender identity is voluntary. What are the assumptions underpinning the formulation of these questions? What are the definitions of ‘sex’ and ‘gender identity’ adopted in the census? What is the effect of making the sex question compulsory while the gender identity question voluntary? The understanding of sex and gender identity underlying the design of the census produces at least two major effects: the flawed configuration of sex/gender as binary and immutable, and the false idea that only trans people have a gender identity.

The questions on sex and gender identity in the 2021 census

On 21 March 2021, the Office for National Statistics (ONS) will launch the 2021 census for England and Wales.

The census takes place every ten years, providing useful informationfor policy-making at the level of government and local authorities. The ONS will administer the 2021 census with the objective of collecting demographic information from all households and communal establishments. For the first time, the census will be carried out primarily online. This is not a minor detail in times of a pandemic. Scotland’s 2021 census has been postponed to March 2022, indeed. If the census appears as a forerunner for the digitalisation of the data-gathering method, the same cannot be said for the way in which it combines the variables of sex and gender identity. This is based on the inaccurate ideas that sex and gender are binary and fixed, and that gender identity is a specificity of trans people only.

For the 2021 census, the ONS has separated two questions on sex and gender identity to address the absence of data on trans people in the official figures, thereby capturing the number of individuals identifying their gender as different from the sex assigned at birth. The UK government acknowledged during the consultation (p. 21) on the Gender Recognition Act that there is no robust estimate of the composition and size of the trans population in the UK, due to the difficulties in identifying and defining (p. 81) trans individuals. The lack of accurate data prevents public bodies from properly addressing trans people’s needs and securing their rights. Identifying social exclusion and reducing discrimination against trans individuals, especially improving their access to health services (p. 9), are the major policy objectives underlying the sex and gender identity questions. In the light of this, the proposed questions for sex and gender identity are:

• ‘What is your sex?’ with ‘male’/ ‘female’ answer and the specification that ‘A question about gender identity will follow later on in the questionnaire’;
• ‘Is the gender you identify with the same as your sex registered at birth?’ with yes/no answer and the specification that ‘this question is voluntary.’

In the Gender Identity Topic Report (p. 4), the ONS defines gender identity as ‘a person’s personal internal perception of themselves,’ by stressing that ‘in contrast, sex is biologically determined.’ Yet the concepts of sex, gender, gender identity, and their interrelation are more intricate than that.

Why sex and gender matter in question design

Under international human rights law, understanding what sex is, what gender is, and how these notions interact is key to assessing whether and to what extent state and non-state actors, such as private individuals, companies, investors, religious groups, and other collective entities, recognise the existence of the rights of  certain people and subsequently respect them. I have argued elsewhere that international human rights law-makers have not provided so far consistent interpretations of the legal categories of sex, gender, and their interrelationship. The lack of a clear conceptual and legal framework is one of the reasons for the absence of a comprehensive sex/gender approach to certain violations in international human rights law, such as unnecessary medical interventions performed upon intersex and trans (pp. 295-297) people. The sex/gender-based breaches and the discriminatory grounds risk to remain covered by the unawareness about the number of differently sexed/gendered individuals. Where properly designed, a census is the preliminary step to identify that a specific group of people exist, so that the violations committed against them come to the surface and receive a specific subjective connotation.

Embracing one specific understanding of sex/gender in the design of indicators (p. 175) such as the 2021 census determines how a certain group become counted, and eventually counts for the formulation of laws and policies relying on the collected information. For example, data on health, housing, education, and employment may shed light on the extent and nature of disparities and disadvantage between and within various groups. The policies and laws adopted at national and subnational levels should, in turn, comply with international human rights standards in the realm of, particularly, sex/gender-related protections. This entails that issues of definition and interpretation of concepts arise both in the international and in the domestic spheres. The methodological choices about the definitions of sex/gender underpinning national surveys impact the (in)visibility of certain sexed/gendered subjects, as the methodology of the 2021 census shows.

