Finance Committee,

Meeting date: Tuesday, May 17, 2005

Official Report 277KB pdf

---

Efficient Government

I welcome members to the 14th Finance Committee meeting of 2005. I also welcome the press and public. As always, I remind people that their pagers and mobile phones should be switched off. We have apologies from Wendy Alexander, who I think will be a little late, and we are expecting Frank McAveety. Apart from them, members are all here.

The first item on our agenda is to take further evidence on civil service reform as part of our scrutiny of efficient government. Members will recall that we agreed to submit a short report to the Public Administration Select Committee's inquiry into civil service effectiveness. We have previously taken evidence from the permanent secretary, John Elvidge, and from the civil service trade unions. Our intention is to finalise our report by the end of June and send it off.

We have two panels of witnesses today. I am pleased to welcome our first panel, which consists of Richard Parry, reader in social policy in the school of social and political studies at the University of Edinburgh, and Professor Robert Pyper, professor of government and public management at Glasgow Caledonian University. Both witnesses have produced extensive work on public management and governance issues in the post-devolution civil service. I particularly thank Richard Parry for agreeing to come along today, as I understand that he has taken a break from a holiday down south to do so.

We have written evidence from both witnesses. I propose to invite them to make a brief opening statement, after which we will proceed to questions, if that is agreeable. Who wants to go first?

I will start, if that is all right. I reiterate the point that I made at the beginning of my written evidence about the importance of the work that the Finance Committee is carrying out. The Scottish Parliament lacks an equivalent to the House of Commons Public Administration Select Committee, but this committee is the next best thing—although it also has to undertake a wide range of scrutiny work, it carries out its other functions extremely well.

I note in that context that the committee is pursuing a twin-track inquiry. On the one hand it is looking in great detail at efficiency issues, but on the other hand it is considering much broader strategic issues to do with civil service reform. As far as efficiency issues are concerned, some of us have been studying British Government and the civil service long enough to think that there is nothing new under the sun. Some of us are old enough to remember the Rayner scrutiny exercises in Whitehall in the early 1980s and we see the same themes coming round again in slightly different forms—the core themes seem very familiar to many of us.

My evidence focuses not so much on efficiency issues as on the broader issues relating to civil service reform and, in particular, the future of the unified civil service and the more interesting possibility of a Scottish public service. I am happy to deal with any questions about my evidence on those matters.

We are looking at two rather different exercises. First, we have the Treasury exercise on efficient government, which has been the big headline grabber in recent weeks. Secondly, we have the whole civil service reform agenda, which is run by the Cabinet Office. However, the two exercises share a theme—the degree of constraint on the Scottish Executive.

We are considering areas of public expenditure and civil service management in which there is some ambiguity about whether there is any role for the Treasury and the Cabinet Office in the work of the Scottish Executive or whether the Executive is free to do its own thing. In each case, a certain amount of home-grown work has been done on the part of the Executive, but we have also seen some reactive work. That is particularly evident in the efficient government exercise, in which the Executive's work in recent months seems to have been driven and heavily influenced by the earlier work by the Treasury. That raises the broad question whether what is being done in Scotland is equal to or as ambitious as—that is the phrase that is used by the Treasury—what is being done elsewhere. There is some ambiguity about whether the Executive is free to act on its own and carry out its own wishes and programmes or whether it is constrained by Whitehall.

I thank you both very much.

My first question relates to efficiency savings and the management approach that is being adopted in seeking out and driving forward efficiencies. The question is the mirror image, if you like, of the one that I asked Tom McCabe at last week's meeting. If one were thinking about how to make a system more efficient, one would first try to simplify the mechanism and strip out programmes and aspects of activity that one thought were less productive or delivered less effectively. Perhaps the second point would be to look at the back-office type savings, which seem to be the predominant issue. Is the Executive missing a trick by not considering more critically what it is trying to do across the board—trying to "do less, better", to use a phrase that Jack McConnell coined for the early part of his period in office—and using that as the mechanism to drive savings, rather than relying on what we might call more technical efficiency savings, which seem to be the main thrust of the package that has been produced?

Richard Parry All that depends on the mix of functions that are carried out by any organisation. The interesting thing about the Scottish Executive is that it does not do very much service delivery; services are delivered through its agencies, such as the Scottish Prison Service, the Scottish Public Pensions Agency and the Student Awards Agency for Scotland, for example. I would have thought that an efficiency strategy would begin in those areas of an organisation's work that were broadly susceptible to routinisation to achieve improvements in productivity.

The Scottish Executive should perhaps have undertaken the exercise by saying, "There are certain things that we run on our own and we will do the work in those areas." In a large body such as the SPS, for example, which employs more than 4,000 people, early work has been done on efficiency. The Executive should ask, "Are we at the end of this? Could we do more? Is there work that should have been done earlier?" All those questions should be asked before the Executive moves on to areas that are either more indirectly under its control—because the services are implemented by the health service, local government and other agencies—or are not susceptible to improvements in productivity because of the nature of the work.

The difficulty with the current exercise, as seen by outside observers, is that it exports lots of the presumed efficiency improvements to bodies that are not run by the Executive. Such bodies are told, for example, "We think that you should be making these improvements and we are altering your financial arrangements." However, there is a degree of detachment in that approach that is not particularly conducive to a good strategy.

Are you saying that the Executive's approach can be regarded as one of financial constriction rather than re-engineering? Perhaps, as you say and as the Minister for Finance and Public Service Reform said at our meeting last week, the Executive has less scope than the Treasury down south, where departments are responsible for major delivery services such as customs and excise, pensions and social security, which can be re-engineered. In a sense, the approach that has been adopted in Scotland has been to set budgetary targets for different aspects of the Executive's functions. Sometimes it is hard to understand why budgetary target A has been set in one department, but a different approach has been taken in another department.

The problem is that the approach that you describe is happening anyway, as the committee has heard from other witnesses. Indeed, that work ought to have been happening anyway—if there had been no great interest in efficiency during the years of devolution since 1999, things would have gone badly wrong. We must ask what is in the current exercise that was not being done before. What work that was being done in the area has the Executive decided to use for the purposes of the exercise? Is there other such work, which has not been brought into the exercise? It is rather hard to work out what it all means.

The other problem is that we are not sure what such an exercise, on the same mix of functions, would involve in England, because, as you said, such a large part of the efficiency savings in England come from departments such as the Department for Work and Pensions, HM Revenue and Customs and the Ministry of Defence, none of which are part of the operation of the Scottish Executive. Therefore, it is extremely hard to say whether the Scottish Executive is doing as well as departments in England or as well as it ought to be doing.

There is an element of trying to keep the Treasury happy in all of this. It is obviously of interest to the Treasury that the Executive is doing work on efficient government; if the Executive said, "We aren't interested in that agenda," that would not fit in with the good on-going relations that the Executive and the Treasury need to have. It is almost possible to say that the current exercise is a minimal one, to leave everyone reasonably happy that at least the Executive is trying, so that ultimately—as is the history of all such exercises in Whitehall—people will conclude that it tried and that it did reasonably well. However, we will not be able to tell the difference between what happened as a result of the exercise and what would have happened anyway.

It strikes me that it could be argued that the Executive has not tried to consider what it is doing fundamentally, across the board. The Executive has tried to ascertain how it can improve procurement and thrash out savings within the existing framework of its activity, rather than question that framework. In a sense, the Executive is engaged in a particular type of efficiency exercise, which is a bottom-up exercise rather than a top-down reappraisal.

I understand what you mean. That is the difference between exercises that are done in the world of business and the private sector and exercises that are done in government. Often, Governments cannot readily choose what they do; obligations are laid on them by statute. It would be convenient for Governments to be able to say, for example, "We are not interested in running prisons," because it is very hard to run a prison service, but Governments do not have the ambitious option of saying, "We'd rather not do that at all." However, Governments can say that they intend to fulfil their responsibilities in a way that is radically different from the way in which they have done so in the past, but I do not think that, in the timescale in which the current exercise is being done, there is room for the fundamental bottom-up work that I agree would be helpful.

I do not want to put words in your mouth, but from what you said and from the memo from Sir Humphrey Appleby that you quote in your written submission, it strikes me that you might be suggesting that there is more than a bit of window dressing in what Government at all levels is putting forward.

All Governments and particularly civil servants need to respond to exercises. The efficient government agenda was an exercise that was launched by the Treasury as part of the 2004 spending review, because the Treasury was concerned that the extra public expenditure that had been thrown into the system since 1999, which of course had borne fruit for the Scottish Executive via the Barnett formula, was not being used as well as it should be. Because the Treasury was concerned by the increase in the number of civil servants and other such matters, those matters were to be examined in an exercise to which all parts of the United Kingdom Government needed to respond. The devolved Administrations also needed to respond to the exercise, but they had to do so at arm's length; they were not in on all the processes of the exercise. To an extent, the devolved Administrations were taken by surprise by what happened and did the best that they could. Whether they could or should have done more in the time that they had is a matter that the committee is considering—I suppose that it is ultimately a political issue.

The convener touched on this issue in his first question. We hear rhetoric north and south of the border that the purpose of the exercise is to release cash savings, which would be ploughed into front-line services. However, at the committee's meeting last week, the Minister for Finance and Public Service Reform felt constrained to defend civil servants and others who work in back-office services, by saying that those people do good and essential work. Is the distinction between front line and back office definable? Do we all know what we are talking about, or do many people work in the grey area that we could describe as either the front line or the back office, depending on our whim? If the distinction is definable, is the definition useful? In the context of what you said about productivity, it strikes me that the productivity of the people in the front line, who deliver a service directly to the public, is often more easily measurable and therefore easier to change than is the productivity of the people who work in advisory spheres elsewhere.

It is not easy to make the distinction between front office and back office and, in any case, the distinction is not particularly useful. I think that the distinction is at the heart of Peter Gershon's report because he began by considering front-line delivery in fields such as health, which is of great interest to the UK Government, education and the police service. In those areas, Peter Gershon perceived an implicit front-office/back-office distinction between people who were delivering services to help the public and people who were not doing so. Also, public service professionals could easily become diverted from their front-office work to deal with back-office issues. In those areas, there might exist the distinction between front office and back office that was made in Gershon's initial work, but the Gershon review ultimately became a big exercise for all parts of government in a way that is not particularly well argued for in his report.

