Data collection and sex disaggregation

[Procrastinatrixx]

Can I get a sense check on this: there’s been so much focus and reporting on single sex services and spaces (rightly so), but what about data collection and medical research? Why is that not more front and centre?

In the workplace, we cannot possibly monitor or track “gender” pay gaps etc unless we can accurately disaggregate data by sex, so when workplaces conflate sex and gender identity, we effectively falsify this data. This includes comparison across categories, eg sex against gender, so comparing women as a group with transmen and transwomen, as discrete groups.

More importantly, medical data must be accurate for mortality purposes, sex specific treatments and outcomes, access, etc. This is literally life and death stuff! Why is it not more of a priority? If NHS services are conflating sex and gender then population based studies and medical research is seriously compromised.

I donated to the Census judicial review, and if I remember correctly I think maybe Professor Sullivan had written on these topics in the past (?)… I also remember various previous issues with academic organisations, journals, and publishers regarding medical terminology for sex vs gender. But in the news reports, interviews, and podcasts etc since the SC ruling, data and medical accuracy is an obvious debate winner that’s not been mentioned at all. Women exist as our own class. We deserve to be counted. That can’t happen if you confuse sex categories or foster an environment where categories cannot be debated.

Or am I missing something? Has this issue been resolved more discreetly in practice already?

It just occurred to me - does the Equality Act 2010 inform data collection (even indirectly)? If not then I can see why this has been sidelined…

Hard agree, it’s a key element of lots of data - because men and women have different needs & motivations. MoJ data from the last several years is near enough unusable for some analysis because of this. Don’t know if the SC ruling will help, but I hope it does

It has been resolved by the SC ruling (see para 239 for the discussion on why biological sex must be applied to data collection for women’s rights), but while the ruling itself applies retrospectively, that won’t untangle the conflation of sex and gender identity in the data that’s already been incorrectly collected over the past few years.

Obviously, going forward, sex needs to be the basis of data collection (unless it’s specifically looking at gender identity issues for trans people), but I do mourn the loss of meaning in all the useless data we’ve amassed.

Certainly for public bodies subject to the Public Sector Equality Duty under the EA, as they can’t show they are meeting the requirements if they can’t even identify people with a protected characteristic (e.g. women).

If you go to the Sex Matters web site you will find they have been working on this aspect for some time.

I think there was a recent update.

The Sullivan report is still 'under consideration' by the government.

Many of us are also worried by the data bill, which may further corrupt sex data by using passport and driving licence information on sex (they are both easily changed with no need for a GRC).

Sex Matters are working very hard on all this.

Oh thanks for the heads up, I’m a sex matters supporter but don’t always keep up with all their work

There is a Sex matters action page about this where you can write to the secretary of stae about it via a form, takes 1 minute

https://sex-matters.org/take-action/write-to-peter-kyle/

The SM template letter does not take into account the SC ruling so i sent an email direct and added this at the top. His contact address is here. You need to include your name and address in a direct email.

Dear Secretary of State,

Below is a template letter from Sex Matters with which I wholeheartedly agree.
I trust given the extremely thorough and thoughtful ruling given by the supreme court on the definition of sex in the Equalities Act it will be without question that biological sex is important for legal purposes and must be recorded correctly (to do otherwise would be a breach of the GPDR / Data Protection Act which requires personal information is accurate)
I urge you to act decisively, and to commit to ensuring that data on sex is recorded acurately and data on sex from unreliable, unclear sources is not allowed to pass through the information gateway.
--------------

The Sullivan Review of data, statistics and research on sex and gender that your department recently published exposes a serious problem: public bodies such as the NHS and the Passport Office no longer hold accurate, clear data on whether people are male or female. They allow people to change their records based on gender self-identification.
You are now passing a new law to create an “information gateway” that will allow these public bodies to share data on sex and gender, and which will mark it as trustworthy for all purposes.
Reliable data about people’s sex is crucial in many areas of life, including patient safety, dignity in care, safety in policing, robust safeguarding of children, and fairness and safety in sport. This is sometimes called “biological sex” or “sex at birth”.
When gender identity displaces sex in records it puts people at risk and harms safeguarding, safety, science, fairness and data protection.
I urge you to act decisively, and to commit to ensuring that data on sex from unreliable, unclear sources is not allowed to pass through the information gateway.
Although responsibility to solve the problem of inaccurate sex records is spread across government, the responsibility for ensuring that the information gateway does not create an official system of “gender self-id” is in your hands.
I look forward to hearing from you.

