Public Administration and Constitutional Affairs Committee
Oral evidence: Civil Service People Survey, HC 575
Thursday 18 May 2023
Ordered by the House of Commons to be published on 18 May 2023.
Members present: Mr William Wragg (Chair); Ronnie Cowan; Jo Gideon; Mr David Jones; John McDonnell; Lloyd Russell-Moyle.
Questions 82 – 148
Witnesses
I: Alex Chisholm, Civil Service Chief Operating Officer, Cabinet Office; Dr Claudia Roscini, Head of Civil Service People Survey Team, Cabinet Office; Fiona Ryland, Civil Service Chief People Officer, Cabinet Office.
Written evidence from witnesses:
Examination of Witnesses
Witnesses: Alex Chisholm, Dr Claudia Roscini and Fiona Ryland.
Q82 Chair: Good morning and welcome to the Public Administration and Constitutional Affairs Committee. Today, we are joined by a number of senior civil servants as we continue our inquiry into the civil service people survey. We are going to be putting questions to them about the design and implementation of that survey. I wonder if they might introduce themselves for the record.
Alex Chisholm: Alex Chisholm, civil service chief operating officer.
Fiona Ryland: Fiona Ryland, Government chief people officer.
Dr Roscini: Claudia Roscini, head of the civil service people survey team.
Q83 Chair: Thank you very much. The opening question is from me and is a rather obvious one: what is the purpose of the people survey in the eyes of the Government? How are the results used within Departments?
Alex Chisholm: We very much welcome the inquiry by PACAC into the people survey, and we look forward to your findings. The survey has been conducted since 2009. It is a census-based survey. It covers 100 organisations in total. Its purpose is to enable us to embed and improve employee engagement within those 100 organisations, and to enable the people who work in them to see what the issues and concerns are in order that they can be addressed. It has proved an extremely useful tool in that regard over the last several years.
Q84 Chair: Overall, would you consider it fit for purpose, or are there any changes that might be appropriate?
Alex Chisholm: I am sure that it can be changed and improved. Indeed, we conducted a review of it last summer for three months and identified a number of potential improvements. We have not yet made those, because, at the same time, PACAC launched its inquiry and we wanted to hear what you had to say before we implemented those changes. Otherwise, we might change in the wrong direction or change twice.
Chair: You are very considerate.
Alex Chisholm: It certainly can be refined. However, we do consider it fit for purpose, and there is a lot of evidence in support of that, which we would be very happy to discuss.
Q85 Chair: How do you evaluate that? We are getting a bit esoteric, which is not unusual for this Committee, but you are carrying out a survey of the survey.
Alex Chisholm: There are a number of ways of evaluating it. First of all, the survey itself includes evaluation. There are three evaluation questions at the end. With 350,000 people completing the survey every year, that is a lot of feedback and evaluation for us.
Secondly, every year, both before and after the survey, there is a consultation with all the survey managers in Departments, which is run by the Cabinet Office people survey team, led by Claudia here on my right. In addition, we seek feedback and input from other stakeholders, including the unions and experts in surveys. We also do benchmark comparisons with other large public sector organisations, such as NHS England, the US federal service or the many other Governments out there who do annual surveys.
Q86 Chair: Ostensibly, the technical guidance suggests that the primary aim of the survey is to embed and improve employee engagement, but it is used by a great number of external parties, from unions to academics, for a variety of purposes, it is fair to say. How are those external interests and voices drawn into the process of the design, such that it might satisfy their use?
Alex Chisholm: With your permission, Chair, let me bring Claudia in on this.
Dr Roscini: As Alex was saying, we consult a number of stakeholders, both internal and external. Just to give you some examples, we try to harmonise the majority of our questions with the Office for National Statistics, so that we can compare what we collect in the civil service people survey with other statistics that are available.
We are part of an engagement group at the OECD, so we are in touch with other Governments that run similar surveys. For example, we included in the survey a six-module question on engagement, led by the OECD, so that we can compare ourselves against other countries.
We meet the unions to get their feedback on how we can change the survey. We are in touch with other public sector organisations, such as the NHS, in order to share good practice. We have also been consulted by other countries that are thinking about starting to run an employee engagement survey.
Q87 Chair: Mr Chisholm, I wonder if you might be willing at all to share with the Committee the review that you have conducted, which you have kindly paused further progress on, given our inquiry.
Alex Chisholm: Yes, we would, and we can write to you, or I can give you a brief summary now of some of the main points.
Chair: By all means, and then feel free to write after the meeting.
Alex Chisholm: The findings were broadly encouraging, in that people felt that it is fit for purpose. They felt it might be possible for it to be refined further in terms of the publication products that we have out there. They are done in a particular type of format, which is accessible, but some people suggested that we might be able to provide a more interactive microsite-type approach in order to enable people to work with the data more interactively. That seemed like a good suggestion.
We had a debate about the regularity of the annual survey. There are a couple of countries out there—Denmark, Switzerland and others—that do it every three years. We felt, by comparison with them, and with the feedback that we got, that once a year seemed about right. We could do it twice a year, but it takes a certain amount of effort to do that. So that seemed appropriate.
We looked at whether it could be complemented by other types of surveys—so-called pulse surveys—and whether people would like that to be done on a centralised basis. The feedback we had from that consultation was that people like being able to do their own pulse surveys, because each organisation has particular issues. The issues across the MoD would be different from issues in the Department of Health, for example. Pulse surveys enable people to measure assessment against their own particular local issues, and then the people survey is that consistent, year-by-year, cross-civil service comparison.
Chair: Thank you. That is helpful, and we look forward to receiving a copy. That is very kind.
Q88 Jo Gideon: Moving to the employee engagement index, why do you choose to design the headline employee index in the way that you do? What does it show?
Dr Roscini: The survey started in 2009. In the same year, there was the “Engaging for Success” report, which identified four enablers of engagement: visible leaders; engaged managers; allowing employees to express their voice; and integrity—the fact that the values that we preach should be visible day by day. We design our engagement index in line with this recommendation.