It is now agreed by everyone, including Sir Andrew Turnbull, who is the head of the home civil service, that it was a big mistake to draw the distinction between front office and back office. It is all too easy for people to argue that transferring workers from back-office tasks to the front office will obviously make the whole organisation work better and lead to improved service delivery. However, even if we leave aside the issues of who those individuals are and whether it is possible to transfer them, the argument rather neglects the way in which organisations work, given that organisations are, ultimately, indivisible. Although one option might be to use outsourcing, whereby back-office work that was previously done in house is contracted out to an outside provider—in principle, that is possible if that is what is desired—it is not possible to divide all civil servants and public employees into those who serve the public on the front line and those who do not. That distinction just cannot be drawn.

The question touches on a theme that has been very much part of the history of civil service reform. Over the past couple of decades, attempts at reforming the civil service have focused on the management of policy and on the implementation and delivery end of the system. That was one of the major criticisms to emerge towards the end of the next steps initiative. It was argued that the next steps principles had focused too much on the front office or the delivery and implementation end of things without being brought home to the back offices in the Whitehall core or parent departments. In the various civil service reform agendas that have emerged, a recurring theme has been to focus on the sharp end of things. The current jargon speaks of back office and front office, but different terms for the same sorts of things were used a few years back. As Mr Morgan rightly says, such agendas focus on people and services that can be identified for analysis rather than on those that are more difficult to analyse. That is an important point.

Richard Parry talked about how savings outcomes might reflect the mix of services, which varies from place to place. Of course, one striking difference is that, in England and Wales, targets of 2.5 per cent savings per year have been set for individual departments for which ministers are nominally accountable, whereas no minimal targets have been set for any department in Scotland.

Secondly, as Richard Parry said, the Gershon process in England kicked off in March 2003 and the report was published some 15 months later. Will he clarify whether, in his judgment, a parallel process was under way here during that time?

Thirdly, given the extent to which the Gershon process relied on external support, will the witnesses say whether, in light of their experience of previous efforts to secure efficiencies that are monitored in a transparent way, they believe that there is a place for targets for which people are accountable? Might advance preparation and external support help in achieving those? Clearly, in those three dimensions, we are heading in different directions, so it would be helpful to have the thoughts of both witnesses on whether the factors that I have mentioned play a role.

On the first of those three issues, it is absolutely true that all departments in England have been required to deliver a 2.5 per cent minimum saving as well as individually defined jobs savings. That seems more rigorous than what has been done in the Executive. On the other hand, those departments are much bigger with a broader mix of tasks, so the targets would seem more reasonable.

On the question whether Scotland's efficiency exercise is as ambitious as England's, I think that that is what needs to be said for the purposes of rhetoric but, given the details, I do not necessarily think that it is the case. However, I think that we should be relaxed about the extra numbers of civil servants in Scotland since devolution, because the Scottish Executive has needed more civil servants in many areas since then. That does not seem to me to be a big deal.

On the question whether the Scottish Executive was on the inside during the Gershon exercise, I am uncertain of the answer. Like all academics, journalists and other outside observers, I felt that, when the spending review was published last July, there was considerable concern, if not incredulity, in Scotland and in Wales, which was interesting. If a single exercise had been carried out, one would not have found Scotland and Wales reacting in that way, but neither Scotland nor Wales knew what was happening. I suspect that, because the Treasury was reluctant to release things in advance, people first heard about the proposals when they read the final paper. I think that the Scottish Executive was taken by surprise. Therefore, I think that the exercise in Scotland is essentially reactive.

In a way, the Executive has done quite well by not going for specific job cuts, which are very hard to handle. The Executive's initiative is about budgetary constraints, where it is hard to know what would have been happening anyway. As we know, the Scottish Executive has had year-on-year underspends in most years.

I absolutely agree that outside expertise is required. We should not simply rely on the personal views of Gershon and of those whom he consulted who had expertise in such work outside Government. An awful lot of credence is being given to Peter Gershon's particular world view, but his is not the only world view.

The committee has relentlessly pursued the issues of transparency and clarity. We live in hope that someone will turn up and say, "Here is the like-for-like comparison that you seek," so that we can move away from the rhetoric towards the numbers. As recently as this weekend, the Executive repeated to the media the desirability of like-for-like comparisons. Indeed, the importance of having such comparisons was stressed to us by the Executive's chief economic adviser, but he steadfastly refused to produce any.

Obviously neither witness has felt qualified to produce such details, but have they any idea how the committee might equip itself with some of those like-for-like comparisons? How might such details be made public? How can we get common definitions so that we can produce the like-for-like comparisons that are urged on us by the Executive? Where we might look for those?

I suppose that one would need to turn to the expertise that is available in the form of Audit Scotland. Clearly, the people who will ultimately audit the exercise will need to be in agreement with the principles that underpin it. Therefore, I imagine that it is important to get Audit Scotland involved. Of course, the committee's special adviser, Professor Midwinter, is also an expert in the field. His expertise is certainly much greater than mine.

If the committee left health and local government to one side and considered everything else, that would probably give a more interesting picture.

My final question is on the role of Audit Scotland. As the witnesses may know, we were told last year that Audit Scotland would audit both the targets and the delivery process. However, in correspondence to the committee, Audit Scotland has now walked away from that commitment by suggesting that it will provide only an independent commentary, which will be selective and non-comprehensive. Although the Executive assured us that Audit Scotland would audit both the targets and the process of delivery, Audit Scotland has written to us to make it clear that it will not perform that role. Do you have any ideas about how that function could be carried out?

To be fair, Audit Scotland wrote to the Executive rather than to the committee.

Audit Scotland said that it would provide only an independent commentary.

It is pretty clear that everyone will want to say that the efficiency exercise has been examined by Audit Scotland. The issue dates back to 1997, when the Labour Government said that its economic assumptions had been audited by the National Audit Office. It was quite convenient for Gordon Brown and other United Kingdom ministers to be able to say that the NAO had considered its economic assumptions and thought that they were fine. Such a process does not amount to an audit; it is simply a question of looking at an exercise, saying that it looks reasonable and argued through and commenting on whether the evidence exists to say that it has worked. I think that Audit Scotland will do its work on the Executive's expenditure area by area. The Executive wants an imprimatur that says that the whole efficiency exercise is good, but I suspect that Audit Scotland does not feel particularly happy to offer that.

Such questions have always bedevilled efficiency exercises. I apologise for again harking back and citing an historical example, but as a student of the civil service and the British Government, I am struck by the fact that the same themes come round time and again. Major problems were encountered when analysis of the impact of the Rayner scrutiny exercises was undertaken in the 1980s and 1990s because there were three different approaches to analysing the so-called efficiency savings. The Treasury and Civil Service Committee, which was a select committee of the House of Commons, had one take on what was happening, the National Audit Office and the House of Commons Public Accounts Committee had another take on it and the Downing Street efficiency adviser, Derek Rayner, and Margaret Thatcher's ministerial teams had a third take. At the start of such exercises, there is no general agreement on what one is trying to measure in detail. Inevitably, one ends up with two or three different versions of the success or otherwise of the exercise. That is almost endemic to such exercises.

My question follows on from that. Both in your written submissions and in your comments today, you have said that efficiency exercises have been tried before and have not been terribly successful. What lessons can be learned from the lack of success of previous efficiency drives that might inform the current drive?

The main lesson is that the civil service is able to conduct such exercises with great ingenuity. It will manage matters effectively and will make everyone say that the exercise has been done and its objectives achieved, but when we examine in more detail what has changed, we might find that things have not changed.

The main problem is that although the civil service is what we are after, such processes must be implemented by the civil service itself because it is the only source of advice and expertise. The lesson that I draw is that one-off efficiency drives are not a particularly good idea and that it is better to work on such matters year by year. The Executive's changing to deliver programme is a reasonable example of how it is working on those matters year by year, by improving expertise, policy making and the service that is given to ministers. That is the way in which things should be done. Efficiency exercises are all too easy for the civil service. Although I am not sure that this is the way in which such matters are dealt with in the Executive, civil servants' instincts are to work out how they can handle an exercise so that, at the end of the day, it does not do them too much harm.

I have nothing to add to that; I concur with it completely.

We are talking about a battle with internal resistance. It is perhaps unrealistic to ask the civil service to reform itself.

The current exercise puts a great deal of emphasis on the possibility of making efficiency savings through procurement and reductions in bureaucracy. Our instincts—and the anecdotal evidence—tell us that that is the likely area for savings. However, responsibility for achieving those savings will rest with the agencies and the funded bodies, such as local government and the health service, that can make major savings through procurement and working together.

In his submission, Robert Pyper mentioned the idea of having a more unified Scottish public service as opposed to a civil service. Does the process of trying to find procurement savings—which might be cross-agency or inter-agency—represent the beginning of a process that could result in the creation of a Scottish public service?

I think that the questions about a Scottish public service are broader than that. It would be wrong to move towards such a model purely on the basis of the scope that it would offer for efficiency savings. That would be rather a narrow base on which to progress the argument for a Scottish public service. Let me make it clear that I am not necessarily arguing that a Scottish public service is a good thing or the answer to all problems, although I think that the idea deserves some consideration as a possible future development for the system of governance in devolved Scotland. There are many issues surrounding the idea that merit close attention.

As I said, however, it would be wrong to argue for a Scotland-wide public service simply on the basis of the scope that it would offer for efficiency savings. If we had a cadre of officials who worked across local government, health, the Executive and the non-departmental public bodies, joined-up thinking on matters such as efficiency would undoubtedly be easier; there is not much doubt about that.

I add that that argument is ultimately about whether we need local government. That is what such discussions often come down to. It is very easy to say how absurd it is that we have 32 local authorities all doing this, that or the next thing and how it would make much more sense if we had only one body, or one grouping of bodies, to perform such tasks as procurement. The argument is not far off from saying how absurd it is that we have 32 local authorities, which is a hard argument to make.

It is not good for the Executive to say that it thinks that local government should do many things that the Executive might not be able to do. Everyone who is involved in the improving efficiency exercise, regardless of which part of government they work in, must conduct the exercise in the fields for which they have responsibility.

In my locality, for example, the council and the health board believe that, by working together, they can make mutual savings—although it is difficult to know whose budget lines that would advantage.

