Publications on the internet
Taken before the Committee of Public Accounts
on Wednesday 23 May 2012
Margaret Hodge (Chair)
Mr Richard Bacon
Amyas Morse, Comptroller and Auditor General, and Nick Sloan, Director, NAO, gave evidence. Gabrielle Cohen, Assistant Auditor General, and Marius Gallaher, Alternate Treasury Officer of Accounts, were in attendance.
REPORT BY THE COMPTROLLER AND AUDITOR GENERAL
Implementing transparency (HC 1833)
Examination of Witnesses
Witnesses: Spencer Chainey, Principal Research Assistant, Department of Security and Crime Science, University College London, Carolyn Downs, Chief Executive, Local Government Association, Jim Killock, Executive Director, Open Rights Group, and Matthew Sinclair, Director, TaxPayers’ Alliance, gave evidence.
Q1 Chair: Apologies for being late: we had a discussion arising out of our rather lengthy-five-hour-hearing yesterday. I want to read a statement in relation to that. There has been some coverage of it in the press, so we want to get something out. I hope you will bear with me. The statement reads:
"The Public Accounts Committee was presented yesterday with written material relating to welfare to work, shortly before we were due to take oral evidence on that subject from the Permanent Secretary of the Department for Work and Pensions. The Committee, on advice, thought it wiser to hear the evidence first in private, evaluate it, put the general issues raised to the Permanent Secretary and draw conclusions. As ever, we will insist that the allegations are fully investigated and the results of the investigation published."
That is completely irrelevant to you guys-it is probably gobbledegook to you-but it related to yesterday’s evidence session. There are a lot of you, and normally the first bit of the session is an opportunity for you, as the experts on the ground, to give us some insights as to the issues that you think we should explore when we take evidence from the accounting officers. It is really in that context that we are asking you to help us. You are all very different, as you have probably realised-you have very different interests in this across the piece.
I am going to start-going across the table-with Carolyn Downs, if I may. How do you, representing local government, think the publication of data has impacted on local authorities? What benefits has it brought? Where do you think things have not gone as well as you wished, and-from our point of view, as we look in the medium and longer term-how will we ensure that we get data around local authorities that will enable us to make proper comparative assessments of the performance of local authorities, which we will have to do from a value-for-money point of view? Whatever you feel-it is a freedom to say whatever you want. Nobody is trying to catch you out or anything like that.
Carolyn Downs: Thank you for inviting us. I should start by saying that the Report from the National Audit Office shows a pretty positive picture in relation to the publication and transparency of data in local government. That stems largely from the culture of local government being one of open decision making and decision making in public-with advice, but from officials in public. The culture exists where openness and transparency is in place and embedded.
The guidance that has been issued by the Department for Communities and Local Government to local authorities, which is non-mandatory, shows-as you will see from the Report by the NAO-a high level of compliance with that. I started in local government 30 years ago, and I worked in the library service. Even 30 years ago, local authorities were publishing their agendas and minutes of meetings, and those were available for the public to look at and take away. That was 30 years ago, and it is now web-based. As I say, to a certain extent, I would have expected a high level of compliance with the non-mandatory guidance, because that has been done for quite some time.
Likewise, when I became a local authority chief executive-that was in 2003-senior salaries were already being published. Members’ expenses were published in the press, and our accounts were as well. On the guidance, I have to say that there is an issue that I think you should consider. I live in the London Borough of Camden, and this is April’s publication, under the guidance, of anything over £500. You will see that that is a considerable amount of information and data. After 30 years in local government, I can understand a lot of this; I have to say, I am not sure how useful it might be to a resident of Camden, and that is an issue that we really need to think about.
Q2 Chair: What would you change?
Carolyn Downs: I think we need to stop being data-rich and start being knowledge-rich. Therefore, we need to get into a position where these data are put in a form which is helpful to the public. If I just pull a bit out: there is an organisation called-something-Security and it does "Repair & Maintenance-Other" for £4,000. I am not particularly sure how helpful that is to a resident of Camden. If you speak to the councils about the hits on these websites, they say that they are very low. Kent County Council has undertaken an exercise with the public asking them for feedback specifically on this transparency information. The public are very clearly saying, "We want it. We absolutely think it is the right thing to do. We think it can help you improve your service provision, but you need to provide it to us in a way that we understand and we can use." That is the No. 1 issue.
You asked me about local authority performance data, and after the Audit Commission intention for them to be abolished we have developed an online site that allows local authorities to compare all data. All those data are already publicly available. We are putting them in a form that we believe is helpful to councils and indeed to their residents-and potentially to the TaxPayers’ Alliance-to use that information to compare both performance and cost of services across all councils.
Q3 Chair: But it is not mandatory, is it? It is voluntary.
Carolyn Downs: It is not mandatory. There are 350 councils, and over 300 are already signed up using it. We have 40 new sites in the past month-new sites and new bits of information being created by councils. So it is not mandatory. My view is that once it goes public-once the IT platform is in place to do that in September or October this year-as a resident of an area that did not have that information I would be asking why that information was not available. It would be very interesting for us to see how quickly the final 40 councils come on at that point.
Q4 Chair: Just one more question, and then I will skip Spencer and come back to him, and go to Matthew Sinclair to comment a little on that. So you want this to be knowledge-rich; I think that is what you called it. Just give us a bit of an idea what that means and what changes you would want to see to enable us to be knowledge-rich. It sounds good to us, but what does it mean?
Carolyn Downs: For example, if you go through this information, a huge amount, as you would imagine-40%-of local authority spend is adult social care spend. I think it would be useful to residents, both as purchasers of adult social care through their personalised budgets and as residents ensuring value for money, to know how many people are being served for that cost. How does that compare with the cost in another residential care home, and how does it compare with the costs of purchasing that might be happening in neighbouring Islington? That is the kind of information that is helpful for the public.
Q5 Chair: And who should decide that?
Carolyn Downs: I think residents should decide what information they want from their council. That is very strongly the view of the LGA.
Q6 Chair: So it might be different in different areas?
Carolyn Downs: Well, perhaps I could give an example on that as well. When I was chief executive in Shropshire, bicycle theft was not really a major issue for the council and for the constituents of Shropshire, but we had to give prescribed data to the Audit Commission on bicycle thefts every year. It was just bizarre. However, my family live in Oxford, and bicycle theft is extremely important in Oxford. It is precisely those issues that I think local authorities and local politicians are in a position to determine-and residents; it should be a two-way conversation. Residents should be saying, "This is the information that we require of you. These are the data that you hold, and this is the form in which we require you to provide them." I think it is incumbent on every good council, and indeed every council, to provide the information that their residents require.
Q7 James Wharton: One of the advantages to the data that have been published is that organisations such as the TaxPayers’ Alliance can collate them and publish things that are useful for giving a national picture. How would you combine the ability of organisations to do that-to look across several councils-with the approach you are talking about, which would appear to move even further away from standardisation of data?
Carolyn Downs: I would not publish the data in such a way as to prevent an organisation such as the TaxPayers’ Alliance from pulling that together and comparing. If they are helpful to somebody, those raw data should be made available. It is for other organisations to manipulate those data. We local governments should manipulate the data in a way that is helpful to residents.
Q8 James Wharton: So what you are saying is that in addition to what is already published, local authorities should be doing more to interpret it in a way that their residents find useful, but you would still publish the data as they are at present.
Carolyn Downs: I think at some point somebody needs to understand the cost. I don’t think the cost of the publication is really-
Q9 Chair: Do you? Have you done any work in the LGA around that?
Carolyn Downs: No, but we will be doing that. We have just commissioned work to do precisely that, to understand the cost and the benefit. I imagine the cost of providing the data might be quite marginal, but if the cost is great and four or five people a year are hitting the data, I just don’t think it is useful.
Q10 Matthew Hancock: It could depend on who those four or five are, couldn’t it? If one is a national journalist and it ends up promulgated further indirectly-i.e. not from the website-then that is important. It has to be a wider consideration than just the number of hits.
Carolyn Downs: Yes, absolutely.
Q11 Austin Mitchell: Is it a substantial cost to comply at the moment? What you are proposing would make it more expensive, wouldn’t it?
Carolyn Downs: What we are proposing would make it more expensive, without a doubt.
Q12 Austin Mitchell: What are the costs now?
Carolyn Downs: We don’t know. Nobody knows exactly. The data that come from Camden are, in my view, on their financial systems-they have just done a feed on to a spreadsheet on their website. My view is that it is not a great cost.
Q13 Austin Mitchell: But that would be mainly in staff time.
Carolyn Downs: Probably in setting it up, but once you have the system in place and are feeding it? I don’t know. I think that information needs to be found out and that there needs to be a proper cost-benefit analysis of the information and data being provided.
Matthew Sinclair: I think some councils are trying to understand costs. Windsor and Maidenhead, for example, recently announced moving its limits down from £500 to £100, because it has been trying for a while to lead on that kind of thing. Its view is that the cost is immaterial. It is a system that exists and is exporting the data. I think its view was that, in terms of cost, it could go to a zero pounds limit and it would add no cost at all. It chose £100 only to avoid it becoming so much data that it was harder for the public to interact with.
There are two comments to make on what Carolyn mentioned-
Q14 Chair: And anything else you want to add.
Matthew Sinclair: Yes, sure. Partly for reasons that Mr Hancock mentioned, I am extremely cautious about using web traffic data. I understand why it has been done in the NAO report-those are the data you have-but I am extremely cautious about that. Some school spending data were released by the Department for Education. I know at least two people used that: The Guardian datablog did it and we did it. Ours saw national newspaper coverage, and The Guardian datablog has a huge readership. That could show up in web traffic for the DfE as two hits, because you go on there and download it. We are not going to send people back to that website; we publish the data on ours, entirely in compliance with the licence.
It is so unreliable, particularly because I think there is systematic unreliability in it. Raw data are likely to be downloaded a relatively small number of very important times, whereas things like the crime maps would be used a lot of times for someone to check their own numbers. It is not just unreliable, but systematically unreliable and will cause us to prioritise the wrong data if we lean on it too much. The only way you could really understand the benefits is to come at it almost from the other direction and go looking for examples of when it is being used.
The critical role centrally, both for local government and for the Cabinet Office in its relationship with Departments and other public bodies, should be ensuring the data is comparable, consistent and there. That is where it can be most useful. I think most of the harms can come when it tries to aggregate data and turn it into analysis. The old area assessments for local government were almost perfectly wrong in that regard, because a lot of the time you couldn’t get a lot of the underlying data on which they were based, but it did give authorities the chance to say, "We have had two stars." What really convinced me that this was a bad idea was when authorities with maximum rankings came to me and said, "We feel we have to do this in order to show we are good authorities, but we think it’s a complete waste of time." When someone who is getting a gold star tells you that that gold star is meaningless, that says something.
Beyond the web traffic, the transparency data can be very important. It is right that some of the risks have been identified, but the process can also be part of mitigating risks from other policy areas. If one method of getting greater efficiencies in expenses payments and things like that is to use procurement cards, the risk everyone has is that that becomes really efficient, and then you would have officials going out and enjoying that efficiency and using it to spend a lot of money very efficiently. Transparency neatly allows procurement cards to go from being something that causes people like the TPA to be quite concerned to something we can actively support and recommend because it can swing policies in that regard.
Our biggest concerns are with hold-outs-Nottingham council, for example, seems to be a systematic hold-out across-
Q15 Mr Bacon: What’s a hold-out?
Matthew Sinclair: They are systematically not complying both with voluntary things and are also being extremely poor at complying with their actual legal responsibilities under freedom of information law. There are some who are overtly non-compliant across the board with transparency principles and transparency law.
Q16 Chair: It’s just one?
Matthew Sinclair: Nottingham City Council stands out. There is no other authority that I would say is anywhere near comparable.
Q17 Mr Bacon: What is the reason for this? Is it just because they think it is not relevant or because there is a festering nest of vipers that they want to be kept hidden?
Matthew Sinclair: I think, in their view, they think it is a waste of time and a hassle, and they do not like it. They have said they are not going to publish until they are forced to. The Information Commissioner has been looking at them on FOI for a long period. We have submitted to the Information Commissioner on our own experience. In local government, our experience is that the vast majority of authorities are working to comply-there are varying degrees, but overwhelmingly they have a relatively positive attitude. There is a hold-out there.
