Culture, Tourism, Europe and External Affairs Committee

Meeting date: Thursday, December 20, 2018

Official Report 673KB pdf

---

Census (Amendment) (Scotland) Bill: Stage 1

Item 2 is our final evidence session on the Census (Amendment) (Scotland) Bill at stage 1. The committee received a wide variety of views on the bill, before and well after the formal deadline that was attached to our call for views. I have been happy to receive late submissions—indeed, as late as I can allow—but there comes a time when the committee must take stock of all the evidence that it has received. We are well aware of the wide range of views and strong feelings about the issues that are raised in the bill, and we are grateful to everyone who has shared their view at stage 1.

I welcome Fiona Hyslop, the Cabinet Secretary for Culture, Tourism and External Affairs. She is joined by four officials. From the National Records of Scotland, we welcome Amy Wilson, who is the director of statistical and registration services, and Scott McEwen, who is the head of collections and operations for Scotland’s census 2021. From the Scottish Government, we welcome Simon Stockwell, who is the head of the family law unit, and Emma Luton, who is a lawyer.

I invite the cabinet secretary to make brief opening remarks.

Thank you, convener. I am pleased to be here to talk about the Census (Amendment) (Scotland) Bill, but before I talk about the bill, I will take a moment to speak more generally about the census.

Scotland’s census will, subject to approval by Parliament, be taken on Sunday 21 March 2021. This will be the 22nd census to take place and the 17th to be managed independently here in Scotland. In 2021, for the first time, the census will be predominantly conducted online.

For more than 200 years, our country has relied on a census to underpin national and local decision making. The census is the only survey of its kind to ask everyone in Scotland the same questions, at the same time. No other survey provides the richness and range of information that the census provides.

Some basic principles underpin decisions around the census. The key things about the census are that it counts for collective decision making, that it has to be credible, that people have to have confidence in it and that it needs to be consistent for the purposes of comparison.

There must be confidence in the census process from all our citizens in order to ensure that they provide us with their personal information. That confidence is two-fold: first, citizens must trust that we will ensure that their data is kept safely and securely, and secondly, citizens must trust that we will ask the most appropriate questions to reflect our society at the time, and that we will do so sensitively. That trust has delivered more than 200 years’ worth of data. It is why we can proudly demonstrate a consistent approach over those years. Some questions have come and gone, but we have always been consistent in our professional approach to the census and in tracking the core data.

I am here today specifically to discuss the bill. As you know, the purpose of the bill is to amend the Census Act 1920 to allow questions on sexual orientation and prescribed aspects of gender identity—that is, on transgender status and history—to be asked on a voluntary basis. The power to ask those questions on a compulsory basis already exists in the 1920 act, and refusing to answer a census question or neglecting to do so is an offence under section 8 of that act. We want to avoid that for individuals who answer the new questions.

The approach in the bill seeks to mitigate concerns about intrusion into private life by making the questions voluntary, as was done with questions on religion when they were included for the first time in the 2001 census. It is important that nobody is, or feels, in any way compelled to answer these important but sensitive questions.

The decision on the need to collect the information has been arrived at through a process of consultation and research. The National Records of Scotland has worked, and continues to work, with stakeholders to understand the needs and concerns of the communities involved. However, the NRS has recognised that more consultation is required—for example, of women’s groups—as the questions are developed. The NRS has communicated that to the committee. That consultation is under way as part of the further stakeholder work that is required to ensure that all users’ data needs can be understood and considered.

10:15  

It is widely recognised that there is limited evidence on the experiences of transgender people in Scotland and currently there is no fully tested question for collecting information. In covering that, the census would take a big step forward to ensure that we can develop the evidence that is needed to support and protect Scotland’s transgender population.

Sexual orientation is asked about in most Scottish household surveys; it is proposed that the sexual orientation question for the 2021 census would mirror the question that is used in other surveys in Scotland and elsewhere in the UK. Society will have changed significantly and rapidly in the 10 years since the previous census, so the 2021 census must reflect that.

Although the bill’s purpose is to provide for voluntary questions on the two issues, we recognise that many stakeholders are focused on whether the questions should be asked and how they relate to other possible questions. The bill is not about the mandatory sex question, but we are aware that there are strong and often very opposing views on whether a question on sex should be binary or non-binary, whether it should be related to birth certificates or legal sex, and whether it should be more focused on self-identification. The bill does not specifically relate to those issues, but it is clear that stakeholders are concerned about those matters.

To enable an element of future proofing on the legal definition of transgender, the bill uses the term “gender identity” to cover transgender status and history. That also assists in clearly separating questions about sex and transgender, so the question on sex can continue to be mandatory. However, I recognise that that has raised the concern that the bill conflates gender identity and sex. I make it clear that the intention behind the bill has never been to conflate sex and gender identity. It is about asking questions to obtain information; it is not about the law on gender recognition or equalities. I am aware that the NRS has written to the convener to indicate willingness to consider how such matters might be perceived in relation to the census and—importantly—to understand the committee’s views.

The census questions that will be asked are a work in progress. The bill is not about agreeing whether the questions will be asked or agreeing the wording of questions. The questions that are to be set will be considered as part of a subordinate legislation process that will happen next year. I assure members that the views and evidence that have been submitted will feed into further consideration of the question that is to be asked on sex in 2021.

On the questions that will be set and their wording, we expect a period of informal engagement with the committee to begin after stage 3, and to continue throughout 2019. As the Economy, Energy and Tourism Committee recommended in relation to the 2011 census, that process is intended to improve on the 2011 census process by ensuring that the committee is given the opportunity to scrutinise the questions for the 2021 census properly before formal consideration of the subordinate legislation. It is intended that the formal census order and regulations will be brought to Parliament in early 2020, but we would like to agree the legislative timetable with the committee and discuss what the committee would find helpful in order to allow the thorough consideration that will be required.

