Treasury Committee
Oral evidence: The effectiveness of gender pay gap reporting, HC 2240
Wednesday 5 June 2019
Ordered by the House of Commons to be published on 5 June 2019.
Members present: Nicky Morgan (Chair); Stewart Hosie; John Mann; Alison McGovern; Catherine McKinnell.
Questions 1 - 91
Witnesses
I: Nigel Marriott, Independent Statistician, Marriott Statistical Consulting; Sarah Gordon, former Business Editor, Financial Times.
II: Hilary Spencer, Director, Government Equalities Office; Elysia McCaffrey, Deputy Director, Government Equalities Office.
Examination of Witnesses
Witnesses: Nigel Marriott and Sarah Gordon.
Q1 Chair: Good morning, and thank you very much to our first panel for being here this morning. My name is Nicky Morgan; I am the Chair of the Committee. I think we are also being broadcast on BBC Parliament, which shows the interest in this subject. I am going to ask our panellists to introduce themselves.
Nigel Marriott: Good morning. My name is Nigel Marriott and I am an independent practitioner who specialises in providing training courses to non-statisticians and working to explain statistical ideas to them, so they can use them in their work. In addition, I am representing the Royal Statistical Society, since I was co-author of their recent report making recommendations for improving gender pay gap reporting.
Sarah Gordon: I am Sarah Gordon. I have recently stepped down as business editor at The Financial Times. We did a lot of digging into gender pay gap reporting and issues around diversity in the workplace.
Chair: Thank you both very much for being here for the first panel this morning. Mr Marriott, we will try to speak slowly and clearly but, if there is anything you do not catch, please let me know and we will correct ourselves.
Nigel Marriott: I will.
Q2 Chair: I will start with some overview questions about whether you think gender pay gap reporting is making a difference and what you say to those who argue that there is no gender pay gap. Ms Gordon, we will start with your perceptions.
Sarah Gordon: On the first question, the numbers suggest that the reporting requirements are not working, in the sense that the median gender pay gap between 2017 and 2018 widened very slightly. The requirements are not there to show the gender pay gap; they are there to encourage companies and employers to think about the representation of women in their workforce. Unfortunately, all the other numbers suggest that brilliant progress is not being made; they do not seem to be giving employers the right nudges, because the number of female chief executives in the FTSE 350 has fallen and the number of women on executive committees has remained pretty static. You are not seeing the follow-through from the reporting requirements that the Government hoped would come as a result. Of course, we are only at the beginning of year 3, so these are possibly early days.
On the second question about whether there is a gender pay gap, one of the muddles that people get into about this issue, as you doubtless know, is that the gender pay gap is not the same as equal pay. If, as an employer, you have a gender pay gap, as most employers do, it does not mean you are not paying women equally for doing the same job as men. It is trying to show something different. The gender pay gap essentially shows that, generally, you have far fewer women than men in the top echelons of your company, school or trust. It is drawing attention to that. That issue is very different from equal pay, although our reporting suggested that some of these employers also had an equal pay issue, which the reporting has at least made a little more transparent, particularly and importantly to their own employees.
Q3 Chair: Mr Marriott, what do you think? Is there a gender pay gap and what are we learning from the figures at the moment?
Nigel Marriott: I am going to pick up on the end point about the understanding of the words “gender pay gap” first. You are right: there is so much confusion. The confusion arises from the words in there, “pay gap”. I believe that, if you are going to use any sort of technical terminology, the terminology has to do what it says on the tin. Sarah has picked up on the key point that this is not actually about pay; it is about the representation of women in higher pay zones and senior roles. In that case, consideration needs to be given to changing the name of this exercise and calling it representation across pay scales. Otherwise, if you say there is a gender pay gap issue, people immediately think that women are not being paid the same as men.
Being a statistician, I have to back up my statements with data, so I have done a couple of quick surveys. First, on Twitter, I searched for “gender pay gap”. Just from the headlines that came up, nine out of 25 had clearly conflated gender pay gap and equal pay, including, I am sorry to say, from prominent names, such as Kamala Harris the United States Democratic candidate, the United Nations women’s organisation and a couple of others I have printed down somewhere. We have that problem of understanding and that can create a barrier in receptiveness to what the goal is.
For me, the goal is to have the concept of diversity, not just gender diversity but all kinds, considered in the same breath as brand image, product positioning and customer perceptions—the things that all organisations and businesses focus on. It should be considered that we need diversity to be successful. If that is the goal of the exercise, I ultimately see success to be when that happens, less so the numbers. Rather, the numbers themselves should not be the endpoint of the exercise. I agree with Sarah that there has been no progress overall but, when you start breaking it down, you can find examples of progress. We are only into the first year of the exercise and I am sure we will be discussing why there has been no progress so far in other questions.
Q4 Chair: The other question at the moment is that only employers with over 250 employees have to disclose and prepare, although others can on a voluntary basis. Do you think that it should be extended to much smaller employers, perhaps with 50 employees or more, or should there be no employee number limit at all? Should we just ask all employers to do this? Mr Marriott, I know that the paper you have just produced showed that some companies were already struggling to produce these numbers.
Nigel Marriott: Yes, but, on the question of the threshold, I am opposed to reducing the limit at the moment. That does not mean smaller companies cannot do it, but my concern is fundamentally statistical. A pay gap is a comparison between two samples. Sample A is men and sample B is women. It is a snapshot at a point in time. If a company were to repeat that exercise in six months, because of people moving around, the numbers would change. Therefore, there is a margin of error in any estimate for a company. If the sample size of one of the groups, whether men or women, is small—certainly if it is less than 50 and, I am arguing, even if it is less than 100—the estimate becomes unreliable.
I did an exercise and I can give you a couple of examples here. If an organisation has a 90-10 split gender-wise—it does not matter whether it is male or female—and 300 employees, I have shown that the margin of error, or what is known as the 95% confidence interval, can be anything from 35 pence in a pound on one side up to 45 pence on the other. By the time you get up to larger organisations of 1,000 or 2,000 employees, that falls to plus or minus 5 or 6 pence. It is purely a statistical concern. If you have a large margin of error for a company of only 100 employees, it could be unjustly accused of having a large pay gap when it is just statistic chance, or an organisation that is discriminating against women could just have been lucky on that day and show no gender pay gap. That is my prime concern. That is not the same as saying smaller organisations should not be doing the exercise. It is whether they should be held to account if they have small sample sizes.
Chair: Ms Gordon, do you have a view on the threshold?
Sarah Gordon: Yes, I do. I feel strongly that the focus should not be on employers with fewer than 250 employees. It is already a significant regulatory burden on smaller companies and public bodies. Certainly, the companies that I talked to in my reporting, as you know, all felt strongly that there should be less regulation rather than more. The issue is also about impact. It is about what the Government want to achieve from these reporting requirements. It is far more effective to focus on what the larger employers are doing and how you are holding them to account or trying to change their behaviour.
We noticed a particular issue about changes between the first and second years. In the first year, companies were extremely worried about the impact of this greater transparency of numbers. In the case of the financial services companies, which I know are your primary focus, the numbers were appalling. They thought, “Oh, the next thing that will happen is that requirements are put on us to do something about it”. Of course, none of that happened. In the second year, 48% of employers made progress in narrowing the gap, but 44% did the opposite. I sense from my reporting that the employers are thinking, “Phew, that was not as bad as we thought it was going to be”. That is the thing to focus on.
The second issue is that the really big companies have no excuse for their behaviour around this. Our reporting found that there are 61 employers with 20,000 or more employees and, of those, one in four did not produce an action plan. There is no excuse for that. From my understanding, Nigel works with small employers that simply do not have the resources or struggle to do this properly. A large company or employer does not have that excuse, so the focus of Government should first be on those larger employers that are simply not stepping up to the plate on this issue.
Q5 John Mann: Mr Marriott, you might like to have a go at this first. Does a company have any discretion in how it makes its calculations for gender pay gap?
Nigel Marriott: I have been told, and this is simply second-hand by someone who has read the legislation in greater detail than I have, that there is discretion. There are two kinds of discretion here. One is the practicality of interpreting who the relevant employees are. What is your headcount, who are the relevant employees and who would you include and exclude in your gender pay gap calculations? When you get down to that detailed definition, people can obsess over the fine details of the wording. “Are we actually complying with the legislation?” This is my attitude, and I teach this in my training course: when in doubt about whether you have the right data, try the calculation with a different dataset and see if you get the same answer. If you get the same answer that you have got a gender pay gap of 10%, regardless of whether you include or exclude those employees, do not sweat the detail; just report that figure. The only time you have to consider discretion or seek advice is when, by doing it one way, you get a different answer to another way.
Q6 John Mann: How are part-time workers calculated?
Nigel Marriott: Part-time workers should be included, because the gender pay gap calculation is based on hourly rate. Whether you work full or part time, your hourly rate should be considered, not your total salary. I would not be surprised if some organisations are getting confused about that and are using total salary as the basis of their calculation. I am pretty certain I have seen one or two examples of that.
Q7 John Mann: Does either of you have any suggestions for us on how the method of calculating could be improved?
Sarah Gordon: Yes, I have a number of suggestions, specifically on the two points that you have just raised. On discretion, one of the problems for anybody wanting to find anything useful from the dataset is that larger companies and employers report in their subsidiary components. BAE Systems, for example, reports as seven separate company blocks. It is perfectly legitimate to do that according to the requirements, but that makes it extremely difficult to interpret the data properly. If an employee wants to understand their overall employer’s gender pay gap, there is no requirement to report that. Some companies voluntarily come up with an accompanying narrative, which often goes with an action plan, although that is a separate issue. That gives a blended number for the whole organisation but, without that number, it is often difficult to understand what is going on, because some of these entities have 250 employees and some have 10,000.