There is another area-I think it relates to this "knowledge-rich" suggestion made earlier. There are still issues with the usability of some data. The Department for International Development has a tool that could be spectacularly useful with the project database, but in reality, trying to work out what these projects do is almost impossible. There is a huge amount of data about audit standards and things like that. Are they using domestic partner audits? That may be of some interest to people in DFID, but I was trying to work out things such as "Supporting the growth programme of the South African Government", which as far as I could see could be anything from funding a civil servant to building bridges to paying individual farmers’ salaries. I tried very hard today to find out what that was.
Q18 Chair: Who should prescribe that? If you are saying that the data out there is not useful, who should prescribe it? Where does the onus lie for ensuring that it is not just data dumping, but that it is useful data?
Matthew Sinclair: The best way to do that is to look at the examples where it is very useful. Business Link’s contracts database is a good example of one that works very well and is easy to use. I have always found that. I have never felt that I did not know what it meant or I could not find out more or it did not have what I was looking for when I have been looking at something in that. I do not think anyone will be able to develop a textbook for useful data. That, to some extent, limits what you can do with councils and other bodies where it might need to be prescribed. I do not think there is any need for a definitive handbook for how a Department does this, because they are so different in what they try and do with information, and there are not that many of them. It is almost a matter of each Department individually working out how to make the data more meaningful.
Q19 Matthew Hancock: Which Departments do you think are the best?
Matthew Sinclair: It would depend on your standard. In terms of the most aggressive-in terms of giving out more useful data-I would say the DCLG has been pushing the boundaries. I think that that is because they have had a clear sense that they need to be doing that if they are going to ask about the standards of local government itself. I am a little wary of singling out a Department, because there are very different-
Q20 Matthew Hancock: You have given us one. You could widen that if you like.
Matthew Sinclair: I agree. I think that that is one where you clearly have a big database that a lot of money has been spent on, as is reported in the NAO Report, and it is not yet at a stage where people could really use it to hold DFID to account. I am struggling, because I do not have any examples that immediately come to mind as being poor performers in departmental terms in the way that we do with local government.
Q21 Matthew Hancock: Can I ask you a related question? Some databases are sold, and you represent the TaxPayers’ Alliance, so on the one hand transparency can help to drive down costs, which supports taxpayers, and on the other hand selling databases, such as Ordnance Survey maps, drives down the costs to taxpayers and Government. There is clearly a balance to be struck there, and it would be very interesting to hear your point of view on where you think that balance lies and whether the Government are getting it right or should be doing more in one direction or the other.
Matthew Sinclair: First, on your previous question, there are actually two other comments worth mentioning about specific issues. One would be the Foreign and Commonwealth Office, where there is a problem not with the FCO itself, but with the fact that so much of the FCO’s operation is abroad, and you will often get responses to FOIs that say, "We cannot answer this. It’s all spread out abroad." There is a challenge in moving from the domestic operation to abroad. A similar issue applies to the Ministry of Justice with prisons, because you cannot FOI a prison.
On selling data, I would err towards giving it away in most cases. There are some cases where there will be a very specific and valuable commercial opportunity to create, and there can be a lot of value for taxpayers where there is a very special interest. Those might be cases where you do want to charge, so I would not want to say, "Always give it away for free." However, I would err towards definitely giving it away, because in terms of the taxpayer value question what you are almost talking about there is that it is something that people have already paid for through tax. We are talking about public information that has been provided for out of taxpayer funds. To then say that there is going to be an additional charge to work with it has the thing backwards. I think it works from the Exchequer perspective more than it does for the taxpayer.
Chair: Carolyn wanted to say something, and then Ian has a question.
Carolyn Downs: I think that I should just respond on behalf of Nottingham city council, which has been named and shamed. The information that they do not publish is this very same information that I just showed you. They say that to make this knowledge rich would cost them £100,000 a year.
Q22 Ian Swales: If I could come in on the back of that-I do have some experience in this area-there is a hierarchy: data, information, knowledge, wisdom. And that is data. Frankly-I have had it done to me and I have even done it myself-drowning people in data is a perfectly acceptable way of not giving people a good knowledge of what you are doing. I am interested in how we move on from fairly useless data, because I have looked at my own areas and it is very difficult to take much action or do anything with it.
In this context, information would be aggregation of that data into buckets that you can actually understand while still being able to see what is behind it. Knowledge is the thing that Carolyn was talking about, which is the benchmarking-type data that starts to help you to say, "What are we actually learning from this?"
I was trying to think what wisdom would be in this context, and I suppose that that would be the impact of policies, such as the example that Mr Sinclair gave in DFID, where you could have all the data, information and knowledge, but it still would not tell you whether it is actually working in the field. I suppose that that would be the wisdom.
In the context of all that, I just wonder-perhaps from the TaxPayers’ Alliance point of view-what you would like to see happen in Government and at what point the cost of doing it overtakes the benefit?
Matthew Sinclair: I think the further you move down the train you are talking about, the less there is the distinctive role for the public sector. Actually providing the data is something that only the public sector can do. Although many in the public sector could provide the wisdom, I don’t think it is a public sector role. We all have an equal role to play in that public debate.
There is a scale in between on that line that you have talked about. My view is that the focus should be on the data side, because although the quality of data is important in terms of comparability, a lot can be done much more quickly than people think in the private sector to turn data into something useful that can be deployed. I would not underestimate how much people can do. People have done a lot with local authority spend data-they have interrogated them to look for savings and have found savings. Equally, it will not necessarily always be a matter of becoming a big public story. If someone can look at data and discover that they can undercut a current cost, that can generate savings.
Q23 Ian Swales: Are you seriously suggesting that we should just dump these massive low-scale data-as you rightly said, Windsor and Maidenhead obviously decided to dump even more data into the public domain-and rely on the private sector, third sector or whatever to make sense of them and suggest what should change and so on, and that the public sector should do no more than that? That seems to be what you are saying.
Matthew Sinclair: I don’t think that it should always do no more; there are areas in which it can do good stuff. Crime mapping makes for an opportunity to do something-although there is an alternative to the crime maps, so there is an argument that even for that there is a rival private sector alternative that could do the job and would save a few hundred thousand pounds. In that case, a very good argument is that they wanted to get a single public source, because it was of such public interest. But to a certain extent, yes, by dumping lots of data out there, you would be surprised how much-
Chair: Ian, shall we ask that of Jim and bring him in, because it is a similar organisation?
Q24 Ian Swales: Linking to that point, how about the benchmarking data that Carolyn spoke about? In my own area, they have recently found out by benchmarking that they are paying more than the authority next door for care homes. All that is going on, plus any public sector organisation worth its salt is bound to have management information that is not put into the public domain. There must be other data that could be made available without costing a lot of extra money. I am trying to find the right balance between what people legitimately want to know and the resources that can be put in.
Chair: I am going to direct that to Jim, just to bring everyone in-and add anything that you want to add.
Jim Killock: To take the basic principle, data get cheaper, processing them gets easier and doing more things gets easier. This is a fact of digital society. The Government and local authorities should not be afraid of putting more information out there and be worrying about who will process it. It will become easier to process it and do good things with it, whether that is for the third sector, employees, users or private sector companies that are republishing things through applications-just see what happens.
The big question underlying this and underlying what I want to talk most about relates to the public data group and how it is managed, which is a point that has been rather overlooked in the National Audit Office Report. The key question is: how do you evaluate the impact of releasing data? You have a hard problem, because until you have released data and people start doing things, you do not know what the benefit is. You have to find a way of modelling the potential benefit of releasing any sort of data, be they the commercial datasets that were brought up shortly before-things like Ordnance Survey mapping and the other trading services-or the raw data that the Government generally produce. You have to decide, and find a way of measuring the impacts.
A few people have done this. Rufus Pollock famously did it, and it helped to inspire a lot of the open data agenda for the Government. But recently, over the past couple of years, we have noticed that the Government have not found themselves a model or a means of adjudicating and thinking about how to advance the data-what to prioritise, what publishing might have what sort of economic and social impacts. They have a bit of a problem, particularly around the commercial datasets. We have a number of companies-Ordnance Survey, the Met Office, the Land Registry and Companies House-all of which commercially publish data that have already essentially been paid for by the public by and large, In the case of Ordnance Survey, they make profits in the private sector, but they also make a lot of money from charging public sector organisations for this sort of information, particularly detailed mapping.
BIS is the owner of those companies, and we think they have recently started operating as a bit of a shop steward for those companies. BIS has basically been saying, "We think they should go on publishing their information for commercial gain, because this is revenue for the Government, and we don’t really know how to replace it." On the face of it, that is a fairly reasonable argument, but the problem is how you know what the economic benefits and the tax revenue benefits might be of releasing some or all of those datasets. How do you assess that? We asked them how they were doing that. We had a meeting with Ed Davey, and he said, "Well, look, we need evidence of the impact. Why don’t you go and get us the evidence?"
We had, at the same time, put in freedom of information requests for detailed information about the business practices of Ordnance Survey in particular so we could look at the demand and what demand might be generated by releasing it for free. They refused point blank to give us that information, and they redacted it in the freedom of information request. So the people who might have gone and done the modelling to give BIS the answer were unable to give the information that BIS had asked for because BIS had told us that they wouldn’t give the information to us.
We are in a situation where BIS have rejected the modelling done by people such as Rufus Pollock to see what the benefits of open data might be. They say that they need the evidence, but they refuse to give us the information to create the evidence, and they don’t produce any modelling of their own. The National Audit Office has picked up on that modelling programme and has talked about it quite a lot. It is absolutely serious, and just about everyone who has spoken about this so far has said, "How do we assess the benefits? How do we understand what is going on?" And the answer is that nobody knows because nobody has done the work.
Q25 Ian Swales: You are raising an interesting area that leads to another strand of thinking. If you take three of your examples, the Met Office, the Ordnance Survey and Companies House-I am thinking about people such as Dun and Bradstreet-all those organisations are effectively competing with commercial players in the field, which is rather unusual for the public sector. If you have organisations that are effectively playing in a commercial field and give their product away, you can well understand that other people might say, "Well, I’m not sure. This is not really fair competition, is it?"
I do not think this idea of selling Government data-maybe Matt might like to do more-is a hugely prevalent thing. Those are specific cases where you are actually working in a commercial arena.
Jim Killock: If they did publish their data for free, I don’t think the companies that are currently competing would go away. They would just take those data, build on them and do more interesting things. I don’t think there is a problem with releasing the data; I think people can do more if they have more information, and that allows the number of people who are able to compete and use that information to increase.
You need some means of adjudication, so I will just explain the current way that this is intended to be adjudicated. BIS owns a public data group that controls these trading funds, and they have a data strategy board that is going to get the profits from these data and will be allowed to plough those profits back into releasing certain bits of data. In order for data to be released, they have to make money out of the privatisation of data and the public-private sale of data. Somehow, that is meant to make more of the information public. If you imagine that you are making millions and millions of pounds on the sale of mapping data and you want to release those data, the minute you do you cut off all the money to get the data released. So the model is broken.
What happened was that the Cabinet Office wanted to release all this information. They persuaded the Met Office to do that, because it all gets paid for by the Ministry of Defence so it is not too difficult to say to the Ministry of Defence, "We are just going to switch some money around, and now we are going to release all the meteorological data." But with the others, people are paying all over the place, and even if it is mostly Government money it is quite difficult to organise because everybody is paying. Local governments are paying and Departments are paying, and trying to rearrange that in Government is difficult.
They were not prepared to go the whole hog and release those data. They have not worked out whether it is economically the best thing to do, but it probably is. The National Audit Office points out that the estimates for the benefits of releasing all these data look like billions of pounds, but so far the Government have managed to find £47 million of benefits over 10 years, or something. So something is going wrong somewhere, and it is probably in the commercial datasets. Because it is difficult to organise, they have bodged it. They have got this bodge mechanism of, "Let’s get the money and then use the money that we are getting from selling the data to pay for publicising it." Basically, it is too difficult so they have come up with a bureaucratic fix that will go nowhere. I think that is something you really need to look at.
Q26 Chair: I am now going to go to Spencer Chainey, who has been handling probably the most successful data in terms of public interest, namely the crime data. I just wondered, out of your experience there, what you think we should be looking at.
Spencer Chainey: If we start by thinking through the anticipated benefits of publishing the crime statistics on the national crime mapping site, the logic was that by publicly disseminating this information you better inform the public, and by better informing them you engage them and get them more involved in local community safety as well as reducing their personal risks. That is the logic. Providing a single source of information, so that we have a single site that is used throughout the country rather than individual police forces doing their different things, is obviously a strength. It improves consistency in the sort of information that is published, and there has been a lot of interest in the site-but you can sense there’s a "but" coming.