I draw the committee’s attention to a couple of drafting points in relation to the policy memorandum. The first is that it incorrectly includes intersex people under the umbrella term “trans”. That was an unfortunate action during drafting in relation to an area that is constantly developing. We recognise that the needs of trans people and of intersex people are different. We will ensure that any future documentation does not include intersex people under the trans umbrella.

The second point is that the policy memorandum says that the 2021 sex question will have a non-binary response option. It should have said that that approach is being considered and tested. The matter will be brought to the committee as part of the subordinate legislation process. I ask the committee to note those two points.

Thank you for the opportunity to speak to the committee today about the bill.

Thank you for that statement, which clarified a number of matters. I welcome your comment that the evidence that the committee has gathered will form part of the NRS’s consultation, because a number of people have come forward for the first time to express their concerns.

I acknowledge your point about the general purpose of the bill’s questions on gender identity and sexual orientation being voluntary; in the evidence that we have gathered, that is not seen as controversial in any way. However, with regard to the drafting of the bill, although you have said that the bill is not about the sex question, you will be aware that the drafting suggests that, after “sex” in section 1 of the bill, we should insert the words “(including gender identity)”. The NRS has written to the committee to suggest that it is open to looking again at the drafting.

Since it gave evidence, the Equality Network has written to the committee with specific suggestions about how the bill might be redrafted. It suggests changing “gender identity”, which it says is not the most appropriate term, to “trans status”, to be included in section 1(2)(b) beside “sexual orientation”. The issue of gender identity would be removed from section 1(2)(a). That suggestion strikes me as being quite constructive: it will probably be welcomed by a number of the opposing groups. What is your view on that suggestion?

We want to hear what the committee’s view is, having taken various evidence. The bill was drafted some time ago, and a lot of issues have become evident now in respect of terms that we use and what they mean.

The bill is about voluntary questions on transgender status and history, so we need to think about how the questions that are currently being looked at would capture both. The gender identity aspect was to have an umbrella term to allow flexibility, but I appreciate the comments and contributions that have been made, which might help to differentiate between “sex” and “gender identity”, which is an issue for some women as well as for equality groups, if we accept that the purpose of the bill is to capture transgender status and history. We shall certainly consider the suggestion, and we will look at what the committee suggests and respond to it.

With regard to the point about the phrase “gender identity”, I think it is about using the term “including”, which is what can be perceived to conflate the terms “sex” and “gender identity”. It is a very serious point that we will certainly look at; we appreciate the committee’s attention to it, and that people have brought the matter to our attention.

Thank you very much. If we assume that we will proceed on that basis, as the Equality Network and you have said, the issue of how the questions are going to be asked is for the future, and the committee will have a role in that.

You will be aware that the Office for National Statistics made its views clear on Friday; it does not think that the question of sex should be a self-identification one, or that there should be a third option on the question about sex. Its equality impact assessment, which was published on Friday, reflects a lot of the evidence that the committee has taken—in particular, from data users. The assessment says:

“The question on sex (male or female) is established in the census, and it is essential to the evaluation of inequality related to that protected characteristic. Consideration has been given to amending the question to reflect a wider range of options, given that there is greater recognition than previously of individuals who reject the traditional ‘binary’ view of sex. Nevertheless, the protected characteristic of sex as defined in the Equality Act 2010, and as relevant for the PSED, is whether a person is a man or a woman. This binary concept of sex is, in turn, fundamental to the Equality Act 2010 definition of sexual orientation and of gender re-assignment, and to the law on marriage and civil partnership and many other matters.”

After a lot of consideration, the ONS has concluded that the sex question should remain a binary question and should not be about self-identification. As I said, we heard that a lot in evidence from data users. Will you take into consideration that there are more diverse views than are in the explanatory notes for the bill?

We will look at what the ONS is doing, and it is looking at what we are doing. I am sure that the ONS will be interested in Parliament’s consideration of the issue. I reiterate that you are straying into the next stage of the process, which is when the questions will be considered. I acknowledge that the committee has received quite a bit of evidence on that matter, even though it is not the subject of the bill.

The importance of the sex question is that it is mandatory, and that we have always had that question. In 2011, it was the same question. Ten years ago, some people—trans people in particular—asked how they should answer it. As you are aware, guidance was produced for the 2011 census that said that the question could be filled in on a self-identification basis.

I am not sure that it is absolutely clear what the ONS plans to do. My colleagues might be able to give us a more up-to-date position. Clearly, a self-identified binary sex question is simple—it is the same as it was in 2011. When we try to define sex, whether legal or biological, we start to complicate the question that is in front of us.

This goes back to my basic principles—there has to be simplicity around this issue. Testing of questions is partly to find out which questions will produce the best responses. At the next stage, I am keen for the NRS to share with the committee its experience of testing different questions to see what will give the best response.

I ask my colleague Amy Wilson whether she has the latest information on what the ONS is likely to do in relation to the next stage, and the content of the mandatory sex question.

My understanding from talking to colleagues at the ONS is that, as the cabinet secretary said, the ONS proposes a binary question in 2021. However, its view is that the question will be consistent with the 2011 census. As far as I am aware, the ONS has not said publicly whether it will be a self-identification question. We will be working more with it on that.

The guidance that you produced in 2011 was online and most people did not know about it. Did the ONS produce the same guidance for 2011 in England and Wales?

Yes, it did. In fact, the guidance that we provided was the same guidance that the ONS developed in 2011.

Thank you. That is very helpful.

