Gender Identity Data Harmonisation Guidance
The Government Statistical Service (GSS) Harmonisation team are currently working to develop a new gender identity data harmonised standard. This is in line with the current GSS Harmonisation workplan, which was updated in December 2024.
This page provides interim guidance for collecting data on gender identity. You can find more information on the collection of data relating to sex on the sex harmonisation guidance webpage.
If you would like to be involved with this work, please contact us at outreach.engagement.community@ons.gov.uk
Policy details
| Metadata item | Details |
|---|---|
| Publication date: | 11 December 2024 |
| Owner: | GSS Harmonisation Team |
| Who this is for: | Users and producers of statistics |
| Type: | Harmonisation standards and guidance |
| Contact: | Harmonisation@statistics.gov.uk |
What we mean by harmonisation
Harmonisation is the process of making statistics and data more comparable, consistent, and coherent by using consistent definitions and questions in data collection. The GSS Harmonisation Team works across the UK to promote and facilitate harmonisation. We provide guidance encouraging the use of standard data collection questions across government departments and the wider public sector, where these are appropriate to meet user needs.
.
Who this guidance is for
This interim guidance is for producers of official statistics and researchers who are currently using, or thinking about including, a gender identity question on their surveys or in their administrative data collection. It will be updated with new guidance as Office for National Statistics (ONS) research on the uncertainties associated with the England and Wales Census 2021 gender identity data, and GSS work to develop a finalised harmonised standard, progress.
.
Measuring gender identity
Measuring gender identity is an evolving topic. On 5 September 2024, Emma Rourke, Deputy National Statistician, wrote to Ed Humpherson, Head of the Office for Statistics Regulation (OSR), to request that the gender identity estimates from Census 2021 are no longer accredited official statistics and are classified as official statistics in development. The change in designation was confirmed by OSR on 12 September 2024. The ONS requested this change in designation after patterns were identified in the Census data that suggested that some respondents may not have interpreted the Census 2021 gender identity question as intended, notably those with lower levels of English language proficiency. More information can be found in the sexual orientation and gender identity quality information for Census 2021, and in the National Statistical blog about the strengths and limitations of gender identity statistics.
The GSS Harmonisation team are now undertaking new work to develop best practice harmonised standards for sex and gender identity data collection. More information about this is provided in our GSS Harmonisation workplan. ONS is undertaking further research to better understand the limitations of the Census 2021 gender identity data, to support users to understand the uncertainty associated with this question. We have therefore archived our previous gender identity guidance, reflecting the new work being done on this topic. This interim guidance page has been created to replace our previous guidance about gender identity. It will continue to be updated as further ONS research is published on sub-national and other smaller group breakdowns of the Census 2021 data.
As there is currently no finalised harmonised standard for collecting gender identity data, this guidance sets out the gender identity questions used in Censuses across the UK, alongside additional information and considerations you should take when considering how to ask about this topic in your data collection.
.
Who this guidance is for
This interim guidance is for producers of official statistics and researchers who are currently using, or thinking about including, a gender identity question on their surveys or in their administrative data collection. It will be updated with new guidance as ONS research on the uncertainties associated with the England and Wales Census 2021 data, and GSS work to develop a finalised harmonised standard, progress.
.
Implementing a gender identity question
Which question to use
There is no agreed best practice for collecting gender identity data yet. The most recent way of collecting data in England and Wales was through Census 2021. Scotland’s Census 2022 asked a question on trans status and history. Some differences in the data collected by each method are expected, reflecting different question wordings. However, these sources provide broadly consistent estimates at a national level. If you are looking to include a question on gender identity or trans status in your data collection, where possible, we would recommend waiting for our work to develop new harmonised standards to conclude.
However, if you are unable to wait for this work to be completed and have a clear need to collect this data, you should consider:
- whether you are looking for data that is comparable with Census outputs for England and Wales, or Scotland; and
- what your individual data requirements are, and whether the England and Wales Census 2021 question or the Scotland Census 2022 question aligns better with these.
The OSR has provided guidance to support official statistics producers in applying the principles of the Code of Practice for Statistics when making decisions about the collection and reporting of statistics and data about sex and gender identity.
ONS is undertaking further research to better understand the uncertainties associated with Census 2021 gender identity data for subnational breakdowns and smaller groups. Updates on the implications of their findings for users of harmonisation guidance will be provided on this page in 2025.
.