You said that like-for-like comparisons could be done if health and education were excluded, but the argument has been made that Executive departments have fewer processing functions than the UK departments have, so such comparisons would have to be done on a department-by-department basis.

I think that I meant that if one was to consider just what the Executive was responsible for and what it had been doing on efficiency in those areas, one could determine whether it was doing all that it could do and whether it was doing less or more than England and Wales were doing. That would at least give one a chance to do an analysis on the mix of work that the Executive carries out. One would find that, in many cases, the things that the Executive does are not particularly susceptible to the Gershon way of looking at the world.

Elaine Murray covered much of the ground that I wanted to ask about, but there are some matters on which I want to be clear in my own mind.

When I read your submissions, I felt that they contained a refreshing cynicism that accorded with my prejudices about the work that the Finance Committee does. It often seems that we are faced with Sir Humphreys, who tie us up in knots. I had rather been hoping—perhaps it is an impossible hope—that you would be able to give us the bullets to fire at the people who come before us. You appear to be saying that the situation has always been the same and that, just as the Treasury has confused people down south for many years, the Executive is now following the same line. You seem to be saying that the Finance Committee's efforts in taking on such people can be compared to those of the boy scouts in taking on the Panther Division, for example. Is that basically what it amounts to—that we really cannot take these people on?

No, not at all. You can ask the questions. When the minister has appeared before the committee, you have asked him some excellent questions that would not otherwise have been asked, and his answer has been to talk about people going to cheese and wine parties. Fair enough. Looking at it from the outside, I think that you have done a reasonable job.

The minister has an audience in Scotland—he has the committee, the other MSPs and the people of Scotland—and an audience in Whitehall, at the Treasury. He has to leave those people happy because if they were to say that the Scottish Executive was not being run efficiently and was not up to speed with the efficiency agenda, as all modern organisations in the public sector should be, that would cause difficulties in the relationship between the Executive and the Treasury. It is easy to be hard on the minister. We have been hard on him and you have perhaps been hard on him in the questions that you have asked him; however, you are not his only audience.

As you remind us in your written submission, when Gershon initially implied that the recommended efficiency savings would be extrapolated out in terms of job losses and efficiencies in Wales and Scotland, that was immediately denied by the minister, who said that we operate differently in Scotland and would not necessarily follow Gershon's recommendations. The minister then went on to say that we would out-Gershon Gershon and be even more efficient; yet, when I asked him last week about potential job losses and so forth, despite the fact that he had noted that it is now a case of managing job losses rather than saying that there will be no job losses, we still heard nothing from him about how many job losses there might be. We never seem to get to the nub of the thing.

To be fair, you are talking about two different ministers. I think that the Executive's view has evolved.

The only figure that was issued by the Treasury was the figure of 20,000 job losses, which included job losses in Scotland, Wales, Northern Ireland and English local government. It was a very vague figure, and no firm figures have been given for Scotland, Wales and Northern Ireland.

It comes back to a point that Richard Parry made earlier. These exercises are, of themselves, not that helpful and inevitably degenerate into debates about what the figures mean. The answer is to avoid such big-bang exercises and instead have on-going scrutiny. In response to Wendy Alexander's point about targets and accountability, I would say that, in context, both targets and making people accountable for the achievement of targets are good things; however, that is better done on a continuing basis than on the basis of these rather artificial big-bang exercises that become clouded in confusion and degenerate into point-scoring exercises.

In his written submission, Mr Parry talks about the 1999 Labour spending review emphasising

"output and outcome targets, not internal efficiency."

Surely the two are not mutually exclusive. Is outcome not the best test of efficiency? Would it not be better for us to focus on outcomes than to pick over the minutiae of the efficiency savings, which will be what the Scottish Executive says that they will be and which, according to reports that we have received from outside the Parliament and our own beliefs, do not withstand audit?

You are absolutely right to say that focusing on outcomes and focusing on improvements in efficiency are not incompatible—they should happen side by side. It is, however, hard for Governments to place an equal emphasis on everything, and it is interesting to note what they choose to emphasise at any one time. It seems to me that, after 1999, Labour at the United Kingdom level was saying that it was going to put in lots of extra resources in an effort to make the health service and education work, so the extra money was a much bigger amount than had been invested in those things in recent history. In that kind of process, we tend to lose sight of how many civil servants there are implementing the policy, as it is impossible to focus on everything at the same time.

It seems to me that, for the 2004 spending review—on which work had begun much earlier, in 2003—the Treasury was beginning to think, "Oh, dear. We've been emphasising these other things. Perhaps we're losing our sense of the internal efficiency of the civil service machine, and we want to look at that"; hence, the Gershon review and the fact that the Treasury has ended up saying that that is the big thing now. The Treasury is extremely fearful that, at the end of the present spending review period, there will be much less money than there has been in the recent past, meaning that the focus will have to be on internal efficiency. That is why the Treasury is saying, during this review period, that that has to be the main focus.

I take your point on consistency. Let us go back to what Professor Pyper said about having a review mechanism that is on-going, perpetual and relentless, rather than having a big-bang exercise. I suggest that there might be lessons to be learned from other places where there is a clear focus on the critical outcome. Today, we read about Adam Crozier and the Royal Mail achieving the very clear-cut outcome of converting a £1 million a day loss into a £530 million profit. Is it possible for the Government to have that clarity of objective and that continuity of focus?

That is a complex question. In some areas such an approach would be possible, although it would be easier in an operation such as the Royal Mail. As Richard Parry said, government is inherently complex and it is much more difficult for the Government to have such clear, measurable targets in every area of operation. We talked earlier about the distinction between the sharp end—delivery by the so-called front office—and the back office. It is relatively easy to set such targets at the point of service delivery; the problem that is bedevilling this exercise is the question of what happens in the parent department—the so-called back-office activities and bureaucracy. That question is much more difficult to answer.

I accept the complexity of the totality, but the situation is less complex if we look at the individual silos of government. The questions are whether we are looking at international comparators that give us some input about others who are doing things better than we are and whether we should look to the Government to have clear-cut macro targets on things such as life expectancy. Should we be moving that forward as a target on which we can start to get—admittedly, in the long term—something that is not big bang but on-going, perpetual and long term and which benefits every taxpayer in Scotland?

I am not a comparative government expert, but I know that it is possible to cherry pick. We can look at the best of the continental European systems and the best of the Australian and New Zealand systems and find examples that put the UK to shame. However, it is not necessarily the case that the full package in all those states is better or that we are comparing like with like. Differences in the constitutional and political systems have to be taken into account, as well as the political culture in some of those states. That brings us back to the difficulty in making reasonable comparisons. Richard Parry has spent some time in Australia recently, so he might be able to shed more light on that.

I do not think that there is much that we could learn from Australia in this particular instance, but devolution has meant that we are now able to look at what is happening differently in Scotland, Wales, England and Northern Ireland. Every Government in the modern world will have long-term and medium-term targets. In health, for example, it does not make sense to have targets for this year and next year: they must be for eight, 10 or 12 years ahead or they are not meaningful.

The interesting thing is not the existence of a range of targets over the various periods but which targets we choose to emphasise. We are still utterly constrained by having elections every four years as a fundamental part of our system. Because of that, the electorate will always ask the Government what it did in its four years. With our elections in Scotland being in an off-year for Westminster, we are always on about what the Government has done and whether it is working. That is not the way to do things in the long term.

A measure such as the efficient government exercise holds a great attraction for Government because it can work by 2007 or 2008. At a UK level at least, the Government hopes to be able to say, "We did what was in our programme. We said that we would lose so many civil servants. The Conservatives said that they would do more, but we have done pretty well and have achieved our targets." All the targets are inherently political.

Do you have any advice for the committee as to what it should do now to maximise the value to the Scottish taxpayer from the exercise in which we are involved?

I would begin by examining the Scottish Executive's activities.

It is quite sad that, 20 years after the phrase "Sir Humphrey" was first uttered, it is still relevant. The trouble is that the ethos of the civil service is to provide services and the more comprehensively departmental civil servants do that within their budgets, the safer they are. That goes against any efficiency savings because if they make the service leaner or more efficient, there are more likely to be holes in it and they will be criticised. My question, which does not go as far as Jim Mather's question on Adam Crozier and the Royal Mail, is whether there is an opportunity for somebody from a different background to look into the civil service—that is not a new idea—to find out where savings could be made. The overriding concerns for civil servants appear to be about safety and service provision, so perhaps a different mindset is needed to reveal savings.

Absolutely. However, the service that the civil service offers is primarily to its ministers. Although the civil service also provides other kinds of service, civil servants basically serve their ministers' needs and interests, because that is the principle under which they work.

It is important that people from the outside look in. It helps if it is more than one person—Peter Gershon did not do his review on his own—but the UK Government seems to have fixed on one individual to examine the civil service and then it will pick up that individual's answer. It can be argued that civil services and public bureaucracies of all kinds need exercises such as the efficient government exercise because that is the only way in which they can change. It can be said that such exercises are not good because they are artificial, but it can also be said that they are the only way in which certain matters can be put on to the agenda, where they would otherwise not be.

In response to a previous question, you raised questions about the need for 32 local authorities. If we are thinking about efficient government, we must ask whether we have too much government since the Scottish Parliament was introduced six years ago. Is there an opportunity to make savings across the levels of government? You mentioned that in your introduction.

There is an interesting history lesson to be given on why we have 32 local authorities rather than 15, 18 or 20, which many people thought in 1996 would be better, but I will leave the history on that to you.

I do not know whether we have too much government, but we certainly have a more complex system of government. Some academic analysts have described it as a differentiated polity or multilevel governance. We have more complex interactions between the different levels of government, but that is not the same thing as saying that there is too much government—full stop. However, more consideration is required to join up government and ensure the quality of service delivery in such a complex environment.

If somebody wants to understand why we have 32 local authorities, they should read the description of the East Renfrewshire Council area and note the houses that are included because, I presume, Allan Stewart knew who lived there.

If the civil service is about managing the political objectives with the minimum of impact on the structure, why should we trust it?

Trust is an interesting idea. The civil service wants to be trustworthy and to have personal integrity, which is important, and most civil servants have it. You have been a minister, so you understand that the important thing for the civil service is to carry out the minister's wishes. If you want to find out whether the civil service is doing a good job, the simplest way would be to ask the ministers. People often have a notion that the civil service exists to serve everyone. There is a certain element of that in the officials' roles, in that they serve everyone's interests, but, ultimately, they give ministers the expertise and political management for which they ask.

