Drawing on insights from his new book, Rainbow Trap: Classifications, Queer Lives and the Dangers of Inclusion, Kevin outlines how a twin-track approach to diversity data can ensure organisations comply with reporting requirements while also reflecting the nuance and complexity of the people represented in the data.
UK organisations collect, analyse and use a vast amount of data about how people identify – but these data practices face a fork in the road. In April 2025, the Supreme Court ruled that the terms ’woman’, ’man’, and ’sex’ in the 2010 Equality Act refer exclusively to biological sex as recorded at birth. This landmark decision, stemming from the case For Women Scotland Ltd v The Scottish Ministers, has significant implications for organisations collecting, analysing and reporting identity data, particularly public authorities, such as local councils and police forces, and organisations exercising public functions, such as higher education institutions and arts organisations.
These bodies are subject to a part of the Equality Act called the Public Sector Equality Duty (PSED), which requires the demonstration of 'due regard' to three key aims:
- Eliminating unlawful discrimination,
- Advancing equality of opportunity,
- Fostering good relations between people who share a protected characteristic and those who do not.
To meet these obligations, organisations must collect and publish data about the people they serve – including those engaged through service delivery, education or research, such as NHS patients, survey respondents, students, alumni, local residents and partner businesses.
While the PSED doesn't enforce rules on the actions of private companies, many collect diversity data to shape policy, demonstrate values and foster inclusion by adopting a trust-based approach grounded in self-identification. The Supreme Court's ruling threatens to throw a spanner in the diversity data machine, unravelling progress we’ve already made and steering organisations towards a biology-first approach to data.
It comes at a precarious time. Human Resources (HR) and Diversity, Equity, and Inclusion (DEI) teams have worked hard to build data systems that encourage openness and respect complexity. But the diversity data machine – built on trust, nuance and people’s willingness to share information about their lives – will struggle to function if people feel that data collected about them is needlessly intrusive or irrelevant to the issue being addressed.
Leaders now face a choice: enforce a rigid biology-first approach to data or take this moment to advance data practices that are inclusive, have the support of the people from whom they collect data and are compliant with reporting requirements.
Here's how.
Build trust through transparency and by demonstrating impact
Too often, organisations ask people to share information about their identity characteristics without clearly explaining why it matters or what difference it will make. Trust is built not just through transparency but by follow-up actions. If you've collected data in the past, show how it has led to tangible improvements. Be specific. Whether it informed changes to parental leave policies, shaped a new recruitment approach or improved access to flexible working, demonstrate the link between what staff told you and what changed.
Transparency isn't just good practice. It's about earning meaningful participation and creating genuinely more inclusive workplaces, not just better reporting.
Ask better questions — not more of them
We often hear about the need for 'accuracy' and 'clarity' – but less about how a biology-first approach to data would work in practice. How will you ask questions about a person’s biological sex, and how do you expect people to respond?
These approaches often ignore the simple fact that people don't always go quietly into the boxes that designers of data systems wish to put them into, and when data systems fail to reflect lived realities, trust breaks down. Worse, insisting on narrow definitions can undermine the validity of people's lived realities and force individuals into a false choice: accept an ill-fitting label or opt out entirely.
Expect resistance if you base your approach on rigid, biology-first classifications. Many people will choose not to respond. Others may provide inaccurate information, rendering the dataset less reliable.
As a way forward, I propose a twin-track approach to sex and sexual orientation questions, the two questions most clearly impacted by the Supreme Court ruling. One track aligns with the definition of sex in the Equality Act, and one track recognises the nuances and complexities of how people describe their gender, sex and sexuality. Importantly, organisations must communicate to those from whom they are collecting data that all questions are voluntary.
Track A: Inclusive, respondent-led questions
Q. How would you describe your gender? (Select all that apply)
This question recognises that, for many people, their gender and/or sex encapsulates more dimensions than biology – it is intentionally inclusive.
- Woman
- Man
- Non-binary
- In another way (specify if you wish):
- Prefer not to say
Q. How would you describe your sexual orientation? (Select all that apply)
This question recognises that, for many people, their sexual orientation encapsulates more dimensions than biology – it is intentionally inclusive.
- Heterosexual/straight
- Bisexual
- Gay/lesbian
- Queer
- Asexual
- In another way (specify if you wish):
- Prefer not to say
Track B: Biology-first questions
Q. What is your sex?
Following the 16 April 2025 Supreme Court ruling and in accordance with the Equality Act, this question is asking about your biological sex or sex at birth.
- Female
- Male
Q. What is your sexual orientation?
Following the 16 April 2025 Supreme Court ruling and in accordance with the Equality Act, this question is asking about the biological sex or sex at birth of yourself and the person you have a sexual orientation toward.
- Heterosexual/straight
- Bisexual
- Gay/lesbian
Separate compliance from inclusion
The Equality Act is a legal floor, not a ceiling. It sets minimum requirements for public sector organisations but does not prevent you from going further. Collecting data in line with the Act doesn't mean you must also abandon the inclusive data practices your organisation has worked hard to build. Doing the bare minimum risks actively eroding trust among those represented in the data – including the many individuals and groups who learn, work, access services, participate in research or engage with your organisation in other ways.
The twin-track approach I suggest ensures you are compliant with the Equality Act’s definition of ‘sex’ while also continuing to capture inclusive, accurate data that meaningfully reflect how people wish to be counted. If your classification system can't accommodate a spectrum of experiences, it fails to reflect the full picture of your organisation.
What to do now: five actions for HR and DEI leaders
- Review data practices to ensure they capture information about the issues you wish to address and are based on participation, not just compliance.
- Update diversity monitoring forms with a twin-track approach to gender, sex and sexual orientation data.
- Communicate with those from whom you want to collect data about why this change is taking place. Be upfront and honest that you need to ask biology-first questions, but there is no requirement for people to respond to these questions.
- Train staff involved in data collection, analysis and presentation on the importance of trust, consent and complexity in data collection.
- Avoid using terms like ‘accuracy’ or ‘clarity’ to describe data practices without thinking critically about whether these objectives are entirely achievable when working with diversity data.
We cannot fix inclusion by tightening the screws on a diversity data machine that is reliant on the goodwill of people about whom we wish to collect data. Instead, we need to lead the design of data practices that respect complexity, centre trust, and understand inclusion not as a box to tick but as a relationship to build.
Kevin Guyan is our Chancellor's Fellow.