The literature on engagement is quite extensive. The “Engaging for Success” report identified more than 50 definitions, so we embraced the approach that engagement is going to measure how people are committed to the organisation’s values and goals, but also how, at the same time, they can thrive in the organisation.
We have five questions that compose our index, which measure pride; advocacy—how much our staff would recommend the organisation as a great place to work; their sense of attachment; how much they are inspired by the work they do; and how much they are motivated for the public good.
We found that our index is also aligned with what other countries are doing. Of course, across countries, for example, the number of questions used to measure engagement varies, from 15 in the United States to five here in the UK, but the broad approach is in line with what we are seeing outside.
Q89 Jo Gideon: What use is it to the civil service and to external users of the people survey data?
Alex Chisholm: For the civil service, the most valuable thing is that it enables us to make comparisons both over time and between units. Just to give a couple of examples of that, for leadership and managing change, which is a very important thing in all organisations, that figure has gone from 37% when we first started in 2009 to 54% today. By measuring over time, we have been able to see the initiatives we take that improve people’s confidence and how well led they have been.
Similarly, if we look at discrimination, which reached an unacceptably high level in 2018 of 12%, that led to a big review by Sue Owen. A lot of steps were taken to try to improve that, and that is now down at 7%. The survey is a guide to action, and the year‑on‑year changes give us feedback as to whether those actions are effective.
It is also very important in enabling us to make comparisons between business units. Within the Cabinet Office, for example, we have an overall figure for bullying, harassment and discrimination, but it shows terrific variety between business units. As well as the initiatives we do, which, so to speak, raise all boats and are of general benefit, we also challenge individual business units. To those that have 0%, we say, “What is the amazing thing that you are doing that makes your workforce so happy and have no such issues?” so we can share best practice. To those at the top end of the scale, we say, “Give me a plan to reform what you are doing locally,” because it is very much at the team level where these issues arise.
Hopefully, that gives you some idea of how we use the change over time—that is why it is so powerful as a tool, because it has been running for all these years and asking broadly the same questions in the same way—and look for patterns and variation horizontally between business units and organisations.
Dr Roscini: Engagement is also a result in terms of how some of the teams that Alex mentioned vary—what we consider are the key drivers of engagement. It is a very useful way for our staff to tell us what they think, because it is not only how they answer these questions, but the way that that engagement has varied throughout the years, which shows us what we have done in other areas that has led to the change in engagement as well.
Alex Chisholm: I should have mentioned, of course, that there are very important demographic pictures there as well—people from particular genders, ethnic backgrounds or age groups, as well as different levels of pay or seniority in the organisation. All of that is telling you, in a very sophisticated way, how different people within your organisation are feeling and what might be driving that.
Q90 Jo Gideon: That covers the civil service element. How would it be useful to external users of the data?
Alex Chisholm: There is a wide group of people outside of the civil service who are interested in the performance of the civil service and Government. High up in that is PACAC itself on behalf of Parliament. I know that it is also used by the National Audit Office and by your sister Committee, the Public Accounts Committee.
Outside Parliament, there are a number of so-called think=tanks—the Institute for Government and others like it—which use this as a core mine of data to try to understand what is happening in Government. Trade union groups are very interested in this and it gives them great sources of data.
Other public sector organisations look at what we do and, on some occasions, learn from that. NHS England, for example, used to do a partial survey—that is, a sample of the people working in the NHS. It then moved towards a census survey, which asks everyone to take part, because it saw that we were getting better results from that and it gave us a more differentiated picture. There are a lot of external users, and I have not mentioned the international ones, but Dr Roscini mentioned that there is a lot of international interest in what we do.
Q91 Jo Gideon: The employee engagement index does not capture areas that responders are typically less satisfied with, such as pay and benefits, and learning and development. Why is that? Is there a risk that we are overestimating employee engagement as a result?
Alex Chisholm: In terms of the biases inherent in the way people respond to surveys, we think that, because this is an anonymous census survey in which everyone is invited to take part, it is relatively free of biases. When you ask people to identify themselves, they are less likely to raise problematic issues and concerns, and more likely to reference behaviours that are approved of rather than less approved of.
The design of the survey does not create those kinds of biases. That means that people are very free to give their ratings. There is also free text, which gives us a lot of additional colour about what people are thinking and what is driving their concerns.
As you rightly say, the single biggest message from the 2022 people survey results was that people were not happy with the level of pay and benefits, because that fell by 11 percentage points across the piece. Again, you can look at the data harder and see that, in some organisations, it rose, and, in others, it fell very sharply. Even within organisations, you will see that different groups of people have different views about that.
As an overall trend, that was the single biggest finding, which has sent a strong message to the Government that we need to address pay and benefits for civil servants.
Fiona Ryland: When we come to action planning, we do not look just at the engagement questions. We look across all nine themes. The actions we take respond to the broad themes that we have, not just engagement.
Q92 Jo Gideon: How would you respond to those critics who describe the index as spin or a way to avoid the real issues?
Alex Chisholm: I do not understand or appreciate that view at all, because you just heard from Dr Roscini how the engagement index is constructed, using a very scientific, well-researched and statistically robust approach. We have compared that every year with alternative methods and taken academic input. I know that you have taken hearings from external experts on surveys, and they have said that the approach we take is credible.
I would almost say the opposite: if you did not have a very strong empirical basis for your views, it would be susceptible to spin and opinion. This is not so susceptible, because it is a very robust set of data, with 150 questions asked every year, going back 14 years now.
Dr Roscini: I totally agree with what Alex said, because we do not look at engagement as an isolated concept. We look at it in line with the nine other main themes that we have as core questions in the survey. We also ask questions on bullying and harassment, discrimination, and staff health and wellbeing, so we look at the full picture. Engagement is, let us say, the headline measure that shows us the results of everything else that is happening across all these areas.
Q93 Lloyd Russell-Moyle: With reference to the 2022 result, can you please tell us when the data were released and to what parties?
