Losing count: consultation on the sex question in the 2021 census

In June 2019 the Scottish Parliament passed the Census (Amendment) (Scotland) Bill, with the effect of introducing new voluntary questions on gender self-identity and sexual orientation in the 2021 census.

The detailed census questions will be agreed by the Scottish Parliament in a separate process that is now underway. In relation to the sex question, parliamentarians are considering how the accompanying guidance to the sex question should be framed;  whether this should based on respondent’s biological sex, their self-identified sex, or, if there should be no guidance at all. The Scottish Government’s current preferred option is to provide guidance that allows a person to answer the sex question in terms of their self-identified gender. This blog considers looks at the consultation process behind this decision.

Question development and stakeholder engagement
Some details of the early engagement process on the sex question are provided in a Freedom of Information response from the National Records of Scotland. This shows that in the formative question development phase, during which the Scottish Government established its preferred option, officials met only with LGBT stakeholders (Equality Network/Scottish Trans Alliance, Stonewall Scotland, LGBT Health). The FOI response indicates that no meetings were held with independent statisticians, census data users, or with other right-holders affected by a loss of data on sex prior to the introduction of the draft Bill. 

Steering the definition of sex away from the legal definition in the Equality Act 2010, key decisions made at this crucial stage included the introduction of a non-binary response option for the sex question (since dropped, on the recommendation of the Culture, Tourism, Europe and External Affairs Committee), and allowing respondents to answers on the basis of their self-identified gender (which remains under consideration)

Minutes from the Scottish Government’s Board of Official Statistics similarly indicate that the Board had no input or discussion on the 2021 Census prior to the Bill being laid in Parliament. This omission is particularly striking given that part of the Board’s remit is ‘to make sure the voices of those using data and statistics about Scotland are heard and influence the way the strategy is developed and delivered’. Ten days after the introduction of the Bill, on 12 October 2018, NRS updated the Board on its plans, although it is not clear in what detail, or if the proposed changes to the sex question were discussed.  

Introduction of the draft Bill   
Since the publication of the draft Census (Amendment) (Scotland) Bill on 2 October 2018, the proposal to put the sex question on a self-identification basis has prompted input from a wide range of stakeholders, including groups representing women’s rights, statisticians and census data users, as well as those groups involved in the initial question development period. Different views have played out in exchanges on social media, through correspondence with parliamentarians, and in the press. 

As as result, the initial imbalance in stakeholder engagement has shifted, as the implications and risks of formally putting the sex question on a self-identification basis have become more widely known. 

Challenging the preferred Scottish Government position, independent statisticians and census data users have raised concerns in relation to the impact on data quality and reliability, and the ability to monitor and address sex-based discrimination. Stakeholders, ourselves included, have raised concerns about the redefinition of sex within the context of Scotland’s most high-profile population survey, what this means for the operation of the Equality Act 2010, and for other data collection exercises. In relation to latter, it appears that the self-identification proposal is already being considered in other major surveys:    

Policy capture 
As we have argued elsewhere, that such a fundamental conceptual change in the definition of sex was developed with reference to only a limited group of interests, without engaging independent statisticians and census data users, or those representing other affected rights-holders, raises serious questions about the policy process.

These dynamics are not new. While those supporting a self-identified sex question cite the introduction of self-identification guidance in the 2011 census as a precedent, it is important to note that the policy process in 2011 was even more opaque. For example the introduction of gender self-identification guidance is not referenced in the 2011 Equality Impact Assessment, nor are we aware of any wider consultation or engagement with census data users or other affected groups.

In practice, this approach has resulted in an impasse. Having failed to fully explore the risks at the question development stage or consider different perspectives, officials now appear stuck as how to resolve the conflicting views, and have struggled to defend the proposed departure from the Equality Act to CTEEA Committee Members. As SNP MSP Annabelle Ewing commented at a recent Committee meeting: “A lot of this discussion could have been streamlined if the NRS had had early engagement with the statisticians” (SP CTEEA Committee col. 13. 12 Septem


Rebecca Bull (@elletorrito) September 16, 2019

Balancing conflicting rights and interests
To break the current impasse, we think that the Scottish Government firstly needs to engage more fully with census data users with a background in data analysis, researchers using population data (including health research), and independent statisticians, and put the purpose of the census at the forefront of its deliberations: that is, as a data collection exercise that provides the information needed to develop policies, plan and run public services, and to allocate funding. At the same time, the Scottish Government needs to ensure that the interests of all affected rights-holders are balanced fairly and respectfully, in the least discriminatory way possible.

Losing count
Our view is that a sex question based on a person’s current birth certificate, and a separate gender identity question provides this balance. Alternatively, the census authorities could revert to having no guidance at all, as was the case in all censuses between 1801 and 2001. Whichever direction is taken, we would suggest that the census is not the appropriate vehicle to either preempt legal reform on gender self-identification, or to set legal precedents. Above all, we would caution that if we lose count of sex as a biological category, we will also lose our ability to chart and address sex-based discrimination.