This is quick, off the top of my head (I used to work in privacy law). Personal data has to be accurate for the purposes for which it is being used. It is the data owners (Data Controllers) duty to keep it accurate and up to date. If they can't they have to delete it. Data about sex / gender reassignment is 'SCD', sensetive data that has additional rules associated with it.

So the SC judgement actually potentially leaves some Data Controllers with a real problem. They have to clean up their data or else bin it. In addition it can't be used for purposes other than those for which it was collected without the individuals explicit consent so it may not be as easy as simply re-labling it.

If Controllers hold incorrect data / are breaking the law then individuals can complain to the Information Comissioner. Complaints are usually about misuse of their own personal data. Not sure that a general complaint could be put in to the ICO about goverment data. I would have thought the ICO would have been invovled with the Sullivan review if it was looking at accuracy of personal data. I will have a look when I get 5 minutes.

Thanks @MixTapeMel.

Honestly, I think in many cases it would be impossible to re-label the data anyway, because they won’t necessarily know who was recorded as female because they are biologically female and who was recorded in the same category because they ‘identify’ as female if the original question was framed in terms of gender rather than sex.

I think either way (if the data remains inaccurate, or if they have to delete it), they will encounter issues in terms of adequately discharging the PSED.

I’m interested in your point about data controllers not being allowed to relabel data. When I joined my previous employer the equality data collection form said ‘sex’ and I said F. At some point my Stonewalled employer changed the label to ‘gender’ in the HR system without asking me. All the equality reporting switched to gender as well. I challenged their practice because they weren’t collecting data on the PC of sex or reporting it. I don’t have a gender either. It’s not a belief I share. Whenever I see it on equality forms now I tick the information refused box.

So @BettyFilous my employer wasn't stonewalled, I worked for a very large US company, and our HR in the UK took direction from them. But here is what I would have said to HR if they'd asked me about the importance of labels.

The fundamental principal is that data has to be accurate for the purposes for which it is used. 'Sex' information is primaliry collected by employers for EA purposes, where it is clearly refered to in law as sex. So the label on the forms should be sex. End of conversation.

Gender has no meaning in law, and it is a loosey goosey term for many people. It's a bad data collection label. The GRA is no grounds to change the label. If (and I say if, the case is it does not), but if, lets say the GRA did mean that if you had a GRC your legal sex changed, then the data point is still sex, so no grounds to change the label.

Now if, in the census, they wanted to collect a data point separate to sex that is gender idenity (for people without a GRC) then that is a separate thing and would warrent a different label so everyone is clear what is being asked and what the data is and that it is different to 'sex' (in the EA sense).

The issue might arise that you you could challenge the lawfulness of collecting gender identity given that asking if you have a GRC is not lawful. For example if gender identity was used to imply who may have a GRC. It leads to a minefield. So if my HR department had asked 'should we have a separate box asking about gender identity' I would have said no. As the only data you need for EA purposes is Sex. I'd like to think that's what I would have done. Who knows?

So self correction - gender identity is not SCD in the UK / EU. Sexual orientation, and information about sex life is, as are some other EA data points (like race and religious / philosophical beliefs) - hence my erroneous assumption.

Why this matters is that an organisation cannot collect SCD unless it meets more stringent conditions for data collection/processing. Yada yada yada. I will shut up now 🤐

Thanks @MixTapeMel I no longer work for the organisation so follow up is moot. When I raised the inconsistency with the EA2010 PCs they told me they were deliberately choosing to get ahead of the law. 🙄 There’s no accounting for stupid.

@Procrastinatrixx After the almighty mess ONS made of the Census the Office for Statistics Regulation produced this guidance https://osr.statisticsauthority.gov.uk/our-work-on-data-about-sex-and-gender-identity/

Prof Alice Sullivan was part of the group that wrote it. Whilst it predates her most recent report and the Supreme Court judgement it is very clear on how to make sure data is accurate.