The second issue is that one of the blatant problems with the data is the decision to exclude partners from the calculations. For example, if you work in a professional services firm, such as Ernst & Young, or a law firm, and you are a partner, so in the most senior rank of that firm, because you qualify as self-employed, the numbers do not have to report on you. All professional service firms voluntarily reported the numbers including and excluding partners, which they did because of pressure from clients and, I would argue, general pressure and pressure from the media. The law companies, for example, had not done this at all and we know there is a serious issue with lack of representation of women in the senior ranks of those firms.
There is another issue about part time and full time. The ONS did some research last year, with which you may well be familiar. Explanations for the gender pay gap were given. Most people think it is because women have babies; it is really not because women have babies. The ONS found that the biggest statistical explanation for the gender pay gap is that, once you start working part time, your career progression does not just slow; it stops. Because more women than men work part time, it is the largest statistical explanation for the gender pay gap. The problem with the data as currently recorded is that, as Nigel says, it is hourly—which is correct, as it should be measured hourly—but you have no way of differentiating between how many employees are part or full time. That would give a far better and more genuine picture of what is actually going on at an employer. There are a number of other ways in which the data could be better represented on the website. There is no measure of progress and no way of comparing your employer, or a given employer, against other employers, the sector or the overall number. They are neither providing a spur to action at the moment, nor giving anybody consulting the dataset a really good picture of what is going on.
Q8 John Mann: Could I clarify your first answer? Take the Treasury, for example, and all the Treasury agencies. Are you saying that it would be satisfactory simply to have one report, rather than a disaggregated report?
Sarah Gordon: No, you have to have the disaggregated numbers but, from those disaggregated numbers, you can provide one number, which is the picture of you as an employer. With respect, the Treasury is not a good example. We are talking about an organisation like Lloyds Bank, which produces several numbers. It is quite difficult, from looking at those several numbers, to understand the overall picture.
Q9 John Mann: I understand that. We would like the Treasury to be a good example. Mr Marriott, do you have any comments? Should the data be audited?
Nigel Marriott: To pick up on the point that Sarah has been making, some of the issues over partners are down to definition. In my understanding, partners have to be included if they are paid a salary. If they are paid a salary they must be included. It is only when they are not paid a salary; that is what I am told by ACAS, at least.
In talking about how data could be improved, I must pick up the point about the group level and when a business is split into multiple units. Yes, those multiple units should be reported separately, but the group should report as well. There is a good statistical reason for that. If the company has a large gender pay gap and is just reporting as a single company, it may want to eliminate that pay gap creatively, without doing anything. The way to do it is to get a statistician in, a dishonest one obviously. If I were to do that, I would say, “Split your company up into groups. Put your low-paid workers into group 1 and your highest-paid workers into group 2. You will have no gender pay gap in those groups”. You split your company into groups with no gender pay gap and do not report at the group level, and you cheer. Once you report at the group level, you will see a massive gap, so it is critical that the group level is reported.
On improving the way that data is presented, the Royal Statistical Society has put out a paper. I will pick up on two things in particular. First, on output, we would like to see a greater emphasis on income quartiles. I have to say that is an incorrect term; they are actually quarters. A quartile is a boundary between quarters. I have to say that, because otherwise statisticians will stick it in my neck. I will stick with the term “quartile”. That is interesting information for me, because for lots of organisations it is 50/50 balanced throughout. Once you get to the top quartile, there are very few women.
Recommendation 6 is that we would like to see the median gender pay gap reported within each pay quartile. Within the top income quartile, you would report your median gender pay gap. That would be really useful, because it would give a clearer indication of potential equal pay issues. So far, I have seen five or six organisations that are already doing that. WHSmith is probably the most prominent one. We tend to see that there is no gap in the lower three income quartiles. Once you get to the top one, there tends to be a gap of some kind, but it is usually down to earning opportunities for men versus women.
The other recommendation for improvement is around explanation and guidance to employers or the public on how the figures are calculated. I have to say that all the guidance published by bodies on how to calculate, particularly the median and the quartiles, are very ambiguous. That includes the Government Equalities Office, the Equality and Human Rights Commission and ACAS. All three have produced ambiguous guidance, which is easy to misinterpret. The big concern here is over the median. If you ask people, as I do all the time in my basic statistics course, what a median is, most people say it is the middle value, which is ambiguous. That invites you to think the middle value is the middle between the maximum and minimum. I can show you the name of a finance company that has made that mistake. It was unfortunate enough to have published that data, and it was clear that was what it had done. It is of course the middle-ranked value. When you stand everybody in a line, it is the people standing at the middle of the line. None of the guidance states that explicitly. There should be graphics and videos that show you how to do that.
Q10 John Mann: It is possible to be inaccurate if you make mistakes. Is either of you aware of any examples of people who have deliberately reported inaccurately?
Nigel Marriott: I do not think you can say anybody has done things deliberately, from my experience of working with non-statisticians.
Sarah Gordon: Not deliberately, no. Neither of us is an auditor or an enforcer. The EHRC has responsibility for enforcing the regulations. One of the things that surprised us was that, as you probably know, we discovered an enormous number of anomalies in the data in year 1. There were fewer in year 2, but were still some anomalies. There are still anomalies in the data in year 3 so far. The EHRC has said it has followed up with a number of employers, but there have been no fines or sanctions. It is extremely puzzling to me that, out of the large number of anomalies particularly in year 1, there were relatively few or almost zero consequences.
Q11 John Mann: This is my final question. We are tasked with dealing with financial services and the financial sector. I noted on the Government website this morning that the largest trade union in that sector, Unite, still has not reported. It is more than two months in. They have a lot to say on many issues, but what do you think the impact on the rest of the financial sector is when the largest trade union in the sector has not bothered to report its gender pay gap?
Sarah Gordon: That speaks to possibly the biggest challenge, which is what the Government want to do with this data and how they want to use it to force or encourage employers to behave differently. At the moment, there is no sanction if you do not report on time. You do not have to say you are going to make progress. You do not even have to publish an action plan. I think Unite, like many employers, feels that it will get away with it. Initially when the gender pay gap reporting requirements were introduced, it was felt, and Ministers said, that transparency alone would force a change in behaviour. One Minister also said the scrutiny of the media would be enough but, sadly, the scrutiny of the media has not been enough.
Indeed, I would argue that, in year 2, there was a sense of the pressure being taken off. I know that the Government have other things on their plate at the moment, which has not helped, but there is a feeling that this is not a priority for Ministers. It is not adequately resourced in the EHRC and possibly not—I know you are talking to Hilary Spencer in the second session—in the Government Equalities Office. There simply are not the resources to audit the information properly, to follow up on anomalies and to make proposals about the next steps.
One of the most surprising things to us at The FT was that, as you know, BEIS had a series of recommendations about what should happen in the next year of gender pay gap reporting. One was that it should become mandatory to publish an action plan, but the Government have said, “No, we are not going to do that”. I do not know whether that is because there are other things going on, but it seems one of the clear things that would make a difference is that you have to not only publish your data, but say what you are going to do about your data.
Q12 Alison McGovern: I want to ask some questions specifically about the financial services sector. Would either of you like to make any initial comment specifically on the world of finance in relation to the gender pay gap?
Nigel Marriott: I certainly could. Compared to other employers, the financial sector has a larger gender pay gap and further to go. For example, I define an organisation as having gender parity if its pay gap is less than 5p in the pound, either way. In 2018, across all employers, I estimated 25% are in that category. In some cases, that is not because they are good at gender recruitment; it is a statistical artefact of their data, so we have to be slightly cautious of that. In the finance sector that figure drops to 7%, so they have much further to go.
They made small progress in 2018. It is small, but at least in the right direction. I can say that from looking at what percentage of companies in the finance sector moved in the right, versus the wrong, direction. There is a net improvement of 3%; there were 3% more employers moving in the right direction than there were going in the wrong direction.
The other thing I did, which I found quite interesting, was to take a sample of finance sector companies and read their gender pay gap reports. I asked basic questions, for example: did they submit their report late or leave it to the last week? Half of finance sector companies either were late or left it to the last week. Did they provide a link to their report on the Government website? Some 90% did. Did they make obvious errors or likely errors? Up to 10% might have made an error; 2% made obvious errors.
Then we look ahead to whether they provide additional information. I found this interesting. Over the statutory information, do they provide additional information, such as breaking it down by job position, doing income quartile gaps by quarters or actually reporting what the pay is—things like that. Some 41% did that. I find that encouraging, given my knowledge of the statistical skills of the HR people doing that. The point is that, if 41% of employers have done that, they are examples that you can point the other 60% to and say, “Look, you can go above and beyond”.
The big question then comes down to whether they have an understanding of why they have a gap today. I judged that about half had a coherent explanation of why there was a gap today and 70% made an allusion to the future, but I am not able to judge whether they have put together coherent action plans. Sarah is probably in a better position than I am there. All those statistics feel slightly better than in other sectors.