Do you believe that that is what civil servants do?

The civil servants derive their positions from the minister. The essence of our system of government is that the civil servants exist to serve the minister in the first instance. They have other obligations, but that is their basic position.

Frank McAveety is right to be sceptical—I say that rather than "cynical"—about the civil service's ability to reform itself. There are two models of civil service reform in the recent political history of the UK. One should be considered with reference to the Fulton report on the civil service in the late 1960s. To simplify grossly, that report was handed over to the senior civil service to implement and the senior civil servants cherry picked it—they took the bits that they liked and discarded those that they did not like. A decade or more later, some significant elements of the Fulton report were unimplemented.

The other model—to which you might be leaning, strangely enough—is the Thatcherite model, in which the approach is to avoid handing over the levers of reform to the senior civil service, to run the system centrally, to bring in one's own people to manage civil service reform and to skip a generation and entrust the levers of reform to young, upcoming civil servants who, because they have careers to make and have ambitions of their own, have a vested interest in change, unlike the mandarins at the top, who have a vested interest in the status quo.

We make something comic out of the matter, but if we read the diaries of any British Cabinet minister of the 1960s or 1970s—which were much more turbulent periods for the nation politically and economically—we see that virtually all of them, whether they came from a right-wing or left-wing perspective, arrived at a conclusion not dissimilar to the second model that you suggested. That model is a more European model, compared to the apolitical and supposedly impartial model of the British civil service that we have created, which is a myth.

There are interesting and worthwhile elements in the change to deliver programme, but I am worried that the civil service does not have the skills mix that is necessary to get to the heart of efficiency savings and to change things. I am also worried that there is no opportunity to bring in people from the other areas of the public sector or from the private sector, or to get civil servants into the public and private sectors at different levels and spend four, five or six years changing things. No matter how hard-working, energetic, enthusiastic or intellectual the ministers may be—each minister probably has an element of all of those—they knock about for only a year and a half to two years and then they are out the door, so why should the civil servants worry about them?

That was heartfelt.

I am not being personal, but the idea that a minister in the Executive or Government can shift the position without having massive support from the very top strikes me as being noble stupidity.

You could do two things: first, you could make it easier for ministers to bring in their own advisers, rather than have the First Minister controlling all advisers.

What would happen then?

That would mean that work that was done in a minister's office by civil servants would be done by the minister's own people. That could bring advantages or disadvantages. In Britain, we have felt that it is an advantage to have civil servants in minister's offices, but it is hard to tell whether it is. If ministers introduced good skilled people, perhaps it would work.

The other question is this: Should a minister be able to choose his or her chief civil servants? That is not easy in the British system, but it is much easier in other systems, such as the German and Australian systems, in which there is a notion that the minister, First Minister, Prime Minister or whatever should be able to say, "I want this person as the head of the department, and that's it." That has not been the case in Britain, and it is not possible under the present rules, but it is arguable that it should be.

How do you achieve synchronicity between ministers and civil servants so that they share broadly the same perspective on how to tackle problems? Unless there is reasonable synchronicity, the introduction of outsiders to the structure will result either in their being overloaded with information, or in the information that they receive being minimised. It might also result in their being given the difficult issues so that they are undermined, in which case the minister will, de facto, be undermined, so it will be, "Ta ta."

The easiest thing would be to leave the minister in the job, would it not?

No, because the reality is that people have to shift because of political dynamics; mood changes might be required. In reality, a senior civil servant will stick around longer than the minister who is looking down the barrel of a gun.

The ideal solution would be to have agreement throughout the civil service and ministerial ranks about the reform process, so that the civil servants and the ministers invest in the process and are all pulling in the same direction. Ideally, that is what is wanted—we have some examples of that. It is not always the case that ministers in British and Scottish Governments are at loggerheads with their senior civil servants; such situations are exceptions, but there must be an agreed agenda for reform if reform is to work. That applies to government as to any other area of public life.

I am a natural optimist on such issues.

Frank McAveety's plea was a heartfelt one from someone who has been a minister, but you must in your research have come across the opposite view from senior civil servants, who must have asked, "How do you deal with ministers when they want things that are undeliverable and contradictory, and they want to badge it in a way that does not make sense?"

The answer that one normally gets from senior civil servants is that they do not mind what political steer they get, as long as they get one and have something with which to work. They do not like confusion and they do not like to work in a vacuum; if there is a vacuum, they will fill it, which suggests that the problem lies with politicians' failing to get their act together.

I have a serious and, I admit, somewhat technical question that has some relationship to Frank McAveety's question. I ask it while being conscious of the fact that we have before us our two foremost experts on governance in Scotland. Does either of you think that the principal-agent framework plays any part in understanding outcomes in government? In referring to the principal and the agent, I acknowledge that different actors have different objectives and that those actors are subject to different forms of accountability, which has a role to play in understanding the process of governance in Scotland. I am astonished that that has not featured and that we have been reduced to characterisations such as "Thatcherite", which—at least in Scotland—remains hugely loaded.

We teach our students about principal-agent relationships. The description is helpful when it is obvious who is the principal and who is the agent, when there is a real relationship and when they are doing the same thing. However, things are usually much more mixed up. The relationships between central and local government, between intermediate agencies and non-departmental public bodies, and between ministers and civil servants, are not as simple as they seem to be, based on a principal-agent distinction. That is the real problem. Maybe it comes back to what Bobby Pyper said about everyone having to agree on the agenda.

Efficiency exercises can be seen as being anti-civil service. The view may be that after years in Government, Treasury ministers are fed up with the civil service because it is not doing enough. Civil servants may then behave in their old-style ways. That situation does not help anybody. There needs to be an element of long-term predictability in the relationship, which is why the last efficient government exercise seemed to be a bit out of line with what was done by Labour at Westminster or in Scotland. It seemed to be a reversion to an old way of looking at things, which was perhaps not helpful.

I do not dissent from any of that. There may be some irony in the fact that although academic models are all well and good, in the real world things are exceedingly complex and the models can take us only so far. However, I agree that clearly defined lines of accountability are good. In practice, given the system of government that we have in Scotland and throughout the UK, and given the proliferation of cross-cutting units and other initiatives, it is sometimes difficult to see where the lines of accountability are in the system. The worst possible outcome for all of us, including service users, is the lines of accountability becoming fudged and there being no mechanisms for securing answers from the people—politicians, civil servants or public servants more broadly—who are responsible for delivering the quality of services that the public demand.

The last question is from John Swinburne.

We have to welcome any attempt at efficient government and civil service reform. The public concept of the civil service is that it is totally inefficient and top-heavy. I am old enough to remember the closed shop in engineering and industry: the civil service is the best example of a closed shop that I have ever heard of. Civil servants can do what they like and the Government cannot touch them. They are protected by their own in-built bureaucracy. For proof of inefficiency, you have only to look around you at this building—£431 million-worth—in which the civil service had quite a bit of involvement. How can any Government shake some sense into the civil service and achieve positive results? At the minute, all we are doing is tinkering at the edges because the civil service is too strong for the Government and has its own in-built protective system. Civil servants will survive when politicians will not.

I think the answer comes from the change of generation in the civil service. As Bobby Pyper said, there is a new breed of civil servants who do not expect to spend their whole working lives in the civil service. They will be brought in to do particular jobs and will then leave and perhaps do the same kind of work elsewhere. The civil service will not be a little in-group of people who joined at an early age and spent the rest of their working lives there. That is the direction of change.

My response to Mr Swinburne's argument is to say that the civil service is made up of individuals: it is important that they are trained, for instance, and that they have a broad view of their responsibilities. Although the civil service will change, there will always be a tendency for civil servants to learn on the job by examining how older people in the organisation do the job. That said, the tendency is not as strong as it once was. Our hope lies in the people who have joined the Executive since 1999 and in those who will join it in the future. I suspect that their way of looking at the world and the role of the civil service is not the same as the older generation's view.

Briefly, the only thing that I admire about the civil service is its excellent pensions system, which I hope will eventually percolate down to ordinary people.

That point is rhetorical.

The image of the civil service as a closed shop and a closed world is not entirely accurate in the modern world. There was a time when the civil service recruitment and promotion system was described as being akin to a velvet drainpipe. Someone came into the system at the bottom and was given a nice, cosy environment in which to work for all their working life. In the past, civil servants were not exposed to the outside world: they could work in the Department of Education, for example, and never go near a school, or they could work in the Department of Trade and Industry and never go near a factory. The assumption was that a person would rise to the top and go on to draw the nice pension that Mr Swinburne describes.

Those days are long gone: the civil service now recruits on the basis of a flexible system that is accessed through various entry points. Senior posts are advertised as a matter of course week in, week out in the quality press. The civil service now has pay flexibilities and recruitment flexibilities and allows secondments to the private sector. Perhaps more secondments should be made from the civil service to other parts of the public sector—I suggest that that would be a good step forward.

The civil service is no longer a closed world. Indeed, that description is a caricature; it is the civil service of a bygone age. It belongs with the image of the pin-stripe-suited, brolly-carrying mandarin walking over a bridge in London to his office. Most civil servants nowadays are women and most work outside London. They do not work directly for ministers in departments, but at the sharp end of delivering services to the public. A contrast needs to be made between the image of the civil service and the reality.

On behalf of the committee, I thank both witnesses for their excellent submissions and for the excellent evidence they gave this morning in response to our questions. I suspend the meeting for three minutes while the PowerPoint presentation is set up for our next panellist. Richard Parry and Professor Pyper are very welcome to stay on if they wish to do so.

Meeting suspended.

On resuming—

I reconvene the meeting. I am pleased to welcome Professor Michael Barber, director of the Prime Minister's delivery unit in the Cabinet Office. I thank Professor Barber for travelling up to Edinburgh to be with us today. I understand that he will make a short PowerPoint presentation on the work of the delivery unit, which I hope will provide a context for our questioning. I see the change in witnesses as a move from the Jeremiah academic to the upbeat practitioner—we will see whether that is the case.

I thank the convener for the invitation; it is a pleasure to be at committee today. My presentation fits into some of the major themes that the committee debated during the past hour with the previous panel. I hope that it will be helpful to the committee's scrutiny.