The interest has failed to be sustained. The figure that is often quoted is that 47 million people have visited the site since February 2011 when it was launched, but over 90% of those were in the first two weeks of the launch. The interest has not been sustained and I think the reason why, certainly from the evidence that we have picked up, is that people scratch their heads thinking, "What purpose does this site actually serve? The information that’s presented on there is interesting-lots of facts and figures on crime-but so what? How can I use this information?"
The information that is published on there is often, at best, six weeks out of date. If you looked on there today, it would be telling you about crime that happened six weeks ago or longer. People generally know what crime is like in their area, so publishing information about what happened six weeks ago does not actually affect you personally in terms of thinking through how you can use that information to deal with a situation at the moment. By not thinking through that purpose, I think the site has been a little bit of a failure. There was a lot of interest in the site to begin with from the amount of hits that it received, but that interest and that curiosity have not been sustained, because the purpose-"What is the point of publishing this information?"-still has not been thought through and realised.
The evidence is that the purpose of publishing this information is firstly to reassure the public that crime is not as bad as they think it is, to tell them what the police are doing about an incident that has happened locally and to give them the facts about what is going on. But it is also to give them information that they can use to influence their own personal decision making-"What can I do to reduce my chances of being a victim of burglary after there has been a burglary up the road in the past couple of days?" The information that is published at the moment does not do any of that, and it does not support any of that. We see it as really being an exercise in ticking the box-of saying, "These data are there; let’s publish them and tick the box of transparency."
What needs to be done? The current site needs to be improved to really meet that purpose. We need to think through what we can do with the information alongside other means of communicating and engaging with the public. How can we use this information in a positive way to inform and engage people? By publishing information in its current form, how can we use other facilities and services, and how can neighbourhood police teams make use of e-mail, social media and door knocking, going round speaking to people and adding some more immediate context to the statistics on the crime mapping site, which are six weeks out of date?
The information centrally needs to be relevant to our own personal circumstance. The focus should not be on publishing more, and publishing data for the sake of it; it should be very much focused not necessarily on more, but on information that is more relevant and that people can actually use.
Q27 Mr Bacon: Who is doing the door knocking?
Spencer Chainey: Neighbourhood police teams.
Q28 Mr Bacon: They go around knocking on people’s doors, telling them what?
Spencer Chainey: Let me give you an example of a scheme that runs in Trafford in Greater Manchester now, which has proved to be successful in two regards: it has reduced burglary by over 25% and it has improved public confidence. If you are burgled, the chance of you being burgled effectively doubles.
Q29 Mr Bacon: If I have been burgled-
Spencer Chainey: If you have just been burgled, your chance of being burgled again has doubled. That risk decays quickly. If you have been burgled, your neighbours are also at a higher risk of being burgled, but that risk decays quickly. When a burglary happens, the neighbourhood police teams in Manchester, particularly in Trafford where they initiated this work, go around doing the door knocking. They say to the neighbours, "I don’t know if you realise, but your neighbour was burgled last night." The chances are they know, because they have seen the police cars outside. They speak to the neighbours and are honest and frank with them, with that information. They then reassure them by saying, "Look, the chances of you being burgled are actually very low"-because they are; they are at a heightened risk, but still the general risk is pretty low-and they give them specific advice tailored to their individual circumstance.
The delivery of that sort of information, through door knocking, using social media and other forms, alongside the interesting facts and figures that are published on the crime mapping site at the moment, makes the information much more useful. It begins to help put the facts and figures on police.uk into a context supported by the other information.
Q30 James Wharton: You gave us a number of possible reasons why the use of the crime mapping website tailed off quite dramatically after the first few weeks. Are those based on research and talking to people who have used it-members of the public-or are they your assumptions?
Spencer Chainey: It is based on both: looking at the site hits, first of all, and at how they have tailed off, and it is based on research captured in the last six months that has asked people about their use of the site, and how that has tailed off. One of the figures is from the British crime survey; it found that in the last year, only 4% of the population has gone on to that website.
Q31 Matthew Hancock: How many people is that?
Spencer Chainey: It is 4% of 60 million.
Q32 Matthew Hancock: That’s quite a lot.
Spencer Chainey: But when you think that the initial-I am sure that everybody in this room probably went to that website-
Q33 Matthew Hancock: That’s about 2.5 million people.
Spencer Chainey: I am sure that everybody in this room went to that website in the first couple of weeks of its launch and were curious about what information was on there. I bet very few of you have gone back to that site regularly to have a look at the information that is published, because although it-
Matthew Hancock: I do actually.
Spencer Chainey: Certainly from the work that we have done, people don’t go back and have a look. The information there is interesting-but so what? How can I use this information?
Q34 James Wharton: I have no doubt, Mr Chainey, that you’re absolutely right that people look at it once or twice in the first week, then they disappear. What I am not quite convinced about is what needs to be done to make them come back, because if there was an easy answer everyone would have successful websites. So I am interested in how you have come to these conclusions and whether it has been done anywhere else. Is somewhere else doing this better? It would be great if we could create a successful website.
Spencer Chainey: There are a few trailblazer sites around the country that are taking forward some techniques and certain ways of improving on the police.uk national site at the moment. Hampshire constabulary and the community safety partnership in Hampshire are taking this forward. They have found that publishing information that is more relevant to people and giving information out daily and supporting that with the sort of engagement that I talked about, through neighbourhood police teams, makes the information that is published so much more relevant to people’s individual circumstances, and sustains their interest more.
Q35 James Wharton: And is that causing a pick-up in users? Is it measurable?
Spencer Chainey: It’s sustaining them. It’s keeping them interested.
Q36 James Wharton: At what level: 4%, 10%, 20%?
Spencer Chainey: I don’t know.
Matthew Sinclair: I think it is important that what we are talking about here is an application, not the data. The crime mapping website is an application using that data and that is where most of the cost comes in. That is why you need to worry about users over time and things like that. It is publishing data that police forces already have from a database they already have from pushing "export". It can almost be a very small number of users and people can experiment with different applications, which might get a different value.
The other thing I would say is that choice in the public sector is when this becomes particularly important. That rather answers the question of when police data will start to become very important. When you start to have democratic choice in the police services, you will be likely to see people finding this data quite interesting. It is hard to think how you could do the elections for police commissioners without a good amount of data on what police forces are up to.
Chair: Spencer, you want to come back on that, and then I will try to draw it to a close so that we can go on to the main session.
Spencer Chainey: I do not think it is just about the application. The data that is published on the police.uk website at the moment is pretty raw, but it is old. It is not relevant enough for people to use.
Q37 Chair: A month old is too old?
Spencer Chainey: It is too old. It is interesting facts and figures, but it tends to tell you what you already know: that crime in your area is low, or, if you live in a high-crime area, you know already that crime is high.
Q38 Matthew Hancock: But you could track it over time, could you not? For instance, coming up to a police commissioner election, I could say, "Is crime higher now in my area than it was when this guy got in?"
Spencer Chainey: But from the surveys that are done on the public, they are most interested in the information that is immediately relevant to their personal circumstances-
Matthew Hancock: Precisely.
Spencer Chainey: -not what happened three years ago as to what has happened in the past few months. It is about looking at it in terms of what has happened immediately.
Q39 Matthew Hancock: But if you were about to go to the ballot box.
Spencer Chainey: It is about publishing data that are much more relevant and more timely.
Amyas Morse: I just want to come back to the point that was made by two of you, which was that sometimes the data that are available off the system do not have adequate descriptors. If you want to be able to use the data, you have got to do a bit more than just lift that off the system. You have to go back, look at the descriptor fields and say, "I’d better put something in that people can use." It is not just about apps, it is about a user being able to look at it and understand it. I am not saying it is wrong-improving basic data quality is always a costly business. That does not mean that you should not do it; it just means that you should be realistic about it.
Matthew Sinclair: To a certain extent, there needs to be a limit at which the public sector leaves that to us. A bit of detective work by a journalist or a think-tank or a politician is not the end of the world. There will be a balance in terms of the descriptors that the public sector provides and we are saying that we can do some of that hunting ourselves. There is of course always FOI as a backstop on that, although part of the ambition here is to reduce the use of FOI.
Chair: Austin wants to come in very quickly and then Jim, you wanted to come back, did you?
Q40 Austin Mitchell: I can see the value of local information for the pushy middle class. Almost certainly, my family are now looking through local sites to find the old people’s home to send me to, and being from Yorkshire, they will send me to the cheapest. I cannot see why local authorities should go out of their way to help you, because you are pushing a cause, which is to shrink the public sector. Therefore, in providing you with information for that campaign, they are allowing you to grind an axe to plant in their own head. You do not do it with any comparators of what the spending is in the private sector or what salaries are in the private sector. Apart from the beautiful row you have created about the salaries of top public servants in local government, what major economies have you produced from the information that you have got out of local authorities?
Matthew Sinclair: I will give you one tangible example, one of the most recent examples. We put out a publication on the cost of the glossy magazines that police forces were producing. Several of them said, "We thought we had to do this. It turns out that we can do this in a much more affordable way. We are going to cut this and change this and we can get better value." There are many examples like that. I know that the ones that get a lot of attention are the bolshie, angry publications that we do, because that is interesting, but there is an impendent value to that. However, the TPA has a perspective and we are not trying to deny that. Fortunately, in a pluralistic debate, we have lots of people with lots of axes to grind, and hopefully we produce a result which is productive. Again, I do not think it should be about helping us; it should be about putting out information. As I said, I am not trying to plead that all this should be done for us. I am quite happy to say that some of that detective work should be left to the people doing this. But other people use these data too-people from all different kinds of perspectives.
Carolyn Downs: I understand absolutely the point that Matthew makes, but the point I would make is that if people are going to use these data, they need to use them accurately and responsibly.
Q41 Chair: You do not think that that happens?
Carolyn Downs: I think sometimes they are used inaccurately and misinterpreted. Then that is spread across the media, and that does damage to the public sector as a whole. I just wonder how helpful that is. I am, and local government is, absolutely 100% in favour of transparency. We are working hard to make sure that the data that we provide are meaningful and relevant to local communities. It is important, therefore, if people are going to use the data, that they are equally responsible with the use of those data.
Jim Killock: On that point, I would defend Matthew a little. It is up to Matthew to make his case, and if he makes a stupid case, it is up to you to refute it.
Carolyn Downs: We do.
Matthew Sinclair: Meet bad speech with good speech-that has to be the way we work in a democracy.
Jim Killock: On the general implementation of datasets being released, I think there is a problem. We highlight, in particular, one from perhaps the other side of the field-the strategic export controls database. Campaigners have been requesting it, in reusable form, since November 2010. It has never been released, but obviously there is a strong public accountability case for publishing it. In fact, campaigners have extracted the information from other sources and published it themselves, having got so fed up waiting for the Government to do it. We have to say that accountability is not always being met in a number of things, and that is a clear case where it should be.
I also wanted to highlight one of the National Audit Office’s concerns around privacy; it raises a privacy risk. You have Tim Kelsey talking shortly about the Government’s work in this area. We would like to ask why Kieron O’Hara’s report on privacy, which is highlighted by the NAO, has been demoted to a mere consultation response to the open data policy consultation. It is submission 119, listed just before that of Kirklees District Council. It is highlighted as a major piece of work by the National Audit Office, but it is being ignored, it seems, by the Government’s policy machine, and by Tim Kelsey in particular.
We would also like to know why the transparency tsar has been pushing some questions that we think have got nothing to do with open data within the open data policy field. This relates to the publishing and use of personal datasets-things like your health data or benefits data. A large part of the open data consultation was around giving commercial access to these personal datasets through anonymised access, so that a company would be able to query the NHS’s health records-as long as you had not opted out, this would all just happen and the privacy risks would all be dealt with by "anonymisation" techniques.
We believe that anonymisation techniques are extremely risky-they are very problematic-which is highlighted in the Kieron O’Hara report. At the end of the day, access to personal datasets through anonymisation, through commercial means, has zero to do with open data, which is about publishing in a reusable form, without licence restrictions, to everybody. You could not possibly do that with personal datasets through anonymisation techniques or not. You just could not do it. Somehow, this has all got lumped into the open data agenda. We would like to know why.
We would also like to know whether the Government, and Tim Kelsey and others, are willing to have a public debate about that policy area, entirely separately from open data, because we think it deserves that.
Chair: Thank you. That was a very informative and good session. Many thanks to all of you for coming. You are welcome to stay and see the evidence. Thank you.
<?oasys [pg6,cwe1] ?>Examination of Witnesses
Witnesses: Tim Kelsey, Executive Director, Transparency and Open Data, Cabinet Office, Sir Bob Kerslake, Permanent Secretary, Department for Communities and Local Government, and Dame Helen Ghosh, Permanent Secretary, Home Office, gave evidence.