Professor Susan McVie from the University of Edinburgh, who is the co-director of the administrative data research centre in Scotland—I believe that she also sits on the Scottish Government’s board for official statistics—said:

“I think that the General Register Office for Scotland got it wrong when it redesigned the census in 2011 and conflated sex and gender identity into one question.”—[Official Report, Culture, Tourism, Europe and External Affairs Committee, 13 December 2018; c 4.]

We went on to discuss how that could affect data in the future, as more people self-identify. We had some clinical evidence on that.

Professor McVie talked about the importance of census data, whereas you are perhaps talking about intersectionality of data and getting down to smaller numbers. Given that people are now able to express their gender identity in another question, Professor McVie thought that this was an opportunity to maintain the integrity of the sex question because of its importance in the Equality Act 2010.

I go back to my principles. Consistency is important, which is why there is an argument for consistency with the 2011 census. However, the census must also be credible, and people must have confidence in it.

On the issue of credibility, the 2011 census was not wrong. Having a voluntary transgender question will help the statistical basis of the census, because it will mean that we can extrapolate projections in relation to population issues and males and females.

10:30  

We have heard much evidence that is supportive of the questions being voluntary, which is the purpose of the bill. However, there have been suggestions that because the questions are voluntary and people might not answer them, not enough data might be collected. To be fair to the witnesses we have heard from, most of them did not think that that was a problem.

Are you satisfied with the voluntary status of the questions? You said that there was limited evidence on transgender. Will the proposed approach provide greater evidence and improve policy making?

I think that it will. To return to my earlier point about the important principles of credibility and confidence, it is for the transgender community to ensure that it is confident about filling in the form and therefore important that the questions are drafted in a way that encourages the maximum amount of completion. Stakeholder engagement with transgender and equality groups has informed the types of question that are likely to be asked.

The content of the questions will form the next stage of deliberations on the census, which we will share with the committee. It is about trying to make sure that we maximise the uptake of the census questions. There was real concern that if the questions were compulsory, they would intrude on the privacy of transgender people. However, the voluntary nature of the questions will respect transgender people and, given that we want people to complete the census form, encourage them to fill in the forms.

On the use of the term “gender identity”, which the convener picked up on, we have received further evidence suggesting that you could change that wording. The cabinet secretary and the bill team have described the wording as a way of future proofing. I am interested in how you see that question developing and why referring to it as a trans question would be too limiting. We have had evidence that “gender identity” is not a recognised term and that there is lack of clarity about what comes under it and what it means.

We expect to be asked about such specific aspects. We will share with the committee the actual text of the voluntary questions, which will be about not just transgender status but transgender history. Given that, the view was taken that we should have a broader, umbrella term. I suppose we are in moveable territory where there is not one clear definition of gender identity or, potentially, of transgender identity. That was what informed having the flexible term “gender identity”, but there is an opportunity to be specific in the voluntary questions. We need to say that the questions are voluntary so that people will not be fined or commit an offence if they do not answer. We do not want that, especially in the personal and private area of sexual orientation and transgender. That was the rationale for using the term that we did.

We also used the term to separate out the questions of sex and gender identity. However, if the bill is perceived to conflate those issues, that does not help us. As I said, we need to have clarity in what we are doing, so I would rather that things were quite straightforward and simple. Although the term “gender identity” might be a useful catch-all umbrella term, if it raises questions around what should or should not be in the voluntary questions, which are about specific information, that does not help.

Okay—that is welcome.

My final question is around the guidance that was published for the 2011 census and where future census guidance will head. I had a look at the Gender Representation on Public Boards (Scotland) Act 2018, which has quite a prescriptive transgender definition because it states that the term “woman” includes someone who has had gender reassignment or is “living as a woman” and intending to undergo gender reassignment. The guidance for the 2011 census was about someone self-identifying as a different gender. I wonder where all that fits with the review of the Gender Recognition Act 2004 that has concluded. I do not know whether we are ahead of ourselves in some areas—or whether that makes sense.

That is an important point, which probably cuts to what a lot of the issue is about. Society has changed hugely in the past 10 years and it continues to change. In part, the census captures where society is at any point in time. Obviously, the guidance for the 2011 census came way before the 2018 act and we do not know what will happen with regard to gender recognition legislation that will follow the Gender Recognition Act 2004. As I said, it is moveable territory, and society and policy have been progressively moving on the issue.

You might argue that the 2011 guidance—as you heard, there was guidance for the rest of the UK and similar guidance in Scotland—was quite progressive when it talked about self-identification. It was also about being inclusive, which is the other point. How do we ensure that we are including people and that they have the opportunity to fill in the form in a way that is true to them? There is a balance to be found in that regard. We cannot retrofit the 2011 guidance—which I was not involved in, as I was not the minister at the time—to fit legislation that was passed subsequently or which is planned in the programme for government. It is not easy.

The census is a point in time. It will capture the information as it is in 2021. It has to reflect society as it is in March 2021, not as it might be in future. That is the point. It is about how we make sure that it is fit for purpose for 2021. Even in the course of the committee’s evidence sessions, various issues have emerged that might not have been obvious in 2011. The self-identified gender issue was perhaps not as politically controversial as it currently is.

Let me speak personally on this: I think that we need to be inclusive and to ensure that citizens of this country can contribute to the census and be part of the collective data. I suppose that we, as the Government of the day, and you, as the parliamentarians of the day, have to determine how we reflect society as it is in 2018—and as it will be in 2021, when the questions will be asked. It is not easy; I am trying to explain why there are differences between what was done in 2011, what was done in the legislation that you mentioned and what might happen in future. If we are trying to future proof our approach, there might be a point at which we have to draw a line and say, “No. This is where Scotland is just now.”