Alex Chisholm: The approach that we take to publication or release of the data is to do it internally first and externally subsequently. That approach has been found, by us and other people who do surveys of this kind, to be the best one, because people internally want to see their data before they read about it externally.
Q94 Lloyd Russell-Moyle: When was it published internally?
Alex Chisholm: We completed the survey at the very end of October, and we released it from the beginning of December. It takes about four weeks or so to do all that quality assurance and data verification, to create all the interactive dashboards and to integrate paper-based with electronic survey material.
Q95 Lloyd Russell-Moyle: When was it released externally?
Alex Chisholm: That happens once we have completed that process, which runs through in a staggered approach to release in December, with more detailed results released after the high-level results, and ultimately giving individual business unit-level results. There are 12,000 teams, so it is quite an undertaking to do that.
In January, the team led by Dr Roscini begins to prepare for external consumption. Again, there are a lot of formatting changes on the 300,000 figures that are published, and that is undertaken by Dr Roscini and her small but very effective team over the first two months of the year. We then published the data on 30 March, along with much other data, as one of the periodic quarterly transparency days on which the Government release data.
Q96 Lloyd Russell-Moyle: How are the unions treated in terms of internally/externally? Are staff members who are union reps on facility time in the Cabinet Office given access to the data when it is produced internally?
Alex Chisholm: Yes. Claudia will want to add to that. As members of staff, they get it, but they also have a particular consultative process, both before we fully release it internally and before it is published externally, so they get two bites of the cherry.
Q97 Lloyd Russell-Moyle: If the data was robust enough in December 2022 and throughout December when you were rolling it out in more detail, but the headlines are produced, why can the headlines not be published at that point?
Dr Roscini: It is important to say that we release a number of internal products, as Alex mentioned, from submissions to data packs and three types of interactive dashboards that cover not only the civil service and the organisational-level findings, but the findings at team level. We have 12,000 teams that receive a report and, in the dashboard, each of these teams can cut its data by demographics. There is an extensive amount of work that the team needs to do also to quality-assure this data that we release internally.
We also share the microdata with organisations that have analytical support. This is individual-level data cut just for the organisation, after they sign data sharing agreements, completely in line with the privacy notice and with UK GDPR. This allows organisations to do analysis in addition to what we can produce centrally.
This takes time. You are correct to say that the headlines are fine, but it is our duty to publish the findings in an accessible format. For example, we cannot publish PDFs or dashboards that include team-level results. We do not publish those. We need to produce accessible ODS files in a different format that, as Alex mentioned, include more than 300,000 figures. Again, that needs to be produced and quality-assured.
Q98 Lloyd Russell-Moyle: You are saying you that cannot give the headline figures in December 2022. I get that you cannot release the detail and the individual dashboards that go down to small work units, because some of that cannot be published without making sure that it is clean and that you cannot identify people from some of the small units, but why can the headline data in December not be published? Are you saying that the only reason that it cannot be is that you cannot get it into a correct format at that moment?
Dr Roscini: It is not only that. As emerged in one of the previous hearings in this Committee, it is very good practice to allow organisations to discuss findings internally and to agree on what happened and what to do next, before publishing the findings and starting to get external queries. Also, our staff would like to have early sight of the findings before they are published.
Q99 Lloyd Russell-Moyle: Is four months an acceptable amount of time to wait, or is that what is required to have those internal discussions?
Alex Chisholm: For internal purposes, that would be too long, because we want to be able to prepare those plans and respond as soon as possible. The chief purpose of the people survey is to inform those decisions, so we try to get that done within four weeks.
Dr Roscini is right to point to it being good practice to do internally before externally. It is also my experience that, when you release very high-level data, it raises a lot of questions. If people then have to wait a number of further weeks before they get answers to those, because the answers are in the detail, they find that a bit unsatisfactory.
We know that the employee engagement index is 65%, which is one percentage point lower than the previous year. That would beg a lot of questions. The interesting part is when you burrow into that data to see what is driving that and what the variation is.
Q100 Lloyd Russell-Moyle: So your view is that it is best to publish publicly all the data at once. Is four months, which is what your gap is at the moment, an acceptable amount of time to wait for an annual survey? It is five months after the survey is done by the time it is made public. Is that kind of time acceptable?
Alex Chisholm: Yes, I would say that it is acceptable, but other people have views on that. A small point is that we do it in two batches. The survey results are published on 30 May, and then the demographic analysis, which requires a lot of extra work, is published in the next transparency returns, which are due in June, so there is a follow-up piece. We could do that all in one go, but that would risk delaying it further.
Q101 Lloyd Russell-Moyle: Your point just a second ago was that you have to release all the data at once, because people have to dig down. Now you are saying to me you that you do not do that anyway, but you release a set of headlines with maybe some more detailed figures in March, and then the demographic data comes along two months later.
Alex Chisholm: Yes, but, when I mentioned that additional detail, it is 300,000 figures. That is a lot of additional detail that is published on 30 March, which would not be available at an earlier point in time.
Just to highlight one other issue to the Committee, you are very aware of the Government’s overall commitment to transparency. The approach that this Administration have taken, like some before them, has been to do that on transparency days, where they go for roughly quarterly returns, in which a lot of information is published together, so that people know to expect that, to aim for it and to work off that. A debate is to be had about whether that is best.
Q102 Lloyd Russell-Moyle: What would the quarterly day before 30 March be?
Alex Chisholm: It would be the end of December.
Q103 Lloyd Russell-Moyle: Is it published internally at the beginning of December?
Alex Chisholm: It is a staggered approach over the course of December.
Q104 Lloyd Russell-Moyle: But you just would not be able to get it out in that session. During the time in which your Department was negotiating pay remit guidance with trade unions, you held information about civil service dissatisfaction on pay and benefits, which you chose not to make available to the unions. You said earlier that the unions would have access to the data, so was there a conscious decision to withhold certain bits of data from unions in that process?
Alex Chisholm: First of all, just to correct the Committee, the civil service does not negotiate the pay remit with unions. The pay remit is a cost control document and is not negotiated with unions at the national level. There are negotiations about how that is implemented at the departmental level, and it is the Department that is the employer in each case. The pay remit for this year was published just a few weeks ago, and several weeks after the publication of the civil service people survey results on 30 March.