Q42 Chair: Welcome. Apologies for running a little late. I have got my papers muddled up; that’s why I’m being a bit hopeless.
Let’s start by saying that there is a lot of success in this and I think there will be general support across the Committee for greater transparency. Everybody thinks it is a good thing. When you established the policy, you had a whole lot of objectives, if we can start with those. You wanted better accountability. You also wanted economic growth. We had a little argument, discussion there about the use of data for economic activity. You wanted to create greater choice and you wanted to improve quality. I feel in a way that, if you are going to make sense of this very complicated policy, you need to prioritise within that. I wondered where your priorities lay. What is the purpose; why are you doing this?
Sir Bob Kerslake: Let me kick off and then Tim can contribute if he wants. All of those three have been important. If there has been any issue of impact and priority, it has been perhaps that you can make progress more quickly on one of those priorities than you can on some of the others. If you take the accountability issue-and I listened with great interest to the conversation before we came on-you can move relatively quickly, by putting data out there and allowing people to analyse it and form their own judgments and open debates. In some of the other areas, let’s say the economic side, it requires more intensive thought and effort and there are more complex issues to deal with in terms of the release of open data and their relationship with trading accounts, such as Ordnance Survey. They are all important. My personal view is that we have seen a quicker impact on public accountability but the economic side will take longer to deliver. I don’t know if Tim wants to add anything.
Tim Kelsey: The original policy started with a big focus on fiscal transparency-budgets and spending-and we have heard a bit about that. About a year ago, we added to that a layer that was all about releasing data.
Q43 Chair: What was the purpose? You put fiscal transparency, to what purpose? What was the aim?
Tim Kelsey: Public accountability, principally.
Q44 Chair: Accountability, okay. Let me ask you a question about that. I tried to enlarge this little thing but I couldn’t.
Tim Kelsey: I saw that yesterday.
Q45 Chair: I am sorry. Others won’t have seen it. I am holding up our tax statement, telling us where our money is spent. I have got the Mr Smith one. That is not the one I got as Ms Hodge. There is something interesting about that. I will point out one thing to you. It goes to the heart of how you present the information. There is a great big light blue thing there that is called welfare. I could not even read it, but it looks like a third goes on welfare. Actually, when you go behind that, the reason that it is so big is because pensions are included. My entitlement to my old-age pension, which is slightly on a contributory basis, is rather different from housing benefit, income support or whatever. The way it looks there is as if a heck of a lot of my money is going on welfare-a third of the money that I pay as a taxpayer is going on welfare-but the way that you have chosen to define it does not provide the sort of data and information to allow me, the taxpayer or citizen, to make a fair assessment of value. It feels partisan.
Tim Kelsey: I think on the back there is another sheet-
Chair: Yes, and even on the back of that sheet it says "old age", not "pensions". If I did not know and I had not gone beneath it, I would take that as being pension credit, and it is just not true.
Matthew Hancock: Old age encompasses pension credit-
Chair: You may know that, Matt, but what I am saying-
Matthew Hancock: Also the welfare state includes pensions-
Q46 Chair: You may feel that, Matt, but from my point of view welfare is benefit payments and that sort of thing, and not the contributory money that I get as a pensioner, and you have chosen to put it in in that way. So fiscal was your great priority, but I worry that the way in which you choose to present the data is of itself leading somewhere-it is perhaps not as open as one would want it to be.
Sir Bob Kerslake: I am happy to go back and check that, but I think that the commonly accepted definition of welfare would include pensioner payments-
Q47 Jackie Doyle-Price: As credited by the DWP.
Sir Bob Kerslake: Yes. I think it would normally be that.
Chair: As a taxpayer or as a citizen-
Jackie Doyle-Price: That answer is very helpful.
Chair: As a taxpayer and citizen, I did not find it that helpful. I think there is a distinction to be drawn between welfare payments and pension entitlements.
Matthew Hancock: I think Mr Kelsey was saying that on the back it goes into that distinction.
Chair: It doesn’t; that is the point.
Matthew Hancock: It does-between old age and not.
Chair: Old age for me is pension credit.
Q48 Austin Mitchell: The point is emotive. That way, if you just look at the pie chart and not at the back, you get the impression that all our money is going to support a set of scroungers-according to The Daily Mail-who are lying around and living the life of Riley at our expense. If it is contributory, and it if it is pensions, it is not in that category.
Sir Bob Kerslake: I hear what you are saying. I certainly do not think that that was the intent. The analysis seeks to show where Government spending goes in relation to different types of expenditure, to allow people to ask questions and to open up debates. Clearly, on a two-page document, you are not going to give full justice to the range of issues about spend.
Jackie Doyle-Price: I think we can trust the public to make an objective assessment of the information put before them.
Q49 Fiona Mactaggart: Have you done any research about what the public understand?
Sir Bob Kerslake: On this particular sheet, I cannot-
Fiona Mactaggart: I do not mean about this particular one. We are talking about mass categories, and I am wondering what your research tells you about what the public understand. Presumably, you would use that to inform the kind of categories that you put spending into in order to avoid inappropriate-
Tim Kelsey: I was saying that the first priority was fiscal transparency, then we have moved on to the transparency of things that really matter to people-that is health, education and criminal justice.
Q50 Ian Swales: There is also a point about materiality here. The Chair makes a very good point that a huge slice of the pie is named one thing. I take my local council to task every year because they, out of their £120 million net spend, have over £25 million down as "other". I would quite like to know what "other" is, when they go down to a tiny figure for some elements of what they spend and yet "other" is a huge number. Sure enough, it does hide some interesting things. I would have thought that if you have a pie chart, such as the one there with more than a third of it as one colour, it is not beyond the wit of someone to say that there is another big number in there, such as pensions, that we would quite like to separate.
Sir Bob Kerslake: I suppose I would ask whether this very conversation is not illustrating the point. You could not have had the conversation that we have just had until Government took the first step to publish something. Having published it, you can then have a debate about whether it tells you the story or tells you enough of what you need to know. But prior to this, people had nothing that told them what their funding was going on.
Chair: I think we all concur with that.
Sir Bob Kerslake: The question that you are raising, which is a perfectly fair one, is, how do people read what we put out? Do we know how they understand it? Can we, ahead of publication, do some work on that? I think that is a perfectly fair challenge.
Q51 Chair: I think there is a point about unintended consequences. It is similar with the crime data. I just don’t know if this is true; perhaps Helen can help. When you published those maps, obviously the intent was to get greater participation. We have heard questions about whether they are useful in their current form and whether they could be improved. What struck me when I looked at the data was, is it an unintended consequence that the insurance industry now makes it even harder for you to get insurance if you are in an area where there is a concentration of crime? Is it an unintended consequence perhaps that even house prices might be affected by the incidence of crime if estate agents decide to do that? Are you thinking about that when you think about how you put data in the public domain?
Dame Helen Ghosh: Indeed; we are thinking about that. Going back to the point that you, Chairman, made about insurance companies, I think they have long studied the data publicly available, and have been for many years, about crime rates. I suspect that all the Members around this table will have had individual experiences that reflect that.
We track through, for example, the crime survey of England and Wales, as it is now called, things such as public attitudes: what did people think when they logged on to the site? Did it make them nervous or more confident about crime? We have seen no evidence that it has any effect on house prices. Again, what a lot of this evidence is giving us is what local people would know anyway. They can already get this information from individual force crime mapping. It was a different kind of form and was often heat mapping rather than street by street, but I do not think that this publication gives us any unintended economic or other consequence.
As Mr Chainey said, the original driver for this was public accountability and engagement. I find myself heartily agreeing with a lot of what Mr Chainey said. At the core, if Members have looked at the site, there are data about crimes in an individual or postcode area. But increasingly, it is enriched a round and about by precisely the kind of connection to the neighbourhood policing team, who is the living face of supporting local people on issues of crime. We get links to the local forces’ tweeting: "Do you want to be tweeted about what is happening?" They are giving YouTube links to the crimes that have just recently happened, and they give you-this is a mandatory part of the site-the link to Her Majesty’s Inspectorate of Constabulary, so if you want to go into the comparative data between your force and other forces, you can instantaneously get links in there.
When I say that I was heartily agreeing, no, just having the crime map is not enough; you need to support it with a lot of other stuff. We have been working with a number of forces on the trailblazers that you talked about, where individual forces like Avon and Somerset, and Hampshire are experimenting with how people can connect more closely.
All of that is very rich, but as I said, the original motivation was, how can we get people engaged? Obviously, with the election of the police and crime commissioners in November, how can we give local people the kind of information that they need to hold local politicians to account?
Q52 Chair: In this conversation, is there data that you have chosen not to publish because of the consequences? This question is probably for Tim Kelsey.
Dame Helen Ghosh: Well, there is personal and security data.
Chair: That is so obvious. What I am trying to tease out is, are there limits to this? Is there data that you have thought there may be an unintended consequence? I am not sure you have entirely answered my question about insurance companies that use them. Are we making it easier for them to discriminate? Is there stuff that you’ve thought, "Actually, we don’t want this in the public domain," apart from the risk register, but I won’t say that?
Tim Kelsey: We started 10 years ago publishing the hospital data, which has allowed us to calculate death rates in hospitals. With that data there was a big anxiety at the beginning that there would be the unintended consequence of doctors picking easier patients and not operating on the more difficult patients. That didn’t happen and, as a result, doctors are now able to save lives that they previously would not have been able to do because of the amount of information that they can use and compare. That was a really important lesson, in fact, for other public services that are now beginning to come on line, because the feared unintended consequences did not really happen.
That 10-year experiment with hospital data has led us to be still cautious but more confident about things like looking at how we can give parents and pupils better access to the data that the education system holds on them, in a way that is both secure but enabling of improvement and looking at outcomes and so on. So, to some extent, I suppose the anxiety would be that publishing data might have some perverse consequence on the behaviour of a public service professional-which hasn’t been borne out-but as a result, going around Government and deciding, in a way collaboratively, what sort of data would make most sense to the profession or to the public, or the manager of a service, has led us to target and prioritise certain areas. But we haven’t really come up against, yet, a data set that we decided not to publish because of that. That hasn’t happened yet.
Q53 Stephen Barclay: Well, there are a couple from our previous hearings that we can perhaps bring out, but just on that previous point, the wider question that I was interested in, Mr Kelsey, is what impact on behaviour you think transparency may have. As someone who is very in favour of publication, a possible risk is developing a risk-averse culture and people being less willing to take risks because of transparency. Is that something you are concerned about, or has any consideration been given to that?
Tim Kelsey: Just slightly going back to the health example, which is actually where I came from, one of the best and most important examples we have is that the current medical director of the NHS, Bruce Keogh, led the cardiothoracic surgeons to self-reporting of their mortality outcomes seven years ago. Seven years later, in properly peer-reviewed fashion, we can now demonstrate that death rates have fallen by about a third in many procedures-attributed by the surgeons to the fact that they were able to have access to comparable standardised data, which previously did not exist. So in that case what we see is both productivity and quality improving directly as a result of publication, and we are now busily building that evidence base in other services and sectors. I spend a lot of my time calming people down when they tell me that perfectly mature and responsible public servants and professionals are about to do something outrageous because suddenly their performance is to be reported in the public domain. It doesn’t happen.
Q54 Mr Bacon: That is an extremely powerful example of how being more open and honest about the way things happen, including mistakes, has enabled improvement. I came across a very similar example with the head of Ford worldwide, Alan Mulally, a while ago, which I would commend anyone to listen to, about the way they were much more open about mistakes within the organisation.
This begs a question-it’s probably for you, Sir Bob-about whether the Government should be more open about publishing gateway reviews. Now, I have had conversations with Mr Watmore about this-sadly, no longer in the Cabinet Office-and I was looking up one that I had with him in May 2011. I was asking about the universal credit, and I asked him if he was prepared to let us have a copy of the gateway review at the starting gate. He said, "I don’t have a problem with that," and I replied, "Great. It seems to me that if there are issues to be flagged up, the more out in the open they are-rather like the Olympics-the more likely they are to get addressed." Mr Watmore replied, "I have always thought that. We have agreed on this in the past."
