Official Report 627KB pdf
Welcome to the first meeting in 2020 of the Culture, Tourism, Europe and External Affairs Committee in 2020. I remind members and the public to turn off mobile phones, and anyone who is accessing committee papers through electronic devices to ensure that they are turned to silent mode.
The first item on our agenda is to take evidence from the National Records of Scotland on the draft Census (Scotland) Order 2020. I welcome to the meeting Pete Whitehouse, director of statistical services, National Records of Scotland; Scott Matheson, senior principal legal officer, Scottish Government; Scott McEwen, head of policy and legislation, Scotland’s census 2021, National Records of Scotland; and Jill Morton, senior business lead, questions and collection instruments, National Records of Scotland.
I invite Pete Whitehouse to make a brief opening statement.
Thank you, convener, and good morning to everyone. I will try not to lean too much into the microphone.
I thank the committee for its invitation to the National Records of Scotland to today’s session, which continues our important discussions on the detail and substance of Scotland’s census for 2021.
As you said, convener, there are a number of officials before the committee this morning, bringing together our legal advice, our senior lead on census policy and legislation, and, in Jill Morton and me, members of the Government statistical service. I was recently appointed as director of the NRS’s statistical services.
I am very grateful to the committee for initiating and working through its informal scrutiny of the work that we are doing to deliver a census order and associated regulations. As you will recall, there were significant challenges in laying the census order for 2011, resulting in the Parliament considering the order on a number of occasions. The current process of informal scrutiny is therefore valuable in enabling the NRS to work through issues with the aim of enabling Parliament to agree the census order for 2021 in a timely manner. The NRS is working to present the draft order to Parliament towards the end of this month, with the aim that the order can be placed into statute by Parliament in April 2020.
Working with the committee to help present an acceptable census order to Parliament for its consideration within that timeframe is central to enabling the NRS to land critical aspects of the census. Those include the recruitment of the field force, the finalisation of questions, which then allows paper and online forms to be finished, and the building and testing of a significant array of information technology solutions. Those tasks all involve procurement, contracts and significant on-going testing.
The consideration of issues such as how to support completion of the sex question, or of the voluntary sexual orientation question, confirms that there are areas where views diverge. The NRS has strongly welcomed the opportunity to bring up and discuss a number of issues with the committee through the informal scrutiny process, so that progress can be informed and recommendations produced.
The consideration of the census questions and associated support requires the deliberation of many factors, including learning from past censuses, the evolving needs of users, the need for consistency of outputs over time and across the UK, and innovation of technical and methodological solutions. Those elements must be drawn together to deliver the timely, high-quality outputs that are required by users, as census outputs begin to be published in 2022 and through 2023, and thus meet our shared objectives of a high-quality, timely census that meets the needs of respondents and data users.
I am pleased to inform the committee that, since the NRS’s attendance in September, we have continued to engage in person and in writing with stakeholders, many of whom are also directly engaging with the committee; we have undertaken a successful census rehearsal in three areas of Scotland to test our processes; and we have provided further written updates to the committee on a range of issues relevant to census design, including predictive text, ethnicity and details of discussions with stakeholders.
My colleagues and I welcome the opportunity to meet the committee today, as we work to deliver a proposed census order to the Parliament later this month.
Thank you for that, and good morning to you all.
The NRS states:
“a binary sex question asked on a self-identification basis provides the best balance in meeting the diverse range of user needs across the full census dataset.”
09:30You mentioned engagement with users. However, we know that a significant number of senior academics who use population data have written to you—and the committee—expressing concern. The latest such letter is from 80 academics, led by Professor Alice Sullivan of University College London. In the letter, they say:
“We are writing to express our concern about the proposed online guidance to accompany the sex question in the 2021 census, which advises respondents that they may respond in terms of their self-identified gender. The guidance acts to conflate two distinct characteristics—sex and gender reassignment, both protected categories under the Equality Act 2010—and will effectively transform the longstanding sex question into a question about gender identity. We are concerned that this will actively undermine data reliability on a key demographic variable, and damage our ability to both capture and remedy sex-based discrimination.”
They go on to point out that, because the 2021 census will be a digital-first census,
“any proposed guidance will be much more visible”
than it was in 2011, when that guidance was introduced without any scrutiny. They also point out that it will affect other data-gathering exercises. The 80 academics who wrote that letter are leading lights of quantitative social science in the UK, and include several fellows of the British Academy and leaders of some of our major surveys. I am interested to know why you are so confident in your dismissal of their views and whose views you have taken in preference to the views of such leading lights.
As you will have heard previously and read in correspondence, the NRS works hard to hear a range of views and to engage in written form and conversation with a range of stakeholders. We fully recognise that there are different views about how the sex question should be presented and supported, and about the role of guidance in that.
In reaching the recommendation that we put forward, which is, as you say, for a binary sex question with self-identification guidance and a separate and voluntary question on trans history and status, we have had to consider past censuses. Certain questions have been asked for a number of years in a number of censuses. We also need to consider the guidance that was in place in the past. In some cases, there was no guidance but, in 2011, as you say, there was online guidance to support expressed user need on how to respond to the binary sex question.
We have gone through evolution on one of those issues. As a committee, you will know that we did some initial thinking on changing the binary sex question and that there was discussion with the committee that changed our position. We have had advice and views from lots of data providers and users and, as I said in my opening remarks, we welcome the views of all our different types of stakeholders.
The role of the NRS is to put in place a census that understands consistency over time, the need for harmonisation and the various needs of data users, and which recognises the guidance that has been in place and the purpose of guidance. For example, following a recommendation from the committee and a discussion with it in September, we took forward further testing—undertaken, on our behalf, by ScotCen Social Research—on the binary sex question, so we know that the majority of people do not look at guidance to answer that question. They fully recognise how they wish to answer it. They use whatever evidence they feel is appropriate to come to their conclusion, and that is consistent with how people respond to what is essentially a self-completed census. This is about people understanding the nature of the census and responding in the way that they feel best represents their sex, in this case.
I am sorry to interrupt, but I just want to drill down on that. Which data users have asked for a self-identification question because of their wish for that data as an output? You have dismissed all those senior academics, such as Professor David Bann, associate professor in population health at University College London, and Professor Mel Bartley, professor emerita of medical sociology—I could list many of them—who use population data. Which data users wish for the sex question to be self-identified?
I am keen not to get into individuals and saying that people have a particular—
But in your letter to me, you say that you have had meetings with data users, so who are the data users who want that change?
We had representations in letters from a number of users.
There were other communications, but they were from academics in other fields. Many of them were from fields such as computer science or literature. I am sure that they are sincere in their views and expert in their areas but, unlike the academics I am talking about, they are not all users of population data, such as sociologists, social scientists and economists. They were a different group of academics.
There is the consistency of how the question has always been asked. It has been asked in a binary sense, with a self-completion approach, and our best understanding is that people will use whatever evidence they feel they need to use in order to answer it.
In 2011, an express need came through to the Office for National Statistics from a group in the population who felt that they needed further advice on how to answer the question. At that point, the guidance was provided—we also put it online in Scotland—and, because the form is a self-completion form and the issue is about people understanding how they will answer that question, the guidance becomes about self-identification.
That guidance has been tested. It is seen to be helpful to people who need guidance. It has no impact on people who, in the past, will have simply answered the question and will continue to answer it without guidance. I recognise that not everybody takes the position that that takes us to, which is that the approach to completion is based on people’s understanding of how they wish to answer that question about sex, which means that self-identification is essentially the place that we get to. We are using guidance to help a particular group in the population who have asked for advice on how to answer that question.
We have also heard from different academics that it is important that we are clear about the guidance that is in place. What we are saying is that, if we ask a question in the census, it is important to data users to understand the basis on which that question was asked. We are therefore saying that guidance, in this situation, will help those people who need guidance with the completion of that question, and it will provide data users with clarity about the basis of that question.