I am pleased that this country has progressed on the inclusivity agenda in different ways, but we must recognise that millions of people will complete the census, so it must be straightforward for them to fill in. We need to take as inclusive an approach as possible, while ensuring that we get the data that we need. The approach in the bill, which is to make voluntary the sensitive and personal questions about sexual orientation and transgender history and status, is the right one. That is the stage that we are at with the census.

You will be aware from the evidence that we have taken that there are a number of feminists who do not believe that self-identification is a progressive move. That is perhaps for another discussion.

Further to Claire Baker’s line of questioning, the objective is to produce data that is helpful for policy makers and service deliverers. I absolutely take the point that was made: if the definitions are not clear, how can we trust the data that is collected?

Well, we have never defined “sex” in the 200 years that we have been doing the census. The simplicity of the question is such that people answer it and we get the data. I think that that is the point. We cannot check whether the form has been filled in incorrectly; we have to respect the anonymity of the individuals who fill in the form.

That is entirely fair, but on the other side, are you concerned that if the definitions are not precise, as you said to Claire Baker, policy makers will say, “The data is interesting, but how can we be assured that it is accurate?”

That is why we have to test the questions and ensure that they are robust. The next stage of engagement with the committee will involve the sharing of the evidence on successful completion rates in the testing that has been done to date. That will be helpful at the next stage of the census development, when you are looking at the questions.

That takes me back to my point about credibility and confidence, not just for the individual but for the data users, as you said. That is essential, because if there is to be the optimum completion rate, people have to know how important the census is—that relates to what you said about it helping service deliverers, and the comments about the communication and advertising campaigns that will take place around the census—and the census must be credible, for data users as well as individuals.

Do you have a supplementary question on that point, Jamie?

Yes, it is on that theme.

Is it a supplementary? We do not have a lot of time.

It would be good to ask the question. One of the problems is that what seems to be quite a simple piece of legislation that addresses a technical issue around the cabinet secretary’s ability to add voluntary questions has opened up a huge Pandora’s box, as part of a much wider social discussion around gender recognition and identification. That is fine, but the committee still has a job to do. It is relevant to what has been said previously.

The sex question in the 2011 census was set out in guidance—albeit not widely promoted guidance—to be based on self-identification, and the question was mandatory and binary. The future census will have voluntary questions around trans history or status and gender identity—whatever the terminology or questions may be. If people answer the mandatory question through the status quo method of self-identification, what difference would that make to the voluntary questions that may be asked? In other words, there has been confusion over how to answer the mandatory sex question, because there were no other options. If there are other options, will that alter the way that people answer the voluntary question?

That is a critical point, which is probably for the next part of the process. Jamie Greene is cutting to the chase of what the issues are. It is about a social discussion—the census is always social. Asking the religious question was controversial the previous time, as was the language question, as some might remember from 2001. The bill and the census are not about recognising the status of any individual—it is not about recognition or identity per se. It is about answering questions for information.

The point about the interplay between the questions is important. My instinct is that simplicity in the questions is helpful. We understand, however, that the voluntary question on trans status and history allows us to be quite clear about what we are asking. Someone might have a history, but recognise themselves currently as in transgender status and as something different from what they were previously. That is what we need to capture. We also think that through the interplay between that question and the mandatory question we will be able to identify numbers; we do not necessarily expect there to be huge numbers, although we do not know, as we have not counted them. The point is to make sure that the statisticians can pull out numbers.

Amy Wilson may have more to add on that critical point.

Mr Greene makes very important points. That is why the testing and the work that we are doing and have done to date are of prime importance. It is about understanding how people interpret those questions and whether having a new question either immediately following the mandatory question or in a different part of the form changes how someone would answer that question. We want to bring that type of evidence back to the committee as part of the consideration, to make sure that we are clear how any changes might affect people’s responses and the consistency of data.

I have a technical question on the bill process. Why are we using primary legislation to allow the Government to add extra questions on a voluntary basis, and subordinate legislation to get into the nitty-gritty of the provisions? Is it because you need more time on the consultation, or is it easier to do it that way than to have everything on the face of the bill?

That is the way it has always been done. We are being transparent by introducing primary legislation. We could have not had a bill, and asked all the questions on a mandatory basis, but for the reasons that I have stated I do not think that that would have been right or fair. The purpose is to ensure that no fines are involved over people answering questions on sexual orientation and transgender issues. There were 15,000 responses to the original consultation and there is a great deal of interest in the census. We are starting to get into issues that are for the next stage, but clearly those points are critical.

One of the concerns of the committee that looked at the questions for the 2011 census—the substance in the census order—was that, rather than having a fait accompli at the end of the process when the committee might have concerns about changes, the Parliament and the committee should work with the process, so that they can inform it and influence what the questions might be, if there are any problems over their content.

That is why we have what I hope is a new and improved process. The issues will have been front loaded, compared with previous times. We want to be open and co-operative with the committee, because we need the committee’s views. To return to Jamie Greene’s point, this is about society. If the committee reflects society’s views, its advice will be very important.

10:45  

For clarity, I will pick up on the issue that was raised earlier. As has been said, the intention of the bill is not about any mandatory question, including the mandatory sex question. There has therefore been some error in drafting where there is language that would refer to the mandatory question. Presumably, therefore, it is in a schedule—I do not have the bill in front of me. However, the intention is that that would be deleted in section 1(2). That is my understanding. If that is not the intention of the bill, the presence of that language in the bill in front of us is incorrect, because that is not what the bill is intended to do. Therefore, would the intention be either for the Government to bring forward an amendment to delete it, or to accept a relevant amendment that is brought forward by a member of this committee to delete it, or otherwise?