Measuring gender identity in the UK censuses
England and Wales
For the first time, Census 2021 for England and Wales included a voluntary question on gender identity. This question was included following in depth research into existing data on gender identity and sexual orientation, conducted by the ONS. This question was developed for respondents aged 16 and over. The ONS has a published sex and gender identity question development report describing how this question was researched, developed and tested.
We recommend you consider the OSR’s review of statistics on gender identity based on data collected as part of the 2021 England and Wales Census before implementing this question in your survey.
The question is available in both English and Welsh. Within Census 2021 it was a voluntary question for those aged 16 and over.
English language self-complete question
Question
Is the gender you identify with the same as your sex registered at birth?
Response options
- Yes
- No, enter gender identity
- Prefer not to say
Welsh language self-complete question
Question
Ydy’r rhywedd rydych chi’n uniaethu ag ef yr un peth â’r rhyw a gofrestrwyd pan gawsoch chi eich geni?
Response options
- Ydy
- Nac ydy, nodwch eich hunaniaeth o ran rhywedd
- Mae’n well gen i beidio â dweud
Most surveys and administrative collections of data are voluntary. It is good practice to tell people at the start of the data collection that they do not need to answer anything they do not want to answer. This means respondents should be able to skip the gender identity question if they do not wish to answer it. When used in the England and Wales Census 2021 this was addressed by providing an instruction to respondents that “This question is voluntary”.
The gender identity question can be sensitive for some respondents and maintaining the privacy of respondents is critical. This means that we recommend being explicit that people do not have to answer the question. This can be done by providing an instruction to respondents as was done in Census 2021, or by offering the “Prefer not to say” option. This is in addition to the implicit option of refusing to answer the question. This is consistent with good practice in data collection for sensitive topics.
Considerations if you are currently using this question
We recommend that you do not make significant changes to existing gender identity data collection unless you have identified strong user need to make these changes. This is because measuring gender identity is an evolving topic. A best practice approach for gathering gender identity data has not yet been established across the statistical sector. The voluntary gender identity question on Census 2021 for England and Wales was the first of its kind. The different approaches used in the UK’s most recent censuses have provided a unique opportunity to learn more about gender identity and the best approach to collecting high quality data on this topic.
The GSS Harmonisation team are undertaking priority work to develop best practice harmonised standards for sex and gender identity data. More information about this is provided in our updated GSS Harmonisation workplan. We intend to publish our research findings for the development of updated sex and gender identity harmonised standards for the online mode in Autumn 2026. We recommend statistics producers review this guidance information once it is published.
As of 12 September 2024, the gender identity estimates from Census 2021 for England and Wales collected using this question wording are no longer accredited official statistics and are classified as official statistics in development. In November 2023 an ONS report on the quality of Census 2021 gender identity data noted that some respondents may not have interpreted the Census 2021 gender identity question as intended, notably those with lower levels of English language proficiency. Analysis of Scotland’s Census 2022 trans status or history data has added weight to this observation. Publishing these data has provided new insights, prompting valuable discussion about the strengths and limitations of the Census 2021 question and data originating from it. While both the England and Wales Census 2021 and Scotland’s Census 2022 estimates are broadly consistent at a national level, this bias will have greater bearing on estimates at lower geographic levels, particularly where there are higher levels of non-English speakers.
We recommend conducting additional quality assurance for any data collected using the question developed for Census 2021 in England and Wales, to investigate uncertainty in the estimates. This could involve exploring the characteristics of those identifying as having a gender identity different to their sex registered at birth, to identify whether there are patterns in the data that could be considered unexpected. If collected, write-in data can also be explored as an indicator of the level of uncertainty in individual responses. If the quality assurance suggests a high level of uncertainty, or that some respondents may not have interpreted the question as intended, the impact of this on the statistical outputs should be explored.
When publishing outputs by gender identity, information about the uncertainty in the data and the implications for onward data use should be clearly communicated to users. We recommend including the following information alongside your outputs, with additional information added based on the findings from the quality assurance of your data:
‘These statistics have been collected using the gender identity question asked in the Census 2021 in England and Wales. It should be noted that in the Census 2021 data for England and Wales, patterns were identified that suggest that some respondents may not have interpreted the question as intended, notably those with lower levels of English language proficiency. Analysis of Scotland’s Census 2022, where the gender identity question was different, has added weight to this observation. More information can be found in the ONS sexual orientation and gender identity quality information report.