You are absolutely right that there were some users in the past, and there might be some in future, who state that the question is being answered in some other way. Our position is the same as that of the ONS, so there is a harmonised approach. Because there is self-completion, because guidance was there in 2011 and because there is a stated need to enable an important part of the population to respond to the question and to seek guidance, which will be important for our contact centres and the people in our field force who will go out to support people to answer the question, we are clear about the basis for it to be asked. That is where we are coming from. It is a practical position that recognises that people answer that self-completion question in a particular way.
I still do not think that you are responding to the concerns that have been raised by that senior group of data users. The office for statistics regulation has emphasised the importance of meeting the needs of data users. You have emphasised the needs of a small group of respondents and their feelings, and you mentioned the independent research that you commissioned from ScotCen. I put on record my confusion about the conclusions that you have drawn from that research.
The question and the guidance were tested on around 2,000 randomly selected members of the general population and 75 trans individuals, who were not randomly selected but who were recruited through various contacts. Of the members of the general population, 3 per cent said that they would not answer the sex question if it was asked on the basis of self-identification and would in effect boycott the question or the census. A larger percentage of the trans individuals said that they would boycott it for other reasons. However, if we work out how the percentages would translate, we find that the number of people in the general population who would refuse to answer is far higher than the number indicated by the percentage in the small trans group. Of course, we do not know how scientific that trans sample was. If it is respondent need that is driving you to the conclusion, it seems a bit strange that you have gone for the needs of one small group, when a boycott by 3 per cent of the general population, which is what your research indicates, would be far more damaging.
I will pull out a couple of points from that. The methodology that ScotCen used is well understood and well regarded. It is a professional organisation that runs many surveys in Scotland and across the UK. I am not here to answer questions on its methodology but, as a Government statistician of 30 years, I can say that, when you are looking at very small population groups in the wider population, it is reasonable to take what is in essence a non-probabilistic approach to getting information. That is different from the situation with the general population, where a probabilistic approach can be taken. With a very small group, if you simply go out and ask 3,000 or 4,000 people at random, you are likely to miss the specific group that you are interested in. Therefore, inviting people in a different way to express their understanding is a legitimate approach that is used by many research companies. I recommend that, if the committee has issues—
I am not questioning the methodology. ScotCen has been very frank about that. Its report says:
“Since participants were not selected at random, the findings relate only to those who took part and inferences to the wider trans or non-binary population in Scotland cannot be made.”
We all totally accept that. My point is based on ScotCen’s findings, which are that 3 per cent of the sample of 2,000 members of the general population said that they would not answer the sex question based on self-identification, and about 61 per cent of the trans sample said that they would not answer. My rough calculation is that, if that was rolled out across the whole adult population of Scotland that is answering the census, 120,000 of the general population would refuse to answer the census, whereas about 40,000 of the trans population would refuse. You have based your decisions on the fact that you think that a number of members of the trans population would refuse to answer the census according to certain guidance.
09:45
Thank you. I will also ask Jill Morton to come in. My second point is that the evidence that we have over time is that people answer that question perfectly well and are fully able to do so. It is one of the best answered questions. Previous censuses show that people know how to answer the question. The NRS has no concern that people in the general population will not be able to answer that question. The guidance is—
I am sorry to interrupt you, but I did not say that they would not be able to answer it. Your research says that, because of their views on self-identification, 3 per cent of people would refuse to answer. You are aware of the wider debate. There has been some discussion, and the committee has had correspondence about it. There might even be a campaign by people who oppose self-identification not to answer the question. Have you taken that risk into consideration?
Jill Morton, do you want to add anything in answer to the previous question?
No, not at this point.
Risk occurs across all sorts of questions. The census is there to meet the needs of a range of data users. We will run a campaign over the coming year—into and through the census period—which will promote the benefit of the census. It will promote engagement and continue to clarify and support people’s understanding of the benefit of the census. We hope that, because people are requested and required to answer that census, we will enable everybody across Scotland to engage in it.
As I said at the beginning, there are always different views on how we put a census into the field. The NRS has listened and we have continued to speak to people. We have had correspondence—whether it has come to us directly or whether it has gone to the committee, to the media and television or to the national statistician at a UK level. We continue to have those conversations. At this point, we are putting forward our recommendation on how we can best meet the needs of data providers and users in order to produce the timely outputs that we require in order to give the benefits of the census. This is part of the informal scrutiny process that allows us to do that.
I am absolutely not ignoring the views of different groups. As an organisation, over the past few years, we have tried to balance all those views, so that we draw together the best census that we can.
We are not going to agree that you have balanced views, but we are running out of time, so I will move on.
I have questions about the testing of the guidance. The process was that two sets of guidance were tested; one set suggested that people have a self-identification response and one was a legal sex definition. If the self-identification guidance that was tested is the one that we have been provided with for the 2021 census, we do not have a copy of the other guidance that was given. Did you test no guidance?
No guidance is in the assessment, in the sense that we understood that the vast majority of the general population do not look at guidance.
There is a link, however; the guidance is there. Did both sets that were tested get presented with the question with absolutely no guidance? That was the situation prior to 2011.
They were presented with the question. The detail is in the report but they were presented with the question, asked to respond and then asked whether they had used guidance in coming to that conclusion. Because the way that people respond can be looked at as they go around the online system, it was clear that almost nobody in the general population looked at the guidance.
We infer from that that the vast majority of the general population do not feel that they need, and do not look at, guidance, so, in a sense, the guidance that exists—
I am sorry to interrupt, but what I am trying to get at is that, prior to the 2011 census, that question was published with no guidance. The NRS argues that it has always been a self-identification question—it makes that point a number of times in the papers that we have received.
It is a self-completion form, which therefore leads us to the view that self-identification is where we are best placed to be. People answer the question in the way that they feel best represents their sex.
You argue that it has always been a self-completion question, which means self-identification. Therefore, why was it felt necessary to introduce guidance in 2011?
A representation was made to the ONS, in the first instance, which then came across to my predecessors in the NRS, that people in the trans community were seeking further advice on how to answer that question. At that time, guidance was provided online. That was done specifically because a group in the population were saying, “We need further advice on how to answer that question.” Therefore, guidance was provided in the context of what would, at the time, have been the expectations of the ONS and the NRS regarding how that question should be answered—by self-identification.
I would like to ask some more technical questions about how the 2011 advice compares with the proposed guidance for the new census order. Under “Help with answering”, the 2011 advice says, “No guidance provided”. There is then a section entitled, “More questions?” That is what the advice on the sex question in the 2011 census, which we have been provided with, says. After a section entitled, “Why is this question asked?”, under “Help with answering” it says, “No guidance provided”. Under “More questions?”, it says:
“I am transgender or transsexual. Which option should I select?”
The fact that the new guidance is entitled “Proposed online guidance” appears to change the nature of the advice. In 2011, it was not categorised as guidance; it was there as information. This time, it will be guidance.
The guidance has changed quite a bit. Why has it been changed so much? Guidance is offered for people who are non-binary, but the section in the 2021 guidance for transgender people is much shorter than, and is phrased quite differently from, the advice that was provided in 2011.
According to some of the other information that you sent us, which included email exchanges and information on meetings that you had with academics, one of the issues that was raised was concern about inconsistency in the wording of the guidance and ambiguity in the proposed guidance for the 2021 census.
Why has the guidance been changed? In 2011, it was not categorised as guidance; it was categorised more as information. In 2021, the guidance for transgender individuals is quite different from what it was in 2011. Why is that?
The guidance that we took into the testing is included in the ScotCen publication that was published on 20 December. We held a series of events with stakeholders to agree the two different versions of the guidance: the guidance on self-identified sex and the guidance on legal sex. A range of stakeholders came along to those events, and we got to an agreed position on the wording and the terminology. Agreement was reached that that was an appropriate set of guidance to help people.
