Select Committee on Treasury Written Evidence

Memorandum submitted by Colin Talbot, Manchester Business School


  The Pre-Budget Report 2006 claims that the government is on track to meet its Gershon efficiency target of £21.5 billion in savings and has so far accrued £13.3 billion. In a written answer the Chief Secretary moreover claimed specifically that £5.5 billion had come from procurement savings, £2.5 billion from productive time and £1.5 billion from policy, funding and regulation. The government also announced a target of 3% pa efficiency savings across central and local government for the period of the 2007 CSR (up from 2.5% pa) and a target for savings from administrative budgets of 5%.

  These are big claims and ambitious targets which deserve serious scrutiny. The NAO is due to issue its 2nd Report on the efficiency programme early next year, so we will have to wait until then to see if the problems with measuring efficiency gains have, in the NAO's view, been overcome since their 1st Report in February 2006.

  The PBR 2006 however offers surprisingly little evidence for the claimed £13.3 billion savings, other than a few examples. The Committee might wish to consider asking for the six monthly reports of the OGC Chief Executive to the Prime Minister and Chancellor to be made public as a way of getting at the data behind these claims.[4] There seems little reason for these reports to remain secret and every reason to see how the government is making its calculations about efficiency savings. They must have arrived at these totals somehow and even if the reports cannot be made public the OGC and HMT can certainly make public how they arrived at the claims made in the PBR and by the Chief Secretary.

  This note seeks to clarify some of the problematic issues surrounding the government's efficiency agenda. It seeks to ensure firstly that these issues are discussed using the same terminology and definitions, which otherwise can lead to great a great deal of confusion and misinformation. And secondly it tries to identify what are the key challenges in achieving the substance and measuring the reality of efficiency savings in government.


  It may appear to be somewhat over-academic to place great emphasis on getting the terminology right in discussions of efficiency but it is actually an immensely practical issue: one person's "efficiency gain" can easily be someone else's "cuts" unless we can at least agree on terms.

  Back in 2001 HM Treasury, together with Cabinet Office, NAO, Audit Commission and ONS issued a small booklet called "Choosing the right FABRIC—A Framework for Performance Information" (Treasury . One of the main purposes of this publication was to clarify terminology so that some of the confusion in discussions about performance, delivery and efficiency could be minimised. Unfortunately the Treasury seems to have forgotten this agreed set of terminology in reporting on the efficiency programme.

  FABRIC made an important and useful distinction between "efficiency", which it defined simply as the ratio of inputs to outputs, and "cost-effectiveness" or "value for money" (VFM), which it defined as the relationship between resources consumed (inputs) and outcomes.

  These definitions are important for several reasons:

    —  They help to clarify the difference between narrow efficiency (input:output ratios) and VFM which are clearly highly important in policy-terms. It is quite possible to have activities which are efficient, but not cost-effective and vice-versa.

    —  This also helps to clarify that some so-called "efficiency" savings are nothing of the sort. For example, proving that the cost of inputs has been reduced (eg in purchasing and supply) is merely an "economy" unless it can also be shown that these reductions have not led to a reduction in the quantity or quality of outputs. Indeed it is perfectly feasible to have economies in purchasing lead to reductions in efficiency (due, for example, to higher failure rates due to inferior supplies being used) and also of VFM (due to, for example, adverse outcomes caused by lower quality inputs). Given that 37% of the Gershon target is due to come from procurement these are vital distinctions to maintain.

  Throughout much of the government documentation on the efficiency programme the concepts of efficiency, economy and VFM are regularly confused.

  This matters because opponents of the government and critics of the efficiency programme, such as the trade unions, will seize on these inconsistencies to argue (rightly or wrongly) that the efficiency measures are actually leading to reductions in services or outcomes. This debate—which is a very important one and should be taken seriously—will be meaningless and sterile unless common definitions are consistently applied. Moreover government itself will not know what is real and what is "smoke and mirrors" if it doesn't stick to rigorous definitions (and measurement procedures).


  For the purposes of this note the Gershon targets split by "work streams" (Computer and Auditor General 2006 p 4) are the most useful.