Q105 Lloyd Russell-Moyle: Was that sufficient time for the unions to enter negotiations from the beginning at that point, or were negotiations already halfway done and dusted?
Alex Chisholm: As I said, it is not a collective bargaining approach. There are no negotiations over the contents of the pay remit. Also, the unions would have seen the content of that well before publication, on a confidential basis, and we do that every year.
Q106 Lloyd Russell-Moyle: So there was no data that was withheld from unions at any point last year.
Alex Chisholm: We do not withhold data from unions. On the contrary, we make a point of showing the data ahead of publication, so that they can see that.
Q107 Lloyd Russell-Moyle: So they would have had it before the December 2022 date.
Alex Chisholm: They would have been part of the roll-out of that data in December, and then we also had specific consultations with them before we published.
Q108 John McDonnell: Just to be clear, the unions would have had that information on a confidential basis. They would not be able to use that publicly in any campaigning.
Alex Chisholm: Ahead of the publication, yes. That is the approach.
Q109 John McDonnell: While they were engaged in dialogue with Government around the pay remit, they would not have been allowed to release any of that information or to use it publicly to in any way influence the Government.
Alex Chisholm: As I mentioned, the remit was settled just at the end of April, and the civil service people survey results were published on 30 March.
Q110 John McDonnell: When were they given the confidential information—31 March?
Alex Chisholm: No. It was published for everyone to see on 30 March, on a non-confidential basis, but they would have had sight of that sometime before that.
Q111 John McDonnell: But they would not be able to use that publicly in any way to influence the Government in their representations.
Alex Chisholm: They should have respected the confidentiality of any information shared on a confidential basis.
John McDonnell: So they could not use it publicly.
Q112 Lloyd Russell-Moyle: Can they use any of that information, in general terms, publicly? Can they say, “We are aware that satisfaction has reduced in the Department”?
Alex Chisholm: I have seen unions sometimes comment ahead of the publication. Of course, with 360,000 people taking part in the survey, and nearly half a million people receiving the results of that in December, a certain amount of view about what those findings would be will emerge into the public domain—no question.
Q113 Lloyd Russell-Moyle: But that would not be a breach of the confidentiality if they were to use some of these in broad-brush terms.
Alex Chisholm: Ahead of publication, we do share with the unions on a confidential basis and ask them to respect the confidentiality.
Q114 Ronnie Cowan: I want to look at how we communicate the results of the survey. When it is published on gov.uk, it is in a spreadsheet format showing both mean and median scores for all respondents, and mean scores by organisation. The published results are aggregated and include only quantitative data, so we do not get the free-text responses, and there is no accompanying commentary or analysis. You used to do that, but, in 2020, you stopped doing so.
In terms of an individual coming to this, do you not think that it would be beneficial to me or to civil servants—maybe those at a lower grade—to have that sort of commentary, so that they can understand the data that you are putting in front of them?
Dr Roscini: We moved to ODS files so that they are accessible to everyone; before that the commentary was a PDF file. We can, of course, improve what we publish, so if the Committee thinks that having a commentary would be a good recommendation and way forward, we can definitely consider that.
Alex Chisholm: Another Member of the Committee was asking earlier about allegations of spin and so on. To some extent, our approach is to publish the data and let it speak for itself. Internally within the civil service, business units will engage with their staff and will say, “Here is the summary. Here is what it tells us about our organisation. Here is what we are going to do as a consequence.” That is very important and powerful, and you would expect that in any organisation.
Q115 Ronnie Cowan: When you talk about business units engaging with their staff, why not just get this information out there on the internet, so that staff can read it themselves, rather than it being interpreted by their management level and telling them what the survey is saying? Why can they not just find it for themselves?
Dr Roscini: Staff across organisations can directly access the dashboard—the interactive tool that we have.
Q116 Ronnie Cowan: Is this the standard dashboard, the intermediate dashboard, the advanced dashboard or the fourth dashboard?
Dr Roscini: It depends. We give organisations the freedom to decide what is the best way to share information with their staff. Some organisations share the intermediate dashboards and others share the standard dashboard with their staff. Staff can look at and play with the findings, apply filters, where they are available, compare themselves against business units, and give their own interpretation to the data.
As you say, sometimes there is a commentary on the data or a message on what the picture is in a specific year, but staff have the possibility to access the findings.
Alex Chisholm: They do very much exercise those opportunities. Daily log-ins on the dashboard peaked at 17,000 in 2022, with members of staff saying, “That is interesting. Let me have a look at it.”
Q117 Ronnie Cowan: Is it the same people going back every day?
Alex Chisholm: If you are interested in seeing that—
Ronnie Cowan: Is it the same people who take that opportunity and spend their lunchtime going back through and looking at the survey for a couple of weeks, or is it 17,000 different people?
Alex Chisholm: I know that it will not be the same people every day.
Q118 Ronnie Cowan: You do not know that. You are assuming that. In my experience of working for companies that run surveys, some people are keen to get involved and will engage, and others simply will not. If I am running a survey of 360,000 people, I want to know how many of those are engaged in looking at the results and at what level they have access to those results.
Alex Chisholm: I am very happy to do some analysis for the Committee and see how many of those are unique sessions and how many are repeat sessions. I am very confident in saying that it will show a small number of repeat sessions with power users. For example, Claudia and the five members of her team will be using it every day, and there will be other survey managers in different Departments doing that. I am sure that the very large majority will be different users on different days, but, if you would like the breakdown of that, we will do that piece of work and share it with the Committee.
Q119 Ronnie Cowan: You also said that the decision was made within organisations as to whether to share the intermediate or the advanced dashboard with people. What are the criteria? Why are we not just saying to everybody, “Here is the information. Here are the tools for you to drill through it,” and trusting them to use their own intelligence to interpret what they are seeing? Aside from that, importantly, why do you not publish an analysis and an accompanying commentary so that they understand what they are reading?