What I will say about the finance sector in conclusion is that it probably has the furthest to go in closing its gender pay gap, because the width of pay scales is enormous. From those who are the lowest paid to the highest paid, which is sometimes called a pay ratio, it is probably one of the largest. That therefore makes the official gender pay gap statistic very sensitive to slight imbalances in the gender ratio at each pay level. You only need a slight imbalance and, statistically, you will end up with a big gender pay gap. If your pay scales are flat, you can have a fairly big imbalance and still have a relatively small pay gap. It is just a point to bear in mind and, actually, I may have a go at this, at some point. Theoretically, you could work out the worst possible gender pay gap any company in the finance sector could have, and compare yourself against that. In some sectors where the pay scales are flat, you cannot get a pay gap greater than 30%.
Sarah Gordon: Just briefly, the finance and insurance sector is, in some ways, the worst of times and the best of times. Historically, it has had very few women in senior executive positions, which is the biggest explanation for why the sector has such a bad gender pay gap. With that said, they vary enormously in their approach to closing the gap. As I said earlier, the most effective thing to do is to focus on the larger employers. Among employers with 5,000 employees or more, in the top 10 largest gender pay gaps are three banks—Lloyds, Barclays and Clydesdale. RBS is the 13th worst performer. These are shocking numbers.
In year 1, and it will not surprise you given that Jayne-Anne Gadhia was chief exec at the time, Virgin Money, which had a shockingly large gender pay gap, was also a brilliant example of how to address this most constructively. It published its numbers early. It was extremely clear. It had two A4 pieces of paper explaining why it had such a bad gender pay gap, and that was because it has far more men than women at the most senior level, in the top quarter, and lots of low-paid women in client-facing roles in the lowest quarter. It also had a very clear action plan and there was a public commitment from Jayne-Anne Gadhia, the CEO, to the actions it was going to take to address the gender pay gap. In a sense, that is the best kind of example of what I imagine the Government want employers to do. Acknowledge the gap. Acknowledge where you are, and then talk seriously about how you are going to address it.
The problem is that the vast majority of employers, including in the finance sector, do not provide that kind of information. As Nigel said, in years 1 and 2, you had a flurry of finance sector companies reporting at the last minute. It is pretty obvious why they do that; it is in the hope that the numbers will be obscured by large amounts of others, so that people like The FT do not focus on particular banks. A lot of banks reported with large numbers of subsidiaries, as we said earlier, which makes it difficult to monitor what is actually going on. One of the points to note is that some of the banks are going in the wrong direction. Not only is their gender pay gap going in the wrong direction—the gaps for HSBC and Barclays increased in 2018 over 2017—but, from a more human perspective, numbers on some excos and boards are going backwards, as in there are fewer women, or certainly no change, year on year.
Nigel Marriott: Can I add two more points that I think will be helpful? When we talk about the finance sector, who are we talking about? It is my perception that, when an employer submits its data, it can choose which industry it is in from a list of SIC codes, as they are known—standard industrial classification. I found that a total of 425 employers have a finance-related SIC code. When I looked into that, 17 were not finance to me, for example the British Red Cross Society, Center Parcs, the National House-Building Council, The Gym Group, and Burger & Lobster restaurants. They are currently sitting in the finance sector.
I then went further to try to find others that have not ticked finance, but are finance. Organisations such as the Bank of Ireland, Coventry Building Society, UBS AG and VISA Europe Services have no SIC code at all, so you do not know which sector they are in. Consequently, if someone tries to analyse the finance sector, they have to do a lot of work to identify the 522 employers that I discovered, in the end. Even then, when you do a mode in-depth audit, some of them are marginal. Quite a few IT companies offer finance solutions. The definition of the finance sector is here.
Secondly, I said there has been small progress. Actually, there has been no progress among large employers of 1,000-plus. What progress there has been is among smaller employers. That could be just statistical chance but, on balance, the direction is positive.
Q13 Alison McGovern: Thank you, that is helpful. I have a couple of brief questions to follow up. First, on the commentary and presentation, do both of you think that the information provided by financial firms is accessible to a layperson? If I was wondering which of the banks I might want to work for and was looking online, would I, as a layperson, be able to understand what these finance firms are telling me, in general?
Sarah Gordon: The answer is that it varies enormously. First, you cannot look at the financial sector as a whole and compare one employer with another. That is one problem for a layperson. One of the initial ideas with the reporting requirements was to help people thinking about whether they wanted to work for an employer. The problem is that, if you go to X bank’s gender pay gap numbers, you cannot say how they compare to the gender pay gap or the sector in general, so you are not really able to make a judgment.
Secondly, all employers have to provide a link to information on their website, but we found with many that it simply went to their company homepage. That does not give you any information about the gender pay gap. Then there is this voluntary element around an action plan. To go back to the earlier example of Virgin Money, if you wanted to consult the information, it is extremely clearly presented. There is a clear pictorial graphic and an explanation in words of why it has a gender pay gap. It is not obscure at all. But, if you go to some others, particularly large banks, the information is extremely obscure.
Nigel Marriott: I did an audit of 50 of them in greater depth. My notes here show that three of them had particularly nice graphics, which I felt were informative. I will give you their names. One I liked was Aldermore Bank plc. AXA Investment Managers Ltd was quite a nice one. Then one organisation, TSB, could have been really good. It tried to show a scenario: “This is where we are today. What if we were there? What would our gender pay gap look like?” That was a great thing to try to do, because it gives you a scale of the possible improvement. It messed up the explanation a bit, which was a pity, but I thoroughly enjoyed the attempt.
There are so many out there. I may have to give you some names. This comes to back to one of my biggest concerns about the whole exercise, which is that the people who are being tasked with doing the analysis and preparing the action plans, not just in the finance sector but everywhere, are the HR professionals and departments. I have to say that, from my experience of working with and training non-statisticians, across all functions, industries and divisions, I feel HR professionals probably have some of the least statistical skills you will ever find. You have basically now asked HR professionals to be statisticians. I cannot think of a greater personality clash than between a people person wanting to work with people and me, who loves my numbers. Now you want them to be me. As per the 10th RSS recommendation, you must look at the training or the skillsets these people have. Some of the graphics, such as showing me gender pay gap as a pie chart, are totally meaningless. Well over a third of them do that, probably because they are copying somebody else who has done it.
Alison McGovern: Nigel, I can think of a worse combination, which is politicians and statisticians.
Nigel Marriott: I am an independent statistician who wishes to be diplomatic.
Chair: You do not have to comment on that.
Q14 Alison McGovern: As a final question, in the women in finance report that this Committee previously published, we looked at some of the reasons that financial services firms were giving for their pay gap. We found that some of the reasons were not terribly credible and there was evidence of unconscious bias. Is there any comment you would like to make on the reasons for the pay gap in the financial services sector? What do you think are credible reasons?
Sarah Gordon: The overwhelming explanation that financial sector companies give, as in the women in finance report, is that they have a lack of senior women. To explain the lack of senior women, they say they lose women at mid-career level. I am simplifying massively, but those are the two big statements. The problem with both of those statements, to me, is that they are effects, not causes. The issue is why there are so few women at senior level and why they are losing women at mid-career level.
I talk to a lot of employers about this data. Most employers saw providing the data as a good thing, not a bad thing, but they then wanted guidance and help in addressing what can often be an extremely intractable issue of why there are so few women at senior levels. For example, they wanted sharing of best practice and guidance. There are enormous opportunities from this data-gathering exercise, not necessarily for the Government, but for someone to look at the best companies, which are making progress, and provide information from them to the companies that want guidance and to do the right thing. The reasons women leave at mid-career level are many, which span not just employer practices but domestic issues. Women are still overwhelmingly doing the bulk of domestic and emotional labour, as we probably all recognise. It is a question about caring responsibilities, not just of children, but of parents.
It then goes into what employers can do. Many employers now tick a lot of boxes—they offer flexible and part-time working—but few of them see that through. It is fine to offer flexible and part-time working, but are you allowing your part-time workers to progress and be developed in the same way as your full-time workers? Are you valuing your part-time workers as much as your full-time workers? We are very much at the beginning of this in terms of what works. There is a lot of fantastic research on this, but most employers reporting their gender pay gaps are not doing some of the basic things. There needs to be commitment from the top. This is not a board issue, and this is one of the problems. This is an executive committee issue. Most chief executives do not regard this as a genuinely significant issue.
I do not know if you saw the wonderful list of the top 10 excuses employers provided of why they do not employ women. One of them, which does not make you smile, is this: “Shareholders do not care about diversity, so why should we?” This is a significant factor, particularly for the financial sector. Until the big investors start putting pressure on the companies in which they invest to change their practices, genuinely make progress and see this as a core operational priority, nothing is going to change. Some investors, Legal & General for example, say they will vote against companies that are not increasing board diversity, but these are baby steps at the moment. We need to see far more activism from the point of view of investors, as well as excos.
Chair: I am going to move on because time is pressing.
Q15 Catherine McKinnell: You have identified a whole load of issues that there are currently and why we are not seeing the progress we would hope to see from the reporting that firms are now required to undertake. The Minister for Women, Victoria Atkins, told this Committee that she believes, “The greatest penalty will be the scrutiny of journalists looking through the list […] and the public stigma once a company has not complied with the law”. From what you are saying, this does not appear to be good enough, either as an incentive or penalty. You probably do not want to talk about penalties, but is it time to move the conversation on and look at how we can provide greater incentives to firms?
Sarah Gordon: I wish it were true that scrutiny from the media had been enough, but it has not been. There are two separate issues. First, when we talk about commitment from the top, there has to be commitment from the Government. It is an issue that this is not seen as a top priority for the Government, particularly for the big employers. They want to be seen favourably by the Government and, at the moment, this is not an issue that impinges on any judgment that is being made. As we know, the role of Minister for Woman and Equalities moves around. It is seen as a secondary role for the Secretary of State in question. It moved from Justine Greening to Amber Rudd to Penny Mordaunt. As we know, Penny Mordaunt has quite an important day job and Minister for Women and Equalities is, frankly, a second string. That level of scrutiny would make a difference.
