It sometimes feels like the only fact we know for sure in this election is that it’s happening on July 4th (with some of the people who heard it first allegedly heading straight for the bookies – kerching!) Everything else is contested. NHS waiting times are going up / going down. Migration numbers are higher / lower. Labour will tax you more / less. But at least we can rely on that venerable, old, trusted source of data, the census. Can’t we?

Well, let’s see, because the latest piece of data to emerge from the 2022 census is a source of some concern it has to be said. The ’22 census, you’ll remember, was the one where the Scottish Government decided to go it alone (no, it did not go well thanks for asking). It also introduced a question on sexual and gender identity despite concerns about what the answers to such a question could realistically tell us.

But never mind, the Government ploughed on regardless and the separate census happened and now the results in different categories are being announced and this week it’s LGBT. According to National Records Scotland (NRS), what the figures reveal is that almost 184,000 people in Scotland identify as gay, lesbian or bisexual and 19,990 people are trans or have a trans history. It amounts to 4% of over-16s being LGBT.

Knowing the figures, we’re told, will help councils, charities, businesses and so on improve the lives of gay people as well as provide information for “equality monitoring”. But that rather depends on the figures being an accurate reflection of the truth and the obvious problem with a question on sexuality, particularly when the form is sometimes filled in on someone’s behalf, is that some people won’t want to answer the question and sure enough more than 8% didn’t.

The other problem is what we mean by the questions and what people mean by their answers. What one person understands by gay or trans may not be what another understands by the words, particularly when you factor in age, religion, education and so on. More than 23,000 people told the census they belonged to “another sexual orientation” which includes pansexual and queer. But what do these terms mean? Some people call themselves queer for example even though they’re attracted to the opposite sex – in other words, hard as it is to believe, the LGBT+ figures for Scotland may include heterosexual people.

Read more

Mark Smith: A Scottish boy abused for being gay. It’s still happening

Mark Smith: Hope and despair. The strange case of Scotland’s ‘protected’ buildings

The implications of all of this are obvious I would have thought, particularly when NRS tell us the figures are designed to help councils and charities target their services at LGBT+ people. In fact, the level of uncertainly means there’s a risk the services will actually be targeted at people, probably younger, who feel confident to answer the identity question while missing those who do not have that confidence or disagree with its premise about multiple sexual identities or the ability to self-identify your gender.

The Scottish Government assures us this is not so and they’ve used modelling to ensure the stats represent the total population rather than just those who completed the questionnaire. But the response rate only just scraped 89.8% after the deadline was extended (at a cost of £6m on top of the £21.6m extra it cost Scotland to go it alone) and the modelling used to work out what the other 10.2% might have said in their answers is politely called imputation but is less politely called guessing.

To make matters worse, that 89.8% figure is the average across Scotland: in deprived areas it was lower, meaning we’re less likely to correctly estimate the level of public services needed in those parts of the country. It’s also significant that, according to the census, almost half of trans people are 16 to 24 and the highest percentage is in university cities. The concern must therefore be that, if we rely on this data, LGBT+ policy will be focusing more on educated middle class people at university who take part in surveys and missing people from more deprived backgrounds who do not.

All of it means, I think, that the gay and trans data from the Scottish census looks very dubious indeed and we should be extremely cautious about relying on it. In what looked like a carefully worded statement, the director of census statistics at NRS, Jon Wroth-Smith, said the survey meant we could now say that almost 20,000 people aged 16 and over in Scotland identified as being trans or having a trans history and almost 184,000 people aged 16 and over identified as LGB+ but that’s true only as far as it goes. It tells us how many people answered the question but it does not tell us with any reliable level of certainty what the size and shape of Scotland’s LGBT population actually is.

I think more humility from National Records Scotland would have helped here. The census was not a success, the response was poor, it failed to meet the Government’s targets, and it included a question on sexuality and gender that introduced random errors on top of the random errors caused by the low response rate.

It also proves another thing: how hard it is to apply apparently cool statistical objectivity to an issue that is subjective and disputed, and for some sensitive and difficult. The truth is that, despite living in apparently more enlightened times, there are lots of gay people out there who need help and support. I’m not convinced the census gets us any closer to reaching them.