Alex Chisholm: It is right that organisations and business units take a view about how much data they provide and what is useful, and they consult with staff and other people about how to do that well. If we did not do that, but we just took all the data we could, dumped it out there and said, “Go figure. Sort it out for yourselves,” we would rightly be criticised for not paying attention to the needs of users.
Q120 Ronnie Cowan: Why do you not publish the free-text responses? They are usually the most amusing.
Alex Chisholm: It is interesting that you should say that—I do not know how, because they are not published, and there are a few sound reasons for that.
Q121 Ronnie Cowan: They are available on the advanced dashboard.
Dr Roscini: The main reason is that we want to protect respondents’ anonymity. We want staff to keep trusting us in terms of how we treat the data, even if the open comments are redacted. We remove the information that can make people identifiable, such as name, job title and so on. It might contain information that people do not want to see released in the public domain. We analyse and look at what people are saying in the open comments, but looking at the comment internally is more in line with good practice rather than releasing individual-level data in the public domain. We treat the open comments as personal data, in line with UK GDPR, and as individual-level data.
Fiona Ryland: The other thing I would worry about is that, if we printed comments, even if they were anonymous, it might mean that people are reluctant to include open comments when they complete the survey. I would want to be really thoughtful about that, because they are used by Departments and contain a lot of good context. I would not want people to feel like they could not be really honest because they know that it is going to be published.
Q122 Ronnie Cowan: There is this line where you are saying it is anonymous. If it has been published internally, and I can see it through the advanced dashboard, what you were saying is that, as an employee, I could be identified by this. It is supposed to be an anonymous survey, and that is why you do not want to put it out in the public domain.
Dr Roscini: There is a difference. When we release the dashboard, comments are not released unless there are 20 comments in a specific team or business unit or a specific cut of the demographic groups—for example, G7, aged between 35 and 40 years old. This is to reduce the likelihood of a survey manager identifying a specific person by what they wrote in the comments.
If a business unit gets 15 comments, they are merged with the upper level—let us say the director level. They are still considered and looked at, but the likelihood of senior leaders or managers identifying someone by their experience is reduced.
Q123 Mr Jones: Dr Roscini, as you know, the Office for Statistics Regulation publishes a code of practice that is intended to ensure that data released is as trustworthy, valuable and high quality as possible. As we have heard, the people survey does not comply with the code of practice. For example, there should be a clear description of the main statistical messages that explains the relevance and meaning of the statistics in a way that is not materially misleading. You do not do that. Have you given any consideration to adopting the code of practice in relation to the people survey?
Dr Roscini: Yes, we have. That was also discussed in the strategy review last year. We do not provide official statistics, but we do comply with the main rules. To give you an example, the project is also overseen by the head of profession for statistics in the Cabinet Office. We quality-assure our findings. We comply with GDPR. We respect anonymity. We apply suppression rules. We have considered whether we should become an official statistics provider.
It is also very important to consider the aim and purpose of the survey—measuring employee engagement rather than being workforce statistics. We collect information on the job and personal characteristics of our staff, but the main purpose of that is to understand whether their experience at work might vary because of these.
There is also the risk that people use the information we have on our samples, which does not correspond to 100% of staff, as workforce statistics, while there are official workforce statistics on the civil service workforce from the annual civil service employee survey.
Q124 Mr Jones: So you have had a discussion with the head of profession as to whether the result should be published as official statistics, but you have decided that, on balance, it should not. Is that right?
Dr Roscini: We have decided for the moment to follow good practice, from both a research and a statistical perspective, but to remain an employee engagement survey rather than official stats.
Q125 Mr Jones: Would it be possible for you to adopt at least some of the recommendations of the code of practice in connection with your survey?
Dr Roscini: Yes. We are in line with them in the approach that we use when we analyse and manage data and so on.
Q126 Ronnie Cowan: Fewer than 35% of respondents to the latest people survey believe that effective action was taken on the back of the previous year’s survey. Why do you think that is?
Alex Chisholm: The question is, “I believe that senior managers in my organisation will take action on the results from this survey,” and the median scores at civil service level have increased by 14 percentage points since the survey began in 2009, which shows that people have increasing confidence that action is being taken.
Q127 Ronnie Cowan: It peaked in 2020, with fewer than 40% of respondents, and we are now looking at fewer than 35% who believe that action was taken.
Alex Chisholm: It is much higher than it used to be, but it has fallen in the last two years. That is correct.
Q128 Ronnie Cowan: Can I ask why that is?
Fiona Ryland: Action planning is a really critical part of the people survey. That is why people give you the feedback. We have been developing some new tools. For example, part of the interactive dashboard is now an action planning template and tracker that Departments can use. We are providing hints and tips for Departments to use around how to develop their action plans as well.
Dr Roscini and her team also support the survey managers in terms of how they develop those Department plans, and there is a knowledge hub where people can go for advice on how to use the dashboards. While there is still more to do, we are developing and maturing the support and tools that we provide to Departments.
Q129 Ronnie Cowan: Am I right in saying that you do not centrally track those actions that have been taken?
Fiona Ryland: That is quite challenging, because there are over 15,000 teams and units.
Q130 Ronnie Cowan: Why do fewer than 35% of civil servants out there believe that their input to the survey is of any value?
Alex Chisholm: There are two ways of looking at action being taken. One is locally and the other is across the whole civil service. When people raise an issue that can be dealt with locally, they want to see the local management say, “Yes, I am disappointed by these results”—for example, on bullying, harassment and discrimination—“and these are the actions that we as a team need to agree to take. We will hold ourselves responsible for that.” Those are the types of things that improve people’s confidence that action is being taken.
There are other issues where the local team does not have full authority to be able to address them. For example, a Department might say, “We have just been merged with another Department,” or someone might say, “We have had a change in my place of work, which does not particularly suit my own arrangements,” or people could raise a concern about pay, which is subject to central control through the pay remit. All those issues are ones where people might say, “I asked and I did not get it exactly back.”