We refer to that as guidance, but the online system that was used in the rehearsal just says, “Need more help?”, and you click a link. That is very similar to the online system that was in place in 2011, which said, “Do you need help?” and provided access to it. I suspect that the difference between what we refer to as guidance and how that appears in the public sphere might just be to do with our internal use of terminology. We refer to it as guidance, but in the rehearsal it appears on the digital platform as a “need more help” button rather than as “guidance”.
The guidance that we gave with the testing was agreed with a range of stakeholders. All that information is published on our website. It includes details of the people who attended, where we got to and what the results were. That was fed directly into the testing, and the guidance that is being used is in the published report.
You talked about “a range of stakeholders”. Does that mean people who have a range of different views on self-identification?
Yes. There were representatives from equality groups, women’s groups and academic organisations. There were some representatives from local authorities, and I think that a representative from the health service came along.
The proposed guidance for 2021 could be seen as more vague than the 2011 guidance in some respects. However, the responses look quite positive, according to ScotCen.
We worked with different groups on a range of different questions to come up with guidance that uses the right terminology, that will not be offensive and that will gather the data that we intend to gather. The guidance is intended to help people who might not find a question straightforward to answer.
How do the sex question and the voluntary transgender questions interact, and can people compare the two?
I cannot find the page just now, but your report contains a chart that shows that, if a transgender individual is presented with the proposed guidance, their responses will be fairly consistent. A transgender male will answer that they are male if they have the self-identification guidance rather than the legal sex guidance, and a transgender woman will answer that they are a woman. Those responses seem to be fairly consistent, and that is how we can expect people to answer, but are the questions cross-referenced in some way so that they can be disaggregated?
That would apply to a range of different questions in the census where people are using multivariate analysis to look at a number of questions together, but potentially yes—if a data user wanted to do that.
On pages 30 and 31 of that report, we are told that:
“When processing the data NRS will not be aware of which members of the trans population have a GRC and in order to analyse the combined data for the sex question and the trans status question, it is important that the trans population answer these questions in a consistent manner and the self-identified guidance allows them to do this”.
That suggests that there would be analysis of combined data and that the two questions are meant to interact with each other.
They could be used in that way if a data user wanted to do that.
It is up to a data user to do that, if they want to.
Yes. However—I may have misunderstood something there—we will not be asking people whether they have a gender recognition certificate.
No, I know that.
We ask individuals to confirm their sex, and we ask a yes or no question and provide a written option for trans status or history. It is similar to the questions on health conditions. General health and disability are used together to produce a—
What I am driving at is that, if the census is being considered by someone who thinks that the binary sex question has to be answered, and if what is being looked for is a biological reading of the male and female population, it would be possible for people who have concerns about that question to look at the voluntary transgender question and perhaps analyse that data in a different way.
People might feel that the data is not consistent because it includes a self-identification question. Some academics are arguing that the data does not give what is needed because it does not accurately tell us what the male and female population is. As the question is self-identifying, some people argue that that changes the nature of the data. Would the voluntary transgender question enable people to analyse that data and take out the information that they think is more consistent?
Potentially, yes.
10:00
We are making two points. One is that the question has essentially always required a self-identification response. The view that it was always based on something other than that is not right. One of the benefits of the census is that people answer the range of multiple questions at the same time, so data analysts and users have the opportunity to look at the interdependencies or links between certain questions. It is not something that we are particularly focused on, but there is the potential to do so.
Earlier, we had a discussion about the fact that some data users and analysts will look to understand the relationships between all the questions across the full census. That is one of the great benefits of the census.
I will pick up on a few points before I get to my own questions. Mr Whitehouse repeatedly referred to the fact that the mandatory sex question involves self-completion. Is it not the case that all questions involve self-completion?
Yes, absolutely.
So, the question is no different in that regard.
No, it is not.
Every question requires self-completion.
The way in which the census is completed—
Indeed; it is all self-completion.
Absolutely.
Therefore, using that phraseology with regard to the mandatory sex question is not really meant to give it any particular significance, because every question is to be answered on a self-completion basis.
I raised it because some of our stakeholders suggested that the question was answered in a very specific way. We are saying that, as with all questions across the census, people come at it with their own understanding of what it is, alongside the availability or not of guidance, and they make their honest and truthful response to the census in that environment. It is a statistical tool in respect of which we are not focused on individual responses; we are drawing together the evidence to provide geographic and population groups and other analyses at a statistical level. That is the point that I am making.
It is good to have that point clarified. Perhaps people listening might have thought that there was some particular thing going on with question 3 that is different to every other question of the census.
Not at all.
The 2011 sex question guidance was the statistical blip, because, from the beginning of the census in 1801 or thereabouts to 2011, there was no guidance at all about the mandatory sex question. Is that correct?
That is absolutely the case.
In that regard, it could be viewed as—
It is not a statistical blip.
Maybe that is the wrong phraseology; I bow to your greater knowledge—you are the statistician.
It is a clarification for a group in society that is saying that it needs additional support in order to answer the question. It is clear that the vast majority of the general population do not need guidance and will not use it. In 2011, a group in the population identified to ONS, and latterly to NRS, that it needed additional guidance. That guidance was then presented in the manner of the understanding of the census that was delivered in 2011.
Okay, but that is the only time that that guidance has been part of the process.
It was the first time that the guidance was specifically put online.
The only time thus far.
I will go back to Claire Baker’s point. On 12 September 2019, I asked your predecessor, Amy Wilson:
“Does the NRS testing also involve a no-guidance scenario?”
Amy Wilson replied:
“It is not specifically for a no-guidance scenario”.—[Official Report, Culture, Tourism, Europe and External Affairs Committee, 12 September 2019; c 15.]
I hear your answer this morning, but I am a wee bit confused. Why did the NRS not give that specifically as an option?
Do you mean the option of having no guidance?
In the testing that you commissioned, why was that not a different head of testing? I just do not understand the answer that you gave to Claire Baker.
In a sense, we have got a response that shows that the vast majority of people just did not look at guidance. We are saying that the ScotCen research tells us about how people access the question. People in general do not use guidance; they answer the question well and they understand what the question is. We know that people have done that in the past.
Do you not feel, on reflection, that including a no-guidance option might have given you a more comprehensive outcome, in terms of considering all reasonable options in your testing? Is that not really the point—that the aim should not be to prejudge but to test across the board within reasonable parameters? It seems a very reasonable option to have thought about, but you rejected it.
We are building our knowledge iteratively. Going back over time, we know that that question in the census has been answered well. We are now trying to address a specific set of questions for ourselves about how we enable all of the Scottish population to respond to that question.
The testing that ScotCen did, which we discussed with the committee, and details of which we have provided, allows us to show—as we know from the past—that the vast majority of the population do not need guidance, will not look at it and are happy to answer that question. There is another group that needs guidance. The ScotCen work looked at some specific types of guidance as part of that conversation.
We have looked at outcomes and the testing of questions, going back over previous censuses, and at the work that we have done over the past four or five years with ScotCen and representative groups, and that has given us the full range of information. I am comfortable that we have drawn out sufficient information to allow us to say that the vast majority of the population does not need guidance and answers the question well without it. Guidance is there to help a particular group in society who expressed a need for that guidance over a decade ago.
I cannot understand why you did not think that that assumption would be worth testing.
As things stand, if the NRS’s recommendation about guidance were to be acceded to, what would be the status of that guidance, or of guidance on the mandatory sex question, as a matter of law?
What do you mean by status?
Would it have a legal status? Would any rights flow from that guidance? If there are other bits of text around the legislation that we are dealing with, we need to know what that would mean.
As with all the questions in the census, guidance allows people to answer the question. It goes no further than that.