Specifically on the gender pay gap reporting requirements, it is important to think about next steps. They fall into two buckets: targets and potential sanctions. Big British business tends to respond better, in my experience, to targets rather than sanctions, but a blended approach would possibly help. First, the dataset has to be tidied up, so that it is more useful and can be used to hold employers to account in a way that it is not at the moment. Secondly, there should be measures of progress. Thirdly, I back up what Ann Francke of the Chartered Management Institute said. It is great that this has provided transparency, but it now needs to be transparency with teeth. Do you hold employers to account if they fail to make progress? Do you give them a one-year or five-year target? If they have not met that target, should there be some form of sanction? Those are really important issues. Honestly, unless one of those or a blended form of target and sanction regime is introduced, the reporting requirements will have been a waste of time.
Q16 Catherine McKinnell: I was going to ask about the enforcement theme, because the BEIS Select Committee also produced a report on gender pay gap reporting, and it criticised the lack of legal certainty and unwieldy enforcement mechanisms currently available to the EHRC. You have said that the dataset needs to be cleaned up and there needs to be perhaps some merger between enforcement and incentives. Unless the enforcement mechanisms become more streamlined, are properly funded and can be utilised effectively, we are never going to get an enforcement mechanism that has the required outcome. Is that something you are concerned about?
Sarah Gordon: I am very concerned, yes. We looked in depth at the EHRC’s role as enforcer. Part of the issue is that public transparency is very low. On the most basic of issues, we do not know how the EHRC knows that all employers that fall in scope of the requirements have actually reported. We made a freedom of information request asking for that list and we were refused that list. It is simply inexplicable to me why that list should not be public. It is a snapshot at April each year. First, we do not even know that the EHRC and the Government Equalities Office have the information they require to do their job properly.
We do not know whether they have adequate resources. We interviewed Rebecca Hilsenrath about 18 months ago and, at that point, it appeared highly unlikely that she had adequate staff or budget to enforce this properly. There are levels of sanction, as you know, but none of them has been used. They have sent letters to employers, but there has been no sanction imposed on any employer, even those that clearly reported totally anomalous data or did not have a director signing it off. The scale of the exercise and the scale of the enforcement do not seem to be matched at all.
Nigel Marriott: Coming back to enforcement in general, I remind you of what I said earlier. Fundamentally, this is a statistical exercise, involving a huge amount of statistics, being carried out by people without statistical skills. It is evident from reading the guidelines from the EHRC and the Government Equalities Office that the guidance on the calculation was not written by a statistician. The kinds of problems that Sarah and I can see in the published data are obvious to us, but suggest they do not have the statistical skills to detect that. It seems that, if they want to enforce, the first question is whether they have statistical skills in house. It only takes one or two people; we are not talking about a massive budget here. It is just someone to focus and say, “Hang on; those numbers look wrong”. This is fairly basic statistics.
Coming on to enforcement, I support what I believe is the intention of the whole process. This is a five-year evaluation project. That does not mean we wait five years before we apply lessons. We can learn lessons along the way. If companies are not providing the report on time or a link to it, you can take enforcement now. I think the message has got out that you have to do that.
When it comes to holding companies to account, I am wary from a statistical perspective. I repeat what I said earlier: particularly for a small employer, the margin of error for any estimate of a gender pay gap is very large, plus or minus 30%. How can you hold that employer to account when a 10% improvement or decline could purely be due to two or three key people leaving? It is not fair on the employer. I cannot endorse that kind of target-setting. You can set a sort of general target, but I do not see punishment from it.
That other thing the Royal Statistical Society is always on the lookout for is the danger of what is ultimately a holistic exercise about diversity being core to the company’s operation and at its heart. Like you say, is it at the heart of Government at the moment? Is it at the heart of the employer? I endorse that goal, but the danger is that you end up with one number. Everything is about one number. We see this in education with league tables: one number dominates everything. The median gender pay gap has its flaws. If that one number becomes the hook on which everybody hangs, I fear that, as we have seen in other fields, people are incentivised to work to that and do not think about the bigger issue. Fear comes into play, as does the potential for gaming and cheating.
Factual enforcement can be done now: have you submitted it? Have you done the calculations right? Have you put in a link to a website? Have you put the correct industry classification on, as I said earlier? You could enforce that stuff now. You need to look at the skillsets of people in EHRC and whether they can actually stop these issues. But, when holding people to account, be very cautious, especially with smaller employers.
Q17 Catherine McKinnell: You have answered the question I was going to ask. Ultimately, the issue is about transparency and complying with the data requirements. It is not enforcing on people a better gender pay gap. That is not the exercise; the exercise is transparency, which will then hopefully bring that about. The question I was going to ask is this: is there a risk that, by increasing the enforcement side, perverse incentives result? I am thinking of an example based on what you have been saying. Would firms be incentivised to employ more men at more junior levels to improve their gender pay gap? That would be a counterproductive output from this exercise. As Sarah said, if firms publish an explanation of why they have such a large gender pay gap, and it is, “We only employ men at senior levels”, the transparency around that is a powerful thing.
Nigel Marriott: I agree. I have seen some good examples from some employers that have a gap but, when you look at their explanation, it is clear and consistent. You should be able to draw a conclusion. I will give you one example. Suppose you are a female job applicant who is looking at gender pay gap data. You see, “They have a large gender pay gap. What is it telling me?” Actually, I would be looking at the quartile information. A graduate is probably going to start at the lower pay scale. I can see that this employer has no pay gap at the lower quartiles; all the issues are in the senior. If I only intend to stay there for a few years before moving on, I would probably think “So what?” If I intend to make a career there, I would want to know more about what they are doing about that top level. There are lots of ways this data will get used.
Q18 Catherine McKinnell: I was not suggesting that is an acceptable explanation. I am just suggesting the transparency around that explanation would make you question the ethos of that organisation.
Sarah Gordon: It can be a perverse incentive if you just focus on the number, but we are an extremely long way from the dangers of that. You are talking about companies in the sector you are looking at with gender pay gaps of 44%. I am not saying this is Nigel at all, but it is a way in which, if you are disinclined to address your diversity problems, these are fantastic explanations or side issues that you can then bring up as a reason to do nothing.
The real challenge in this whole issue is how to persuade the unpersuaded. There are employers that are really trying to do the right thing on this, and the right thing because it is the best thing for business as well as the just thing to do, which are trying to increase diversity at all ranks and not just gender diversity, but all forms of diversity. But then there are still an enormous number of employers out there that do not get this. They do not get why it matters and they are resistant to doing anything about it. That, to me, is a far greater issue and danger, at the moment, than the danger of us overusing a gender pay gap target and then it encouraging perverse avoidance mechanisms.
Q19 Chair: The final question we wanted to ask was about wider diversity. Of course, we are talking about gender this morning. The Government have recently consulted on ethnicity pay reporting as well. I just wondered, from your work, if you had a view on whether this is a direction that the Government should start to go in, or should they concentrate on getting the gender pay gap data better collected, better enforced, better utilised before wider diversity pay gap issues are examined? I do not know, Ms Gordon, whether that is something you have covered in your professional career.
Sarah Gordon: We have looked at it. There are very interesting issues. So far, the research suggests that, although there are some ways in which focusing on greater gender diversity does filter through to other forms of diversity, there are also ways in which it does not; it is not a silver bullet. The problem with the ethnicity pay gap—and Nigel would know far more about this than I do—is that it is very much more difficult to collect than the gender pay gap. There are issues of self‑definition around gender, as you know, but on ethnicity, on the whole, very few employers classify employees by ethnicity. Indeed, it is voluntary whether you tell your employer your ethnicity, so there are some very serious practical problems. Personally, I think this needs to be addressed from lots of different angles. I do not think that an ethnicity pay gap would be as useful as thinking about access. This is really what we are talking about socially, improving access to these jobs, to the universities that provide people to the biggest employers. That would be, frankly, an effective area for Government attention.
Nigel Marriott: First of all, let me make it quite clear I am completely in favour of the principle of trying to assess ethnicity pay gaps, disability and whatever other protected characteristics we are interested in. The principle I agree with, but I cannot recommend the replication of the gender pay gap process for ethnicity or disability at all. I am afraid I must recommend against that. You must use a different angle.
I intended to make a submission to the Government consultation that ended in January, but unfortunately I got the date wrong. I could not put together a coherent alternative and I think more research needs to be done. I have three reasons why I do not like the rolling out of the existing process for the gender pay gap, one statistical, one data and one ethical.
The statistical reason I have raised already: that a gap is a comparison between two samples. If one sample is small, especially if it is fewer than 50, it is unreliable. When we come to ethnicity, it varies enormously across the country. I believe, if you get up into Cumbria, it is 98% white British, so even if you have 1,000 employees you can only expect, on average, 20 ethnic minorities. Compare that with London, where I like to look at it by parliamentary constituency. There are 58 constituencies that are minority white British, 40 of which are in London. One possibility to consider, if you want to try an ethnicity pay gap process, is perhaps devolving it to the London Assembly. That is a personal idea. I am not saying that is a recommendation, just an idea. If it is does not work in London, it is not going to work outside, basically.