The wording of the bill can be amended at stage 2 and we will obviously consider the advice of the committee on that. The wording in the bill does not affect the mandatory questions at all; it provides for the opportunity for the question on gender identity to be asked. I ask my legal support to correct me if I am wrong, but my understanding is that it could be asked about either in relation to the names, sex, age schedule—schedule 1—or in the schedule that talks about the different things that can be asked. It is therefore an either/or question.

However, with regard to your interpretation that the language is not helpful for those who think that it conflates sex and gender identity, I think that the word “including” is the problem. If it is clearer to have it in another section, we would consider that.

Okay—so the Government would be open to doing that.

The committee has received a lot of evidence and there are very strong views on a number of issues. Turning to the issue of the mandatory question, a consistently clear view from many people who have given evidence—including before the committee—is that sex does not permit a non-binary option. Professor McVie, who has been quoted by the convener already, said in evidence last week:

“Sex is about either biological or legal sex—whichever you decide to use—whereas gender identity has non-binary options. Sex does not have non-binary options.”—[Official Report, Culture, Tourism, Europe and External Affairs Committee, 13 December 2018; c 6.]

I understand that the National Records of Scotland is currently testing questions on non-binary options and I therefore wonder what informed that approach. It seems that many people were not even consulted by the National Records of Scotland; people who have expertise as statisticians, data users and a number of women’s organisations were never spoken to. If the National Records of Scotland did not speak to those people, how has it reached the stage of testing non-binary questions? Who has informed that process thus far and what will change going forward?

Last week, I asked the data users and statisticians whether, if they were now contacted by the National Records of Scotland, they would be willing to work with it. They said that yes, they would be, but they have never been contacted about the matter.

I will bring in the NRS but I make the point that we must make sure that we involve the stakeholders that would have a view on the actual content of the actual mandatory question. Obviously, that is the stage that we are at just now. That would include a lot of the people who have given evidence to the committee. I know that the NRS has communicated to the convener that that is now happening.

On why we have been testing that question, the consultation on the topics asked for people to come forward with their data needs for 2021, and there was an expression, as part of the consultation, that the question in a binary form may not be inclusive enough for some people to respond to.

Up until this point in time, throughout the process with the bill, we have not been made aware of other concerns around that; we did not have anybody who came forward as part of that consultation. We therefore followed up on the needs that were expressed as part of the consultation. Now that we have been made aware of the concerns, we are getting in contact with people who have expressed views on that issue. You are right: so far, we have been responding to what has been expressed as a data need, but there are new or previously unexpressed needs, and we will follow up on those.

Part of the advice that we will take from the committee is who we should involve. However, as the responsible minister, I will want to make sure that the consultation is as wide as possible.

There is a genuine issue about whether we should refer to binary or non-binary in the mandatory question. To return to Tavish Scott’s point, we have never defined sex. It could be defined as biological sex or relate to legal sex and so on. We have to be aware that anybody who is answering that question may have difficulties in doing so; they also have the right to privacy.

Indeed. However, as I said, the evidence that the committee has received raises significant concerns from many perspectives. A consistent theme in our evidence is that sex does not permit a non-binary option. Therefore, I find it a wee bit surprising that, notwithstanding what would presumably be at least an accepted view in society today among many people with strong views, including experts, data users and statisticians, it perhaps did not occur to the National Records of Scotland that that might be an issue. Anyway, we are where we are, and it is good to hear that the National Records of Scotland will now speak to a much wider cohort of people who have useful contributions to make to the debate. The committee looks forward to being involved in that process.

Claire Baker mentioned the voluntary nature of the question on gender identity. It has been suggested by at least a couple of witnesses that to simply equate gender identity with trans and trans history would exclude other possible individuals. Therefore, possibly having a non-exhaustive list under that broad category or umbrella might be more helpful, if we recall that the purpose of the census is to collect data that can be useful for planning purposes, including for health and other public services. What thought has been given to that? That view has certainly been expressed to us in evidence.

When following up the consultation, which raised the issue of gender identity, it was apparent in subsequent conversations with stakeholders that the actual need that was being expressed was to do with collecting information on transgender people, and that is certainly the way that the work has gone so far. However, again, as I said in response to the previous question, other issues have been raised. In our testing so far, a broad question on gender identity has not necessarily tested well; it has been less understood. Again, that probably reflects the issues to do with terminology and definition.

As we start to follow up the issue with more stakeholders, we will try to establish—if we can—what that data need is. There could be a range of needs. We will bring the matter back to the committee, to explain what the need is, any proposals that come back and why we are proposing the questions that we are.

The NRS says in its letter to the committee that it has been testing a non-binary question. Given where we are now with the evidence before the committee, will you consider pausing that? You still have to hear from a host of other people about the efficacy of such an approach.

We have been testing non-binary questions and binary questions. We will reflect on the evidence to the committee in relation to what needs to be tested. I think that we will continue to test a range of questions. To pick up on a point that was mentioned, the issue is partly to do with the interplay of the questions. Therefore, we need to look at how any proposed question, whether it be binary or non-binary, interplays with other proposed questions.

There is concern that, although you had 15,000 responses to the consultation, the evidence that we have is that a huge number of groups that represent women—and women comprise 52 per cent of Scotland’s population—feel that they were not adequately consulted. There is a wee bit of a need to go back to the drawing board; that is important.

Cabinet secretary, you mentioned that sex has never been defined but, in the 200 years of the census, I do not think that it would have occurred to most people that there is anything other than male and female. When I was growing up, I certainly never thought of anybody as being other than male or female. I appreciate that society might have changed in terms of how people self-identify, but the evidence that we have had was fairly conclusive that people are dimorphic and are born either male or female. Therefore, I would think that the first question should ask what sex a person was when they were born and then there could be voluntary questions.

You have talked about consistency with the 2011 census, but I am concerned that two rights do not make a wrong. If it was wrong for the question to ask about self-identification, should we not change the guidance now to make it much more specific so that people know what question they are being asked?