What Mr Kelsey was just saying with his example with Sir Bruce Keogh was in exactly the same vein. It was a different area, but it was exactly the same point: learning from mistakes. Something I have always been puzzled by is why Government does not learn from its mistakes, particularly, but not only, in the area of IT, where things go wrong again, and again, and again, and again. I have come to the conclusion the reason the Government does not learn from its mistakes is because it doesn’t have a learning curve. If you don’t have a learning curve you are not going to learn. In the case of this particular example of the universal credit gateway review, Mr Watmore arranged for it to be placed in the House of Commons Library and I went and got a copy of it, and I happen to know that a particular computer journalist then applied to get hold of all the gateway reviews, and they were refused, including the one that had been placed in the House of Commons Library. So there is still a culture of intuitive, instinctive secrecy that militates against the very helpful phenomenon that Mr Kelsey was just describing.
Sir Bob Kerslake: Yes. Actually, we are looking at this very specific issue as part of the civil service reform plan. We have not yet published that, so I cannot say exactly what will be in the plan because we haven’t finalised it yet.
Q55 Stephen Barclay: When is it due?
Sir Bob Kerslake: It is due in June. I am very sympathetic to the publication of the RAG ratings.
Q56 Mr Bacon: Are you? That’s good, because the Cabinet Office structural reform plan monthly implementation update originally stated that gateway reviews would be published, and then it somehow got downgraded to a draft. From what is publicly available at the moment, the position of the Government is not to publish gateway reviews. You sound like you are saying that that is going to change. Is that right?
Sir Bob Kerslake: Watch this space, but I am sympathetic.
Q57 James Wharton: I broadly welcome in principle the idea of publishing information, but there are lots of risks. We have already briefly discussed some of the risks that exist, and I am just a bit concerned about the "Transparent Government, Not Transparent Citizens" report that was commissioned. I understand that the Cabinet Office hasn’t formally responded to that report, which obviously highlighted a number of areas of concern. Is the Cabinet Office going to respond formally to that report?
Sir Bob Kerslake: Tim will say a bit more, but I think the intention is to pick this up as part of the White Paper that we are due to publish.
Tim Kelsey: This is the Kieron O’Hara report. There is a bit of misunderstanding here. We very much applauded what Kieron did. Basically, we are planning a White Paper on the right to data very shortly, which will, I hope, reassure people that we listened very carefully to what Kieron had to say. As we move into a world of very advanced technology, so much data that is part of this public data movement is potentially, theoretically susceptible to interrogation by that emerging technology. We are taking privacy extremely seriously.
Sir Bob Kerslake: You will see a proper response in the White Paper. On that point, openness to publication of data more generally does not mean that we shouldn’t be absolutely rigorous about personal data. I wrote jointly with the Information Commissioner to every local authority chief executive pointing out the importance of this. Indeed, they are at risk of quite significant fines if they get this wrong. So I don’t personally see any mismatch between openness of data wherever possible and a very rigorous approach to personal data and privacy.
Q58 James Wharton: Have you quantified within the Cabinet Office the potential risk of the Government’s overall push towards more transparency? I have been looking at the report. Fraud attempts to a value of £7 million directly related to transparency releases have been found in local government. Obviously, that is not a huge number across the scale; but, then again, that doesn’t necessarily mean that every instance has been identified. What do you assess the overall risk potentially to be?
Sir Bob Kerslake: My personal view is that I don’t think there needs to be a risk associated with transparency, particularly if you have well run authorities. Part of the work that was done there also found that other local authorities that did have good anti-fraud policies saved something like £20 million from the way they operated things. So I don’t think this is an issue of transparency, per se, creating the risk, but it will create risk if you have an organisation that hasn’t done the right things in terms of having a clear strategy to tackle fraud. That is really the key issue, not the transparency itself. Weak policies on tackling fraud are what create the problem.
Dame Helen Ghosh: We have been thinking about this, as I am sure other Government Departments have. We have thought about how that fraud could operate, for example, in our large data sets. As Bob says, we have thought about how it could happen, and we have concluded that, as it is audited by the National Audit Office, it could only happen if our fundamental financial management systems are in any way weak. They could look down our list and see when particular payments are made for particular things and then decide to send in a fake invoice pretending to be so-and-so security services on the right date to see if we pay them rather than the real invoicer. We have thought about how that could happen, and we have checked our systems to make sure that it couldn’t happen. I imagine across all Government Departments, under NAO guidance, we have all been thinking about that.
Q59 James Wharton: To follow that question up in one more way, if you have identified that transparency itself should not give rise to a risk, the risk predominantly lies where systems behind that fail to function. Have you quantified the potential cost or risk of systems failing to function linked to the transparency? There is still a risk, even if it is indirect, that comes from this process.
Sir Bob Kerslake: There is quite a lot of work being done on quantifying the potential risks around fraud and the actions needed to safeguard against that both in central Government and local government. Quite a lot of work is being done on this. I am happy to share that.
Q60 Ian Swales: I suspect that, rather like the Public Accounts Committee, a lot of these databases do their best work simply by existing. Those of us who are part of the parliamentary expenses database know that all too well. My question is: having got these databases, one has to look at the information in terms of how it is described and who is using it. I will give two examples. I think the medical database that Mr Kelsey spoke about earlier is a really good example where some good work happened because professional people were using the information, understood what it was and made things happen. However, I go back to your comments around schools. I would say that a lot of the schools’ information is used by people who do not actually know what it means. This goes way back to leagues tables. That does cause, first, members of the public to take ill-informed decisions. Secondly, I totally do not accept your assertion-there are a lot of negatives in this bit-that it does not cause people within schools to do things to make the databases look better, like not entering pupils into exams, for example. It is a long question, but my real point is that people need to have a certain amount of knowledge to use some of these databases. The school database is a particular hobby-horse of mine. I am wondering how we can make sure that, as we release information, it is not misinterpreted and we take account of who is likely to use it and what they are likely to do with it.
Tim Kelsey: This goes back to the previous conversation about how do we create a knowledge economy around the data. That is a really big challenge for this agenda. The emphasis of the current programme, which is to get the data out there and to improve their quality, is right. But I absolutely see that we need to do much more to make sure that around that we create tools and applications, and that we work with the private and third sector, as you said earlier, to make it more sensible to people and make knowledge the net result. In the case of schools, the Department for Education will shortly publish one of the Prime Minister’s transparency commitments-the national pupil database-precisely in an effort to release much more granular data to help power a sort of information economy and to help other people build their apps and so on, so that people can be better informed.
Q61 Ian Swales: Let us take an example because, as I understand it, a decision has been taken not to publish value added information. One of the most valuable things a parent can do is to forget what the average level of the intake of the school is and to look at what is the school doing in terms of improving the attainment of its intake. Yet, as I understand it, the Government have decided not to publish those data. Why would that be?
Tim Kelsey: The national pupil database will precisely allow people to develop measures of value in teaching, so I am not quite sure-
Q62 Chair: Can I just check you on that one? Sorry to interrupt you on that, Ian, but it has been one of the bees in my bonnet. As I understand it, we do not have any data at the moment on the academy schools. But, when we get the data, they will be by trust-ARK or whatever the trusts that run several schools are called-so you will not be able to disaggregate it to get the information on a school-by-school basis. As I understand it, I think that was what we were told.
Tim Kelsey: Let me take that one.
Ian Swales: I have not heard that one. There is the value added, but I have not heard that one. Most of the academies of the trust that provides an academy in my constituency are in Essex, so it is not going to be very interesting to see an overall performance figure.
Sir Bob Kerslake: Can we take that back? I have not come briefed with the detail of the education thing. I just want to make a general point. This is really quite important for me. The Government’s impetus at the beginning was to kind of show and not tell, to get on with being open about data and publishing it. If you move forward, each Department is now charged with doing two quite important pieces of work. One is a quality data plan, so how are they going to raise the quality? Secondly, they are all going to have to produce open data strategies about how they are going to make the data more useful to the user. It is perfectly right for you to challenge on individual areas, and we have built in something that should push them to carry on improving what they produce.
Q63 Ian Swales: My final point on this would be that if Departments have got their strategy, to what extent are they talking to users and potential users of the data to find out what they are doing with it and what they might want to do with it?
Sir Bob Kerslake: They should be, and if they are not, that is a legitimate challenge either for you or for their Select Committee.
Q64 Matthew Hancock: So far, we have talked mostly about outcome data and a specific case about contracts. Can I ask how this push for more transparency and the benefits of more transparency on outcome data intersect with the existing framework for transparency on what I would call process information that mostly comes through FOI rather than publication? In the evidence earlier, there was a strong argument from the LGA that-if I can try to paraphrase it-outcome data is better than sometimes partial or sometimes inadvertently misleading information that is presented in a more analysed form, including minutes from meetings and what have you. What is your view of the intersection of those two?
Sir Bob Kerslake: For me, open data is only one part of the transparency agenda. It is an important part, but it is not the only part. How decisions get taken and the openness to access to that is just another part of the transparency agenda. I do not think that by putting a lot of data out there we answer the whole of the transparency question. FOI is often not about the data; it is about how the decision was reached and so on. I think that that is perfectly compatible. In other words, it is another part of the transparency agenda.
Q65 Matthew Hancock: It is another part in the question of making sure that there are not unintended consequences from publication, which, largely, we have not found in the outcome data publication. How widespread are the concerns about publication processes?
Sir Bob Kerslake: On the FOI side, the principal concerns have been around, I guess, three things really. One is to ensure that advice to Ministers is kept separate, but that you have openness in almost everything else. Secondly, you protect absolutely personal interests in relation to FOI and, where you absolutely have to, some of the commercial issues as well. So if you look at where the tensions come on FOI, leaving aside the "small p" political tensions that there are in FOI, they usually centre around advice to Ministers, personal information and the boundaries on commercial information.
Q66 Mr Bacon: I know it is a rather radical thought-I am sure the answer will be no-but what do you think of the case for publishing all advice, including all advice to Ministers? Ultimately, when Ministers make up their minds about things, they have to do so on the basis of facts and opinions, including their own political opinions. They have lots of sources. They do not just have the civil service. There are lots of think-tanks out there. In fact, Charles Clarke told the Public Administration Committee some years ago that the civil service was having trouble keeping up with the ideas that were flowing out of think-thanks, there were so many of them, and it has become much more professionalised in the last 10 to 15 years.
If decisions are taken in the light of all this, presumably they ought to be defensible based upon the Minister’s own political point of view and taking account of all the things that he or she has had to take account of. Why would it be so horrendous to make all this advice public?
Sir Bob Kerslake: It would be a very radical constitutional change, as I am sure you are very aware.
Mr Bacon: I didn’t come into politics to leave things where they were, Sir Bob.
Sir Bob Kerslake: The key argument that militates against that move is that it would have the consequence of impacting on the ability of the civil service to speak truth unto power to give very clear and direct advice to Ministers. It would weaken the processes of Ministers making their decisions-that would be the key concern. The potential consequence might well be-frankly, I saw a bit of this in local government-the veneer of that advice being fully available in the public domain, but that actually other routes are created for advice to be given to Ministers. It is a very profound change. It is clearly not a decision for me, as head of the civil service, or, indeed, as permanent secretary at CLG. It is a profound change in the way the relationship works between Ministers and officials. The consequences would be quite profound for the way in which advice was given to Ministers.
Q67 Matthew Hancock: What does that mean?
Sir Bob Kerslake: I think it would influence, as I said earlier, the ability of civil servants to speak truth unto power to give unvarnished advice.
Chair: There would be less on paper.
Sir Bob Kerslake: There could be, yes.
Q68 Chair: Can I put to you an equally radical, but really important issue? Our job is to follow the taxpayers’ pound-you have heard me bore you to death on this one. It is the information-Steve may well want to raise it-about taxpayers’ pounds that are going into areas where we cannot follow them: private providers providing public services; Network Rail, which I thought Steve might want to raise; and LOCOG. I can think of endless examples of issues that have crossed our agenda. PFI contracts are another really good one where we are trying really hard to assess whether there is value for money. An FOI extension or some contractual obligation to provide the data so that we can see both probity, A4e yesterday, and value for money, PFI or any of these other contracts, would be hugely beneficial to your accountability agenda. Go on-respond to that one.
Sir Bob Kerslake: I will. I will deal specifically with the issue about contracts, because quite clearly we are moving into a world where more is commissioned and less is done directly. My personal view is that we have to build into the contracting processes the intentions around open government. There has to be clarity in the contracts about the expectations on the contractor in terms of providing information. That has to be part of the deal. There will be information that is commercially sensitive, but that is distinct from what you would require if you are going to fulfil your intent, if you like, in terms of open government. That is my personal view. This is a changing picture. What I think we are learning is that being very clear what you require by way of information is now something that is being built into contracts, but wasn’t built in as clearly in historic contracts.