‘Similar respondent error may have occurred during the data collection for these statistics so comparisons between subnational and other smaller group breakdowns should be considered with caution.’
Alongside communicating the uncertainty in the data, we recommend that you help users by giving examples of what the statistics can and cannot be used for. These will depend on your findings from testing and quality assurance.
Presenting and reporting the data (outputs)
Recommended data presentation:
In the table [data] represents where data should be inserted.
| Variable | Data |
|---|---|
| Gender identity the same as sex registered at birth | [data] |
| Gender identity different from sex registered at birth | [data] |
| Prefer not to say | [data] |
| Spontaneous only: refusal or did not answer | [data] |
Showing 1 to 4 of 4 entries
Gender identity the same as sex registered at birth
This refers to people who selected “Yes”. It includes male when registered male at birth and female when registered female at birth.
Gender identity different from sex registered at birth
This refers to people who selected “No” and whose gender identity is not the same as the sex they were registered at birth. This is inclusive of a range of gender identities, including:
- binary male or female genders when not the same as registered at birth
- non-binary genders such as those on a continuum between male and female
- non-gendered identities, which are neither male nor female
Prefer not to say
This refers to people who responded with “Prefer not to say”.
Spontaneous only: refusal or did not answer
This refers to people who have either refused to answer the question, or more generally, did not give a response. This should not be presented as an option to the respondent, but it can be recorded. This is useful to infer the acceptability and usability of the question.
Scotland
The National Records of Scotland (NRS) are responsible for the census in Scotland. Scotland’s Census 2022 included a question on trans status and history.
You can find information on this testing in the NRS Sex and Gender Identity Topic Report. NRS published a quality assurance report on the topic data captured by this question in June 2024.
Question
Do you consider yourself to be trans, or have a trans history?
Response options
- No
- Yes
- Prefer not to say
For those who respond ‘yes’ to the trans status question
Question
If you would like to, please describe your trans status (for example non-binary, trans man, trans woman):
Response options
This is an open text response
- Prefer not to say
- Not Applicable
Most surveys and administrative collections of data are voluntary. It is good practice to tell people at the start of the data collection that they do not need to answer anything they do not want to answer. This means respondents should be able to skip the gender identity questions if they do not wish to answer them. When used in Scotland’s Census 2022 this was addressed by providing an instruction to respondents that “This question is voluntary”.
The gender identity questions can be sensitive for some respondents and maintaining the privacy of respondents is critical. This means that we recommend being explicit that people do not have to answer the question. This can be done by providing an instruction to respondents as was done in Scotland’s Census 2022, or by offering the “Prefer not to say” option. This is in addition to the implicit option of refusing to answer the question. This is consistent with good practice in data collection for sensitive topics.
Presenting and reporting the data
The Chief Statistician to the Scottish Government has published guidance for public bodies on the collection and publication of data on sex, gender identity and trans status.
Northern Ireland
The Northern Ireland Research and Statistics Agency (NISRA) are responsible for the census in Northern Ireland. NISRA did not ask a question on gender identity in their Census 2021 due to a limited user need for a question on gender identity.
.
Question placement
If you are including questions on sex and gender identity in your data collection, we recommend placing them next to each other and adding a guidance note to the sex question to tell respondents that a gender identity question will be subsequently asked. You should ask the questions in the following order: sex and gender identity. This is because testing has shown that placing the gender identity question after a sex question provided better understanding by respondents and improved data quality, this was the case for both the Census 2021 for England and Wales and the Scotland 2022 Census.
.
Guidance for data collection
How and where people will answer the questions
Privacy is an important issue in answering a question about gender identity, which is considered a personal matter.
If you are asking the question in a face-to-face context, aim to ensure respondents’ answers cannot be seen by others. For example, if you will be asking the question as part of an interviewer-led household survey, respondents could be offered the chance to complete the gender identity question themselves. This would enable them to keep their response private from the household and the interviewer.
Proxy response, or completing the question on behalf of someone else
You should aim to get an answer directly from the respondent themselves. This is part of good data collection practice, and we would recommend this for the collection of gender identity information.
The topic of gender identity is sensitive and private. You should carefully consider whether it is appropriate to enable proxy response. We advise putting in place appropriate ways to maintain privacy and protect respondent confidentiality.
.
Contact details
You can contact the GSS Harmonisation Team by emailing Harmonisation@statistics.gov.uk.
Related