I appreciate your points. That is your view, but other people have different views. I do not think that we can say that the 2011 census was wrong, because I do not think that it was. However, you can have an opinion on whether that should have happened. That is the point of consulting the committee.

On what we do in the future—this goes back to Jamie Greene’s point—we have to reflect society as it is. Of course, the issue of transgender people would not have occurred to most people over previous decades—although it might have occurred to transgender people—so there was no understanding of how transgender people might engage with the census. Although that understanding was not there, there would have been transgender people answering such questions over many years. Currently, society more generally is more conscious of the existence of transgender people.

The point is about trying to capture the information as it is in 2021—we are not even there yet—and how society views things. However, we have to make decisions now because the timetable for processing the issues means that we have to have the evidence and so on to produce the regulations that the committee has to scrutinise. That is why we are not defining it and just having a male or female question.

As Kenneth Gibson said, there is biological sex, legal sex and so on and there is not one definition. It is only nowadays that we are starting to identify that. If someone has a gender recognition certificate, their legal sex will be different from their biological sex. Therefore, it is not right to say that we should never consider what male and female are because we do consider it. My instinct is to keep it as simple as possible and be consistent in what we have asked for the past 200 years.

I never said that you should not consider what people feel their sex to be. I said that we should identify people according to the sex that they were at birth. We are talking about the Equality Act 2010 and the protected characteristics of sex. I will quote the briefing:

“The Equality Act 2010 refers to ‘sex’ in binary terms – ‘man’ or ‘woman’ (s.11). It defines ‘woman’ as ‘female of any age’, and ‘man’ as ‘male of any age’ (s.212(1)). The use of ‘sex’ in these definitions is generally understood in the biological sense. However, if a trans woman has obtained a gender recognition certificate, she will receive a new birth certificate stating she is female. This is despite ... being born biologically male.”

The 2011 census seemed to ignore that and people could self-identify. The reason why so many women’s groups have given evidence to us to express their concern about that is that they feel that there is a threat to women in a whole host of areas—we do not have time to go into those, but the evidence that we have taken goes into detail. Their evidence is that self-declaration can threaten women, particularly women who get intimate care from people who may be trans who are self-declaring but who are still biologically male.

I appreciate those points. We cannot retrofit where we were in 2011 and we have to deal with where things are now. The issues that have been brought to the attention of the committee are far more alive and current than they were in 2011. That is where the responsibility of the census is at any point in time. The question is where society stands now and what we can do to capture the information that we need. In public policy, we now have the equality duties in legislation to ensure that public bodies are dealing with equalities issues. We have a different situation now from the one in 2011 in a variety of areas.

This might be straying into the next stage of the consultation, but we also have to ask whether there is an opportunity to keep the clarity and simplicity of the male/female question and have a voluntary transgender status and history question or, if we do not want to conflate gender and sex, whether it is in the interests of some women’s groups to have a non-binary “Other” option for the mandatory question to separate the gender issue from the sex issue. Having a non-binary option might also be in the interests of the equality groups, as people could then say what their view is and choose “Other”. There are arguments for the non-binary option from the women’s groups’ perspective, because it separates out sex and gender, and from the equalities perspective.

11:00  

The problem might come with people saying, “Hang on, you are either male or female, and that’s it, so we shouldn’t have a third option.” All that I am saying is that there are different views from different perspectives in society. Our responsibility, collectively as Parliament and Government, is to steer us on a road that captures enough information to make the data credible and to give confidence that we are counting in a way that is meaningful for those who have to use the data. I am not saying that it is easy. When we come to consider what should be in the mandatory question, we will consider whether we just have a binary question and rely on a voluntary question to give us a perspective on people who have a different gender identity from what their sex is defined as.

With respect, I am not aware of any women’s group that is suggesting that the question should be anything other than binary. If there are such groups, they have not been in touch with the committee. Women’s groups seem to be pretty consistent as far as I am aware.

You have to reflect on the evidence that you have been given. However, as the convener said at the start, the call for evidence to the committee was about the bill, which is about the voluntary questions.

If people have a choice of “Male”, “Female” or “Other”, they might decide to tick “Other” just for a laugh. I am not being facetious. In the 2001 census, more than 400,000 people put their religion as Jedi. If the choice is “Male” or “Female”, we will get a much more accurate figure, and we can then have voluntary questions about orientation and gender, on which I think everyone on the committee agrees.

That is an important point that goes back to the point about credibility and confidence. The testing is really important, because we need to see how people respond and whether there is a temptation for people not to treat the questions as seriously as they should. The census is serious, given the information that it provides us.

Is there a danger that adding a third option in the sex question would mean that the census data would not be compatible with the previous data that we have? The whole idea is to ensure that planning processes are in place. We need to have correct data, because it is used in many sectors, especially the health sector, to provide services.

My point about the need to be consistent is important but, as the committee has heard in evidence from the national health service, it uses the sex information at a very high level. We anticipate that, even with the voluntary transgender question, only a small number of people will fill it in and that will not necessarily affect some of the global data that we need through the sex question. Not being consistent would not necessarily be an issue in that regard.

Again, there is a statistical issue about what can be projected, so Amy Wilson might want to come in.

One area that we have not talked about yet—partly because, as we have said, it is for the next stage—is about what we would output. If we ask a non-binary question—that is the big if and is obviously something for the committee to take a view on—we do not propose to produce outputs on a non-binary basis. In our conversations with stakeholders, we have always been consistent that it is about allowing people to respond in a way that reflects how they identify but that we will still produce outputs on a male and female basis. We have discussed with stakeholder groups the fact that we would randomly assign people back into the male and female categories because, as the numbers are expected to be very small, that will not affect the statistical distributions. That is seen to be acceptable, as long as we do it randomly and do not try to establish through other information that has been given whether we believe people to be male or female.