Q69 Matthew Hancock: But isn’t commercial confidentiality often used as a catch-all, rather like the market sensitivity catch-all in releasing statements to the Opposition ahead of a financial statement that always seem to cover things that wouldn’t really move the markets very much? Commercial confidentiality likewise, if there is no structure in which it is judged, can easily be used in a way that, as you explained earlier, has blocked things becoming public that, in some examples, have led to more cost-effective decisions.
Sir Bob Kerslake: Personally, I think if we are not careful it can be used as a blanket catch-all, which is why you have to be very clear in the contracting process what your expectations are about data.
Q70 Chair: Can I just get it clear from you: are you saying that policy is going to change on this now?
Sir Bob Kerslake: I am saying that it should be part of good procurement practice to specify the information that you want. Whether or not that is a change of policy, I think it is, if you like, learning from the new world we are now in.
Q71 Mr Bacon: Can you give me an example of something sensitive? You said that there will always be things that are sensitive. Can you give me an example?
Sir Bob Kerslake: If the data were about the key performance of a service, we should not say that that is commercially sensitive. If, on the other hand, some of the data that people were seeking were about the commercial position of the company or whatever, that would be sensitive.
Q72 Mr Bacon: Plainly, asking, "How much money are you making out of this, Mr Contractor?" is sensitive in some ways, but that is information that the taxpayer ought to be able to know. We have had the major contractors group-the builders-in front of us. They have acknowledged openly what their margin generally is on big PFI projects that they have been involved in. This is all helpful information.
Matthew Hancock: And might drive down costs.
Mr Bacon: They might prefer that if it were not published.
Sir Bob Kerslake: I didn’t say, "How much money are you making out of a contract?" I said, the detailed financial position of that company, which would be a sensitive issue.
Dame Helen Ghosh: This is only suggesting a line of inquiry that the Committee may like to pursue: clearly, the issue of what is genuinely commercially confidential is something that has been pretty strongly tested through the FOI regime. My understanding is that we have reached a position-thinking of the FOIs that pass over my screen, not my desk-at which the definitions are quite narrow. The kind of issue is: "And how have you priced that particular bit of the contract, or that bit?" If a rival were to see it, it would enable the rival to undercut company A.
Matthew Hancock: That might sometimes help taxpayer value for money.
Q73 Mr Bacon: Dame Helen, that might be a good thing. For example, we know that G4S, the security firm, has just increased the amount that it is proposing to charge for management fees for its role in the Olympics from £7 million to £60 million-an increase of £53 million. Many of us struggle to understand how that type of increase can be justified. It is entirely possible that someone else could come in and do it for less, so it might be exactly that type of information that we do want to have in the public domain.
Dame Helen Ghosh: And under certain circumstances you do get that kind of information. I was simply replying to the point that there might be circumstances in relation to the transparency or FOI regime in which you would not want to release bits of information for the sake of economic growth and prosperity. The example of how that has been tested through the FOI regime-
Q74 Chair: I am just wondering whether you are offering to give us a bit more information that we can understand rather better.
Dame Helen Ghosh: I have to say, I believe that we, and Mr Stephens, have given you a great deal of information about the reason for the G4S contract increasing.
Tim Kelsey: Practically speaking, I am certainly observing a much greater willingness of private companies providing public services to engage in the transparency agenda. That is partly because everyone wants an outcomes-based conversation. For example, next month, ATOC and the bus companies will publish for the first time real-time train and bus-running data covering x-number of hundred thousand bus stops in England. That is an initiative that they have taken to liberate the data in order to-
Q75 Chair: Financial information?
Tim Kelsey: This is to do with the real-time running data. Finally, through an open market in applications, people will be able to find out whether they have time to stop for a cup of coffee on the way to the bus. This has been a really big issue, because those data are collected by private companies, but in the context of a public contract. I observe, in a number of markets, almost a rush to determine how transparency can be of benefit in those sorts of contexts.
Mr Bacon: That’s because it’s flavour of the month, though.
Q76 Stephen Barclay: Can I just pick up on some transparency from previous hearings? Sir Bob, I asked you a month or so ago about your personal objectives and whether there would be transparency on them and those of other permanent secretaries. You said that you would go away and consider that.
Sir Bob Kerslake: Again, we are considering that as part of the civil service reform plan, and I will come back to you on that.
Q77 Stephen Barclay: So we still do not know whether we can see your personal objectives.
Sir Bob Kerslake: Personally, I would not have a problem with it, but I need to confirm that across the piece.
Stephen Barclay: Thank you.
Dame Helen, in a February hearing, we learned that four fifths of police forces were found by the NAO not to have used new technology more efficiently. You said that you would give us a breakdown on those forces, but the note provided failed to do so. Why was that the case?
Dame Helen Ghosh: I am not aware that we decided not to give it to you; it may be an oversight. I think what Nick Gargan had in front of him was a table that showed what kinds of technology people were using. I am very happy to take that away and come back to you on that.
Stephen Barclay: That is a totally different point. Mr Gargan wrote to us quoting legal advice that contradicts that given by Speaker’s Counsel.
Dame Helen Ghosh: That is on a different issue, I think.
Stephen Barclay: Absolutely. There are a number of issues, which I am happy to pick up separately. I have six for which the answer provided in the note does not address what was asked in the hearing.
A wider point from a number of previous PACs is the ability to have transparency when like can be compared with like. There are issues with that within Departments, and foundation hospital trusts are an example. There are also comparisons between Departments, and certain things lend themselves more to that, such as back-office data, consultants and contractors. Mr Kelsey, do you see your role in transparency as putting the information out or for others to do that, and to what extent will there be resources within Departments driving that as part of the accounting officer’s duties? The evidence we had from Sir David Nicholson, for example, was that he did not see it as his role to be looking at the comparators between foundation hospital trusts and identifying outliers on similar things.
Tim Kelsey: Certainly a big part of the agenda is to ensure that standards are clear for different datasets so that we can create that standardised environment in which people can sensibly compare data. Clearly, the Cabinet Office is not, as it were, running individual delivery Departments, and it is in their power to determine what the specifics are. But a push to standards is very much a key part of the agenda.
Q78 Stephen Barclay: So your role is to ensure standards in common definitions, for example.
Tim Kelsey: Yes.
Q79 Stephen Barclay: That is absolutely logical, but perhaps I could give you two examples. We asked one Department about consultant spend coming down. It said that it was coming down, but that contractor costs were going up, and we never really got to the bottom of what the distinction was between the two. If we look at the current Government, they had a recruitment freeze on all staff except for essential front-line services, but each Department was free to determine what an essential front-line service is. If you want to, you can a drive coach and horses through the definition.
You have a standard definition, but very wide interpretation by Departments. Again, from a PAC point of view, should we be coming to you in terms of the application of standard data, or is it free for each permanent secretary to interpret those definitions in a way that could take them outside what might have been intended?
Tim Kelsey: Whenever possible, an effort is made to get consistent definitions, and to hold Departments to those definitions. The Treasury has done a lot of work to get absolute consistency in the information that is produced on spend. If you take the quarterly data summaries, a huge amount of work has been done to ensure that you are comparing like with like. That is a key part of where we must continue to push to get absolute consistency. You are quite right. How do you make your comparisons?
On your specific question about recruitment, I think it is a slightly different point, because what is a front-line service will vary between different Departments, which are hugely different in what they do, and how they do it. I would have been slightly wary of an over-rigid description of front-line services that got in the way of Departments running their business. To some extent, there will be situations when you leave it to individual Departments, but on a lot of the data stuff that we have just talked about, quite a lot of effort is being made to get that more consistent. I think there is a difference.
Dame Helen Ghosh: My experience-Bob may correct me on this-is that on the issue of what is a front-line service and, as Bob says, precisely what is the definition of a contractual consultant in particular circumstances, each Department has had to agree with either the Treasury or the Cabinet Office what that is. We are not allowed simply to make it up. Certainly in the 2010 spending review, a lot of work was done on what counts as a front-line service for each Department, and agreement was reached, from which we cannot move. As Bob says, it may have to be different from Department to Department, but it is an agreed definition that we work through.
Q80 Mr Richard Bacon (South Norfolk) (Con): What about the possibility of Departments finding different ways through the back door to circumvent the rules? I am thinking particularly of the Ministry of Defence, and the suitably named FATS contract-the framework agreement for technical support. When we asked the permanent secretary at the Ministry of Defence about this, and particularly how many of the people who were being given work under the FATS contract were ex-MOD employees, she did not know, and I had the distinct impression that she did not want to know.
Q81 Stephen Barclay: Because it came out of a different budget. It didn’t come out of the front-line staffing budget, so it was a way of using a different part of the budget to address the costs-a classic case of moving costs from one part of the budget line to another.
Q82 Mr Bacon: I know you guys are good at moving things from one column to another. We talked about that the other day in relation to the £364 million in the regional growth fund, which Mr Smith, who is not here today, referred to as Mr Donnelly’s side account.
Sir Bob Kerslake: I shan’t rise to the bait on that.
Q83 Mr Bacon: That aside, is there not an in-built tendency in Departments to find ways to work their way around these rules?
Sir Bob Kerslake: In my experience as somebody who used to work as a director of finance, there is no limit to human ingenuity on these things, but I do think genuinely that more effort has been made to get consistency than I have seen at any other time in my experience with the civil service.
Chair: I agree.
Amyas Morse: I want to follow up on something that you said and that perhaps was implied by Mr Kelsey as well. I want to look at what happens when you are dealing with private sector suppliers requiring access to information as part of the contractual terms. You quite rightly said there is an avenue there. Is it true that Departments, when they are putting contracts in place, as many of them do all the time, have very clear guidance that they should be seeking those contractual terms, or is it a local decision as to whether they are going to do that? For example, with those bus companies, was it really their initiative or did you push them?
Sir Bob Kerslake: My personal view is that there is a changing picture on this, Amyas. If you go back a few years, it would have been very much down to local discretion. We are now trying to get-
Amyas Morse: More standardisation?
Sir Bob Kerslake: Consistent standards on contracts. We have learned. I remember that in Sheffield we did a number of contracts where, truthfully, we didn’t specify the information we required as well as we should have done. I am sure that was true of central Government as well. But by having procurement expertise and setting that alongside the open data capacity we have, we can improve on that and get consistency across Departments.
Amyas Morse: So you would give them guidance or you would expect there to be guidance at least in the Departments?
Sir Bob Kerslake: I think personally that the guidance on procurement ought to pick that issue up if it doesn’t already. I haven’t got it in front of me, but I am pretty sure it does pick it up.
Chair: I hope that you will have regard to our views in that guidance, because we are attempting to ensure that there is value for money.
Amyas Morse: Can I just ask this about the bus companies? Was it their idea?
Tim Kelsey: There was a bit of push, but actually not as much as I thought. I would also say that the open data strategies-these chapters in the business plans that Departments are building-do have, as it were, a chapter heading that relates very much to this point. I think you will find, and be quite reassured by the fact, that Departments are waking up to the idea that transparency extends beyond just their borders.
Q84 Ian Swales: Can I clarify one point? We use the expression "following the public pound" a lot in this Committee. We are talking about private companies’ contracts and so on. We in the Committee have a problem with the bit in between, the arm’s length bodies, which are actually public bodies, so can you give us some comfort that those organisations will have the same standards applied or ideally better standards, because they ought to be more under our control? You have the Departments and you have the private sector, but there is this chunk in between. I am thinking of the BBC, Network Rail and so on. We get frustrated because there are lots of organisations where we don’t seem to be able to access-where the data is not available to the same extent.
Sir Bob Kerslake: I can’t talk for all of them. Tim may have a better idea about this, but I do know in relation to CLG that we have specified the same information on spend, for example, and we are expecting the same levels of compliance on-
Q85 Ian Swales: Transparency?
Sir Bob Kerslake: Transparency, yes.
Q86 Chair: You weren’t here for the LGA evidence, but obviously it will be in our interests post Audit Commission to be able to do comparisons between local authorities. By the way, you were praised in your absence-with your departmental hat on-for being the best Department in terms of publishing data. They publish financial data. What we will require is performance data that we can compare. At the moment, as I understand it, that is not compulsory. There is no requirement to do it; it is voluntary.
Sir Bob Kerslake: I am surprised to hear that. Certainly the data for the major arm’s length body is published-
Q87 Chair: Not for the local authorities.
Sir Bob Kerslake: I beg your pardon. No. What we are seeking to do on the local authorities, as you heard from Carolyn, is to look at a collaborative model across local-
Q88 Chair: I understand that, but if Nottingham City Council decides that it does not want to let us know some performance data on the quality of care homes or something else we may be looking at, we will mind that.