There are two data issues. The first data issue is that gender, at the moment, for the purposes of this reporting, is binary. More widely in society, the discussion is whether it is indeed binary, but for the purposes of this exercise it is binary. When we come to ethnicity, it is not binary. Are we going to say, “White British versus non‑white British” or “white versus non‑white”, or are we then going to split the ethnic minorities into various categories, which will exacerbate all the statistical issues I have said? On top of that, we move into the ethical issue of recording that information in the first place. From my own experience of market research surveys and, indeed, staff satisfaction surveys, when you ask people questions about ethnicity, 20% will not reply, and that suggests a motivational issue. I did see a report by the Government Equalities Office on this issue and, I have to confess, I thought it was far too sanguine about the prospect of employees voluntarily declaring their ethnicity. That is not my experience, and that could introduce bias.
Q20 Chair: I have one final question. Ms Gordon, earlier on, you said it has to start from the top, so it has to be the exco. In terms of accountability, would something like the senior managers regime, where a person, a senior executive, is directly accountable for the gender pay gap, make firms or an individual think even harder about what they are reporting?
Sarah Gordon: If it is any other senior manager than the chief executive it could be counterproductive. It needs to be the chief executive.
Chair: Right at the top.
Sarah Gordon: Yes, and there should be accountability under Government regimes and financial accountability within companies.
Q21 Chair: Can I thank you both very much indeed for your evidence? We could have been here for much longer, but we are very grateful to you for your expertise this morning. Thank you for your time.
Examination of Witnesses
Witnesses: Hilary Spencer and Elysia McCaffrey.
Q22 Chair: Thank you very much, to our second panel, for being here this morning. I am going to ask you to introduce yourselves, and I know you heard the earlier evidence session as well, so we will follow up with some questions from that.
Hilary Spencer: I am Hilary Spencer. I am the director of the Government Equalities Office.
Elysia McCaffrey: I am Elysia McCaffrey, a deputy director at the Government Equalities Office.
Q23 Chair: Thank you both very much for being here. As you will have heard, there has been quite a discussion about the Government Equalities Office already in our earlier session and we will give you a chance to respond to some of those issues. First, we are into year 3 of gender pay gap reporting. It would be helpful to hear how the Government Equalities Office is tracking gender pay reporting, how you are highlighting to companies where perhaps the data looks rather dubious or incomplete, and the relationship you have with the Equality and Human Rights Commission on this issue. Ms Spencer, shall we go with you?
Hilary Spencer: Yes, I was going to say, I can talk a bit about tracking and ask Elysia to talk a bit about data and accuracy of data, and then I will talk about the EHRC.
In terms of how we are tracking the data, I suppose the important thing is to say that we are in year 3, but year 3 has only just started and, as other witnesses have said, there has not been a significant change in the vast majority of organisations that have reported their gender pay gap between year 1 and year 2. I thought it might be helpful to explain a bit more about why we think that is the case.
Q24 Chair: We are going to come on to some of that. It would be really helpful to understand a bit of the tracking. We will come on to some of the reasons why, but how is the GEO approaching this process? I suppose that is the best way I can put it.
Hilary Spencer: Yes, fine. In the first year of introduction of gender pay gap reporting, as you will know if you were following it, 90% of organisations reported in the last month before the deadline. The data that they were reporting drew on the year before, so then the next year’s worth of data, the snapshot data, was three days later. The relatively small changes between year 1 and 2, we think, are related to that.
How are we using the data now? Now that we have two years’ worth of data, which gives us more of a sample to start drawing on, we combine the data that comes through gender pay gap reporting with the ONS data that we get on the overall gender pay gap in the country. Those are our two principal sources of data for gender pay gap and tracking whether it is improving.
As we enter this third year of reporting, we are now trying to work out what are the sectors where we think we could have the biggest impact or where we think we should focus most of our energy and attention. The finance sector, of course, although it is a relatively small proportion of the overall workforce, about 3%, has the worst gender pay gap of all the sectors in the UK and is one of the biggest wealth‑generating sectors. We are keen to focus more in this third year on the finance sector, so we welcome what you have done on that. Other sectors we are looking at are the biggest employers of women. Health, education and retail, between them, employ just over 50% of working women in the UK. One in five working women work in the health service and there is a fairly significant gender pay gap there, and similarly in education and in retail, so those are big areas of focus for us. The final set of sectors is looking at those sectors where they do not have a very high representation of women at all, such as construction and engineering.
Gender pay gap reporting has allowed us to try to identify those sectors in more detail. As a specific example, in the education sector, where, as you will know from your time there, the workforce is predominantly female, the gender pay gap is bad. Some 98% of the early years workforce is women and only 20% of vice chancellors are women, so you have a drop‑off through the education sector. Multi‑academy trusts make up nearly one‑third of the public sector reporters for the gender pay gap and there the gender pay gaps are terrible, in the main—not all of them. There are some good examples. Gender pay gap reporting has allowed us to focus quite hard on those multi‑academy trusts as we develop our plans.
Q25 Chair: Ms McCaffrey, you were going to come in on the data and what happens next.
Elysia McCaffrey: Absolutely, and it was really heartening to hear earlier a view that people are not maliciously reporting the wrong data. That chimes with our experience as well, in that where mistakes have been made they have been honest mistakes. Our commitment in year 1, and we refined our guidance in year 2, was to make sure people have all the tools they need to report in the right way.
In the first year, before the deadline, we put in place a system that would automatically flag if somebody was reporting something that was really unlikely. We are not the enforcers on this; that is the role of the EHRC, but we see it as our job to help people get it right. In the first year, we had a system where, if people reported perhaps a 0% mean or median or if there was a discrepancy between some of the quartile information in the reported gender pay gap, they would receive an email to say, “Are you sure? Here is the guidance again; check that you have got this right”. We know from a number of people that they found that very helpful. There were a few people who came back and said, “We are sure; we have checked it; we do have this”, which was fine, but we wanted to make sure people had been set up in the best possible way to report the right data.
In the second year, we introduced an error message to show people where they had clearly made a mistake. We know that people have made some mistakes. We keep our guidance under review, and we will listen to Nigel’s feedback from earlier. We are already in contact with Nigel on various other things, so we hope to continue to make it easy and prevent people making mistakes in their data.
Q26 Chair: You mentioned the EHRC is the enforcer. What is the relationship between the Government Equalities Office and EHRC in the context of gender pay gap reporting?
Hilary Spencer: As the sponsor body for the Equality and Human Rights Commission, which is independent from us, as you will know, our job is to make sure as many people as possible, in advance of the reporting deadline, know what they need to do, are aware they need to comply and have taken all reasonable steps to calculate their data accurately. It is our job to help them do that. At the point at which organisations have either not complied in date terms or, subsequently, it turns out they have not calculated it correctly, it is the EHRC’s job to enforce that.
Q27 Chair: Is it right, as reported in The Guardian in February this year, that no companies have yet been fined for failing to comply with the legislation?
Hilary Spencer: That is right.
Q28 Stewart Hosie: You have a workplace and gender equality programme, part of a £1.1 million initiative to research and help employers reduce the gender pay gap. How are the results of that research published and what practical help has the office provided to employers, particularly financial institutions, so far?
Hilary Spencer: Quite a lot of the first year of gender pay gap reporting was helping people with the practicalities of the reporting; quite a lot of the second year was helping them with the accuracy of their calculations. Through the second year, we have noticed lots more organisations asking us, “How can we close the gender pay gap?” One of the reasons we set up two programmes, one of which is the workplace and gender equality programme, and the second is a gender and behavioural insights programme, is exactly to try to address those two questions.
The WAGE programme has looked at family‑friendly policies, which we have published some guidance about, and women’s progression in the workplace, which we have also published. We have some work underway on women in low-paid and low-skilled roles and we have three upcoming research themes. Some of those are generic and will be relevant to aspects of the financial sector, particularly the customer‑facing elements at some of the banks. The gender and behavioural insights programme draws on quite a lot of behavioural economics, so some of the international research that looks at different ways that you can promote gender equality in your organisation. Quite a lot of that will be extremely relevant to the finance sector, and I am really pleased to say that Zurich is one of our big six partners in that, with which we are developing some large‑scale trials.
Q29 Stewart Hosie: How much of the gender behavioural research has been shared with financial institutions?
Hilary Spencer: Last summer, we published a guide called What Works in Improving Gender Equality, which is publicly available. We have spoken to teams at the Treasury regularly about this; they put it out through the women in finance charter. Elysia and I have both appeared at various finance sector conferences; we have worked with the Women in Banking and Finance group as well. It is publicly available and we have done quite a lot to get it out there. We would like to do more to publicise it.
Q30 Stewart Hosie: Do you have any plans to tailor research to specific sectors? You spoke earlier about how there are lots of women in low-paid jobs, where there are very few women in engineering and construction, for example. While that is interesting, it does not help sort out the problem, so what additional research are you doing, sector by sector, to find out what is going on and what needs to be done?
Hilary Spencer: As I said, quite a lot of the behavioural insight stuff will be more sector specific as we move forward. We have a programme of large‑scale trials in six different organisations where we deliberately picked different sectors. The constraint for us really is time and budget as we are trying to work out where we focus our research. As you will know, the finance sector has quite a significant range of initiatives working in this area, so part of our job is trying to bring together some of those different initiatives and make the best use of all the work that they are doing, as well as commissioning any additional research.
Q31 Stewart Hosie: In the absence of that kind of sectoral research at the moment, can you tell us whether the barriers to reducing the gender pay gap are the same across every sector? Do you know that?