We have already spoken about last year’s findings, where the biggest fall was in pay and benefits, and a number of civil servants will have compared the remit that was issued in March of that year, for 2% plus 1%, with the pay agreed over the course of the summer for some other public sector groups, as well as the cost of living, with a lot of pressures that came into the wider economy. All of that will have contributed to people saying, “I asked to be paid more and I have not been. Therefore, I am a bit disappointed by the action taken.”
Fiona Ryland: There is also an opportunity to be very clear about actions that have happened as a result of the people survey. Departments often put improvement plans and action plans in place, but you need to be clear that, “This is as a direct result of what we heard in the people survey”, so that people can make that connection.
Q131 Ronnie Cowan: Why does the Cabinet Office not co-ordinate the activity so that other organisations and HR representatives are able to share examples of actions undertaken? Why are you not co-ordinating that?
Alex Chisholm: We co-ordinate the survey and have compared doing that centrally and allowing Departments and bodies to do it themselves. The last time that we measured that, it was showing several million pounds in savings, and I am sure that that would continue to be the case. We also co-ordinate the management of the contract and of the communications, the quality assurance, or the analysis, and the presentation. That is done centrally.
We do not try to co-ordinate the action taken. As I mentioned, there are 17,000 teams or identifiable business units within the overall survey. Of those, 14,000 receive a report because they are big enough to be able to deal with the confidentiality issues that Dr Roscini mentioned. Notwithstanding the undoubted abilities of my Department, the Cabinet Office, co-ordinating the work of 14,000 teams would not be efficient and effective, and we do rely on the local responsibility of good organisations to say, “These are our results and our findings. This is the action plan that we need to put in place.”
We support that by making it easy for people to generate action plans. It is a feature of the interactive dashboard to enable you to do that with templates, with hundreds of tips and pieces of advice, with mentoring sessions and with training. We support and enable the action planning, but we do not try to take responsibility for 14,000 dynamic action plans, which, of course, are changing through the course of the year.
Q132 Ronnie Cowan: You are filibustering now. We are running out of time. Fiona, have you made any attempts to integrate people survey data with the broader range of HR data that you hold, such as on recruitment or absence?
Fiona Ryland: The people survey data is just one set of data. You are absolutely right. We have workforce statistics. We are piloting a new recruitment survey. We have absence figures. There is an opportunity to do more of that.
Q133 Ronnie Cowan: Are you doing it?
Fiona Ryland: Yes, we are doing it. For example, we would look at what the people survey data tells us, lined up against exit interviews, for example, in terms of reasons for leaving.
Q134 Ronnie Cowan: Has this been successful? Is this something that you will progress? Do you see a future for it? Are you going to develop and grow it?
Fiona Ryland: There is opportunity to do more in terms of how we share and compare data. A lot of data is held in Departments, so we need to think about how we do that as well. There is an opportunity to do more, but, where we have different sources of data, we do try to see whether they are telling us the same thing.
Q135 John McDonnell: Can we just talk about the results a little bit? The survey this year saw the results fall in satisfaction across all themes, which you have acknowledged. Were you surprised at that? Do you have any information on or assessment of what is driving this?
Alex Chisholm: First of all, was I surprised? The honest answer is “probably not”. Heads of Department had a discussion beforehand about what we expected. We were very conscious that the results had held up very well with the previous year, notwithstanding the undoubted pressures and strains of operating in Government during the covid era. To tell the truth, it surprised me that we had held up so well and that it was still at a record high. I was not surprised that it fell by one percentage point year on year. It might have fallen by more, given the level of turbulence, pressures and challenges that people felt.
To your second question, on the key drivers within that, I have tried to emphasise how you get a lot of variety within this survey, but, if you are trying to extract two things, the two most striking findings are on pay and benefits, which fell to a historic low level and by 11 percentage points—there is no question that that is a very substantive trend—and on leadership and managing change, which fell by four percentage points, which is not as great, but is a substantial change. Otherwise, the moves tend to be in the 1% to 2% area. Those would be the two main factors driving the overall issue of engagement.
You then get variety in different organisations. In the Cabinet Office, we fell quite a bit on our sense of organisational purpose, and I am happy to comment further on what we have done about that or how we attribute that. I saw some other organisations where that was not an issue. Just looking across the piece, I would pick out pay and benefits, and leadership and managing change.
Q136 John McDonnell: We will come on to the Cabinet Office, as you said. I just want to understand a bit more about leadership and managing change, and the factors that drive that sort of drop.
Alex Chisholm: There are a number of things. When you look at the data, the thing that matters the most is local leadership. The reason I say that is that, looking across an organisation, you see an average result. In the Cabinet Office, for example, we have 40 business units, with some areas that are knocking it out of the park and that have fantastic results, and others that are absolutely bottom of the table. Local leadership matters more than departmental leadership or, indeed, civil service leadership, being perfectly modest about my own influence.
That said, there are some trends there, and the management of change in a lot of organisations fell. It is particularly noticeable, incidentally, in London-based policy Departments. Factors that contributed to it are the sheer rate of change and the pressure of going through the gears.
You may remember that Deputy Prime Minister Oliver Dowden, when he was before PACAC a couple of months ago, said that, in the Cabinet Office particularly, the pace of change had felt quite relentless, with the pressures of being the lead Department for Brexit and for the covid taskforce, dealing with Russia/Ukraine and, in the meantime, having G7 and COP26. We have had changes in political leadership as well as part of that. Managing change has been an issue.
I would also pick out whether, when change happens, people feel fully involved in that. There are planned and well co-ordinated changes where you tell people beforehand why and how it is going to be done. You continue to communicate before and ask their views, and you tell them afterwards as well. During that whole process, people feel, “I have bought into that change.” When you read on the front page of a newspaper about a new Government statement, which seems to affect you personally, that is less positive for the change process.
Q137 John McDonnell: You mentioned announcements. The survey was being done soon after or during the period when there was an announcement of 91,000 job cuts. Was it that time?