  This breakdown is important because the different work streams have very different dynamics in achieving and measuring successful efficiency improvements.

  A basic issue here is how labour or capital intensive the areas of work are. As a general rule, it is much easier to achieve substantial net efficiency and productivity gains in areas of higher capital intensity than in those of greater labour intensity—although of course the initial capital costs are much higher.

Table 1


(£ billion)

of target
Actual as of December 2006[5]
Percentage of target met
Procurement7.9637% 5.570%
Productive time5.1624% 2.449%
Policy, funding and regulation3.66 17%1.541%
Corporate services1.51 7%

  Thus it can be assumed that those areas of more capital intensive activity—procurement, corporate services and transactions—are going to be in principle areas where efficiency gains are easier to achieve—with the right investment—than in the area of productive working. This seems to be at least partly born out by the partial figures we have which suggest that 70% of the procurement target has been met as against only 49% of the productive time target.

  The productive time efficiency gains are mainly targeted at very labour intensive public services. Education is expected to produce £1.3 billion and health £3.25 billion—which together comprise 88% of the total productive working target.

  These are areas where technical solutions have limited scope for improving individual productivity. Indeed, the recent emphasis on:

    —  successful outcomes in education and health (and some other fields);

    —  on the need for much more "co-production"—ie actively engaging "users" in the production of services and outcomes (eg in health promotion); and

    —  and addressing so-called "wicked issues" across organisational boundaries;

can all demand the increased use of labour power—for example longer consultations with GPs; more counselling type activities; more intensive pupil engagement; etc These may all increase VFM whilst appearing initially to reduce narrow efficiency.

  The crucial point here is that a "one size fits all" efficiency agenda—including an across the board 3% pa target across al departments and areas of activity—is simply unrealistic. Some areas can make much greater efficiency gains—with the right investment—whilst others will struggle to achieve anything like a 3% pa annual gain and may even appear to drop (more on this below).

  Although efficiency in the public sector is clearly different in many respects from productivity growth in the economy as a whole it is interesting to compare the expected growth in efficiency with expected productivity growth. Most importantly, the public sector as a whole is much more labour intensive than the UK economy as a whole. It would therefore probably unrealistic to expect similar gains in productivity or efficiency in the two—the public sector would almost certainly lag the private sector.

  According to the PBR 2006 trend growth in productivity in the UK economy is running at 2.4% pa (in the period 2001Q3-2006Q3, p 39). Productivity (output per worker) is not the same as efficiency (input:output ratios), in fact it is a subset of efficiency. But especially in relation to the productive working targets it is fairly close.

  The Committee may therefore wish to ask HMT to explain why it is reaslistic to expect a growth rate of 3% pa in public sector efficiency in the context of growth in national productivity of 2.4% pa?


  The NAO's first report on the efficiency programme identified major concerns about the measurement of reported efficiency gains. These are reported in generic terms, but the issues of differentiated efficiency work programmes raised above also apply to measurement issues.

  Put simply, there are some areas where at least some aspects of measurement are much easier than others. For example, purchasing areas are likely to have relatively good long-term data about the costs of various inputs. This is unlikely to be true of the measurement of working practices, where data is unlikely to be available from before the various efficiency initiatives started.

  However, in all cases the biggest problem of all is measuring outputs accurately and especially the quality of outputs, an issue which the Atkinson Review and the ONS have been struggling with for some time. Although progress has been made, no-one would claim that this problem has been "solved" and there are still both technical and political disputes surrounding the measurement of outputs. Moreover the approaches developed by Atkinson and the ONS are designed for macro level analysis of outputs in relation to whole sectors (eg education) and are not fine-grained enough to address more micro-level issues of efficiency.

  Without such agreed metrics measuring any efficiency gains is going to be at best problematic and probably impossible in most cases. It is very difficult to see, for example, how the reported efficiency gains of £2.4 billion from productive working—which must have come mainly from health and education—can be squared with existing ONS estimates of changes to productivity in these sectors which appear to be flat-lined or reducing, depending on the assumptions used.