Sir Bob Kerslake: I do see your point. We have sought actively to reduce the demands we put, from central Government, on local authorities and to create the space for them to do this themselves. Let’s see where Local Government Inform gets to. The feedback so far is that they are getting pretty high levels of buy-in from local authorities, but the proof of the pudding will be in the eating. We will have to take stock then.
Q89 Austin Mitchell: The Prime Minister set out admirable principles, as I regard them, in speeches and his letters. That is in contrast to Tony Blair, who found the whole business of openness to be creating a rod for his own back. It certainly created a rod for our backs as MPs, from which I do not think we have recovered. Let me ask this amiably-this is for the Cabinet Office-because I am in favour of transparency and openness: as Richard pointed out, the culture of secrecy lingers still, and that culture of secrecy is amazing and often impenetrable, so why don’t we just make the Prime Minister’s principles mandatory and have the Cabinet Office as the enforcer seeing that Departments observe them?
Tim Kelsey: On the Prime Minister’s letters, those commitments are mandatory. Obviously, they are agreed with Departments. I agree that they represent a very big step forward in terms of making key datasets available, so they are mandatory. Departments have agreed them and they have to deliver them by a fixed date, and that is set and published. At the moment, even colleagues at the NAO acknowledge that we are meeting those commitments-23 out of 25 of the Prime Minister’s commitments have been delivered on time, as proposed.
Q90 Austin Mitchell: Have you had problems with non-compliance?
Tim Kelsey: To be honest, on occasion, but not serious ones. As some will know, I was a journalist originally, and I have been genuinely surprised by how seriously Departments have taken this agenda, even though it is frankly demanding. It is not, at this stage, a matter of cost, because most of the data are already there, but just needs to be pushed out, but there are cultural and other challenges. I have not experienced serious problems with compliance yet.
Q91 Austin Mitchell: Do not begin sentences with, "To be honest"-leave that to us as the politicians.
Let me turn the question to local government. There is now a code of recommended practice for local authorities. Some are not complying, so why don’t we make that mandatory?
Sir Bob Kerslake: I have two points. To finish off the last point, the NAO looked at compliance, and the compliance levels with Departments are very high-over 90% on what their requirements are.
On local government, we have gone for a code of practice that is voluntary at this stage. We have said that we will review it and will look at the option of making it mandatory. We have not concluded our view on that. Clearly, if we made it mandatory, it has impacts in terms of burdens. At this stage, if you take the spend data for example, we have only one authority that has chosen not to produce those spend data, so we have kept open the mandatory option.
Q92 Austin Mitchell: Are you satisfied that the voluntary system, as it were, will work with an ingrained culture of secrecy?
Sir Bob Kerslake: The answer is so far, so good. What is quite interesting is where some local authorities-not very many-have not produced them regularly, what you find is that the local media have then held them to account for not producing the data, so there are other ways in which they get held to account, if you like. As I say, so far, actually the compliance levels with local government on a voluntary basis are very high.
Dame Helen Ghosh: I was just going to talk about the Home Office’s non-compliance. We are compliant with every single requirement, I think. The one that we are in debate on, and it is interesting that it was picked up in some of the hearings with previous witnesses, is the value of the additional effort required to give-I am not quite sure how they describe it-"plain English" explanations of our spend. We put out extremely lengthy tables of all our spend over £25k, and they explain quite a lot. A specialist would almost certainly read it and work out what we were spending all our money on.
The question is, and it is at the heart of some of the recommendations that you make in your report, what would be the additional value of me employing people to say, "This payment from the Criminal Records Bureau to such-and-such police force was to do so and so"? At the heart of that one, there is quite an interesting value-for-money issue about how much more we should enrich that kind of bulk data, as opposed to going for the knowledge-rich data that Carolyn was talking about. We would say crime maps are increasingly knowledge-rich. In terms of the bulk data for every transaction over £25k, how much richer can we usefully make that and is it worth our investing money in it?
Q93 Chair: One of the key questions is: are you doing a cost-benefit analysis on all this stuff?
Tim Kelsey: Right. That was the central constructive point that came out of the Report for me-that it is vital and certainly a very big part of the agenda moving forward that we do that. You will have noticed that we have not quite launched but are beginning to set up something called the Open Data Institute, which Tim Berners-Lee and Nigel Shadbolt will be running. That will launch formally in October, and a big part of its responsibility will be to develop those economic models-benefit, business case-by dataset. That is an important piece of work that has not been done. People have speculated as to the economic value of this stuff, and to be honest there are some interesting angles but there’s nothing really solid that’s out there.
Q94 Chair: What is the time frame for that bit of work or for developing the model?
Tim Kelsey: Yes, it will be a model, which I think will be ongoing. I imagine the ODI will work with Departments to look at particular datasets and help them through the business case for their release or development. That is part of its ongoing-
Q95 Chair: You mean the moment you consider whether to release a new dataset-is it October when you’re setting it up?
Tim Kelsey: The ODI launch is in October.
Q96 Chair: So from the autumn onwards we can expect some sort of cost-benefit analysis before you release new datasets. Am I hearing that right?
Tim Kelsey: Yes, you are. At the moment, it is good practice in Departments to do that impact assessment in any case, and many do it extremely competently, but what we want to do is to enhance the capabilities in the public sector to do that in a way that would satisfy you in asking, along with the NAO, this very important question: what really is the benefit of all this when you stand back from the detail?
Q97 Jackie Doyle-Price: I want to come back to the issue of local authorities. We have had much discussion, which has illuminated the fact that too much information is actually the enemy of transparency and good decision making. With that in mind, how confident are you, Sir Bob, that the LGA’s Inform tool will deliver useful information to the public, bearing it in mind that it is really being led by the local authorities themselves?
Sir Bob Kerslake: From what I have seen so far of the work they have done to develop it-at the moment they are doing it internally and then they will publish it-it looks good. It looks to me as though they are collecting the right sort of data. My personal view is that they will do a first run of it and then it will have to improve; it won’t be right the first time. I think it is about 800 measurements they are looking at across the piece. They are much better placed, I have to say, to get to a view about what is useful data to publish than I am, because they work on this on a daily basis. The evidence so far looks encouraging. We will see what it looks like when it comes out. My guess is it will be reformed and reviewed over time.
You said the bulk data is the enemy of knowledge. Personally I think you need both. I think personally there is no contradiction between putting data out there and allowing people who are interested to analyse that data themselves and, alongside that, having processed data whereby people can make comparisons. Both are entirely valid. In some ways, having the first is a bit of a safeguard against me choosing to release what I think people might like, as opposed to what they really want.
Q98 Jackie Doyle-Price: I think the key to that is that for a lot of this data, the public are not the audience. It has to be assessed by a trust of proxies, of which you, the Department, are one. I suspect elected members in authorities are as well. There will be a comparison element in that. With that in mind, have you given any thought to how you will analyse that data, perhaps to review the performance of individual local authorities?
Sir Bob Kerslake: What we have tried to do is to establish the minimum data requirements that we need as the Department. Indeed, we have done that across Government as well. There is a clear set of data requirements on spend and other things-the things we say we need at CLG-and we have deliberately tried to cut that back. We can clearly draw on Local Government Inform like anybody else can, but I think we will supplement it with more qualitative data about the position of individual local authorities and where they are. But our focus is on creating systems by which the public can make those comparisons. It is not for us to make the judgment on their behalf.
Q99 Jackie Doyle-Price: My concern is this. Transparency is a tool to aid accountability. We used to have committees in local government in which difficult questions would be asked, even though papers were sanitised for public consumption. Now-you alluded to this earlier-what looks like being open actually isn’t, because a conversation is taking place behind closed doors. Now, because you have cabinets and leaders, a lot of that challenge is happening behind closed doors, which makes this openness and transparency much more important. Dragging elected members into really understanding the data and comparisons is crucial and where the Department needs to play a key role.
Sir Bob Kerslake: I am very comfortable with that. In an authority that is well run, the scrutiny committees can be quite powerful-I have certainly experienced that myself-but the transparency agenda enhances it. There is no doubt in my mind that the fact that people know that transactions over £500 are published influences behaviour in terms of what they do and don’t spend money on; I am pretty convinced about that. There is evidence that the fact that salaries are published has brought down some of the higher salaries. That exposure does have a powerful impact. The comparative data is the next bit in the story. We need to put a lot of time and effort into seeing how that works.
Q100 Nick Sloan: A point of clarification, if I may. Will the ODI’s research and evaluative remit include the accountability and the service improvement stuff, or is it focused on economic affairs?
Tim Kelsey: To be clear, the ODI’s focus is to incubate business in order that it can deliver enterprise value from open data. Its purpose, therefore, will be to identify datasets that will have that effect and make the case for their release. Of course, productivity comes into that, but it is principally about economic and enterprise value from the public data.
Q101 Nick Sloan: Are there any systematic ways proposed to look at improvements in accountability and how you might assess that or, indeed, pick up open data’s contribution to service improvement and choice?
Tim Kelsey: I think they’re wrapped up together a bit. If you are stripping out accountability on its own, that’s a thing on which the Cabinet Office, moving forward, will continue to develop a programme holding Departments to account for delivery. That is all about the open data strategies and how those are produced. Our role is to challenge Departments on demonstrating how they are themselves developing models of impact, related to accountability and choice. The ODI’s principal focus is on enterprise value that you can measure in pounds and pence.
Q102 Fiona Mactaggart: Dame Helen, I was going to ask a question about a parliamentary question that I recently asked about how long the longest time was between the promulgation of a decision of an appeal tribunal and the issue of an entry clearance, but I realised that I was going to ask you the wrong thing, because I realised that the reason why I had got an answer from the Minister saying, "I can’t give you that information" was that your Department is not the Department that holds the date of promulgation of those decisions. I should have asked you what the longest time was between the date on which the Department started working on one of those decisions and the issue of the entry clearance.
Dame Helen Ghosh: Yes.
Q103 Fiona Mactaggart: The reason why I raise this is that I am concerned that we do not have clear enough-I can see Mr Kelsey knows what I am talking about-rules about certain kinds of data. There is an interesting, real-time bit of data about people’s lives that we know we need about buses, but it is obvious when a bus starts and when a bus stops. It is slightly more complicated when something has got buried-I know that the Home Office does not deliberately bury these things, but they get buried; you know and I know they get buried-for people to know whether my piece of burial is an extraordinary one or whether it is usual. The key thing about this-it is highlighted in the Report-is that we need clearer rules about what is data, what is information. I am not sure that I have heard, either through reading the Report or from what you have said, that there is a strategy for getting those clear rules. I don’t know what it is and I would like someone to tell me what they think it is.
Dame Helen Ghosh: Can I just say, as a debating point, I’m not sure that you ever could get rules quite as you describe? In preparation for this hearing, I was looking through my list of datasets that we publish. There are 200 or perhaps 140; it depends on how you define a dataset. It struck me that they are a set of data that we collect because they happen to be, for a variety of reasons, the performance targets that we set, for example, for the UKBA in the instance you describe, the UK Border Force, the crime and policing group or police forces, so we measure them. Therefore, I don’t think Tim could sit there giving me a rule for needing to have a dataset that answers the question you describe. What we need-Mr Vaz is working hard on this in another part of this building-is a sensible set of performance measures that genuinely enable us to manage our businesses better, and to answer that kind of question from a customer’s point of view. That is how we would-at the end, there will probably be 300 or 400 datasets-be collecting the information you describe.
We collect information, and it is often as much in response to a political crisis five years ago as it is to something that helps us to manage our business now. It is something that I am very aware of, and the Home Affairs Committee is very aware of. Rob Whiteman and I are looking at the roots of why we have that data, and not another set of data, and why that is our performance target and not something else. I do not think it is rules that would enable me to answer your question, but others may not agree.
Q104 Fiona Mactaggart: It might not be, but it seems to me that paragraph 8 of the report is the critical point.
Tim Kelsey: Hopefully, I can provide a bit of reassurance, because I understand exactly what you are saying, and there are two bits. First, we need a common glossary of the words-what is open data, and what is public data-and there is a degree of confusion about that out there. We will publish that glossary with the White Paper next month as a starter for 10 on crowd sourcing and so on, and what we mean by these things, so that we can have a reasoned and proper debate about not so much the rules, but the definitions. We must be clear about what we mean when we talk about open data, public data and so on.