Does it have no legal status at all?
It has no legal status; it is advice to people—
With respect, perhaps Scott Matheson might want to offer an opinion on that, as it is his area of expertise.
The guidance would not have the force of law. The Census Act 1920, the census order and the census regulations will be the relevant legislation, and, if any rights and obligations arise, they will be set out in those.
As Pete Whitehouse was saying, the guidance will assist people who are being asked to engage with the census through completing the returns. It is an adminicle of evidence: it is there in the background and gives context to the legislation. The fact that the committee is scrutinising it here while informally scrutinising the order is all part of the context and background, but it is not a document from which rights and obligations flow. It does not have the force of legislation.
Okay. That is helpful. Given that the guidance gives context to the legislation, I presume that, prior to making its decision, the committee will be provided with the latest up-to-date version of the guidance.
The guidance development is an on-going process, so—
Yes, but if, as Mr Matheson says, the guidance gives context to the legislation, presumably we cannot decide on the legislation until we have seen the final draft of the guidance.
I do not think that that necessarily follows. In previous decades, the order has been made and—
Yes, but if the committee wished to proceed on that basis, presumably there would not be a problem with that—or are you refusing to give us the final text of the guidance before we make our decision? I do not know; I am asking the question.
I am sorry—I am slightly unclear about the order in which we are doing that, so I will ask both Scott Matheson and Jill Morton to come in. The order goes to the Parliament and we develop the guidance, but when is the guidance finalised?
Later this year, towards the end of the summer, it will all be finalised and signed off. For much of it, we just have to make small tweaks, and then we will proof it and check it with—
Would it take until the summer to do that if, as you say, it just needs final tweaks?
One of the problems here is that the Government routinely issues guidance on legislation. In doing so, it is trying to help the population who need to engage with it to understand that legislation. If there were to be a requirement that the Government must always publish such guidance and make it available to the legislature before legislation is enacted, that would seem to involve a chronological problem: we cannot issue guidance explaining legislation that has not yet been made.
I am asking about the final draft of the guidance.
When I say that the guidance provides context because it has been considered by the committee, I am referring to the fact there will have been background material in the papers that the committee has considered between September and now, and it will also be available in the formal stages when the instrument will be laid in draft and considered by the Parliament.
There is nothing special about the census in that; there is just a general statement about the extent to which the proceedings of the Parliament can legitimately be taken into account in interpreting legislation. That does not mean that absolutely every paper that is available to the Parliament will have a bearing on the proper interpretation to be put on legislation, but it is part of the context in which it is meant to be viewed.
I thank you for all your answers. The committee will reflect on those, because the point is so important.
If I may, convener, I will ask two brief final questions. At the committee’s evidence session on 12 September 2019, the convener herself put to Amy Wilson the proposition that, in referencing self-ID, NRS was seeking to jump the gun—in effect, to usurp the role of the Parliament, which in due course will have an important debate on the matter. NRS’s response to that proposition was not clear, so perhaps I could put it to you again today. Is it the case that NRS is seeking to jump the gun on the self-ID debate, thereby usurping the role of the Parliament?
From my perspective, we are presenting a question that is well understood, with guidance that helps a group of the population to answer that question. We are not jumping any gun or trying to change or get in front of all the other conversations, discussions and work that the Parliament and others are doing in this space. We are saying that, from 2011, the question has been answered in a particular way and we are providing guidance to support that. In essence, that is where that census work starts and finishes.
Okay. On that basis, and given the importance of the debate—across the spectrum, people have very strong views on all the issues anent it—would it not therefore be appropriate to include that language, to isolate this approach as being specific to the 2021 census and having no wider import? Would it not be appropriate to include language to that effect in the guidance itself, or perhaps in the order or elsewhere, so that it is made clear that a Scottish public agency is not seeking to usurp the legislative processes of the Scottish Parliament?
I am not entirely clear how that would look if it were contained in the guidance.
Well, you could give it a wee go and see what it looks like.
The point that I would like to make is that the census is just that. It is about asking questions and gathering data; no rights or obligations flow from it. It is not as though someone who completes a census form and ticks the box to indicate their sex in response to the sex question means that, for legal purposes, that is what they are. It does not give them rights to be treated any differently than they otherwise would be.
It is an entirely separate legal question from that of the Scottish Parliament’s debate, which will happen in due course, about whether the law on gender recognition should be changed. To my mind, they are so separate that I do not understand—
10:15
It might therefore not be a problem to make that point very clear, from a legal perspective, because that debate is still to come. The committee could reflect on that issue further and we could always write to you after that.
A number of members raised the issue of the interplay between the Census Act 1920 and the Equality Act 2010, which we discussed at some length at the evidence session on 12 September. For example, everybody on the committee was 100 per cent signed up to, and supportive of, the inclusion of the two new voluntary questions on sexual orientation and on transgender status. Those questions were included, further to the policy memorandum, on the basis that their inclusion in the 2021 census was needed to discharge the public sector equality duty.
There would surely have to be consistency between what we are doing further to census legislation and the 2010 act across the board. Otherwise, it begs the question of the stated basis for the inclusion of the voluntary questions—which everybody supports.
Mr Matheson, do you have any thoughts on that, as our resident legal expert on those matters?
I am grateful for your directing the question at me. Ultimately, this is a statistical matter, as it is about generating the data that are required.
With regard to the legal position, I do not understand what is being got at. Is it that the census order would be ultra vires—beyond the Queen-in-council’s powers in making the order in council—by being written in the terms that it is? If my understanding is correct, the question is whether it is appropriate to be gathering the data in the way that is proposed.
I am very conscious that I have used up a lot of the time. Other members may wish to come back to that. Obviously, if part of the census is expressly stated to be informed by the Equality Act 2010, that begs the question about the other parts of the census, where relevant issues exist with regard to protected characteristics. It would be difficult to argue that those other areas are not also expressly related to the Equality Act 2010.
I am conscious that I have had a good innings.
I want to return to the no-guidance question. Given that the guidance on the sex question is controversial, it strikes me that one way through the issue would be not to have any guidance. Mr Whitehouse, in answer to Annabelle Ewing, you stated that the majority of people do not look at the guidance. However, that does not answer the question about what people would do if there was no guidance and whether that has been tested.
The evidence from the testing that has been carried out over a fairly long period is that the vast majority of people will not require guidance on the sex question. In the run-up to 2011 and in discussions with certain groups in the population after that, guidance was sought. Therefore, to enable people to respond to the census, and because of the importance of the question for data users and its use in all sorts of facets of public work, guidance is provided.
It is clear from conversations that we have had that having no guidance at all is not helpful, either to the groups who feel that they need guidance to answer the question, or to data users, who may have a different view on how the question has been asked or on the guidance, or lack of it. The guidance helps to achieve a full and complete contribution to the census and it clarifies for data users the basis on which the question has been asked. On that basis, they can therefore do the analysis that is appropriate to using what is in essence a statistical tool.
Just to be clear, are you saying that you have not tested or asked ScotCen to test a scenario in which no guidance is provided?
ScotCen asked people to respond to the question and then asked whether they used the guidance, and the vast majority of people in the general population did not feel that they needed access to guidance. In previous censuses for which guidance was not made available, people will have answered the question and the general population will have done so happily and to a high level of quality and consistency.
There is a group that has been keen to get further support, and guidance has been developed to enable that to happen. That guidance is important for people who are phoning in to our support centres and for people who are out on the streets to help people to answer the question. The guidance is there to enable people who need it to respond to the question.
Early on in the development of the question, one of the first pieces of testing that we did was to ask the sex question and see how people responded. It was cognitive testing rather than a large-scale test that is posted out. We presented people with the sex question with no guidance and asked what they thought the question meant. As we have found across other bits of development work, different people had a different understanding of what the question meant and a small group said that they would have to be given guidance because they did not know what they were being asked. It is not true that we have not done no-guidance testing. That was not included in part of the latest testing work, because of the aims of that piece of work.