  This problem is compounded by the weakness of allocation systems—that is allocating the cost of inputs against specific outputs, even where these can be adequately measured. The NAO has rightly pointed out that the systems needed for this sort of analysis (activity costing systems) are relatively scarce in government. Even when they are put in place, back-dating them to create base-line data for analysing efficiency gains is often not possible or too expensive.

  In the area of productive working all of these problems apply—weak or non-existent and problematic measurement of inputs; serious allocation issues (which work contributes to which outputs); and finally problematic measurement of outputs.

  It should be noted this only deals with the main difficulties in measuring narrow efficiency. VFM measurement is even more problematic. And as noted previously, narrow efficiency gains can have detrimental effects on VFM. For example, when some NHS hospitals began using a particular artificial hip joint the initial analysis of narrow efficiency was positive—cheaper inputs for the same quantity and quality of output. It was only several years later that faults in these hip joints emerged which led to severe losses in VFM as the outcomes for the patients deteriorated badly.

Gaming and Unintended Consequences

  The issue of "gaming" in relation to all kinds of performance data have been explored in a number of scrutiny and research projects. Whilst this is often exaggerated, one thing we do know is that the more narrow and focussed any measurement is, the more scope there is for both gaming and unintended consequences.

  The Treasury and OGC seem to have now recognised these potential problems in relation to the efficiency programme and its relative autonomy from the "normal" business systems. The need for "mainstreaming" efficiency is currently more clearly recognised, but it is unclear how far the programme to date has produced gaming and/or unintended consequences. Not is it clear that the approaches of HMT, OGC or the NAO will expose such problems.

Realism, Efficiency and Public Value

  This note has been a plea for more realism in how much efficiency can be both achieved and measured in the public services. Unrealistic promises and claims of success have a high negative impact generating cynicism amongst staff, managers, users and the public. It does not suggest that efficiency gains are either impossible to achieve or even impossible to measure, but the difficulties of both need to be recognised.

  Also, a more nuanced and differentiated approach is needed—recognising the dynamics and characteristics of different areas of activity. As I pointed out to the Committee when the Gershon Report was first published in July 2004, there were in fact very different levels of relative efficiency savings expected from Whitehall departments (ranging from 1.2% to 17.4%).

  No explanation for these variations has ever been given and they may indeed relate to the realistic possibilities for efficiency gains in different areas of work. So despite the apparent "one size fits all" policy, in fact there has been a great deal of variation -what is not clear is whether these variations are due to realistic analysis of efficiency possibilities or some other reason. Indeed, the fact that there are very big differences between the two departments where the expected biggest gains from productive working practices are to be realised—6.9% in health but 12.4% in education—suggests there are clearly other unknown factors at work.

  The Committee may want to ascertain whether the new 3% pa target announced in the PBR is an across the board target or is as variable as the Gershon targets, and if so why?

  There is an increasing discussion in the UK of "public value" which seeks to marry concerns about efficiency and VFM with issues of trust and legitimacy in the public domain. Especially in relation to co-production in areas such as education, health and criminal justice trust and legitimacy are crucial ingredients to achieving real efficiency and VFM in producing outputs and outcomes. People will not work with organisations they don't trust.

  In terms of increasing the trust and legitimacy of public services, which is crucial to the value which the public attributes to public domain activities, the unintended impacts of unrealistic targets and claims can more than outweigh any genuine and objective efficiency gains. Paradoxically, rather than increasing public confidence in public services in the context of the expanded spending of the past few years, a lack of realism in the claims about efficiency gains can actually undermine public trust—perhaps a more serious negative outcome than any positive efficiency gains.

December 2006


Comptroller and Auditor General (2006). Progress in Improving Government Efficiency (HC802-I. London, National Audit Office.

HM Treasury (2001). Choosing the Right FABRIC—A Framework for Performance Information. London, HM Treasury.

4   The author has made an FOI request to OGC for these reports which as yet has neither been accepted nor declined. Back

5   Based on Mr Timms written answer on 7 December 2006, col 435. Back

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2007
Prepared 25 January 2007