Hilary Spencer: Some of the barriers are similar. We do an awful lot of research in this area, across a whole load of different sectors, so I am just trying to work out how to convey it sensibly and succinctly. We know that the vast majority of organisations do not have as many senior women as they have senior men. Even in sectors like education and health, which are heavily female‑dominated sectors, we still do not have as many women in leadership roles. That issue of women in leadership roles is common pretty much across all sectors.
There are other issues, though, which vary quite a lot by sector. If you look at the lifecycle of getting people into your organisation, how long they stay and what they are paid, recruitment varies. Quite a lot of parts of the finance sector do very well with getting lots of women in at a junior level, but then they do not manage to help them progress to middle management or to more senior roles. There are other sectors where they are much better at getting women to middle management, but then struggle to get them higher up in the organisation. There are other sectors where geographical location has a big impact on pay overall. For example, if you are a big retailer and you have lots of branches all over the country, but you have your head office in London and you have quite a lot of men in your head office, your gender pay gap will be very high.
To some extent, there are relatively common themes, but how they precisely work out in different sectors does vary and that is the subject of quite a lot of our ongoing work.
Q32 Stewart Hosie: In terms of the information that you do have, the help you have provided, the research you have done, what evidence is there to show that employers in financial institutions are using that information for positive change given, as you said, it is the worst‑performing sector and this is not a new problem?
Hilary Spencer: With the work that you have ongoing on all of this, you will be aware of this too. There are leading lights in the finance sector of organisations that are really taking this seriously, that have all the things that previous witnesses talked about in place: they have leadership from the top, clear targets and a range of programmes, initiatives and policies in place. There are others that really are not taking this seriously at all: “We already have a woman on our board; why do we need any more? It is because they have children and they cannot stay in the office”.
There is a spread across the finance sector. There are organisations that will always have taken it seriously and our job is probably to understand more about what they are doing and publicise it. There is probably a group of organisations in the middle, for which the fact of gender pay gap reporting and the women in finance charter will now have raised this up their agenda and their awareness. Our job, as the Government Equalities Office, and yours and others’, is to try to move that middle group of players more into the territory where they are taking proactive steps. Then there is a group of organisations that do not take this seriously, and that is where some of the shareholder pressure or public pressure arguments come into play.
Q33 Stewart Hosie: I will come back to that. I have two brief questions. What other information will the workplace and gender equality programme be publishing?
Hilary Spencer: Our intention is to publish everything that we produce. We have established a network of 240 academics whom we work with, and our intention is to publish everything that we produce.
Q34 Stewart Hosie: We know, we have heard and we have taken lots of evidence that says, if you get this right, it improves business performance. Have you communicated that to businesses? Do they know this to be the case?
Hilary Spencer: We were discussing this recently. McKinsey has been producing reports for years saying that there is really clear evidence that having a more diverse board, executive committee and workforce leads to greater profitability. It published some research recently that pointed to a 15% increase in productivity or profits if you even had one woman on your board rather than none. There are organisations like that out there making the case for diversity as a productivity or profit measure. Our view is that, generally, businesses are more likely to listen to organisations like McKinsey or others out there than necessarily the Government. But, at every opportunity, and in the range of speeches, publications and documents that we put out we continue to try to make the case for diversity from a business perspective, not just a moral perspective.
Q35 Stewart Hosie: As a final question, given where we are, notwithstanding statistical anomalies, and progress has not been great, do you think the workplace and gender equality programme offers value for money?
Hilary Spencer: It is relatively early doors on the particular programme. It has a pot of £1.1 million that we have allocated to it. We have not spent all of that; we are still working it through. It is a programme that we have generated to try to provide research to respond to the questions people have asked. Each piece of research that we commission is evaluated and approved by a research committee in a Government Department, so it is assessed on a value-for-money basis. Given the stage we are at with the programme, yes, I am confident it represents value for money.
Q36 Stewart Hosie: It is value for money in UK Government terms, but what if it is not delivering the goods?
Hilary Spencer: As I say, it is early doors yet. In the gender and behavioural insights programme, the other research programme we have got going, the What Works guidance has been extremely popular and I do think that is starting to prove its worth.
Q37 John Mann: Is there any consideration of extending reporting to smaller firms?
Hilary Spencer: Yes. As previous witnesses have said, there is a number of ways we could take reporting in the future. One option is to lower the threshold, another option is to ask for more data on gender, and another option is to ask for data about different characteristics.
Q38 John Mann: You are actively considering this.
Hilary Spencer: Yes.
Q39 John Mann: Is there any timescale for any changes?
Hilary Spencer: The formal review of the gender pay gap legislation is five years after it was introduced, which is 2022. However, our view is that there is a pretty crucial window between years 3 and 5 where, if year 1 was people getting up to speed with their reporting requirements and year 2 is trying to start making a difference, you want to start seeing changes between years 3 and 5. In our view, there is probably an argument for making some changes to the reporting regime from year 3 onwards or what will now be year 4 onwards.
Q40 John Mann: So you will be coming back to Government with proposals on that soon.
Hilary Spencer: Yes, we are actively working on that. We are putting advice to Ministers on ways we could take this going forward, including questions about ethnicity pay gap reporting.
Q41 John Mann: You have these companies, PwC, Nationwide, Coventry, where there is no SIC code. Why? Whose fault is that?
Elysia McCaffrey: When we designed the system, we designed it really to look at individual organisations rather than to look at sector bases. That is why we decided to use the Companies House SIC codes and we import those over. We also work with organisations—
Q42 John Mann: So these are Companies House flaws.
Elysia McCaffrey: I am not saying it is a Companies House flaw. We need to have a look at those exact cases and why that data is not there for those.
Q43 John Mann: I am saying it looks, to me, like it is a Companies House flaw. Could you correct me if I have the wrong impression?
Elysia McCaffrey: I am not confident that it is a flaw in that area. We have exported those codes over and, if they are missing, we should go and have a look at why they are missing. I know that last year there were more and, in some areas, that was because they were charities or public sector bodies that just would not have an SIC code.
Q44 John Mann: Is that discussion actively ongoing with Companies House?
Elysia McCaffrey: Yes, it is.
Q45 John Mann: What is the timescale for resolving that, given that this anomaly has been repeated?
Elysia McCaffrey: We would like to get it resolved before the next deadline, absolutely.
Q46 John Mann: Why was it not done in the first year?
Elysia McCaffrey: It was not something that we viewed to be a really significant issue because we were not using that data to look across at sectors. We wanted to make sure we were reducing the burden, as much as possible, for businesses and not reinventing the wheel on anything, which is why we used those existing codes. As Hilary said, we are keeping all of this under review. We want to keep improving things, we want to make it better and, if people are using the data and the way that we have collected it does not quite work, we keep that very closely under review. We were really very pleased and surprised that in the first year we had about 4,600,000 unique views on our website.
Q47 John Mann: What is your attitude to those organisations that have not managed to get their information on to your website, but have stuck it somewhere on their own website instead?
Elysia McCaffrey: If people have put it on their own website and have not reported to our portal, they have failed to meet the regulations. They are marked as not having complied.
Q48 John Mann: If someone has not done it from 5 April this year, when are they going to get a kick from somebody?
Hilary Spencer: They have already.
John Mann: So the fact that they still have not done it—
Hilary Spencer: You asked our view: we take a dim view of it. As I outlined earlier in terms of the division of relationship between us and the EHRC, a large part of our job is, before the deadline, making sure that organisations know they are in scope, and that they are clear what the legal requirements are and that they must do it. In the vast majority of cases that has been effective. We have had an extremely high compliance rate for the first year of any reporting regulation of this type. There are still a small number of organisations that have not reported this year. The EHRC is pursuing them. It has written letters to a number of organisations on that and, in some cases, has issued them with the terms of a review.
Q49 John Mann: Where someone sticks on their website a mean figure and a median figure with no explanation of what “mean” and “median” is, what is your attitude to that?
Hilary Spencer: If they fill it out on our—
Q50 John Mann: No, on their own. Would you regard that as being deliberately helpful to people?
Hilary Spencer: I am sure there are a range of reasons that people choose to report some of their data in different ways. Some organisations are very helpful about it, because they want to be transparent. Some are not very helpful about it, because they do not want to draw attention to it. Some are not clear about it themselves.
Q51 John Mann: Would putting down a mean and a median figure without explaining the difference between “mean” and “median” be, in your view, transparent?
Hilary Spencer: If they reported it on the Government gender pay gap portal, it does explain there what the mean and the median is, so it is possible for someone to find their data.
Q52 John Mann: If someone was giving their excuses on their website about the gender pay gap, what should the balance be in how much space they are giving to their excuses compared to the detail of their action plan on what they are going to do? What would you regard as best practice?
Hilary Spencer: This is one of the things we are putting together advice for Ministers on at the moment, this question of action plans. There has been, in some sectors, a call for making action plans compulsory in some way so we have been exploring what that might look like. In our view, best practice for an action plan would be that there is a clear explanation of what the gender pay gap means, why the organisation believes that gender pay gap exists in the organisation and some clear, targeted next steps about what it is going to do to make a difference. Again, the What Works guidance I was referring to earlier has some very clear steps in there that are proven, with a comprehensive evidence base, to have a positive impact. Our view would be that best practice is for organisations to adopt as many of those as are relevant for their issue.