In a brief presentation of about 10 minutes or so, I intend to explain how the delivery unit works and what impact it is having. Obviously, the committee is much better placed than I am to know whether the model is appropriate for the Scottish context. I am very excited about the delivery unit; however, as I have been responsible for its development, I am also totally biased.

I have two more things to say by way of introduction. The origin of what we are doing in the delivery unit—indeed, it was implicit in much of the debate over the past hour—is that citizens of a modern country are generally quite willing to pay taxes but only on the condition that they see the benefit of doing so in real service improvements. They want to see those benefits not in five or 10 years, but preferably yesterday and certainly today. They also need evidence that services are improving.

The dilemma that modern politicians in all developed countries face today is that, although they need a long-term strategy in order to be successful, they must delver short-term results, otherwise no one will believe in them. The aim of the delivery unit is to help people through that dilemma.

The unit has about 40 people who consider a wide range of substantial Government programmes in health, education, crime, asylum and parts of the transport system. By definition, we are great simplifiers, whereas the job of many people in Government is to complicate things; as the previous witnesses said, Government is a complex thing. That said, somebody in Government has to simplify and make sense of things. That is what we try to do.

At the beginning of the previous United Kingdom Parliament in 2001, when the Prime Minister established the delivery unit, the process that was adopted was broadly that the Prime Minister debated with secretaries of state for health, education, transport and the Home Office his top priorities, and they then agreed what the priorities should be. That set the agenda for the delivery unit's work; for example, we did not work on the whole range of major tasks for the Department of Health but specifically on a handful of them, which I will come to in a minute. For each priority, we worked out—in collaboration with the Treasury—what the measurable goals would be and what the deadlines for delivery should be. The goals were often public-service agreement targets that had a clear definition of what would constitute success. For each goal, a set of indicators was needed; it is important to know whether you are on track towards achieving goals. On one level, our work is very dull, but it is about getting the kind of data that you were discussing earlier so that we can say whether we are making progress or not.

Having established the priorities with the Prime Minister and the secretaries of state, we developed a system with the characteristics that are shown on the slide that is on the screen now. I will take committee members quickly through the priorities. For each key goal, we asked officials—starting with the relevant permanent secretary—about their plans for achieving their goals. We said, "Your minister has agreed with the Prime Minister a priority for this Parliament. We want to know how you plan to achieve it." We asked them explicitly not to write a plan for the delivery unit but just to show us the plans that they would use in their departments. We did not want to create bureaucracy.

As I said, we are simplifiers, so there were two things that we really cared about in their plans. The first was the key milestones or the key steps on the route to achieving the outcomes and the second was the trajectory. We knew the goal and how getting there would be measured, so we also wanted to know the current data point and the line that connected it to the goal. The concept of the trajectory is now in widespread use; indeed, the word is often uttered by the Prime Minister when he questions secretaries of state.

Trajectory is to do with the progress that is being made towards achievement of a target and being asked about trajectory requires that the civil servants who are responsible for programmes think about the relationship between their actions and the outcomes in the real world. They have to connect their actions to real outcomes.

Having established plans with milestones and trajectories, we report to the Prime Minister on the key data streams, which we usually do monthly. We never ask for data that a department does not need for itself and we do not invent bureaucratic requirements; we ask only for data that the departments think they need for delivery of their own programmes. We then turn those data into something that the Prime Minister can consider.

Every two or three months in the previous UK Parliament, the Prime Minister held stocktaking meetings with the key secretaries of state. Reports of progress were made on the various trajectories, and discussions were held on what was going well and what was going less well. If something was going less well, discussions were held on what would be done about it.

Periodically, in what we call a priority review, we review with the departments whether what we see in the data is actually reflected on the front line, and whether what the data appear to be saying is connected to what people in a hospital, a school or a police force are actually saying. We check the reality of delivery and we try to do it quickly. It might take a month from deciding to carry out a review to getting the outcome. We are embedding that review process in Whitehall so that departments can do it for themselves.

Obviously, with many big programmes going on at any given moment, some things are not quite on track. When that happens, we try to prompt problem-solving action. We always offer assistance in solving problems, which departments often accept. We are very focused: we ask people, "What are you trying to do? Are you on track? If not, what are you going to do about it, and would you like us to help?" Periodically, we summarise for the Prime Minister and the Chancellor of the Exchequer the progress on delivery on the Government's key priorities. At the end of my presentation, I will show committee members how we do that.

First, I want to take members through three stories that illustrate practically the process that I have just described. The first story is on health waiting times. The figures that you see on the graphs that I have provided are from the Department of Health in England. In March 2003, there was a target that nobody should wait for more than 12 months—that target is shown by the black line—which was pretty much achieved. In March 2004, there was a similar target that nobody should wait for more than nine months for elective surgery. The red star that is shown in the bottom right-hand corner of the graph on inpatient waiters represents the public service agreement target for December of this year. The big red vertical bars on the graph represent the number of people who were waiting for more than six months.

Members can see that getting those figures down was a substantial challenge. Until early 2003, there was really no progress towards achievement of the key target. At that point, we agreed a trajectory with the Department of Health—the trajectory is shown in the graph as a broken red line—and we said that we would believe that the department was on track to hit the ultimate target if its figures came within a target represented by the red circle on the graph, which represents the 2004 milestone. As members can see, the department achieved that. We now have the figures for the most recent financial year; the department is now on track and will hit its target, probably in November.

We can track progress but, more importantly, we can analyse what the department did to achieve its targets. A range of factors helped. The first was the introduction of choice. In England, if patients were waiting for more than five months, they were offered a choice of hospital. They could keep the original hospital appointment or, if they wanted to, they could go somewhere the operation would be done within six months, which created an incentive for hospitals to speed up the process.

A second factor was the new treatment centres, competition from which has also been an incentive to improve. Thirdly, the introduction of three-year funding in March 2003 enabled primary care trusts to plan better. Since April 2004, the kind of choice that I described a moment ago has been available right across the country. All those factors are incentivising progress towards achievement of that very important target.

I move on to the lessons that we have learned on reducing waiting times. There was absolutely consistent priority on that particular target all the way through the previous UK Parliament, which is being sustained, as it will be until the target is achieved. Choice and contestability have undoubtedly made a difference. Also, building of capacity—the big investment that we are all familiar with—really helped, as did the ability to plan long term.

I want to emphasise the last lesson, which was that constant management of performance—examining data and holding people to account at all levels in the national health service—is fundamental. Public accountability through the star-rating system has made an enormous difference. It was also important to get behind and really understand the data because that allowed us, for example, to pick out the fact that progress in orthopaedics was less rapid than it was in other specialties, which allowed us to focus specifically on solving that problem.

The second story that I want to tell is on accident and emergency services. Until 2003, very little progress had been made towards the target of a maximum waiting time of four hours—the target was not being taken seriously. In August and September 2002, we carried out a review with the Department of Health and, as a result, accident and emergency services were included in the star-rating system and hospitals became publicly accountable for delivering the target. A new focus was placed on the data and Alan Milburn made a series of speeches on the importance of the target. That target was, as a result, pushed up the priority list; committee members can see on the appropriate graph the effect that that had on the data. In addition to the pressures of accountability, incentives were introduced in the second half of 2004, which also improved progress. The target has now just been achieved.

Again, management of performance, interventions, checking and use of data, and the help that was offered to hospitals that were having problems all underpinned the achievement of the target. Ministers in the Cabinet were delighted when the issue was mentioned in "EastEnders" on 26 November last year. Committee members can see a bit of the dialogue on a slide on that. I think there was Cabinet resolution that such an endorsement was worth 10 ministerial speeches—[Laughter.]

The key lessons that have been learned in meeting the accident and emergency target are on the next slide. Again, we see the importance of sharp accountability and the right incentives, and the importance of engaging the delivery chain, which is jargon for everybody who is involved in delivery of a target. Doctors, nurses and everybody else became much more motivated as progress towards the target was made. Also important were tailored support and performance management right through to the end.

The third story is on secondary education in London and it illustrates a rather different set of themes. The London challenge strategy considered performance in five boroughs that lagged far behind the rest of the nation—Hackney, Islington, Haringey, Southwark and Lambeth. Some committee members will be familiar with those boroughs. As you can see on the slide, in 1997 their performance in attainment of general certificates of secondary education was nearly 18 percentage points behind the rest of the nation—a huge gap.

One of the Treasury's floor targets, which was introduced in 2000, was that no local education authority should have performance on that indicator of below 38 per cent. When that target was introduced, I thought that although it was a noble aspiration, it was heroic. However, it was not just achieved but exceeded, for a number of reasons. There were interventions from central Government in four of those local education authorities on the ground that they were underperforming. In four cases, parts of or whole education authority services were contracted out. In three of the four cases, that contracting out worked well.

A programme called excellence in cities, which encourages schools to collaborate in dealing with difficult pupils and provides various financial and other supports to schools, was introduced. The London challenge programme targeted those schools that were far below the floor target, provided expert assistance and brought major progress, which got steeper in the past couple of years, as members can see from the slide.

The next slide shows key lessons. There was strong leadership from Stephen Twigg, the minister with responsibility for London schools in the previous Parliament, supported by good professional and official leadership. There was constant focus and prioritisation throughout as well as tailored support and clear accountability—everything was published—and the impact of contestability was monitored.

I will now summarise and show members how we do our delivery reports across the whole of government. We wanted to develop a framework that would allow us to compare how the 20-odd priorities that we have at any given time—they have mostly stayed the same over the past four years—are working and the likelihood of delivery of quite different things. For example, we wanted to compare the likelihood of delivering on health waiting times with the likelihood of improving the performance of the railways or of reducing crime. The next slide shows the framework that we developed.

We decided that four factors really made the difference and members can see them listed on the left side of the slide. First, we ask how difficult the thing is to do and what is the degree of challenge; the more difficult it is, the harder it is to deliver. The second question is how good the planning, implementation and performance management are. Those are the kinds of things I have been speaking about. We ask whether the department has a good plan and whether it is implementing that plan well. Does it have a system for managing performance against that objective? The third question is what the capacity to drive progress is. Does the responsible department have the money, the relationships and the skills not just in the department but throughout the public service to ensure that that occurs? The final question is what stage of delivery the departments have reached. Have they just published their white paper and introduced the legislation? Are they in early or late implementation, or have they thoroughly embedded the measures? We consider those four questions and reach a judgment—I will show members a slide of how that looks in a minute. Then we take account of recent performance and, if it is good, that will be a factor in giving people more optimism.