Alex Chisholm: The announcement about the reduction targets of 20%, 30% and 40% was in July, and the survey was conducted in September/October, but it is fair to say that it would have been fresh in the memory.
Q138 John McDonnell: You have touched on this a bit, but I just want to get to how you respond to the results themselves. As permanent secretary at the Cabinet Office, how do you respond to the results, not just in relation to the Cabinet Office but beyond? You have mentioned the multiplicity of all these units and individual leaderships. Although you are not co-ordinating the response, how do you monitor it in order to find out what is and is not effective, as well as the effective use of these surveys?
Alex Chisholm: If I can use the Cabinet Office as an example, we got our results, which, to be absolutely straight with you, we were disappointed by.
Q139 John McDonnell: There was less satisfaction in the Cabinet Office than anywhere else, was there not?
Alex Chisholm: We were at the bottom of the table. That is correct. It is not just about being competitive. This is a strong message from our workforce saying that the management had changed, and they had felt insufficiently involved in that and discombobulated by the rate of change. Their sense of attachment to their organisation was not that great. It was strong to their team and to their own work, but the wider Cabinet Office was not something that they related to so strongly.
That is partly always a feature of the Cabinet Office, because we are the central point of Government. Our work and our priorities have to change, which had been very much a feature over the last year, with people being told that the main priority of the Department and the Government was Brexit, then necessarily having to roll into covid as being the dominant priority. We moved from having no specialist covid staff to having several hundred people working in the covid taskforce.
That would have been true as well for people’s sense of the wider Cabinet Office and attachment. There are groups that are very focused on achieving particular goals. We were the Department responsible for COP26. At the time that that survey was done, a lot of people were still completing the presidency year for COP26. Their answer to the question, “How strong an attachment do I feel to the Cabinet Office? Will I be here in a year’s time?” should have been “not very” and “no”, because that unit was to be disbanded at the end of the presidency year, in December.
There are some particular factors, but the thing that we have taken very seriously is how the message from our staff is very strong, first of all, around pay. As a result, in common with other leaders in the civil service, I have been part of this overall process of discussion about what the remit could be next year. You will see in the announcement 4.5% plus 0.5% for lower grades, which is a considerable increase on the amount from the previous year and a little bit ahead of the forecast level of inflation according to the Office for Budget Responsibility. That was something we contributed to.
As soon as that remit was announced, we went through it with our staff. We had an all-staff call. We said, “This is what it is. This is why it is. This is how we are going to make sure that you can get paid at the first available opportunity,” because we knew that people want the money more than the words.
We have also looked at some of the other benefits that people get and said, “Can we improve those?” They are smaller things, but they do make a difference—discounts, allowances and things like that. Are people aware of the value of their pensions? In some cases, less so, and particularly younger staff. We are trying to make people very aware about how much the employer contributes. There is all of that around the pay and benefits area.
Around the wider change in the Cabinet Office, we have a programme with the not-very-original title of “A Better Cabinet Office”, which we launched at the beginning of this year. That has six workstreams, which are devoted to addressing each of the main themes coming out of the people survey. People said progression is an issue, particularly linked to location, so we have had a big body of work saying, “Can we give people more flexibility about the extent of progression that can take place within particular locations, professions and business units?” People also say, “I am not really happy with some aspects of my workplace,” so we have a big programme of improving wi-fi, and refurbishment, in some cases.
All of that has been driven by the results that we have had. I do not want this to appear as something that is happening in a quiet space. We have a weekly meeting on this. We have an all-staff call every month. We have a call for all of our senior civil service leaders as well about their responsibilities there.
On things that we are dissatisfied by, such as bullying, harassment and discrimination, we wrote to all the bottom performers, requiring plans for remediation as part of that. Staff have been thoroughly engaged. We have 250 volunteer change agents and a very high level of engagement with the programme.
Your other very good question was about whether this is making a difference. As well as doing the annual people survey, we do a pulse survey to try to see whether we are making progress within the year. We have seen from pulse survey results that identification with the organisation’s purpose and objectives, which was a relatively weak area within the Cabinet Office, has improved by nine percentage points since we started this campaign. We had a big focus on that, so it is, in a way, a perfect illustration of how you use the survey to drive action and improve people’s feelings.
Q140 John McDonnell: Are you getting feedback from the other Departments about the action plans that they are putting into place that you can monitor at all?
Alex Chisholm: As discussed with your colleague, we do not monitor the 14,000 action plans like that, but those things that are common themes across the piece—pay being a very obvious example of that, but also some things around the tone of engagement and the way we try to involve our staff, and some initiatives around diversity and inclusion and the implementation of the civil service diversity and inclusion strategy—are done on a cross-cutting basis.
Q141 John McDonnell: So you can get common approaches agreed.
Alex Chisholm: Yes, exactly.
Q142 Lloyd Russell-Moyle: Reported levels of bullying, harassment and discrimination remain broadly stable in the latest people survey. They did, of course, fall in 2019-2020. How did you achieve that fall and what has led to the stagnation?
Alex Chisholm: Just to add some specific detail to what you said, so that people can see the picture, the results in 2018, exactly as you said, showed that the level of discrimination was 12%, and of bullying and harassment 11%. They have both fallen to 7% last year and this year, which is great. There has been an improvement. We would like to go further, but it is a substantial improvement.
In terms of factors within that, when those poor results came out in 2018—again, this gives a good illustration of how the survey helps drive action—Dame Sue Owen, as she is now, who was permanent secretary in the Department for Culture, Media and Sport at the time, was commissioned by the then Cabinet Secretary to do a review across the whole civil service of what our approach was to dealing with bullying, harassment and discrimination. She wrote a report that was published, and a whole set of actions have been implemented since that.
We are now doing a review five years on of any remaining gaps, so we will have a fuller answer to your question in due course, but a difference seems to have been made across all parts of the cycle. For example, in terms of whether people understand how they can raise a concern, that has been improved with lots of speak-up campaigns and alternative routes for people to raise issues and so on.
