Stage One debate on the Census (Amendment) (Scotland) Bill: MBM briefing



It is planned that the 2021 Census will gather data for the first time on sexual orientation
and on the transgender population, using voluntary questions. The census is the ideal way to gather information on relatively small populations like these. Doing so voluntarily reflects the sensitivity of the information. The Bill deserves support to enable this.

The Scottish Government also proposes to replace the compulsory male/female (M/F) sex
question, which reflects how sex is defined in law, with an “M/F/other” question. The
Government proposes to combine sex as defined in law and gender identity in a single
question. This is because some people regard their sex as defined in law as failing to reflect their gender identity (note: gender identity may also be referred to as lived identity). This proposal is not on the face of the Bill, but would be in regulations, likely early next year.

The Stage 1 Report recommends instead that the census should keep the ideas of sex and
gender identity distinct, and ask separate questions about each. This recommendation
avoids confusing the ideas of sex and gender identity and recognises that both matter. It
would make the most of the opportunity in 2021 to improve our understanding of the
whole population.

The unique value of the census

  • The census is a once-in-a-decade opportunity, with a substantial cost (£64m in
    2011): getting the data right matters.
  • The purpose of the census is to provide robust data for public planning, policy-
    making and research.
  • The census asks everyone, so it allows analysis at a level of detail that sample surveys cannot. It is used by a huge number of organisations to inform their strategic planning and to understand how Scotland works.

What data do we need?

  • Evidence presented at Stage 1 showed that users need data on sex, for instance to
    inform the allocation of resources to health boards, or to examine how far the
    interaction of sex, education and employment has changed over time.
  • This usage is consistent with an extensive body of evidence that shows biological sex remains relevant to individual lived experience, irrespective of wider societal change.
  • Sex-based discrimination is commonplace. For example, there are robust
    explanations on how employment inequality can relate to sex-based discrimination
    on the basis that a woman might get pregnant, as well as actual pregnancy and
    motherhood. The introduction of schemes to address ‘period poverty’ in Scotland
    also underlines the relevance of biological sex.
  • Under the Equality Act 2010, sex (defined as either M or F) is one of nine “protected
    characteristics”. The Act requires public authorities to eliminate unlawful
    discrimination and advance equality of opportunity based specifically on each of the protected characteristics. They need reliable data to monitor this.
  • The Office for National Statistics has already decided that the 2021 Census for England and Wales needs a binary sex question for public bodies to uphold their duties under the Act.
  • International research shows that sex is the census data set most often cross referenced with others.
  • As the numbers declaring a gender identity different from their sex as recognised in law increase, knowing more about their experience also matters.
  • No evidence presented during Stage 1 demonstrated that gender identity has become the only thing which determines people’s lived experience, and is the only relevant concept on which to gather information.
  • The better the data we gather on sex and on gender identity separately, the better we will be able to understand the operation of each separately, and together.
  • Collecting data separately on sex and gender identity will therefore do most to meet the varied requirements of data users.
  • The government should continue collecting reliable data on sex until there is compelling evidence that it is no longer needed.

Data quality

  • Consistent data are needed for longitudinal analysis over time. Data on sex as M or F have been gathered since the first census in 1801.
  • If the ‘sex’ question becomes, in practice, a mixed sex and gender identity question,
    we cannot predict how it would affect the data.
  • Small differences at the whole population level could be significant for particular
    sub-groups, perhaps most obviously by age.
  • Collecting data on those identifying as neither men nor women as part of an
    “M/F/other” question pre-judges that there is nothing to be gained from knowing
    how these respondents as a group are drawn from the M/F categories.

Collecting data on sex and on gender identity

  • Continuing to collect clear data on sex is not in conflict with recognising that for
    some people their sex as defined in law does not reflect their gender identity.
  • Early question testing showed that collecting data separately on sex and gender
    identity could be confusing. However, only one question format has been tested,
    which had an obvious weakness*. NRS intends to undertake further testing of formats for collecting data both on sex and gender identity.

Ethics and Confidentiality

  • Access to census data is strictly controlled. It is a criminal offence to unlawfully
    disclose census data: a person may be fined by up to £10,000, sent to prison for up
    to two years or both.
  • For those who have a Gender Recognition Certificate, there may be specific privacy
    issues in asking for the sex shown on their original birth certificate. For that group, which is known to be very small, this can be dealt with by only collecting data on sex as shown on any re-issued birth certificate.
  • It has been argued that asking for a declaration of M/F sex is unethical because it infringes respect and dignity for those who identify as neither a man or woman. That argument must extend to all those who do not identify with their sex in law.
  • The implications of accepting that ethical argument go far beyond the census because it is likely to be widely cited as a precedent in other contexts.
  • Any such precedent could affect the practical operation and monitoring of the Equality Act 2010 in relation to sex as a protected characteristic, because the Act relies on organisations being able to ask about and gather data on legal sex.
  • Lawmakers therefore need to consider very carefully the precedent that would be set by accepting the argument that asking for data on legal sex is of itself unethical.
  • None of the 47 countries covered by the European Convention of Human Rights as yet collect data on gender identity in their census (ONS, 2019 para. 5.5). There is therefore no indication that using a census to gather data on sex in its legally-recognised binary form is regarded as raising fundamental issues of human rights.


The Census (Amendment) (Scotland) Bill has tapped into a much broader philosophical debate about how sex should be defined, legally, socially, statistically, institutionally and medically. It is a debate raising issues of substantial public interest, which the Parliament
has had little opportunity to consider. Given the sensitivities and potential implications, we think the census is not the right place to pre-empt this debate and/or to set legal precedents.

The debate here is not about whether to give people the opportunity to declare their gender identity: it is about whether also to continue collecting data on sex consistent with its existing legal definition, in its own right. Capturing both types of information would tell us the most about the population as a whole and meet the purpose of the census as a data-gathering tool for governments, policy-makers and researchers.

The Committee’s preferred approach recognises that both sex and gender identity may be relevant to people‘s lived experiences. In a debate which can feel characterised by uncompromising either/or choices, this offers a pragmatic solution which acknowledges
multiple perspectives.

*It asked all respondents if they were male or female and, straight after, if they identified as a man/a woman or in another way. There has been no testing of an approach which asks everyone their M/F sex, but only asks about gender/lived identity as part of the voluntary transgender question.