We are working on that and we will continue to do so to consider, if that is the question that is asked, how we will deal with that in the outputs.

It is vital that you identify that. By doing that, at least you try to tackle the situation. Otherwise, the issue will become much more complicated and you might open up another can of worms in many respects.

That is the crux of the issue—not for the bill but for the next stage. If we are not going to use non-binary data as non-binary data and are just going to use binary information, why ask the question in the first place? The point is that it allows people to respond who might otherwise not feel comfortable about responding. That is the decision that has to be made. To go back to the principles behind this, we are counting people for a purpose in the census. If we are not counting them for a purpose, why do it? It comes back to the societal point—should the census be set up in such a way as to allow responses, even if we are not necessarily going to use that response information for a data purpose?

Keeping it simple is more reliable and reflects the usage of the data but it does not necessarily reflect how people might want to contribute when filling in the census in the first place. I would welcome the committee’s advice on what your priorities are. That will help us with the next stage, which is the mandatory questions.

That is quite important.

Good morning, cabinet secretary. You have mentioned a couple of times that, even prior to 2011, there has not been a specific, narrow definition of the sex question, whether in relation to a biological characteristic or any other specific characteristic. It is important to draw out the fact that the guidance in 2011 was not a departure or a change but rather a clarification of the way in which the census has always operated.

That being the case, I wonder whether you have seen the evidence that has been submitted by a coalition of national women’s equality and violence against women organisations, including prominent, well-respected organisations such as Scottish Women’s Aid, Engender and Rape Crisis Scotland. Those organisations say:

“We ... have a long history of deliberation on the interrelationship between trans equality and rights and women’s equality and rights.”

Quite unlike the earlier suggestion that a huge number of women’s organisations were expressing concerns and felt uninvolved in this discussion, they say:

“While we are each engaged with the Census process in our work, our view was and remains that the proposals as consulted upon will have little impact on gendered-data gathering and analysis.”

However, they also say:

“we were concerned to hear calls to the Committee to reverse the practice of the Census 2011 and Census 2001, and mandate respondents to describe their ‘sex at birth’.”

They are concerned about that, saying that they are not aware of any problems with the current approach and that

“trans individuals responding to the question on sex with details of their lived identity”

is appropriate and does not cause problems. They say that

“In most instances, this will accurately reflect how a broad range of public bodies and providers of goods, facilities, and services will understand and treat them”,

and that a departure from that approach would cause problems by breaching rights to privacy.

Does the Government agree with that analysis and share that position? In particular, does the Government agree with the suggestion that, for most trans people, their lived identity is most closely relevant to the way in which most data users of the census will engage with them and treat them in respect of services?

I understand that. I have not read that evidence to the committee, but I will make a point of doing so. It runs somewhat counter to the points that the committee has made about the evidence and representations of some women’s groups. That is part of the NRS consultation on the next stage, which is on the mandatory questions.

It comes back to the point about lived experience as opposed to biological or legal sex. That is why the guidance was the way it was in 2011. However, we are not dealing with the 2011 census; we are dealing with the preparations for the next census. The census does not lead public opinion; the census has to reflect society as it is just now and ask questions that maximise the response rate so that the data can be used. It is important for the Government to make sure that that happens.

I appreciate the points that have been made and I think that we will have to get to the nub of those—what are the issues that different organisations, and women’s groups in particular, want to get at and how much concern is there about self-identification and what it means? However, the bill is not the place to resolve those issues—they should be resolved somewhere else. That is why we must be clear about what we are doing. We want the census to count people and we want it to be credible and instil confidence. Those are the primary aims. I am not taking a Government view on whether people are right or wrong to say such things, I am saying that we almost have a guardianship of the census and it is my role to make sure that it is credible, that people have confidence in it, that it counts people and that there is a consistency so that we can use the data properly.

Indeed. It is a 10-yearly population-level snapshot. It is not about recording individual information as a national identity database would, which I think most of us would oppose on principle. If the Government continues, as I suggest that it should, to stick to the historical situation in which the mandatory sex question is answered by each individual honestly in their own terms, rather than being defined in relation to one characteristic, and the Government also decides not to have an “Other” option, how should a non-binary person answer the binary question? Is there not a fundamental problem that, if we maintain the position that people must answer the question on sex honestly in their own terms but do not include an “Other” option, we inevitably risk people feeling that they have to give inaccurate information?

That is the point that Alexander Stewart made about allowing responders to respond accurately. As you have said, I think that most people would want to answer the question based on how they live, as opposed to their sex at birth or their legal sex. That is why the extrapolation of the voluntary question on transgender status would allow us to make sure that the information is as accurate as possible. That is the balancing act that we will have to achieve and we have not determined how to do that yet. That is why we are engaging with the committee to get its view and doing so earlier than we did for the 2011 census. There is something to be said for keeping the question as was previously but new issues have been brought to the attention of the committee and we have got to hear the committee’s views on that balance, in particular, the convener’s point about women’s groups. I have not seen the collective evidence that Mr Harvie referred to, but I presume that it was submitted to the committee.

It is on the committee’s page of the website.

As far as I understand, the suggestion to change the reference to gender identity and to refer instead to trans status was first made in the letter from the NRS of 5 December. The idea of being open to that change was in that letter. If there is active, on-going consideration of that, would there be any potential problems if amendments were made to the bill that restricted the kind of question that can be asked? Is there a change that has to happen, or do we need to resist making restrictive changes to the bill that would close down that consideration of the options?