I presume that all the other questions have guidance.
Yes.
So, am I right in thinking that it would look anomalous if there was no guidance for one particular question?
Guidance will have evolved over many decades to meet the needs of data providers, to provide information to data users on the basis of the question and to enable people to answer it. I do not know, but I can well imagine that, if we went back a few censuses, the guidance was not there for everything, but we have iteratively built on that so that we are now comfortable and confident that we are enabling a full completion for this important product.
In our previous evidence session with you, which I am sure that you have reviewed, we discussed that in some detail. It came out that the new guidance that was introduced in 2011 was accessed by very few people. There is a big difference between 2011 and 2021 in that it will be a digital-first census, so the guidance will be much more prominent.
In fact, one thing that came out in that evidence session is that even lesbian, gay, bisexual and transgender advocacy groups were clearly unaware of the guidance in 2011—that is how difficult it was to access—and that is the reason why some of those LGBT groups argued for a transgender question. At that point during your consultation, one organisation—I believe that it was Stonewall—argued that we need a transgender question because gender and sex are different things and people could not answer the sex question honestly. Another point to bear in mind is that big difference between 2011 and 2021 in that we now have a transgender question that people who feel strongly about their gender identity can answer. That means that the confusion about the sex question no longer exists.
The evidence from the testing and the representation that we have had is that the trans status and history question is well understood and valuable and that it provides important information, and that the guidance is valued and required by particular groups in society to answer the sex question. That is the information that we have and that is what we are trying to respond to.
You keep saying that data users have asked you for clarity and that they want guidance. However, to go back to my original question, I talked about Alice Sullivan’s letter and the 80 academics, including Professor Susan McVie, with whom you have engaged, who are all concerned about prominent digital-first guidance that erases biological sex as a characteristic. You have dismissed them, but you are not telling me which data users have requested that clarity and that self-identified gender identity be conflated with biological sex.
The committee’s papers include representations from groups of academics. We listened to those academics, and they have different views.
No. The academics who use population data are very clear that they want a biological sex question. Are you saying that you have dismissed them to listen to another group of academics, most of whom are not social scientists who use population data and who might include, for example, professors of literature, of queer legal studies or that sort of thing? I am sure that they are experts in their fields, but they are not social scientists who use population data.
I am not dismissing anybody’s views and I was clear right at the beginning that we welcome the contributions that people have made on the issue for a number of years. We are trying to build a census that delivers on that. As has been noted in the conversation, the reason why I talk about data users relates to how people understand the basis of the question when they do their academic or other work. The information that we have is that it is not helpful if there is no guidance. Even if people do not necessarily agree at the end of the day with the basis on which the question was asked or the guidance, they need to know how it was asked and what guidance supported it. That is what we are trying to achieve, because otherwise people might infer something that is not necessarily the case.
Okay. We are not getting any further forward on that. We have questions now from Kenneth Gibson.
I was just looking up what queer studies actually are, being an innocent in this particular area. However, my initial question is: can NRS explain why the table on page 30 of its submission outlining the expected responses to the sex question by population refers to “cisgender man” and “cisgender woman”? Is NRS aware that “cisgender” is a contested and politicised term to which many people object and with which many people are completely unfamiliar?
Sorry, but which paper are you looking at?
It is the table at the bottom of page 30. The heading just above the table says:
“How NRS expect people to answer based on alternative versions of the guidance”,
and the table uses the terms “cisgender man” and “cisgender woman”.
I am just trying to work out what we are responding to in the paper.
It is your sex questions recommendation report.
It is section 6 of the paper that was submitted to us.
So, it is our recommendation report. Sorry, but can you ask your question again, Mr Gibson?
I just want to know why you have used the expression “cisgender” in “cisgender man” and “cisgender woman” in the report, because “cisgender” is a contested and politicised term to which many people object and with which many people are completely unfamiliar. To be honest, until six months ago, I had not heard the term and I did not realise that I was apparently “cisgender”. I wonder why you would use a term that I do not believe is widely used in normal discourse among the general population.
We can come back to you in more detail about how we have presented that table. As Jill Morton said, we try to use language that is understood and not seen to be in any way pejorative, demeaning or insulting. We try to take an approach that reflects our need to be fully appropriate in the area. Personally, I am not entirely clear why we have used certain language, but we will come back to you on that.
However, as always, our intention is to be fully respectful and to ensure that the language that we use is understood, accepted and recognised by the groups that we are talking about when we use certain terms. If we have not done that, I apologise and we will do better. However, I understand that those terms are understood and are not seen as demeaning or insulting. If they are seen in that way, we will come back on that.
10:30
They are contested by some people and misunderstood by many. I do not understand why you would not just use the words “man” and “woman”.
This morning, an element of frustration has seeped out from members of the committee. Maybe I should not speak for others, but it seems to me that, from the start, NRS has had its own agenda on the issue, regardless of what other people think. For example, the convener talked about the letter from 80 academics. NRS did not originally want a binary question. It was only after the evidence was presented and this committee was overwhelmingly in favour of a binary question that it has been changed. Since then, you have evolved to say that a binary question is being used in the rest of the UK, so that is probably a good thing.
With regard to the arguments that we have been having about guidance and self-identification, from my perspective—and possibly that of colleagues—it almost seems that NRS is fighting a rearguard action. As an organisation, you have been dragged kicking and screaming into having to ask a binary question on sex.
I am sorry if that is the way that it is perceived, but that is not the way that NRS is working or has worked. Over the past few years, my colleagues have been trying to develop questions that respond to user need.
As I said at the beginning, it is a huge benefit to NRS to be able to engage with the committee and with the organisations that talk to you individually or as a committee in order to hear those voices and get advice that can help us evolve what we are doing.
The dialogue around the non-binary question has happened. There has been consideration of that and we are moving forward. I do not see me or my organisation operating in the way that you characterised. As I said at the beginning, I see us working with you, with the organisations that come to you and to us, and with our colleagues in Northern Ireland and the ONS to build a census so that, when the order goes through to Parliament and the regulations follow, they are broadly recognised as delivering the requirements of the census. We are having these discussions because it is valuable to have them. We had them in the past. You made your contribution and others made theirs, and we responded. That is legitimate and appropriate, given that that is the way that the process was formulated.
I fully accept that—you have explained yourself frankly and openly. However, do you accept that, as the convener said, the issue of self-identification has caused considerable concern among the people who use the data? There is frustration that NRS seems reluctant to take on board the views of those who most need to access and use the census data, as opposed to other people, who might have given evidence to you, but of whom we are not fully aware.
We are hearing from a range of users and data providers and we are trying to respond to that range of views—
I am sorry to interrupt. You explained that to the convener, but we do not know who the people are on the other side of the argument. Those 80 academics suggested that the approach is not appropriate, but we are not getting hard information as to who the people who have a different perspective are. We are hearing views, but not who is putting forward those views.
There has been representation in the press. You have seen representation from other academics. There are views as to who speaks most eloquently and appropriately about the issues. We hear from different stakeholders, including data providers and data users. Our job is to put in place a census that people can respond to, and guidance allows people to respond to that question and data users to understand its basis. That is what the census is there for.
I absolutely recognise and appreciate that there are different views and that discussions will continue beyond the census into social surveys and other work in future. There are different views, but we have to put a census in the field in March 2021, and this is our recommendation on how best to do that.
I want to focus on what I think the intention is, which is to get more people to complete the census. I have been a bit confused about the line of questioning so far, because I thought that we were here to make sure that the census is completed accurately by as many people as possible.
In answer to Annabelle Ewing, you said that all the questions are self-identification because, at the end of the day, the individual who is completing the census has to fill it in according to what they think is the right thing to do.