That also relates to the critical importance of delivering an inventory of public sector data, on which we will comment in the White Paper. That has been missing, and people rightly complain that it is very hard to find out where data are, what data are available and, if they want to build a business, what the opportunities are. We need collectively to engage on that initiative. Perhaps those two things will help gradually, or hopefully quite quickly.
Dame Helen Ghosh: None the less, we have to collect the data about the things you think are interesting. You could look at my inventory and say, "That’s very odd; why don’t you collect x?" That is a way that transparency can get us to collect the right stuff.
Q105 Fiona Mactaggart: And that is the process. I am very interested in what Mr Kelsey said, and it is an important first step, but then there is another step within Departments, which is, what is the process for giving attention to something? If you run a train, you know that the process is: when does the train go? It is as simple as that. The information is there and collected in one place. If you run a complex system, you might know when someone applies for entry clearance and when they got it. Presumably you will have those data, but you might not. It seems to me that we need to work out what the slices of cake are, and the users of the service need to be engaged in what those slices are. Otherwise, there will be a problem of too much data that no one can understand.
Dame Helen Ghosh: Indeed, or the reverse of that, which we have all had experience of. If we are busy collecting data that is completely irrelevant to our business, and no one actually wants to know about, that wastes a hell of a lot of resources. That is the other side of the problem.
Q106 Chair: The final bit of that little jigsaw is that it is almost impossible for anyone to access it. The stats in the NAO Report of people who have visited www.data.gov and then go beyond the home page is really telling. I have tried to look at employment figures and so on, and I may not be the most digitally literate-we will come to digital literacy-but it is pretty impossible to access. I do not know what to do. It is crazy. Not only might we not be getting the right data, but even if we had the right data, the likes of well-versed, educated, knowing-what-I’m-looking-for me find it impossible to get in there.
Tim Kelsey: No. data.gov is about to be relaunched as an extractive application, which will make it much easier to download data automatically in a way that currently is not the case. At the moment data.gov is a library of links. I agree. I think everyone is agreed.
Q107 Chair: And have you money to do all this, to give us a website that we can navigate easily?
Tim Kelsey: The great joy of this, luckily, is that we are doing it at a time when the costs of this sort of technology, the availability of open-source applications, means that the amount of money involved in relaunching data.gov, the total cost, has been around £140,000 for that redesign and the relaunch next month-it has been in beta test with the developer community and others, and they are very pleased with it. So, pretty good value, I think. I must just make the point that even now, although it is relatively not as good as it could be, data.gov remains the largest data resource of its type anywhere around the world. So we should remember that those 8,000 datasets that do work on data.gov constitute the largest community of open data anywhere. But certainly constant improvement, and as I say data.gov is a key focus of that.
Q108 Mr Bacon: Dame Helen, I would like to return to the question of crime maps, because they have been held up as one of the examples of something that is working. When the crime map thing was first launched, I put the postcode in but, unlike Mr Hancock, I confess that I have not gone back regularly. One of the reasons that I did not go back regularly is that when I put my postcode in, the site was reporting crime in a village many, many miles away, which bore no relation to anything that might be of interest to me.
I have just gone back on to the site now, with my own postcode, and it has a pretty little map of my area, including Harleston, the local market town three miles from us. When you look at the contact details, however, you get the message that no contact details are available for this team, and then there is the address of a police station in Fakenham, which is 43 miles away. Not only are there lots of police stations nearer than that, but we have a very nice little police station in Harleston, the market town that I was speaking of; it is staffed by volunteers and is usually open. When you go into "Meet the team", you get the names of some police officers, some of whom I have met, but there is no indication of how to contact them. And when you go to "Get involved", it says that no event details are available for this neighbourhood and to please contact Norfolk constabulary to request that they add this information.
If the crime maps are to be enriched, what is the process by which you envisage that this will happen? Is it, as it were, a top-down one where people in the Home Office are scanning them and pointing out the inadequacies, or are you hoping for a bottom-up revolution as the citizens of Norfolk rise up to demand higher quality information? Because you might be waiting quite a long time. The citizens of Norfolk are quite busy and, when they look on this site, it tells them to contact Fakenham police station, except there is no phone number for Fakenham police station and they know that Fakenham police station is miles and miles away. How is this enriching that we are talking about and which you specifically spoke about going to happen?
Dame Helen Ghosh: This, of course, comes back to a theme that we have discussed on a number of occasions: the extent to which the Home Office is any longer in a position to instruct police forces around the country on particular issues. Curiously, I did exactly the same thing in preparation for this hearing: I looked at my local site, OX1 4LT, and at the Norfolk site actually, as well as Newmarket, because I was under the mistaken belief that it was in Cambridgeshire although in fact it is in Suffolk. So when I looked at my TVP or Thames Valley Police site, I observed that I had pictures of my beat officers, I had tweet opportunities, I had how I could contact people-
Q109 Mr Bacon: Isn’t it possible that they had checked your postcode and made sure that it was all pretty pukka for you?
Dame Helen Ghosh: I think it is just that Sara Thornton is a very advanced chief constable and she had made sure. I was indeed very disappointed when I came on to Norfolk, I think, and you just got blank shadow faces of the beat constables in wherever it was I had picked-Norwich probably. I just got shadows and none of that same rich material.
I think that I have to give the answer that I have given previously. What we have done in this instance is to set certain national standards-as we do, as you know-across a whole range of policing, and then we must use the pressure of your constituents, and the comparisons that constituents will see from what kind of information they get on other police forces. I do not think that we are in the business of requiring that kind of information on crime map sites, apart from the Her Majesty’s inspectorate of constabulary material, for example. I had also looked at the comparisons in crime between Norfolk and Suffolk to see how you were doing.
Q110 Mr Bacon: We are the safest county in the country.
Dame Helen Ghosh: Yes; actually you were both quite safe, I think.
The point I would make is that the move into social media, the use of YouTube-that kind of sense of local communication between local neighbourhood police forces and the community-is something that has come on by leaps and bounds in the past, probably, only two years. I think that is reflected in the crime maps for most places, and I can only urge your constituents to make that point to the chief constable of Norfolk.
Q111 Chair: I have two final questions. One is for you, Sir Bob. All these are data about public services, and the people who are most dependent on public services are poorer people; that is the truth. Yet, we know-a little stat I managed to get off the thing-that currently 8.7 million adults have no access to the internet. When you look at who they are, of course they are the old, the disabled, the poor. All those sorts of indicators are there. In a way, you have done a sort of blitz on putting data out there, but you are not touching the people for whom that would be most beneficial in terms of choice, accountability for public services and all those things that really matter.
Sir Bob Kerslake: I think there is a lot to do on extending access to broadband. I know Martha Lane Fox has done a lot of work on this, in trying to raise the level of access to broadband.
Q112 Chair: But 8.7 million is the ONS 2011 figure.
Sir Bob Kerslake: There is clearly still a long way to go, but it is moving rapidly up. Rather than say that, because we cannot get everybody access to the internet, we will not do the transparency; the focus should be on what steps we can take to drive up the level of access to the internet.
Q113 Chair: Or should you do the transparency in other ways?
Sir Bob Kerslake: Well, I think there are options to do things in other ways, but the whole direction of travel of the Government is to try, if you like, to move to a model of digital by default and then to overcome the barriers of people’s access to the internet, rather than to run two or three systems of access to information. Yes, we still have a journey to go, but I think we should focus in on increasing access rather run a whole range of systems because we cannot make progress on that. Quite a lot is happening in Government to drive that agenda.
Q114 Chair: I completely support increasing access-increase access, of course-but 8.7 million adults in 2011 is a heck of a lot.
Sir Bob Kerslake: It is quite a lot.
Q115 Chair: If you look at the tables I have-I do not know; what shall I do?-it is 76% of over-75s and 42% of over-65s. If you look at income brackets, it is 31% of those on an income of less than £10,000, so it is really concentrated on the poorest. I am all for Martha Lane Fox doing all she can and all of us doing all we can to have increased access to the internet, but-
Sir Bob Kerslake: I have two last points. I understand your point, and we can and should look at other ways of enabling access, but quite a lot of this information, as we have demonstrated this afternoon, is available on people’s phones; that is one point. Secondly, you can and do have places where people can access the internet-in libraries and other places. I am not saying that that is a complete answer, but I think we should resist the temptation to keep open every channel indefinitely in the future, simply because it will increasingly become an expensive and unaffordable model.
Q116 Mr Bacon: On that point, would not one of the simplest things that you could do, in relation to the upcoming 4G licence, be to increase the coverage that is required of providers? I think that, at the moment, it is at 98%. You could push that number significantly higher and, as a result, have mobile phone providers making permanently smaller profits.
Many of my constituents, particularly the poorer ones, have only mobile phones; they do not have any other form of telephone. In a very rural area, the difference between 98% coverage and 100% coverage is the difference, for many of my constituents-including me, by the way, in my own house-between having coverage, which I have in my constituency office but not in my home, and not having coverage.
That would be a very simple piece of clever regulation. Admittedly, it would damage in the long term the overall profitability of mobile phone providers, but probably-no, certainly-not in a way that would stop their bidding, because this is a huge revenue stream that they want access to. However, it would radically improve things for everyone, because everyone has access to mobile phones.
Last summer, I was in Tanzania working for VSO. It is one of the poorest countries in the world, but it has 40 million people and 20 million mobile phones. It is now universal around the world. That simple measure-increasing the level of coverage that is required-could have radical effects.
Sir Bob Kerslake: I am not closely involved in or close to the detail of the 4G negotiations, so they would not thank me for-
Q117 Mr Bacon: Why don’t you just wade in anyway and raise the bar?
Chair: You are the head of the Civil Service.
Sir Bob Kerslake: I hear what you say on that point.
Q118 Ian Swales: You raise a very good point, Margaret. I think we have to be realistic. In the days when I started work, computers were just coming in, and when you got data, it was so exciting. These days, it has to be a pull rather than a push agenda. In other words, you have to make data available for people to pull it when they want to look at it. That raises a question that I had not thought about before about whether local authorities should put that information in their libraries on paper, for example.
Sir Bob Kerslake: If you look at local authorities, quite a lot have done things like that to make data more accessible and more available, principally through digital routes, but including libraries and other things.
Q119 Ian Swales: Nobody wants that amount of stuff pushed through their letterbox, but access, maybe via libraries, is an interesting extra thought.
Dame Helen Ghosh: One of the things that struck me in preparation for this, and in talking to Tim and others, is that-many constituents of people around this table will benefit from this-transparency has driven public service improvement because professionals have had data. It is easy to concentrate on whether every individual citizen has had data and how they get access to it, but if you look at the service improvement, the most powerful message that came out for me was that it was the consultants looking at the medical data or the social services professionals looking at the data. It is them looking at risks to children in different local authorities. That has also been an important part of the story.
Tim Kelsey: One example of a really good and sustained transparency platform has been NHS Choices, which is often not discussed in this context, where, for the first time, comparative data on a whole variety of hospital services has been published. The dynamic there is that a large number of people-14 million unique users a month-use that directly online, but crucially, that has engaged with the whole GP community to recognise that you can do something as simple as saying, "We’ll enable GPs to print off this stuff." They can have that conversation in the GP surgery, but without having to spend a fortune on toner by making it printable in black and white. Simple things such as that have done a great deal to improve the inclusivity and engagement with the data and the information outside of online engagement.
Q120 Chair: I think we are going to have a vote in a minute. I don’t usually do this, but I sat in on the Chief Secretary’s statement today on people being employed by the Civil Service on pay-as-you-earn. I just wanted to check, Mr Kelsey-I am sorry if it embarrasses you-that you are paying PAYE.
Tim Kelsey: I certainly am.
Q121 Chair: You are on the books, so all this other stuff that you do, you do separately. My briefing tells me that you are still being paid by McKinsey and that you are linked with a senior NHS Commissioning Board role.
Tim Kelsey: No. To be absolutely clear, I am a civil servant. I left McKinsey because I actually believed in this so much that I took the pay cut and hopefully did the right thing. I do not have any other employment. I am just a regular old civil servant.
Q122 Mr Bacon: So you do not do anything for McKinsey any more.
Tim Kelsey: No.
Q123 Mr Bacon: Sir Bob, was Mr Watmore on PAYE?
Sir Bob Kerslake: He was a civil servant employed in the normal way. If you look at the information released today, we have indicated where people weren’t in that position. There are pretty small numbers, but the aim is to move to a point where any permanent civil servant is on PAYE.
Q124 Chair: Good. I will end by saying that I am a great admirer of what you achieved on Dr Foster’s agenda, so it is good to have you on board.
Sir Bob Kerslake: He was on a standard employment contract.
Chair: Thank you very much.