That is why we will have to reflect at stage 2 on what changes can be made in order to give clarity, which is what everybody wants. We have always been quite clear that the questions that we want to ask—unless the committee is about to tell us otherwise—are about transgender status and history. If we want to ask voluntary questions about those, we might be able to be explicit about that. That is exactly what we will ask about in respect of the drafting. The amendments at stage 2 could limit things to that. I do not think that that would limit us in any other way.

The irony is that, for the mandatory questions, we can ask anything—there is far more flexibility in their scope. All that we are trying to do is to protect people from committing an offence or being fined when they answer those questions. I do not think that limiting the questions in the way that was suggested earlier by the convener—limiting them to transgender status and history—would unnecessarily restrict us, unless our lawyers tell me otherwise. However, we may need to respond to the committee on that in writing.

11:15  

We should consider that further down the line. Discussing the effect of specific changes or suggested amendments to the bill is not for this moment in time. We should consider those things going forward into stage 2.

Just for clarity, in the Equality Network’s submission to the committee, it suggested including trans status, but it said that it did not think that that would pin down the question and that the consultation process was open to further developing that question. That did not seem to be a concern for it.

We have the caution of lawyers in the room. However, that is my understanding, as well.

Good morning, cabinet secretary. I was on the committee 10 years ago when we went through that process.

Ah!

There had to be one.

All will be revealed then.

I have found the process to be helpful. Earlier on, you mentioned front loading. I genuinely think that the approach has been helpful. The committee has been teasing out particular issues for the next stage. I simply wanted to make you aware of that.

A number of points have been raised in the evidence. Last week, we heard about issues relating to the guidance that was available for 2011 and directing people to the online guidance. Has any consideration been given to that so that, when we get to the census, there will be an improved process and more people will be made aware of the online guidance?

Different questions come and go, and the guidance is very important for particular groups. You will remember the language issue. A lot of stakeholder groups tried to encourage people to answer the question on that. There was quite a lot of activity around that new question, what it meant and how people should engage. We would also want that on the voluntary question on transgender in particular in order to encourage people to answer the question so that we have that information.

On your point about guidance, the process has identified that, although the bill is straightforward in a sense because it is about the voluntary aspects, the next stage will be more complicated because it will be about the mandatory questioning. That will be very detailed. The guidance will sit alongside that. The process and the evidence sessions have shown that the 2011 guidance evolved during the process of that census development, but it was clearly very important.

On our transparency and engagement with the committee, it is really important that we ensure that the guidance is given the attention that it deserves and that, in relation to potentially controversial issues or otherwise, the guidance is given as much importance as the questions. Your advice on that is well taken.

I have a question on a different aspect. You mentioned that the census will mostly be online. I take it that there will still be paper copies and copies in various other formats, including Braille, for other people in society, particularly those who may have a visual impairment.

Absolutely. To pick up on that last point, we will continually test things with individuals and groups to ensure that all the information that we provide is accessible and easy to understand. We are working alongside equality groups and groups that represent people with different needs to ensure that what we do is appropriate for them.

That is helpful. I posed that question because I chair the cross-party group on visual impairment.

That is very important. Obviously, there are different aspects of the census. We are focusing on particular issues relating to the voluntary questions on sexual orientation and transgender, and we have touched a great deal on the mandatory question on sex. However, there is an awful lot else around the census, so we offer to keep the committee up to date on the whole project.

I am sure that we would appreciate that.

Have you finished your questions, Stuart?

Yes.

Patrick Harvie mentioned the submission that was made by a number of women’s organisations, under the umbrella of Engender, in which they raised concerns about how public bodies collect data on protected characteristics. They believe that there is a lot of confusion about the fact that sex is a protected characteristic. As Kenneth Gibson said, it is defined as such by the Equality Act 2010, as is gender reassignment. In the equality impact assessment for the bill, the NRS dealt with sex and gender reassignment under the same heading and talked mainly about gender reassignment. That contrasts sharply with the ONS’s equality impact assessment, which dealt with those two protected characteristics separately.

Do you acknowledge that there is some work to be done across the Government and in your department to ensure that every area of Government and all public bodies adhere to their obligations under the Equality Act 2010 to deal with sex as a protected characteristic independently of gender reassignment?

The answer to that is yes. Across Government, we can always improve, as the exercise on the bill has identified. However, I would say that the equality impact assessment is about the bill, which is about gender identity, not sex. It is not about the mandatory question on sex; in effect, it is about gender reassignment, gender status and sexual orientation, which will be captured by the voluntary questions. The equality impact assessment relates to the bill that is in front of us; it is not about the census process generally.

I welcome the fact that you intend to go back to various groups to gather views. You will be aware that there are differences of opinion even within lesbian and gay communities—for example, there are many lesbian and gay people who have concerns about self-identification. There are also women’s groups that disagree with some of the comments in the submission that Patrick Harvie mentioned, and they have made their own submissions to the committee. Can you assure us that you will take all those views into account and that you will not go only to what we might call the usual suspects, who are funded by the Government and whose money depends on them taking a particular position on such issues?

I hear what you say. It is important that the consultation that we have when we get to the next stage—the determination of the content of the final questions—is as wide as possible. You are right to say that many people have different views and opinions, but even though we want to take on everyone’s views and opinions, we will still need to come to a judgment at some point. I hope that I have given the committee a sense that, as the guardian of the census, I must ensure that the census is credible. There must be confidence that it will count people in a consistent way.

We will not be able to please everybody. The census provides a snapshot in time, and we will need to determine the questions on the basis of where society is at that moment in time. Although, as the minister who is responsible for the census, I will have to make the final decisions, I will need to take on board the committee’s advice. The committee can help me in that task, and I appreciate its consideration.

Thank you. With that, we move into private session.

11:23 Meeting continued in private until 11:36.