If I am not mistaken, your objective is to get as many people as possible to fill in the census as accurately as possible. Page 30 of your report says:
“in 2011 there was a 0.8% non-response rate for the sex question.”
Am I right in assuming that the guidance that you are providing now for the census is seeking to bring that percentage down to even less than 0.8 per cent? Am I correct in thinking that you feel that that is the appropriate way to proceed?
That is correct. The way to get the best-quality data from a census across all the questions is to have all the people answer all the questions. A primary aim is to maximise the response by households and individuals—
I understand why Donald Cameron asked his question, because it is the guidance that seems to be causing the controversy. If you removed the guidance, would you get a better response rate? In your responses to members, you seem to be saying that you want the guidance to be there because you feel that it will promote more accurate responses and a higher response rate. Is that right?
Yes.
Yes.
Thank you.
I want to stick with the points around testing. I am interested in the general population test and the test with the specific group or sub-group of trans-identifying people. Did those who identified as trans in the general population test answer the questions in a significantly different way from the self-selecting trans group? I accept that, in a general population group of 2,000, the trans sub-sample will be very small and that there are questions to be asked about statistical significance. I am interested to know whether there were any differences or whether the trans sub-sample of the general population group essentially answered or responded in the same way that the people in the trans-specific group did.
At the moment, I am not aware that there was any difference. We can come back with a specific answer to that question, but I think that you are right that the number of people of trans status in the general population survey was going to be very small. We can check that. I am sorry that we cannot answer that question.
That is okay. That would be useful, and it might address concerns that have been raised about one of those groups being a random sample and the other being a self-selected sample. If it seems that both groups are responding in the same way, that would be useful in addressing that concern.
I apologise if my next question has been answered in the papers that you have already sent. When you were testing guidance for the question based on recognised legal sex, did it explain what legal sex is? Someone’s birth certificate and passport could give different sexes, because they would need a GRC to change their birth certificate but not to change their passport, and they are both legal documents. There is no single legal definition of sex. I am interested to know how, during the testing, you explained to people how to answer on the basis of recognised legal sex.
The guidance is set out on page 101 of the ScotCen report. That is all the information that the respondent has. I am not sure whether it would be useful for me to read that out—I suspect not.
As you have mentioned it, it is probably worth doing so, just for the sake of putting it on the record.
It says:
“How do I answer this question?
The answer you provide should be the same as your birth certificate. If you have a Gender Recognition Certificate (GRC) you may record your recognised legal sex.”
There is then some other stuff about the trans status question. I stress that that guidance came out of conversations with the same stakeholders with whom we simultaneously developed the self-identification guidance, so everything that we produced was agreed with them.
Thank you. Are you aware of whether the ONS and the Northern Ireland Statistics and Research Agency have done any testing on the same questions? If so, how does that compare with your testing?
To my knowledge, they have not done the type of testing that we have recently completed, but they do a range of testing across a range of questions. As far as I am aware, ours is a fairly unique piece of testing work both in the UK and internationally.
Given that those bodies have come to the same conclusions as you have on how to go about asking such questions, it would be interesting to look at any research that they might have done. However, perhaps we could ask the Scottish Parliament information centre to look into that.
I turn to the sexual orientation question. I presume that the auto-complete feature for the other response is there for data consistency reasons. I wonder whether it was brought in because there had previously been inconsistency when questions simply had an open text box for answers. Will you explain why it is there? Am I correct in presuming that the issue is consistency in responses, or is it something else?
Consistency is definitely a part of it. For a range of reasons, the auto-complete feature is a piece of functionality that we have on our digital platform across nearly all the questions that have a write-in answer. One reason is that we now live in a digital world. People are used to completing forms digitally—for example, for buying insurance, making passport applications, shopping or booking holidays—and auto-completion is a reasonably standard feature in those forms. From a respondent’s point of view, it makes it easier for them to answer the question. From our side—this is very important—it means that we will obtain consistent data that we can code at the point at which we collect it. That introduces enormous efficiencies into all our processes for gathering and processing census data, and allows us to hit our published key output timelines. It speeds up the entire process.
The consistency issue is not about the responses that people might put in—most of the write-in answers also have a free-text option whereby people can type in anything that they want. It is simply about cleaning up data at the point that it is being entered, so that we do not have to go back later and check it for spelling errors, abbreviations and so on, all of which that feature removes. For us, achieving such efficiencies in the system is the aspect of our approach that has had the greatest impact. However, it is also about improving respondents’ experience and making it easier for them to answer in a way with which they are familiar.
That is useful. Thanks.
I have a final point. I am sorry to jump back to the sex question, but I had noted something and I then forgot to ask about it. We have previously mentioned the way in which the sex question was asked before 2011. That was done without guidance and on the assumption that people were answering on the basis of self-ID. Have you or any comparable agencies—such as the ONS, which functions under similar circumstances—done research on how people understood that question under the previous system, when there was no guidance?
There has been no explicit research. During the discussions on the 2011 practice, guidance was requested. We and the ONS took the position that that was how the question was being asked and answered, and we issued the guidance accordingly. However, we did not go back to ask people how they answered that question in 1991, for example. As I have said, the majority of people in the country understand the question and know how they wish to respond to it.
10:45
I have questions about unrelated issues but, because of the time pressures, I will write to you with them.
The whole purpose of the census is to get good-quality outputs. You have identified through the testing that you have done that few people in the general public access or read the guidance before they answer questions, including the sex question. The guidance is there to support individuals in order to get the best and most accurate answer from them at that time. You have learned that lesson from testing. What other lessons have you learned? You are about to hold rehearsals and create a public awareness campaign, which will be crucial in ensuring that people engage. The census will be done in a different way, as it will be more electronic than it has been in the past, and that might create barriers. How will the lessons that you are learning impact on the public awareness campaign?
We have done a rehearsal, as you have noted, and that has allowed us to draw on experience. We went to 70,000-odd households and got a fairly good response in a voluntary test scenario. That told us about how the systems work, whether we can put a question online, whether people can get a question on paper, and whether they can phone the control centre and get advice.
All that learning is being drawn together, and we will produce reports on it in March and beyond. It is feeding into the discussions that we are having in the NRS with our programme board, executive management board and strategic board. It also tells us about individual bits of questions, such as whether our coding was right and whether it worked functionally. Gaining all that information was the purpose of the rehearsal.
One of the great benefits of being able to talk to our international colleagues and those in the rest of the UK is that we can learn from what they are doing. For example, we can see how various approaches—such as that in Northern Ireland, in which a paper form has initially been sent out to certain groups in the population, or sending a postcard in advance to tell people that they will receive a form—have helped with census response rates. We can see how certain field force use that we did not use in our rehearsal in Scotland makes a difference to how people engage with a census. We will pick all of that up.
We have consultants who help us with our communications strategy and engagement approach, including how we position television and other media campaigns and how we work with influencers in society to engage people at the local level. Those might include community groups and others who are already engaged in communities, such as groups that help people who need additional support to respond to the census.
All that has to come together over the next year so that, when we go out in March, we make the best of our learning, our rehearsal and what others are doing, and we do better than we might have done. It is an iterative process, because we will learn from whatever we do in the future.
You have identified that there are sections of the community that might require support and assistance. It is important that you engage with them, because we do not want individuals to be turned off. Through our work on the sex question and other issues, we have found that some individuals and organisations have strong views on what is taking place, which has upset them. As we have heard, some individuals might boycott the process and not respond to the census because they are not happy about what is happening. At the end of the day, you need to manage that successfully so that we get the quality that we want. How you achieve that is vital. In reality, there is still a lot of work to do over the next few months to get to that outcome.