The system enables us to take the 20 priorities and put them in a league table. The next slide shows a made-up league table, but we use exactly that framework. Members will see a list of goals down the far side of the table. The degree of challenge is low, high or very high. The quality of planning, implementation and performance management is judged to be green in the case at the top of the league table and red in the case at the bottom. The capacity to drive progress is graded in the same way. The stage of delivery in all those cases, as of last July, was rated at 3 or 2 and that enables us to make a judgment about whether something is likely to occur and then we can rank order it. The system is used internally and is sensitive stuff, but if one puts the league tables in front of the permanent secretaries who are responsible for those targets, as I do from time to time, one can be sure that the person at the bottom of the league table will go away and do something about it.

The next part of the system is a single page explaining how we reached the judgment that I described. In the top corner of the example page, members can see that it includes progress against trajectory, so that people know what their target is and whether they are ahead of it or behind it. We make an analysis of progress and ask why that progress is occurring. We ask, "What do we think success will look like six months from now? In six months, when we do this exercise again, what do we expect the departments to have achieved?" Finally, we ask what it takes to achieve that success and what we will do to assist people.

Both the league table judgments that I showed members and the pages of the delivery report are agreed with the department, so that the report becomes a brief statement for the Prime Minister of what the department intends to do on that key priority over the next six months. That one-page summary of performance and future plans is negotiated with the department.

One can also use public data for public purposes and the system tells us whether targets are ahead of 1997, which is an important factor for the Prime Minister, and whether the current direction is positive, flat or negative. Members can see from the next slide that the current direction is largely positive.

However, one reason for the big efficiency drive that members debated with the previous witnesses is that, according to Treasury figures, the big growth in total managed expenditure will flatten out over the next few years, but public expectations will not flatten out—they will continue to rise. The public's demand that we keep delivering ever-improved services will, if anything, intensify as that spending curve flattens out. The critical question is: what does it take to get continuous improved performance, even when the rate of increase of investment in public services cannot continue at the rates of the past few years?

From looking at the kind of stories that I just described to the committee, as well as many others, we believe that, if we get them right, three things will bring about transformation. The first is to get the right mindset. The questions that Frank McAveety raised about synchronicity between politicians and civil servants are crucial to that. Have we got the guiding coalition and are people all lined up behind the goals? Is there a shared vision and do people have a sense of ambition? Do they believe that those big targets can be achieved? In Whitehall, we are always the people who believe that targets can be achieved, however heroic they look, and we try to bring out that sense of ambition. Although clarity about priorities is difficult in government, it is essential. We also look at effective performance management, which members saw illustrated in each of the stories.

Incidentally, one of the things that targets do when they are published is to bring consistency through changes of minister. For example, when Alan Milburn ceased to be Secretary of State for Health and John Reid took over, John Reid did not say, "Actually, I won't meet that six-month waiting times target." Instead, he carried on with it. However, in the previous 50 years, at the time of ministerial change, it was common for there to be a new white paper, a new policy and a relaunch. All those things are not very exciting, but they are essential in getting things done.

Finally, we look at the content of reform. In general, when the list of reforms on the slide has been applied, results have followed.

Thank you; there was a lot in your presentation.

I thought that you made an extremely interesting presentation. I am tempted to ask whether you have made it to the First Minister, as it seems to be a rather different approach from what we have seen in Scotland in recent times.

You seemed to concentrate on three specific success stories when you talked about your targets. Is it important that you limit the numbers of targets in any period? Is your system possible if you select more targets? The other question that Frank McAveety has put in my ear is whether you have had any failures.

The three stories that I told were success stories and their point in the presentation was to provide evidence for the final slide. I intentionally selected successes.

It is important to limit the numbers of priorities. The targets have evolved since 1998; the Treasury and ministers in London have accepted that they definitely had too many targets at the beginning and the targets have become fewer and much more focused. The design of the target is important too. Government has generally got better at all that.

What I describe can work only on a small set of priorities and, by definition, if something is a priority, there is a lot of other stuff that is not a priority. It involves a great deal of discipline and rigour to deal with that, not just at Government level, but at each department level. You are quite right about that.

It has been much more difficult to make progress in some areas than in others. There have been no outright failures but, for example, between 2000 and 2003, after the big improvements in literacy and numeracy in primary schools in England, there was a plateau that was very frustrating for a while, although we had a jump off that in the final year of the period. Some aspects of the criminal justice system took a long time to improve from quite a low base, but they are now improving, and trying to improve the number of removals of failed asylum seekers has caused quite a lot of frustration. Some things have proved tougher nuts to crack than others have, but I would not call them failures. However, if the lessons that I have described are consistently applied, there will almost always be progress, which is interesting. Two years ago, we thought that we had a good idea and a theory; the approach is now demonstrably effective.

Are there any implications for what Frank McAveety said earlier? It is clear that Cabinet ministers have signed up to and fully supported the initiative. Is there any evidence of less unanimity and less support in the civil service for Cabinet ministers and the Prime Minister in driving the targets?

At the beginning of the delivery unit process, there was quite a lot of scepticism among senior civil servants about whether things would work. As the previous witnesses said, senior civil servants have seen units and processes come and go and some of them may have thought that the approach that we are discussing would come and go. The delivery unit's deal with Cabinet ministers and the Prime Minister is that we will get their bureaucracy to do what they want it to do anyway, so we are aligned with the political aspirations that are set out in the Government's programme.

Without exception, senior civil servants—certainly those with whom we work—now speak highly of the delivery unit. They do so for several reasons. First, in the end they have welcomed the consistent pursuit of a small number of priorities, which makes their lives better. They know what the Prime Minister wants all the way through a Parliament. We are now involved in the same process for the current Parliament.

Secondly, we have helped senior civil servants to solve their problems. Once they came to terms with having to provide us with the data and with not doing the traditional bureaucratic thing of holding back data and information, and once we began to have a real debate about what is and is not working and what to do about that, they welcomed the problem-solving support.

Thirdly, the delivery unit does not want credit for anything. If something works, the credit will go to the department's secretary of state. Therefore, senior civil servants are given clear priorities, problem-solving assistance and credit when things work.

That probably captures matters. However, in describing the model in question, you used the words "contestability" and "sharp accountability", which is a lovely euphemism—I know what you mean by it. Do you detect that there is an appetite for such things in parts of the UK other than Scotland? Is there a desire for such things in the devolved institutions? Such a desire would mean tough political challenges for those who look at things from a Labour perspective.

Tough challenges are involved and tough political thinking is required. However, introducing clear accountability makes a difference. It seems to me that holding people accountable for their results is fundamental.

I started with a big political dilemma. People want a long-term strategy, but short-term results must be delivered; otherwise nobody will believe you. Unless there is clear accountability in a system, such results are hard to achieve. Whenever politicians want to deliver results within reasonable timescales—I do not mean only in the UK—they will be driven towards sharp accountability, which includes making data public. Recently, I was struck by the discovery that school results are now published annually in Norway. Five years ago, I did not think that that would happen, but people in Norway have started to publish them annually.

Contestability is essential as, although part of the explanation for our public services' relative underperformance in the post-war era is underfunding at different points within that 50-year period, much of it is to do with monopoly control by professionals. If those monopolies are left in place, we will be faced with great difficulties in trying to achieve results.

What you have said is interesting. On what Ted Brocklebank said, some of what we are discussing informed the changing to deliver programme, so the Executive has done a similar exercise.

You mentioned that there are 20 priorities. How are those priorities decided? Obviously, there is a danger that priorities might skew other aspects of service delivery. For example, waiting times and times that people spend in accident and emergency might become all-important rather than improving general public health, stopping people smoking, getting people more physically active and so on, which are more difficult objectives to achieve in the short term.

The unit is the Prime Minister's delivery unit, so we ultimately answer to him, and he ultimately decides what the priorities are. In 2001, at the beginning of the previous Parliament, the Prime Minister and the relevant secretaries of state discussed what the priorities might be—they discussed which departments would be priorities and the priorities in those departments. We also discuss priorities with the Treasury at some length and we have reviewed priorities each year. The core of the programme stayed the same for four years, but we changed priorities at the edges annually. We are now trying to establish priorities for the current Parliament and again I expect that the core of the programme will stay much the same but that variations will occur.

You are right to draw attention to the downsides of prioritisation. Picking priorities means giving less attention to other things. When a priority is set, the delivery unit tries to anticipate what will happen and we have tried to embed such an approach in the disciplines of delivery in government. We ask what might go wrong if a priority skews everything and try to work out unintended consequences. For example, when the street crime initiative began in 2002, the police said that burglary and car crime would increase if they were made to focus resources on street crime. We said that that was a plausible hypothesis and that we would check what happened. When street crime figures started to fall, we discovered that car crime and burglary fell in areas in which street crime was falling fastest, as police effectiveness is an element in driving the figures for all three areas. If people say that data will be skewed, we say that we will check. We must then judge whether we can take the cost of the negative impact, if there is one.

There is a league table for the 20 priorities and people can see who is at the bottom of it. What happens if a department is always at the bottom of the table? If its performance on a project does not improve, will the minister be sacked? Will civil servants be shuffled sideways? What will happen to them?

The good news is that nothing has always been at the bottom for any length of time. Departments will be at the bottom of the table as a result of the absence of some of the factors shown on the slide. Let us start with the mindset. It will be hard to make progress if there is the lack of synchronicity that Frank McAveety mentioned and if a minister thinks one thing, a civil servant thinks another thing, a set of people in a particular public service think another thing again and there is a lot of conflict. If a priority is at the bottom of the table, it will receive intensified focus, which should lead to progress.

Obviously, some matters take longer to deal with. I did not mention dealing with railway performance, which has been a struggle, although in the past six to nine months, there has been steady progress and performance levels are now regularly above Hatfield levels. That has involved aligning the rail industry with the factors shown on the slide. The rail industry is very diverse and, under previous legislation, had not been organised in the way that one would generally plan to organise an industry. However, things are now being fixed and improved.