The 2021 census requires respondents to declare their sex (sex question) and, if they wish, whether they identify their gender identity with the sex assigned at birth (gender identity question). The former should be answered relying on the birth certificate or on the gender recognition certificate in case of respondents who have legally changed gender. Ten years ago, the 2011 census also contained a sex question (‘what is your sex?’) with ‘male’ or ‘female’ as possible answers. However, unlike the 2021 census, the earlier census allowed individuals to answer according to the gender they felt themselves to be. A note accompanied the question: ‘You can select either “male” or “female”, whichever you believe is correct.’ One could select female or male based on their self-identification, regardless of the details contained in the birth certificate and without the need to have a gender recognition certificate.

For trans advocates, this change is a setback in the human rights record since it obscures trans existences. The 2011 census allowed respondents to provide an answer according to their inner sense of the self rather than the determination made by others on their birth certificates. The 2021 census also envisages the sex question and the gender identity question. But the way in which the census is structured makes the legal sex question outweigh the gender identity component.

The ONS believes (pp. 4-5) that this specific structure will allow statisticians to calculate the proportion of trans people to the overall population. Drawing on a 2016 workshop, the ONS has stressed that gathering information on sex and gender identity would be useful to detect discrimination and to inform policy. Data would: ‘assist with allocation of funding and specialist services; provide education and awareness of gender identity; [and] be beneficial to have official, reliable data about the size of the trans population’. If this objective is commendable in theory, the method used is inadequate in practice.

In fact, the architecture of the 2021 census draws a clear line not only between trans and cisgender people, but also within the trans community. Who will ultimately answer the gender identity question? Most probably, the respondents will not be cisgender people, who identify themselves with the sex/gender indicated in the birth certificate, but trans people. A large number of trans people will have to answer both the sex question according to the sex/gender assigned at birth and the gender identity question to specify they do not identify with that determination. The census can be a powerful tool for trans rights advocates, also for those trans respondents who have taken formal measures to legally change their birth certificate and identify themselves with the gender of the gender recognition certificate. For instance, answering the gender identity question could be instrumental to make sure that trans livesare counted, and count for policy-making.

The lack of data on trans people might lead to conclude that a flawed census is better than nothing. Some inaccurate data would probably be better than no accurate data for the purposes of one-step-at-a-time advancement. But the configuration of the 2021 census has significant implications in practice.

Does the ONS know that we all have a gender identity?

 ‘What is your sex?’, the sex question asks, by providing two options: ‘select either “female” or “male.”’ The second paragraph specifies that: ‘If you are one or more of non-binary, transgender, have variations of sex characteristics, sometimes also known as intersex, the answer you give can be different from what is on your birth certificate.’ The question mixes sex characteristics and gender identity, by considering intersex among ‘transgender’ and ‘non-binary’ gender identities.

The fact of having specific (inter)sex traits does not determine the individual’s gender identity as ‘intersex.’ Intersex is not necessarily a gender identity: intersex persons can have as various a range of gender identities as non-intersex people. While it is true that intersex may be a gender identity, it remains nevertheless unclear how a person with sex characteristics which do not fit the typical male and female binary could provide an answer different from ‘male’ or ‘female’ to the sex question. ‘Male’ and ‘female’ are the only options available, even where the second paragraph seems to open the question to ‘other’ answers beyond the binary. For the ONS, sex is binary while the gender identity question is presented as a way out – a device to solve all the issues arising from the binary structure of the sex question itself.        

The main distorting effect of the 2021 census stems from the fact that the respondents to the optional question on gender identity will most likely be those who do not identify themselves with the sex/gender written in their birth certificates. In so doing, the ONS approach conveys the inaccurate idea that only trans people have a gender identity. But we all have a gender identity, regardless of the gender we identify ourselves with. Further, the designers of the 2021 census link the binary answer to the sex question to the gender identity question. The issue of gender non-conforming identities is reduced to one of binary trans identities. By reflecting the dominant conception of binary sex (male/female) and gender (woman/man), the 2021 census limits the possibilities of clear representation of non-binary respondents.

Please continue to read the second part of the article.