Do people see that action has been taken as a consequence? Again, since five years ago, confidence in seeing action being taken has gone up a lot. Have they seen that across the civil service as well as in their organisation and their team? Again, yes, that has been the case. Is there more visibility on those improvements? Is there more transparency in reporting? Yes, there has been.
Is the case handling seen as something that people have confidence in? Yes, that has been improved. One of the recommendations from Sue Owen’s report was to establish a centre of excellence independent investigation service, based in the Home Office, as it happens. That has been put in place and people do have more confidence in that.
There are a whole load of recommendations and actions, but it is a perfect illustration of a problem and actions taken to solve it, and then you re-evaluate and say what has and has not worked. It mostly has worked, but we will have a fuller answer in due course.
Q143 Lloyd Russell-Moyle: Have staff absence and sickness followed a similar pattern to what the survey is reporting to us?
Alex Chisholm: No, they have not. Central Government staff absence, which is measured as a percentage of hours lost, is 2.1% according to the latest published figures from the ONS. That is historically the lowest figure that it has ever been. Interestingly, it is, for the first time, lower than in the private sector, quite considerably lower than local government, and very considerably lower than within the health sector.
Q144 Lloyd Russell-Moyle: In the Cabinet Office, are you performing better or worse than other Departments in terms of absences? I will come on to the staff survey in the Foreign Office in a second.
Alex Chisholm: We do not have that to hand.
Q145 Lloyd Russell-Moyle: That would be an interesting comparator. It would be interesting to see, while those absences have continued to reduce, why the feeling of bullying and harassment, which is often a key indicator of absence, has not reduced.
There was a respect and inclusion review undertaken last year, which you have discussed. There was some controversy about the unions’ participation at different times, but it is now being published. What actions are you going to take to implement this and further drive down the feeling of harassment or bullying?
Alex Chisholm: Just to give a little bit of background for the Committee to make sure that everyone understands fully, the review was commissioned in October 2021, and we published it internally in July 2022. It was a very substantial piece of work. To give it some independent credibility and to enable us to benchmark with other organisations, we enlisted the help of Ipsos and the Business Disability Forum—two different groups—for different aspects of the respect and inclusion review.
We also worked very closely with staff groups on that, with the unions and also with our board. That process was led by the lead non-executive director, Anand Aithal. The report has 61 recommendations, which is a very large number. We have accepted and are committed to implementing all of those. Some were quick wins and some take a longer time.
They have some similarities to the things that I was talking about around the Sue Owen report, in the sense of all those fundamentals about how confident you are to raise a concern, for example. We have had a particular focus on that in the Cabinet Office, encouraging people in staff sessions at team level, but also creating fair treatment confidants as a way of trying to achieve that.
We have also introduced an option for external investigation, as well as external decision makers, in order for people to feel that there is no sense in which local management might be influenced in the wrong direction. We have committed to giving a much higher level of reporting on cases, handling cases more quickly and trying to achieve early resolution, and bringing in mediation where appropriate. Those are the types of activities.
It was warmly supported by staff. People thought that it was a good response. That was published in July, and we are in the process of implementing it.
Q146 Lloyd Russell-Moyle: Will you implement every single one of the recommendations?
Alex Chisholm: Yes, that is our commitment. Not only will we do that, but we want to do it with effect. We want to see that it is helping to build people’s confidence in the organisation in which they work, and hence further pulse surveys and next year’s people survey will be very telling.
Q147 Lloyd Russell-Moyle: Just going back to the staff absences, when you write to us about what they are in the Cabinet Office, it would be interesting also to know the breakdown for race, gender and the usual characteristics. Particularly around race, it would be interesting.
Alex Chisholm: If you would like it around grade as well, there may be differences there.
Q148 Lloyd Russell-Moyle: That is important to see. The levels of bullying, harassment and discrimination were particularly high in the Foreign, Commonwealth and Development Office surveys. What has driven that? As the central Department, what are you doing with colleagues in the Foreign, Commonwealth and Development Office to address this and support them to learn from your successes?
Alex Chisholm: You are right to say that the first responsibility is with the leadership of FCDO, but we, as the Cabinet Office, also co-ordinate and take an interest in these matters. Certainly, I take that responsibility seriously, as does Fiona.
Speaking with the leadership of FCDO, they are absolutely clear that they are not satisfied with those results. It has been persistently high. It is not just a very recent phenomenon. We are interested in whether it could be because of the machinery of Government changes or the excessive pressures of trying to deal with the Russia/Ukraine situation, Afghanistan or whichever challenge they were facing, but it has been high for quite a long time.
There have been some special features historically in the Foreign Office particularly, because over half of the people who work there are locally engaged, so general approaches across the Foreign Office are less effective than making sure that the way in which they deal with people in the high commission in Islamabad is as good as the way in which they deal with people in the embassy in Madrid and so on.
The commitment that they have very publicly given to their staff is to focus on and tackle that. They launched a new dignity and respect at work policy in December; it is a bit like our campaign—ours is called “A Better Cabinet Office” and theirs will have a different name. They are really driving a focus on that. They recognise that they have some persistent issues. I have discussed it directly with the permanent secretary, who is 100% committed, as is the executive leadership, to trying to fix that.
Fiona Ryland: I have spoken to the team there as well. They are focusing on improving standards in leadership. They are gathering case studies for their heads and deputy heads of mission training. They are focusing on the support given to people who bring complaints, but also on action being taken and cases being resolved. They are going back to their management board on a bi-annual basis in terms of some of that case information, but also, when moderating performance ratings of senior leaders, they are using the people survey scores to aid in that.
Chair: Thank you very much indeed. That concludes our questions this morning. On this subject matter particularly, it is very much in order that I should thank our three witnesses for coming this morning and, indeed, pass on the wider thanks of the Committee to the civil service. I know that some of the topics that we have touched on have perhaps shown an occasional lack of appreciation, and I would wish that this Committee place on record our appreciation of our civil service.
Alex Chisholm: That is much appreciated. Thank you very much.
Chair: It is timely, given the topic. I thank our three witnesses and look forward to receiving anything in writing in addition to what they undertook. Thank you very much indeed.