My final observation is that many of the goals are delivered by agencies rather than the Government. What levers ensure that, for example, health boards, local education authorities or those in the railway sector deliver what you want them to deliver? Getting those in the railway sector to do so is more difficult.

That is a difficult question. The national health service, for example, is in effect a unified service, but the implications behind your question apply even to it, as doctors and front-line managers must still be persuaded that things are worth doing. The levers are clear objectives, persuasion, money and accountability, which is a strong lever. If primary schools' literacy and numeracy results are published, they will be focused on. Inspection systems and good relationships are levers.

Above all, the moral purpose must never be neglected, although it is easy to miss. We can persuade public servants that higher standards of performance and greater equity are goals in which it is worth investing time and energy. Teachers want to do that. Doctors and nurses believe in a national health service that provides higher standards and greater equity. We must set that in the framework of what we are trying to do—to build public services that are fit for the 21st century, meet rising expectations and deliver greater equity. That is a cause to which people will sign up. We can get a guiding coalition if we build it round a belief, rather than just a target.

So much of it is about leadership.

Absolutely. That is a very good word to use.

A box on one of the slides showed the actions that you decided should be taken at one review meeting, when things were not going right. Often people are not sure what actions need to be taken, because they do not know how to solve a problem. Have you run into situations in which people think that they have an insoluble problem? It is not that they are not focusing on the problem but that they are not sure what to do to make things work.

We are talking about big, complex programmes. In some parts of reform, we are at the cutting edge, so we run into problems that have not been solved before. There is a set of questions that we can ask in that situation. In the past, the fact that people were not sure what to do might have led them to sweep data under the carpet. However, once people put the problem on the table, they must do something about it.

If someone from a department asks us what they should do about a problem with which they do not know how to deal, we ask them whether there is any variation in the service. The A and E data by hospital may show that some people met the target three years ago. We need to ask what they did to meet the target, as that may provide a solution. We must also ask whether there is international best practice. If we want to know how to teach reading, we can learn from many programmes that are in place worldwide, some of which are working and some of which are not. If we want to know how well schools in other countries are performing, we can make international comparisons and seek to establish what the countries that are solving problems are doing.

Although we may run into problems that we have not solved before, there is a set of questions that lead us to the answers. Where is the best practice within a service? Where is the best practice internationally? Where is the best practice in other industries that may be relevant? The delivery unit has been able to bring to Whitehall a set of learning from across programmes, which we have tried to summarise. I often ask people in the Department for Education and Skills whether they have considered how the Department of Health is managing A and E performance, for example. The answer is out there somewhere, and we must try things in order to find it. With early data, it is possible to refine solutions if they are not working. It is important not to give up trying.

How do you reward ministers who have delivered? I listened to your presentation and heard about the wonderful progress that you have made in the health service. However, the man who oversaw that is now in charge of guns and so on in another department.

People prefer to call it security.

Rewarding ministers is not part of my job. You need to address your question to the Prime Minister.

How much are the efficiencies that have been made in the English health service due to the fact that there are better number crunchers, accountants and statisticians south of the border than up here?

I do not know whether that is true. However, although sometimes number crunchers, accountants, bean counters and so on get a bad name, they are essential to the running of effective modern public services. One of the things that struck me most vividly when I started at the delivery unit in 2001 was how many programmes were spending very large sums of public money without there being any really good data that allowed us to monitor implementation. I want to speak out for the bean counters.

As you mentioned bean counters, we will take a question from Jim Mather.

As a former bean counter who has been in therapy for many years in other spheres of activity, I would like to move from the theme of numeracy to that of literacy. Can you expand on three key terms that appear in the column headed "Bold Reform": "personalisation", "contestability" and "vibrant supply side"?

Personalisation is about changing the take-it-or-leave-it public services that were developed in the immediate post-war era: "Here is your national health service. Everybody liked it at the time. This is the way we do things and if you don't like it, tough." We have moved to a system that is designed round the patient, the consumer and the customer. There are many examples of that, but the best example in the current transformation programme is the work that is being done with people who have long-term conditions. In the past, the health service treated someone who had diabetes as a series of episodes. Each time they went into hospital or social care they were a new episode, whereas now the service is designed round the patient. Staff can say, "You are Michael Barber; we've met you before, we've got your records and we know about you. We are trying to support you to manage your diabetes in your own home." That is a good example of personalisation.

Contestability is about opening up to alternative providers an element of a service that is underperforming. That might involve someone coming in from outside the service, as in the contracting out of education services in Islington by the local education authority, or it might involve an alternative supplier from within the public sector. Some schools in England have taken over nearby schools that were underperforming and improved them. A vibrant supply side goes along with contestability; one cannot open up services to alternative suppliers unless there are people who are willing to provide that service.

When you talked about contestability earlier, you talked about not leaving monopolies in place. Is that purely at the local level—for example, in a school—or do you have other monopoly targets? If so, have any of them been tackled yet?

The introduction of treatment centres in the health service has been important. Some of them are provided by the NHS and some by external, independent providers, which have undoubtedly provided an element of contestability. The idea that there should be alternative providers of services is now a key part of the emerging vision of secondary education in England.

In making that happen, you challenged a number of orthodoxies. Was the programme purely the Prime Minister's idea? What was its genesis?

The Prime Minister provides important leadership, but the genesis of the programme is where I began: enormous sums of public money are invested in public services and the people who, as taxpayers, are paying for that investment expect to see results. In the post-war era, when the NHS was created, one could say to people, "We have a vision of an NHS. You might not see results for a while, but it is well worth investing in." At that time people would take that on trust, but we do not live in that era and people are impatient to see results, not just in public services but in other services. The programmes that I have mentioned are the things that appear to be likely to deliver results within a reasonable timescale. It is helpful if one's long-term strategy is designed in such a way that it will deliver the short-term results that I mentioned earlier.

Beyond your presentation, which was excellent, and the case-study material, which is impressive, what other work do you have that is knowledge-transfer ready and could be passed to other entities?

That is an interesting question. There is a lot behind our work that we might call knowledge-transfer ready. I showed you a diagram on how we assess the likelihood of delivery and behind that is a set of questions that help us to make decisions. We have an assessment framework that explains how we do that.

Incidentally, Sarah Smith, who runs the Scottish Executive strategy and delivery unit, came to visit us for a week or so and has been introduced to the delivery unit approach, although I am sure that she came back here and improved on what she learned from us.

The methodology for doing a review in a month is simple but powerful. You have Audit Scotland, and in England we have the Audit Commission, which often takes two years to produce a report. We asked the commission how long it takes for it to know 90 per cent of what is in the final report, and the commission said that it takes a month; we just do the month and get 90 per cent of what we need to know, and then we act on it.

Have you given your presentation to the Scottish Cabinet or senior managers? You mentioned that Sarah Smith had been down, but she is the messenger. Have you been asked to deliver your presentation?

I am not sure whether the permanent secretary has seen the presentation, but he has been involved in discussions on delivery and how we have been doing it in Whitehall. I have not given the presentation to the Scottish Cabinet collectively.

The challenge for a committee that is charged largely with scrutiny is how we put some of the issues on the table in Scotland and reflect on our choices, so that our response on efficient government is not exclusively about the minutiae of figures, much as some of us are interested in those. As Jim Mather said, the existence of devolved authorities has led to knowledge transfer. There are areas in which we think we are out ahead—for example, smoking in public places or bus travel for pensioners—but your presentation suggests that we can learn something about how to deliver improvements in public services.

I hope that you can share the non-proprietary information in your presentation, because that would be helpful to the committee. It would also be helpful if somebody in your unit could do a bit of work with our advisers or support staff on the three examples—rather than all 20—that you mentioned and the key levers for driving improvement. That would let us put some positive suggestions on the table at the end of the process, so that other parties to the debate could think about the possible lessons and incentives. For example, on in-patient waiting, you highlighted the choice of hospital after five months and the importance of hospitals having an incentive to comply. I presume that that relates to activity funding of hospitals, which we still need to think about in Scotland, along with funding of primary care trusts, star ratings and so on.

I agree that devolution can bring about knowledge transfer and learning across UK boundaries. That is beneficial, although we have not seen enough of it. In preparation for this meeting, I examined the data on detection rates in Scottish police forces, which are way ahead of English ones. When we performed a one-month review last autumn of how to improve detection in police forces, I did not send somebody to Scotland, so shame on me, because I should have done so. In fact, when I get back, I might try to do that, because the figures are impressive. We would be happy to collaborate with a member of the committee's staff. It would be no problem for somebody to come down or for somebody from the unit to come up to discuss the background to the examples.

One of the difficulties that you face is in making progress throughout England. Presumably, the size of the population and the range of agencies involved make the changes more difficult to implement than in Scotland, with its smaller population and the potential for the Government to drive through changes. Do you share the perception that some of the lessons that you have talked about could be applied more quickly in England? We have a lot of diversity in Scotland, but I presume that we could be more focused.

Basically, I agree with what you are saying, but there is a dilemma. Scale is an interesting issue. Although Scotland evidently has a smaller population, which is an important factor in some of the services that we are talking about, there are still often parallel agencies and the same number of lobby groups. The influence of well-organised parts of the status quo in a small country is sometimes more powerful than it is in a bigger country—things can cut both ways, as the intensity of those relationships in a smaller country can be stronger.

I agree that, for some of the health and education programmes, on a smaller scale and with the right design, it should be easier to achieve what I call faithful implementation. It should be easier to find the model that will work at school or hospital level if there are fewer front-line units. For example, an A and E policy in England will perhaps be required to influence 200 A and E departments, whereas there will be many fewer such departments in Scotland—I do not know how many. Similarly, a primary school literacy programme in Scotland will be on a much smaller scale than one that covers 20,000 schools in England. To secure faithful implementation among 20,000 schools and 190,000 primary school teachers in England is a challenge. In Scotland, implementation should—at least, theoretically—be less of a challenge. Nonetheless, the other forces will also apply in a small country, as we see in places other than the United Kingdom.

One of the most striking parts of your presentation was your comment at the beginning that you see your role as being to simplify what needs to be done and to focus on that priority. Is that not what the Government needs to do if it is to do anything at all? Complexity will mean inertia.