Q53 John Mann: If we take the largest trade union in the financial sector, on its website, it does not have any action plan at all, it has a mean and a median figure without any explanation and it has not submitted, despite the requirements, to your website. What hope can there be for change in the sector when the largest trade union in the financial services sector has not yet complied with your requirements, and has not done anything even vaguely on best practice on an alternate platform? What sanction ought to be brought in to make sure that organisations like that comply?
Hilary Spencer: There are two issues there. One is in terms of its compliance. When you mentioned in the earlier session Unite not having published, I do not know whether you saw, but there was a flurry of people rushing for their iPhones. On its website, it has not reported for 2019, but it is not required to until next year. It reported for 2018 and for 2017, so at the moment it is not in breach of the technical requirements about reporting.
In terms of the content of its action plan, our view is that any organisation should set out clearly what it is trying to do to close its gender pay gap. An organisation like Unite, I would suggest, might want to be an example of good practice on that and I would hope that members of that organisation might place quite a lot of pressure on it to do more.
Q54 John Mann: In terms of the requirements and the differences with best practice, on the day that the Unite union could have put the last year’s information properly into the public domain on your website, instead it chose to put it, in a very misleading way, in my view, certainly without an action plan, on its own website. Despite the fact that you are saying that that is still compliant, if somebody has bothered to put it on their own website simultaneous to the time when it could have been reported directly to you, is there not a weakness in the system here in terms of defining best practice? Should that not have been on your website, if it has been put in the public domain anyway, using the structures, the explanations, the systems you have, so that the world can then do a proper comparison?
Hilary Spencer: Yes. It might be best if we perhaps followed this one up in writing, to make sure we give you the details of this precise case. Our understanding is that it has now complied with the requirements for last year, as it was supposed to, but I just want to make sure we are giving you the right information, so we will follow that up in writing.
Chair: Thank you very much.
Q55 Alison McGovern: I have some questions about the data, which are probably for you, Ms McCaffrey. Before I come to those, Ms Spencer, I want to ask you a few quick-fire questions about cross‑Government understanding of this issue. When was the last time the Chancellor of the Exchequer asked to meet with you?
Hilary Spencer: To my knowledge, the Chancellor of the Exchequer has never asked to meet with me. He has met with Penny Mordaunt, as the Minister for Women and Equalities, I think, within the last three months, but I would need to check; I know they have met recently.
Q56 Alison McGovern: Were you required to provide information about this subject, the gender pay gap, for that meeting?
Hilary Spencer: We did.
Q57 Alison McGovern: Have you been required to provide information for the Chancellor of the Exchequer about the productivity challenge in the UK?
Hilary Spencer: We have not, no.
Q58 Alison McGovern: Have you been asked to provide information about how labour market reform might assist with the productivity challenge in the UK?
Hilary Spencer: We have not, no.
Q59 Alison McGovern: Have you been asked to provide information about the Government’s industrial strategy and how the gender pay gap might feed in to changing the industrial future for the UK?
Hilary Spencer: We worked closely with BEIS on the industrial strategy, to make sure that gender pay gap and other gender‑related workforce issues were included as part of that work.
Q60 Alison McGovern: That was to BEIS, as part of the industrial strategy, but not to the Treasury.
Hilary Spencer: No.
Q61 Alison McGovern: Have you been asked to provide information for the Treasury about how investment in childcare might assist with the gender pay problems that we have in the UK?
Hilary Spencer: Yes.
Q62 Alison McGovern: What information have you provided about childcare?
Hilary Spencer: We have provided quite a lot of information in terms of analysis, research, statistics. There are a number of cross‑Government official‑level groups, which look at that issue and others that affect gender. We have been active participants in those.
Q63 Alison McGovern: Okay, but you have not been asked to provide information for the Chancellor of the Exchequer directly.
Hilary Spencer: No.
Q64 Alison McGovern: Have you been asked to provide information about public transport and the impact of public transport investment on women in our economy?
Hilary Spencer: No.
Q65 Alison McGovern: Has the Government Equalities Office been asked to provide or provided information about sexual harassment in the labour market and how that might impact women in our economy?
Hilary Spencer: Yes, we have been asked to provide information about that to the Home Office, the Ministry of Justice and BEIS. I do not believe we have had a specific request from the Treasury.
Q66 Alison McGovern: Thank you. That is all very helpful. Ms McCaffrey, do you know how many times the data on the gender pay gap has been downloaded?
Elysia McCaffrey: I do not have that figure to hand. I know how many people have visited our website, if that is useful—unique visits. That was 4,600,932 in the first year. That is very specific, is it not? In the second year, it was 2,412,165.
Q67 Alison McGovern: Could you find out for us how many times the data has been downloaded and write to us about that?
Elysia McCaffrey: I would be very happy to. Yes, of course.
Q68 Alison McGovern: Okay, great. You mentioned earlier the proactive flags in the system if you feel there might have been an error in the data. Are there any plans for developing that system for years to come?
Elysia McCaffrey: Yes. We keep looking at how easy the system is for people to use. We have changed the portal quite a lot from the first year. For example, the way that the data is displayed back to people we have changed, following a lot of testing we did to check that people understood it. We keep that under review and we want to make sure that it is useful. We also introduced a comparison tool this year, so that when people download data they can compare their organisation with other organisations, based on certain characteristics. We keep developing that to make sure it is useful for everyone.
Q69 Alison McGovern: Are there any plans to bring in checks, for example, as part of a firm’s audit? Are there any plans to involve checks in the quality of the data through that mechanism?
Hilary Spencer: Yes. Again, we are thinking about, after the third year of implementation of this, what changes we might usefully bring in. One of the questions is whether we could include it as part of companies’ annual reports, so that it becomes audited as part of their wider accounts. We have been looking at quite a lot of those options.
Q70 Alison McGovern: Would you be able to update us in future as to where that gets to? It will be quite an important change if that happens.
Hilary Spencer: Yes, we would be pleased to.
Q71 Alison McGovern: I just want to ask you about bonuses in the financial services sector. As far as we understand, from two years’ data, there is no great decrease in the inequality in bonuses, and that is a significant problem, as you will have heard from the previous panel. What is the GEO’s plan to deal with that specifically?
Hilary Spencer: As I was saying right at the beginning, the snapshot dates are a year before you are legally required to publish the data. Even if, as an organisation, you calculated your data in the first year, you realised you had an appalling gender pay gap and you desperately wanted to do something about it, the data on which you are then drawing for your second year’s reporting is three days later. Even if you really wanted to do something about it, you probably were not going to see much difference between year 1 and year 2. As you say, we are trying to think about what we now do as we move forward on this.
We have been having discussions with Women in Banking and Finance and the Treasury about whether there is anything we could do on bonuses. One of the gender and behavioural insights pieces of work that we are doing is looking at this question of bonus allocation and how remuneration is decided. That is at an early stage, but we are conscious it is a major contributor to the gender pay gap in the finance sector.
Q72 Alison McGovern: As yet we do not know. That is basically what you are saying. We know that in financial services they have this massive problem with gender pay gap. In part, that is to do with where women work in financial services, but it is also to do with the fact that, if you are lucky enough to be a woman who makes it in financial services, the chances are you will not be there arguing for yourself to have a big bonus as well. At the moment, you are saying the Government Equalities Office does not know what to do about it.
Hilary Spencer: Essentially, the evidence that we have had is—you are right—that part of the question is which precise roles you have. If women are currently clustered, in general, at the top levels in finance organisations, in HR, legal or comms, they are not typically the roles that attract significant bonuses, as you say. Also, the more senior you are, the greater the proportion of your pay is a bonus.
Q73 Alison McGovern: Ms Spencer, to be clear, you are a civil servant; you are not responsible for this. Ultimately, that is the politicians, which is why I asked you about the Chancellor of the Exchequer. But what we are saying is that there is a big problem in our labour market. The people who earn the most in our country and who are able to accrue the most bonuses are men. The Government Equalities Office does not really have a plan for what to do about it and the Chancellor of the Exchequer has not asked to see you anyway. Really, the Government, at least at a political level, are not really bothered, are they?
Hilary Spencer: I think it is probably best if I do not comment directly on that.
Q74 Alison McGovern: That is fine. As a final question, my colleague Catherine asked the previous panel this, but the Government’s view, as explained to us by Victoria Atkins, Minister for Women, was that the greatest penalty will be the scrutiny of journalists being able to look through the list and the public stigma, once a company has not complied with the law. But we heard from Sarah in the previous panel that that is not really the case either, and that penalty of transparency has not really been significant for some firms. Is that not the case?
Hilary Spencer: The level of public scrutiny is, in large part, what has contributed to very high levels of compliance with reporting. For the first two years of a new reporting requirement we have had extremely high compliance, up in the high 90s, so I think public scrutiny about producing something has been effective. As a number of us have said, I do not think we were going to see significant changes between year 1 and 2, for a combination of reasons, so, in some ways, that question is still open.
Q75 Alison McGovern: Right, but that is not quite the point being made here. We were told that the Government felt that there would be some sort of penalty, that people would feel a sense of public censure. The evidence we have heard is that people submit these reports in a cluster at the end of the requirement process. They are basically using that cluster as a shelter from any public censure. Is it not the case that there is not a lot of public censure from this transparency and firms are basically getting away with it? Yes, they have to put their figures out there, but there has not been a lot of public censure, especially for financial services firms, which, frankly, have some shocking results. Is that not the case? That idea of public censure has not turned out to be real, has it?