Absolutely. NRS is very focused on having those engagements and helping people to provide their time to enable their answers to those questions to meet the needs of data users. There will be a big campaign on the benefits of the census, which move beyond individual questions to service provision, the allocation of resources and everything from how we understand Barnett allocations to what might happen in a local authority. I take your guidance on that. That is very helpful.
My first question is brief. Has any census had a 100 per cent response rate?
No, not that I am aware of.
I would not have imagined so but, nonetheless, I thought that I would ask.
My next question follows on from earlier questions from colleagues. The final paragraph on page 29 of your report states:
“NRS are aware that some non-binary respondents do not feel that they are able to respond to a binary female/male question honestly and this may have an impact on census response for this group of respondents.”
On the work that you still have to undertake to encourage people to complete the census and to engage with more people, have you given any further thought to what you plan to do to achieve that, or are you still trying to work out what you have to do to get more people to feel comfortable about completing the census when it arrives?
We are still working through what we need to do on the specifics. As I have previously said, part of the work that we need to do over the coming year is to engage with all groups across society to help them to understand the purpose and the benefit of the census and their engagement with it.
What we have done for 2021 is quite different from what was done for 2011. We talked about that earlier. We have had an open discussion and informal scrutiny of our approach in a way that has never happened before, and that has been hugely helpful in drawing out different views and understandings, and it helps us to put in the field a census that we think best enables the data to be gathered and used for the purposes that it needs to be used for.
Once Parliament is able to agree the order and the regulations, we can say, “This is the considered view of Parliament on how we will do the census.” We must then work with everybody across Scotland to help them to feel that it is valid and relevant to them. We will work hard to ensure that. Where there are still concerns, we will do our level best to help people to feel that the census is an important thing for them to contribute to.
I go back to the question that Mike Rumbles posed earlier about the 0.8 per cent who did not respond in 2011—that is mentioned on page 30 of your paper. That might not sound like a huge figure but, when it comes to doing the research and the planning for finance and resources, it could have an effect on the outcomes. The work that you have to undertake in a fairly short space of time is immense. If you want to get the figure of 0.8 per cent down—I am quite sure that all members of the committee want it to decrease—there is a huge amount of work still to be undertaken.
There is.
I have a final question, which goes back to the question that Ross Greer posed. There appear to have been multiple versions of the trans and non-binary survey, which Jill Morton commented on. I am keen to get further information on that, particularly because the wording in the guidance is different, depending on the version.
There was one sex question with two versions of guidance, so everybody, no matter how they came into the survey, was asked the same question: “What is your sex?” We had to recruit in a different way in order to ensure that we had a sample size from the trans community that was large enough to allow analysis, so there were slightly different questions at the start for that group. However, in all the analysis that has been presented, all the participants, no matter how they were recruited, were faced with the same questions and the same sets of guidance. That bit of the survey was presented in the same way. Is that what you are asking me? I am not sure whether I have answered your question.
I want to understand why there were various versions and why the text in each version was slightly different. I accept that you will have different versions so the text will be different, but I want to understand why the wording was chosen for the guidance.
We were trying to replicate census conditions, in which people are faced with a question and, if they choose to do so, they can get some question help. The version of the help for the sex question that was based on self-identified sex says, “You can answer this how you feel.” There was a separate version around the legal sex question, which people must answer with what is on their birth certificate. Everybody got those two pieces of guidance. How they got that guidance was randomised to make sure that the order did not influence anything. Everybody saw those pieces of guidance and the same wording.
The wording that we used was derived from engagement with a wide range of stakeholders. We sat in a room and agreed the guidance. Everybody who was present and those who were not present had a view on and contributed to that. There was consensus that that would be appropriate guidance to use. I apologise if I misunderstood, but everybody saw the same question and the same versions of the guidance. There was no difference in relation to who they were or how they were recruited.
Okay, that is helpful. Thank you.
Thank you. Further to Ross Greer’s line of questioning, I have a supplementary on the use of predictive text around the sexual orientation question. You address that in your letter to the committee—you point out that the predictive answers to some questions are based on well-established lists, such as those on occupation, which are non-controversial. However, with the sexual orientation question, you consulted stakeholders. Some of the terms are not familiar to most people, including many people in that community—you will be aware of the response to that.
In your letter, you indicate that you have still not made a final decision on that. Will you consider the submission made to the committee by the LGB Alliance? It has two main concerns. First, sexual orientation, as defined in the Equality Act 2010, might be undermined and trivialised. Secondly, in relation to the predictive text, if someone answered “demisexual”, for example, we would not discover whether they were gay, straight or bisexual—they could be any of those things, but we would not get that information. Will you consider the views of the LGB Alliance? They were submitted to the committee after the row broke out.
11:00
The guidance and the predictive lists are not all fixed. On the sexual orientation question, at the moment the four categories of straight/heterosexual, gay or lesbian, bisexual and other are well regarded. Questions that are based on them are asked in social surveys in the UK and elsewhere and work well.
If I am right, one of the LGB Alliance’s arguments is about the “other” category.
There is no controversy about the categories. They are defined in the Equalities Act 2010—the committee wrote to you about that. The problem is with the suggestion that there are other sexual orientations. That is the disputed issue.
The LGB Alliance’s voice has been heard. Those categories have been well tested.
If you can indulge me for a minute, I will explain how people engage with the predictive lists. If someone gets the form on paper, they write in what they want. If they decide that the first three categories are not the ones that they want to tick, they go to “other” and write in what they want to write. That is how it is, and that is fine. For anyone who goes to the paper form, that is what they will do.
Someone who goes online—and we expect that the vast majority of people will go online—and decides that the first three boxes are not where they want to put their tick will go to “other” and will start typing in the term that they wish to use. We have heard from stakeholders and from people who have expressed a view that there are certain terms that people are likely to use. For practical processing reasons and, as Jill Morton mentioned, to enable us to quickly analyse and process that data so that we can publish the first results in March 2022, predictive text is helpful.
These are not the NRS’s terms; they are not things that we have come up with on our own. They are terms that have been suggested by groups that represent and advocate for people who are likely to be in that category. I do not want to put a percentage on this, but it is going to be a fraction of 95 or 96 per cent.
That does not really answer the question that I put, which was based on the LGB Alliance’s point.
We are still considering what should go into the predictive text. It is not for us to put in terms that we have decided; its use is to help us process the terms that people who wish to tick the “other” box are likely to use. It is a processing and functioning approach—nothing else. It is the same for all the other predictive lists.
People object to the terms in the census. It is our most important data-gathering exercise; it has status. There are clearly some people who feel that having those terms in the system, if you like, is unhelpful and damaging. Will you take that on board?
I absolutely hear that.
We will move on to other areas of the census that have caused concern among committee members.
What concerns me—I do not want to say that it concerns me more than what we have been discussing—is the ethnicity questions. I raised this issue in our previous evidence session. In the ethnic group question—question 18—you mix up geography and colour. I understand why you do that: you are trying to get the best, most accurate, result from as many people as possible. People will answer the question in different ways, so I understand why it would appear inconsistent—in my view, it is inconsistent, but I understand that.
I am perfectly happy with question 17, which is about national identity. It is really good. Regarding national identity, it asks whether someone is Scottish, English, Northern Irish, Welsh or British or holds another national identity and, if so, it asks them to say what that is. That is perfect, because you will get an accurate response.
I tested the questions just on the basis of how I would fill in the census. If I go to question 18 on ethnicity, I am confused. I do not know how to answer it. It says:
“As regards ethnic group, whether—
(a) White and, if so, whether—
(i) Scottish,
(ii) Other British”.
I am Scottish and British, so I am not going to answer Scottish because I also consider myself British. How do I answer? I ask because, in other sections of the same question, someone who is Asian can answer that they are
“Asian, Scottish Asian or British Asian”,
an African can answer that they are
“African, Scottish African or British African”,
and an Arab can answer that they are
“Arab, Scottish Arab or British Arab”.