I work directly for the Prime Minister, who has a lot of things to do apart from delivering public services, although that is probably his most important objective. We have put a lot of work into simplifying and clarifying so that he can very quickly pick up what is happening and make well-informed decisions. That is about the presentation of data and their clarification.

The basic insight that you are describing is absolutely right. Modern government is, potentially, so complicated that one can be overwhelmed by it. We need to work hard to get a few clear, simple priorities and try to deliver them without being deflected. Of course, other things will happen and the Government machine has to do all the things that it does. However, we cannot deliver big change in a reasonable timescale across everything; we have to be determined in focusing on those objectives.

Let us pursue that point slightly. Some of the policy examples that you gave are problem driven—there is an identifiable problem and you are trying to find a practical way of dealing with it or a series of steps that will allow you to make progress on it. Does the same model work in the context of anticipated needs? I am thinking about demographic patterns of change, how public services anticipate potential change and how they respond to it, rather than about here-and-now problems and how we can fix those.

Part of the answer to having an anticipatory programme lies in the unit's approach, but that also requires strategy, and this is not about strategy. Let us say that we wanted a five-year to 10-year programme for some Government somewhere in the world. A strategy would be needed, so we would have to take account of demographics, the other changes in the world and what we wanted to do. We would then need to move from strategy to design. Once we knew what we wanted to achieve, we would have to decide what design of system we would need to deliver it. Then, when we had a strategy and a design, we would get into the planning, delivering and monitoring. To do what you are talking about, in addition to plan to delivery and delivery to outcome, we need strategy to design and design to plan.

Obviously, your focus has been on ways of identifying problems, managing performance, assessing data and moving forward to delivery. Reflecting on your experience of that, what do you believe politicians have to learn about their role in this context?

That is a good question. First, to go back to your starting point, I would say that it is important that politicians are clear about their priorities. One of the previous witnesses said that his research had led him to realise that civil servants do not mind whether a programme is left wing or right wing—they just want to know what it is. Clarity about priorities is important and should constantly be reinforced.

Secondly, it is important that politicians should consistently pursue those priorities, even when the going gets tough. The civil service should be held to account for results. Until the election, John Reid held a monthly meeting in the Westminster Department of Health with the director of delivery and other key people at which they would go through the kind of data that I was showing you. That meant that he was personally holding civil servants to account for the delivery of the targets on which he knew that he, on behalf of the Government, was resting a huge amount of credibility.

Thirdly, it is important to push people to find solutions. In the world of bureaucracy, there is a significant tendency to look for incremental change and modest progress. In fact, however, many things are achieved as the result of big ambitions. It was ambitious to say that no one will wait for more than four hours in an A and E unit and it was ambitious to say, in 1997, that 80 per cent of 11-year-olds in England will achieve level 4 in national curriculum English. That sort of selective use of ambitious targets, related to political judgment, jolts the system into action.

Finally, all the stuff in the presentation about understanding what makes systems work and letting people get on with it is important.

Do you find that politicians have the right background to understand the data and the process in a way that will allow them to operate effectively? Might that require a different type of politician from the one to which Frank McAveety referred—the 1960s and 1970s model who might be good at inter-party fisticuffs, but might not be geared up for the sort of exercise that you are talking about?

I am not sure about the answer to your question, but I will tell you what I think. The bean counters, on whose behalf I spoke up about half an hour ago, have a responsibility to learn how to present and communicate data—even complex data—in a way that makes sense to ordinary people, among whom, in this case, I include politicians.

For example, most people in this country have a rough idea of what the inflation rate is. They do not need to know how it is calculated to make sense of it. If inflation is 2 per cent or 5 per cent, people have a rough idea of what that means. We spend a lot of time taking extremely complex data and turning the information into graphs so that the Prime Minister can understand it. There is an issue about presenting data in a clear way while staying faithful to the message of the data. Politicians need not have degrees in statistics, but they have to be able to ask for the data and push for the people who are responsible for the information to produce data that make sense and drive action.

Given the demographic trends, older people are now more important for elections than they have been for generations. If there had been a different result a fortnight ago and there was a hung Parliament, in which one of the potential coalition partners was holding out for free personal care for the elderly in England, what light would your data shed on the situation and would they influence politicians?

I have no idea; I have not even looked at the data in question. It would be good if the data were available to make the decision. I am sorry to be ignorant, but I have no idea what the data would show.

I ask the question because that is perhaps a rollerball that will come your way sometime soon.

I read an interesting article by George Walden, the former Conservative minister, in the New Statesman. He talked about sharing a whisky with John Smith prior to John's death and of the conversation turning to the ridiculous position of mortgage tax relief—which is a political no-brainer in electoral terms. George Walden asked John Smith, "Why don't you just be brave and say that you will abolish it?" to which John Smith gave the wonderful answer, "I will, if she will." That is the sort of political debate that we are engaged in today. We need to move people towards accepting more sensible solutions to the long-term use of public finances, whether for free personal care for the elderly or for mortgage tax relief. Both those policies are popular with the electorate but they are political dynamite.

Those are the sort of dangerous shoals in which politicians swim, as it were. Such decisions are difficult political judgments to make. I am not in a position to answer the question, but I can say that politicians need a system that enables clear outcomes to be set and data to be used to demonstrate whether those outcomes are achieved. That is likely to improve the degree to which the public are informed about the debate on such issues.

Prior to doing my present job, I was a professor of education. I cared passionately—and still do care—about the standards of literacy and numeracy that are achieved in primary schools. I care much more about that than I do about class size. Class size might be a contributory factor and it is in the popular category to which the member referred, but all that we should really care about is results, which are what affect life chances. If a child can read at seven, it will affect their earnings at 37. The debate around some of those issues will improve if a system such as the one I described can be put in place.

I will return to the theme of efficient government. Your presentation is all about the delivery of outcomes. How does one build efficiency into that approach? How does one ensure that the outcomes are achieved with the minimum of input?

That is the most important question to be asked in relation to the development of government systems over the next five years. That said, I am not yet sure what the full answer is, although I will give my speculative thoughts on the matter.

In the previous Parliament, the delivery unit made two assumptions about the defined targets or goals. They might seem bizarre, but they were that every target could be achieved—or at least we should start off with that mindset—and that there was enough money to achieve them all. We brook no debate on those things; if anyone wants to debate them, they must go to the Treasury to do so.

Members can see the curve on the graph that shows total managed expenditure. If it were to show a forward projection of public finances, the question is whether, as people's expectations rise, their impatience levels would rise. People come to expect ever-improving public services, but will that be possible without the rate of increase in investment that we have made over the past few years? The question is a fundamental one. The answer to it will require the delivery unit to connect with the work that is being done to follow through the Gershon efficiency review in a much more integral way than we have done in the past year. The job is one for the delivery unit, working with colleagues in the Treasury and the Office of Government Commerce, which is the part of the Government machine that is responsible for the efficiency review. A new Cabinet secretary will shortly be appointed and a key job for that person will be to think through how the connections are to be made.

If I speculate a bit further forward, we will also have to get into the academic debate about public value—what it is and how it is measured. The academic literature is interesting, but most of it does not go to the level at which one thinks, "That is how I would do that in Government." We must take the leap from the public value literature to getting some definition. Generally, public value is about more than delivering targets; it is about delivering public and customer satisfaction and thinking through the capacity of services.

I am sure that, as in England, many of the reforms in Scotland involve investing in the long-term capacity of services, such as improving the supply of teachers or transforming the school buildings stock. Those reforms do not affect next year's targets, but they are hugely important for public value, because they will give value to public services 10 to 15 years from now. We must find ways of bringing that information into the equation, otherwise we will get odd debates, such as the one about delivery on waiting times versus the £100 billion that we are spending on health over the next few years. People say that the spending is not productive, but the debate is a false one, because we are not measuring all the outputs that would be captured by public value.

I am keen to follow up on that exceedingly interesting point. In the private sector or commercial world, there is a wonderful expedient to address those issues—it is called the balance sheet. Assets are posted to the balance sheet, where their value can be seen. In fact, if things turn out particularly well, assets on the balance sheet can be revalued. Has any thought been given to how incremental investment in the ability to deliver can be fairly, reasonably and consistently reported?

As you say, in the private sector, shareholder value is a way of measuring a company's current performance and long-term prospects—that is relatively straightforward to achieve. However, with public value, there are all the dilemmas that I described in answer to the previous question. You asked whether there is any progress on measuring public value. There is early thinking about how to do that, but we are not as far on as we need to be. To speculate, my view is that we should measure five aspects of public services. The first is whether they deliver outcomes in the short and medium term. The second is whether they are developing capacity for the long-term health of the service—for example, whether they have the correct people, buildings, skills and leadership. Thirdly, we should measure what the public, particularly the users of the service, think about it. Fourthly, we need to measure the implementation of reform and ask whether the system design is improving—we need a way of checking whether reform of the system is working. Finally, we need to measure the amount of money that is being spent to get the improvements. The first four are measures of outcomes of different kinds and the last one is a measure of an input. Somewhere in there is a mathematical equation that would allow us to work out productivity in a much more refined way than we do at present, but I have not quite got that done yet.

International think-tanks are beginning to consider international competitiveness and to drill down into various areas. I presume that if the international arbiters push us further up the league table for school infrastructure, which is a key component of competitiveness, we will have a result to show.

That is absolutely right. International comparisons will become much more important in public services during the next decade. At present, we cannot have a debate about the economy without international comparison—a debate about inflation or unemployment is not serious unless it involves international comparison. Over the next 10 years, the same will happen with public services. Such comparisons are needed at the macro level—we need to consider how our education system compares to that in Finland or the United States—and also at the detailed level. I cannot remember where it was published, but I recently read a study that compared urban transit systems and the amount of time that drivers spend on task. International comparisons at the macro and micro levels will become incredibly important in the next decade. A lot of learning, which we discussed earlier, will come out of that.

We have exhausted our questions. I thank Professor Barber for his interesting presentation and his responses, which give us food for thought. The suggestions could be pushed into our changing to deliver system, which seems to be too broad. In a sense, you are pointing us clearly towards a focused and data-driven approach, which accords with the committee's prejudices.

I repeat that we want to make a submission in June to the Public Administration Select Committee's inquiry, given that we have taken evidence on the issue from a variety of sources. Some of the issues that we have discussed will feed into the efficient government approach and our budgetary scrutiny for future years.