Hilary Spencer: I do not know. Look at the press coverage. There were probably slightly fewer political things going on last year than this year, which might have affected the media coverage a bit. If you look at the media coverage last year on this, there was a lot that was very much in that spirit: “This organisation has an appalling gender pay gap and this is not acceptable”. The combination of the BBC equal pay issues, MeToo and, as the previous witness said, work from investigative journalists or members of the press resulted in quite a lot of very critical press for some organisations that have not done enough.
Q76 Alison McGovern: The best hope that women in Britain have today for getting change is that Brexit goes away, so that the issue gets some attention. In summary, that is what you are saying, is it not?
Hilary Spencer: If you look at what is happening in organisations, lots of organisations have now, as a result of this, taken action in terms of setting up diversity committees; it has gone up their agenda; they have changed the way that they focus on this issue in their board discussions. There is quite a lot of work happening internally.
Alison McGovern: But it would be nice if the Chancellor of the Exchequer cared enough to ask you for a meeting.
Q77 Chair: Following up on that, in the earlier session we heard, and I know from personal experience, that you are asked to be Minister for Women and Equalities while you are also doing another Cabinet‑level job. The point that Alison is making is one that was reflected in the number of women on FTSE 100 boards, which required ministerial telephone calls to chief executives to ask them why they were still all‑male boards in some cases.
I suppose this is the question following from Alison’s: are you aware of whether the Minister for Women and Equalities has made any calls, written letters or made approaches to companies where the gender pay gap is large and has not moved, or whether, in this case, the Chancellor of the Exchequer and his Ministers have, again, made approaches to financial services companies where the gender pay gap has not moved or is still extremely large? What is the ministerial involvement in this?
Hilary Spencer: The Minister for Women, Victoria Atkins, has made contact with a number of organisations where the gender pay gap is bad. She is also due to write shortly to all the public sector organisations in scope, to ask them to produce an action plan. She has had meetings with a number of organisations in her sector, in the Home Office, to encourage that. We have also been working, at official level, with Government Departments, which have been pushing quite a lot of this out through their networks. There are numerous examples of Ministers who have been writing to their sector at different points, who have included references to the gender pay gap, either to reporting or to the need to take action to reduce it, in their communications. I believe that there is a significant degree of political support for this and the Prime Minister herself has been a very strong backer of this work.
Q78 Chair: In terms of GEO resourcing, earlier on, it was mentioned in the previous panel and you have also made reference to limits on resources. If you had more resources, a bigger gender pay gap team, would you be able to do more on this space?
Hilary Spencer: I like to think of it as a small but perfectly formed team, full of high‑quality staff. Resources in the Government Equalities Office have increased significantly over the last few years. We now have 120 staff. We have a fixed gender pay gap team, but we have also expanded a lot of our work on things like the research and what can be done to solve the problem, not just reporting it but what actions people can take. At critical points, particularly on gender pay gap reporting, we have moved people from other parts of GEO to provide additional support on this, so we have been able to flex the resource we have to mean that we have driven high levels of compliance.
Of course, if I ruled the world, I would love to be able to do much more work on a whole set of these issues. Part of our job, part of my job, is trying to match the resources we have to the priority areas that we need to address.
Q79 Chair: Do you have your own parliamentary team or do you rely on the parliamentary team of the Department that you are in?
Hilary Spencer: We now have more of a hybrid model than perhaps has been the case. It is important for the public‑facing functions like the press office that they have some alignment with the ministerial team in the Department that the Secretary of State is in and a private secretary as well. We have a team there that reports to us for management purposes, but is co‑located with the Secretary of State in whichever Department they are in.
Q80 Catherine McKinnell: Alison touched upon the issue of media scrutiny of the data, but there is another aspect to this: how easy do you think it is for a layperson to read, process and make a meaningful use of the data that is available?
Hilary Spencer: As Elysia referred to, we have made a number of changes between years 1 and 2 and will continue to do that as we move forward. In year 1, the priority was to get a number that was clean and clear for people to look at, so the very simple GOV.uk website did that. We found, when we tested comprehension with lay members of the public, that it was not completely clear. They obviously knew that a large number was worse than a smaller number. We then tested a variety of ways of displaying it, with several hundred people, to see what gave them the best comprehension of it. The diagram we now have on our website is the result of that research.
Q81 Catherine McKinnell: You think it is resolved, so you think a layperson could quite easily interpret the data.
Hilary Spencer: It is clearer than it was. As others have said, there is still an issue around the conflation of gender pay gap and equal pay, which we need to do more to address.
Q82 Catherine McKinnell: It is clearer than it was, but is it still clear for a layperson? If they were looking to work for a particular company, do you think they would be able to look at the website and make an assessment of how that company provides opportunities for women within its organisation?
Elysia McCaffrey: We have been told by lots of people that they have used it in that way and that they found it easy to do that. I understand our team recently received a “thank you” message from somebody, who we do not know, to say, “This helped me make a decision about which employer I wanted to go to”, so we do hear that it is useful.
Q83 Catherine McKinnell: Do you have any plans to try to improve that further? Are there currently plans?
Hilary Spencer: We are still testing it. You are asking two slightly different things, I guess. We have been testing: “If you look at the website can you tell what is a big gender pay gap and what is not? Does that make sense to you?” The question about whether this organisation is a good employer for women is slightly different, because you might have a terrible gender pay gap but be putting in place loads of good things to tackle it, which might mean that you are a good employer to join. You might have a low gender pay gap because you are just a great employer, but you would probably want to have a combination, would you not? You would want to look at what was on the gender pay gap website and then at what the organisation said about what it was doing about it.
Q84 Catherine McKinnell: Does your website provide the ability for laypeople to easily access all that information?
Hilary Spencer: It does not currently and that is one of the questions: should we ask it to include more?
Q85 Catherine McKinnell: Ah, right. Another issue that has been raised is ensuring that this data is used meaningfully and it makes a meaningful difference. We are not seeing the progress that we would like to see and need to step that up. Is one of the issues with that the lack of enforcement at the moment? What are your views on where the balance should lie between public scrutiny, transparency, incentivisation and, on the other hand, enforcement when it is not being complied with?
Hilary Spencer: I am not trying to dodge these questions. These are really live questions for us at the moment, about exactly what we think the balance ought to be going forward. It has been important for the first three years of introduction that the job has been helping organisations do what is essentially quite a complicated statistical calculation, as the previous witness set out. For lots of organisations that has been quite a challenge, so a big part of our job has been making sure people knew they needed to comply and helping them comply to the best of their ability.
As we move through into subsequent years, we would expect people’s fluency and capability in calculating their gender pay gap to improve and, therefore, our role needs to evolve to match that. While in the first year you might excuse errors as genuine mistakes, as we move through and this becomes a more established system, people not complying either by the deadline or in terms of the content becomes less reasonable. Therefore, there is a question for us about whether we ought to introduce stronger sanctions at that point.
Q86 Catherine McKinnell: At which point is that, sorry?
Hilary Spencer: The legislation, as I said earlier, has a formal review at five years. The question for us is whether we should, for years 4 and 5, consider changing some aspects of the way it is reported.
Q87 Catherine McKinnell: You must be looking at that now.
Hilary Spencer: Yes.
Q88 Catherine McKinnell: What are you considering now in terms of enforcement?
Hilary Spencer: We are considering a whole range of things. One is whether we require companies to publish more information about what they are doing—so their policies. Another is whether we ask them to publish more data. You could ask people to look at numbers of women on boards or retention post maternity leave; there is a set of things you could ask for there. There is another set of questions about whether the powers of the EHRC might need to be changed. Some of our colleagues are at another Select Committee hearing at the moment, giving evidence on whether the enforcement powers of the EHRC, more generally, in relation to the Equality Act, ought to be increased. These are really live issues for us.
Q89 Catherine McKinnell: It would be helpful to receive an update on progress in terms of decisions that are being taken, because a key issue that seems to keep arising is whether the EHRC’s processes are currently adequate, whether writing a letter is enough. I am pleased to hear that you are not just waiting until the end of the five‑year period and that there is going to be, hopefully, some action taken to deal with what are already quite clear violations. But it would be useful to get a bit more detail about what enforcement you think is appropriate and that you will be coming to some decisions quite soon on that, to ensure that we get that balance right between transparency and being able to enforce it. The stark warning we had from previous witnesses was that the whole effort is rendered meaningless without it being properly enforced.
Hilary Spencer: Yes, and we really want this to work. We have invested a huge amount of time and energy in this and we really want gender pay gap reporting to work, hence the evolution that we have tried to make. There are some things that we can change through our systems and processes; there are things we can clarify in the guidance as we find them. Some of the things we are asking people to report are in legislation, so we would need to find legislative time to change the regulations there. But we really want this to work and we are committed to doing it, hence trying to keep in play a range of options and taking into account the views of experts like the previous witnesses on this.
Q90 Catherine McKinnell: And getting the interest of Government at the most senior levels focused on this as well.
Hilary Spencer: Indeed, and the work that this Committee has done on this has been incredibly helpful for us. I do not think any Government colleagues would thank me for this, but if other Select Committees were to ask questions about things like the representation of women in their sectors that would also be quite helpful.
Q91 Chair: Duly noted, and I will suggest that to other Select Committee chairs. Finally, on Catherine’s question, will you publish a consultation or some sort of discussion paper when you have come to some conclusions about changes that might be needed to regulations or enforcement?
Hilary Spencer: Yes, that would be our proposal. If we are changing the regulations, it is sensible to consult.
Chair: Can I thank you both very much indeed? It was short but sweet. That is very helpful evidence and, yes, we will continue to return to this area. We will keep a close eye but, for now, thank you for your evidence this morning.