However, that is not an option for someone with my ethnicity. I literally do not know how to answer that question. I raise the issue only to ask whether the information that you get from the way in which people like me answer that question will therefore be inaccurate.
The question of ethnicity and national identity and how those interrelate is a complicated and sensitive issue, as you say. From our perspective, it has been evolving for a number of decades, and it continues to evolve. There are different views about how one ought to ask questions when the complexity of how we understand our ethnicity and national identity and all the interrelated aspects continue to evolve. We recognise that.
The question has been broadly developed. It is part of the social surveys that we run week in and week out, and it is asked in population surveys up and down the country. The response from those surveys shows that people know how to answer it and that it is largely understood and accepted. I am not saying that everybody knows how to answer it, but it has been well tested.
There have been discussions, which will continue over the next few months and years and beyond, about how certain wording and phrases might change. You mentioned the section that says “Asian, Scottish Asian” and so on. That used to say “Asian Scottish”, but it has been flipped because doing so has been seen to enable people to respond to that question. As I say, in the social surveys that are run week in and week out, people are able to respond to the ethnicity question. There is also a write-in box, so if the tick-box options do not fully represent what people want to put they can write that in. That will include a predictive list.
I understand all that and I am sympathetic to what you are saying; I am just trying to make sure that we get it right. You have done a good job overall, but what I find disappointing is that question 18 is not as good as it could be.
You got it absolutely right in question 17, where you give all the options. May I make a suggestion? People like me would be happy if you were able to replicate that approach in the question on ethnicity. I want to be able to say that I am British. It asks whether someone is Scottish or other British, which implies that they are either English or Welsh. That is what causes the confusion. I do not think that I am particularly unusual. I think that a lot of people might be in the same position of wondering whether to write in at the bottom that they are British. It would be much better to say, as you do in the previous question, “If you are white, are you Scottish, English, Welsh or British?” and then continue the list. That would solve the problem and enable you to get the best and most accurate information possible, which is what we are trying to do, is it not?
Certainly. Thanks for your advice on that. As I say, the question that we have tested, which is used week in and week out, works, but perhaps we need to think about how we can put guidance around it to point people to that option. We know that “British” is in the predictive text for that section.
I do not think that I am unusual.
No. I am very happy that you have raised the issue, and we will look at it. However, as we understand it, the question functions.
This is what feedback is about, is it not?
Absolutely, so thank you very much.
I have a quick follow-up question on the same subject. I appreciate that we are looking at the order, not the census itself, but can a respondent select two options in the question on national identity?
Yes.
They can select all that apply.
The concerns of the Sikhs in Scotland association, and your response, are covered in the written submission that you have made to the committee. I would like further clarification around the concern that I have raised about how Muslims are asked to answer. For the first time, Muslims are being asked to identify their denomination, if that is the right word. I have concerns around how clear that is, and I wonder what further work you have done on that.
Following the previous session, we have engaged with stakeholders who have an interest in the data that we gather on Muslims. We have just about agreed a set of wording for that which better reflects what I understand that you are proposing to us, which is that, rather than just say “please write in”, we should specify that we are asking for a denomination or school. That work is in hand, and we anticipate that that will come forward.
Again, in your written submission, you say that you are considering what the potential responses under “other” might be. For “Christian”, you have 49 suggestions of various denominations and churches. For “Muslim”, you have two suggestions. Am I to presume that, at the moment, those are Sunni and Shi’ite?
That is correct. It is a slightly more complicated section, and there are a number of ways that we could have taken it, which all lead to different complications. For the rehearsal, the decision was to keep it at that level, largely because that is the user need that has been identified to us. However, it is a free-text response; people can type in anything that they choose.
As long as it is clear that that is what is being asked, I am sure that anyone in the community of Muslims in Scotland who are not Sunni or Shi’ite—it will be a very small number of people—should be able to respond.
Yes.
On the same area of the order—paragraph 18 of Schedule 2—for the other Asian, Scottish Asian or British Asian ethnic groups in subparagraph (c)(v), and for the African ethnic groups in subparagraph (d), will there be a free-text option?
Yes; both of those are free text.
What are your plans with regard to the Sikh Federation UK? In your submission, you said that there had been a judicial review that it had lost and that it had not been given permission to appeal. However I understand from media coverage that it is still saying that it intends to appeal. Are you keeping a watching eye on that?
Yes. We are talking with our colleagues in ONS, who are closer to this, as the appeal—if there is one—will be to the English courts. We are keeping an eye on developments in that area.
Okay. To finish, I want to go back to the issue that a number of members have raised with regard to who the data users were that had asked you for self-ID guidance. You said that you did not want to name names. Are you talking about the submission to the committee of 20 September 2019 from a number of academics—I think it was around 50—who wrote in support of self-ID guidance?
If that was the one with, as you say, around 50 signatories, then yes. That is one of the groups of people who have expressed—
We know that you have been lobbied by lots of stakeholders and campaigners on behalf of the community of relevant respondents, but that is a different thing from independent data users. If we talk about data users that you have responded to, are we talking about the academics who signed the letter of 20 September? Are those the people you are referring to?
If you are asking for full details of all the academics who have been in touch with us, we can come back to you on that.
It is not just academics, though, it is data users who are independent researchers and are saying that they want a biological sex question, a legal sex question or a self-ID sex question. You have said that you have been swayed by data users who favour self-identification. However, you were not able to tell us who they were. Are you talking about the 50 or so academics who signed the 20 September letter?
11:15
Those are some of the people who expressed a need for that.
In contrast to the letter from Professor Alice Sullivan, which is signed by senior academics—mainly professors—who are social and economic researchers, the 20 September letter is signed by people who are not as senior, and, in most cases, it does not say what their expertise is. I looked some of them up. I will not name them, because that would not be fair, but they include researchers into medieval literature, materials chemistry and computer studies. Is it fair to give weight to that group of people? They are not data users; they are just academics who feel strongly about the issue, as opposed to Professor Sullivan’s group of 80 senior social and economic researchers, experts in medical sociology and people such as Professor McVie, who sits on the Government’s advisory group on statistics. I am confused as to why you are favouring that other group of people, who do not have that expertise.
I do not agree with the idea that we are judging and therefore finding certain voices. We are saying that that is how the question has been asked and that is how guidance has been used. We are clarifying that, and there is a need for guidance.
Some of the discussion that is happening, which is hugely helpful as part of the scrutiny process, involves pulling out examples of where people have made an assumption about how questions might have been asked in the past. Therefore, clarification of how that question is being asked is helpful, and guidance is necessary in that regard.
We know that other data users, be they in the health system or elsewhere, see the use of the information in their context and know that they will use other information, as will many academics. Therefore, a number of the people who were part of that group of 80 will use other sources of operational and management information in order to do the important work that they do.
According to the advice from some of the people that you mentioned, it is critical that, where we use guidance, the guidance is clear, so that data users can understand the basis on which they use data. The census is a tool that does a set of things. There is a vast amount of administrative and other social data that others gather. They use that for all sorts of other purposes. They are all part of adding to the wealth of knowledge.
I do not think that we will get any further on that.
Earlier, you said that the sex question has always been a self-identified response. Was that the case in 1921?
I said that people respond to that question in the way that they feel best reflects the way that they wish to answer it. Therefore, self-identification becomes our sense of how people are answering that, when we look across the whole population.
In 1921, a 25-year-old woman answering that question did not have the right to vote. It did not matter whether she identified as a man, she still would not have the right to vote. Therefore, sex is important; it is not something that we can erase simply by changing our gender identity. Would you agree with that?
I am not sure what you are asking me to debate now. I am trying to talk about the census and putting a credible product into the public domain, so that people up and down this country can engage in it.
Thank you for coming to give evidence to us today. We will now have a brief suspension.
11:19 Meeting suspended.Previous
Attendance