Christopher pollitt a background paper for the world bank consultation exercise

Download 84.71 Kb.
Size84.71 Kb.




Institutional details
Christopher Pollitt

BOF/ZAP Research Professor of Public Management

Public Management Institute

Katholieke Universiteit Leuven

Parkstraat 45 (bus 3609)

B-3000 Leuven



Let me immediately confess that this paper was written for another purpose. It is very much an academic overview, not written with any prescriptive or advisory intent. What is more, it is focused on the developed world, not at all on developing countries. However, when Nick Manning invited us to participate in the World Bank’s consultation exercise it did occur to me that this might nevertheless make useful background reading. Some of the observations apply even more to many developing countries than to the developed world. There is also perhaps some corrective value in remembering that, far from public management reform having been routinely successful in the developed world but being beset with difficulties in the developing world, it appears frequently to have been problematic in both. Humility is undersung as a management reform virtue.
I have added, at the end, a few possible ‘lessons’ that might be drawn from this analysis.
Has there been a pattern?
The answer is - yes, there has been a discernable pattern to the public management reforms which have swept over so many of the European, North American and Australasian countries (not to speak of elsewhere) in the three decades since 1980. Six prominent elements of this pattern are set out below.
First, public management reform itself (hereafter PMR) has become a more prominent political topic. It has emerged from the dusty backrooms of administrative change to feature, sometimes at least, as a major agenda item in party manifestos and government programmes (Pollitt and Bouckaert, 2011, chapter 1). It has acquired its own national and international networks and communities, which embrace not only public servants themselves, academics and a few politicians, but also, with increasing voice and influence, management consultants (National Audit Office, 2006; Sahlin-Andersson and Engwall, 2002; Saint-Martin, 2005).
Second, the New Public Management (NPM) ideology of treating government as if it were a business has informed many changes in many countries (Pollitt and Bouckaert, 2011, passim). More recently NPM has been partly diluted and partly merged with other streams of reform ideas. It is not, as some have alleged, ‘dead’ (Dunleavy et al, 2006), but it has evolved (Hood and Peters, 2004). We will come to all that later.
Third, the widespread popularity of NPM ideas does not mean that the developed world has converged on the NPM paradigm. In fact the impact of NPM ideas has been highly variable, for reasons that we will enumerate. And other, rival models have gradually become discernable, although each seems as slippery and potentially inconsistent as NPM itself (Goldfinch and Wallas, 2010; Pollitt and Bouckaert, 2011, chapter 1).
Fourth, we should note one very common element. It is that reform talk is often much more ambitious than the formal decisions that eventually get taken, and that there is then frequently a substantial implementation gap between what was decided and what finally gets put into operational practice. This is captured in the now widely-used stages model of talk, decisions, practice and results (Pollitt, 2002; 2007a; Goldfinch and Wallis, 2010). This feature carries the strong implication that in our assessments we need to be careful to treat like with like: to compare talk with talk and practices with practices, and not confuse the different stages. The pattern of talk may well be a poor guide to the pattern of practice.
Fifth, evaluations of PMR have been ambiguous or unconvincing in many cases where they have been attempted. In even more cases - in fact probably in most cases - no credible evaluation (and certainly no independent, scientifically-designed evaluation) has been carried out (Pollitt, 1995; Pollitt, 2009; Pollitt and Bouckaert, 2011, Appendix B). PMR has frequently been more an act of faith than a piece of evidence-based policymaking. The recent explosion of international comparative indicators of good governance (and the like) has not fundamentally changed this state of affairs.
Sixth, instances of cycling or alternation in PMR are not uncommon (Pollitt, 2008, pp51-59). Decentralising measures are followed, after an interval, by centralizing measures. Intensive and detailed performance measurement is succeeded by a phase during which reformers say that measurement and audit must be lightened, and operational management freed from the unreasonable burden of monitoring. This is certainly not a universal pattern, but it occurs fairly frequently (and has recently been sanctified in management-speak as ‘re-balancing’). It appears to be connected with the long ago-observed fact that public administration faces a number of inherently difficult trade-offs, where opting for one direction brings benefits but also, inevitably penalties, so that, after a while, the penalties seem to loom large while the benefits become too familiar and taken for granted, so that a new generation of reformers turn back to the opposite direction (Hood, 1976; Hood and Jackson 1991; Pollitt and Bouckaert, 2011, chapter 7; Simon, 1946)
Deepening the answer – 1 – the growth of PMR
Why has PMR become such a prominent topic? Few writers have focused on this question, but we may hazard a few suggestions.
Almost every party with a hope of power that contests elections in Western European and North American states now feels obliged to say something about PMR. Each has to aspire to ‘run things better’. Historically speaking that is a recent development. In the 1950s and 50s party manifestos were not full of PMR, they were full of promises of substantive new programmes and polices. Management reforms (or ‘machinery of government’ changes as they were sometimes called) were usually regarded as rather dry, technical issues – not something either headline writers or most politicians needed to take much interest in. However, as a number of political scientists and sociologists have noted, since the 1970s politics in the West have become less ideological and substantive but more ‘technical’ and ‘presentational’. The welfare state has been built, and now the job is restraining its unaffordable further growth, or even managing a gentle decline. Public management has become not simply an adjunct to other policies, but a policy in itself (Barzelay, 2001). How things are to be organized becomes almost as important as what things are to be done. And how things are to be presented is also crucial – policies can no longer be expected to speak for themselves, they must be designed and presented so as to catch the current public mood, as revealed by professional polling and focus groups, and channeled and amplified by the mass media:

‘One of the key implications of this emphasis on government as technique is to contest those models of government that wish to view it solely – or even mainly – as a manifestation of values, ideologies, worldviews, etc.’ (Dean, 1999, p31)

Management, of course, is identified with both dimensions of technique – first, with how to squeeze every last drop of output from every dollar spent and, second, with how to market policy ideas and decisions in ways that will maximize acquiescence and minimize opposition.
In addition, power is draining away from the nation state – upwards, to international networks, and in some cases downwards, to regional or local bodies (though this should be understood in a nuanced way, not as a zero sum game – e.g. Held, 2004). So for national politicians ‘technical politics’ means projecting the idea that you can do more with less; that you can cut out waste and increase efficiency. Hence, during the 1980s, the huge explosion, on both sides of the Atlantic, of concern with ‘management’ in general, and ‘efficiency’ in particular (Pollitt, 1990). In a book which celebrates the coming of age of public management as a policy question, Barzelay wrote that:

‘beginning in the 1970s, the potential for policy change in the area of public management increased, as economies suffered stagflation and public perceptions of bureaucracy became more negative’ (Barzelay, 2001, p1)

Since that time we have seen a new community – one might almost say industry – come into being. It is a cross-sectoral, international network of public management ‘experts’, people based in governments or agencies or consultancies or occasionally in academia, and who more and more frequently move between these different roles. They swap ideas, compete for contracts and attend conferences together. They draw on an extensively common and rapidly changing stock of generic management ideas (Sahlin-Andersson and Engwall, 2002). They feed on a continuous flow of lucrative contracts from governments (National Audit Office, 2006). More and more governments have set up one or more specialist management reform units, such as the Prime Minister’s Public Service Delivery Unit (UK), the Public Management Department of the Finnish Ministry of Finance, the French Directorate General for State Modernization, the Norwegian Ministry of Government Administration and Reform, and so on. Members of these organizations may themselves have consultancy experience or they may become consultants afterwards, trading on their experience gained near the heart of government reforms (for multiple examples of this interpenetration, see Barber, 2007). More profoundly, these units and departments help to institutionalize ‘modernization’ and ‘reform’, continuously putting forward programmes and targets, drawing attention to new management ideas and techniques and generally keeping the rest of central government ‘on its toes’ (for a vivid account of how intrusive this can become, see again Barber, 2007). Such international networks are, of course, immeasurably facilitated by the advent of the Internet. Nothing like this existed in the 1950s or even ‘60s.
All this has constituted a kind of ‘multiplier’ effect. A community has grown up in whose interests it is to create new ideas and techniques, and therefore further reform. There is nothing necessarily sinister about this, even if it can often be construed as a form of self-interest. It is simply that more and more people take up public sector roles after some training in ‘management’, and more and more consultancies depend on winning and subsequently sustaining contracts to facilitate reform (Saint Martin, 2005). For example, the UK public sector spent approximately ₤2.8 billion on consultants in 2005-6, a 33% increase on what the level had been only two years previously – in fact central government spent more on consultants per employee than did comparator private sector firms! (National Audit Office, 2006, pp5 and 15).
Deepening the answer – 2 – the business of government
The language of public administration has changed enormously over the past three decades. One main trend – at least in Anglophone countries – has been towards the widespread adoption of titles, terms and concepts from the world of private sector business. Thus departments and agencies which wish to do something now have to present a ‘business case’ for it; numerous senior public service positions are now designated as ‘chief executive’; organizations are expected to ‘benchmark’ and ‘business process re-engineer’; civil service training curricula are saturated with sessions on ‘leadership’ and ‘innovation’ (Pollitt and Op de Beeck, 2010). One peak in this trend came in 1997, when the then US Vice President, charged with a major reform of the federal civil service, published a glossy, cartoon-festooned booklet entitled Businesslike government: lessons learned from America’s best companies (Gore, 1997). It is easy to forget how foreign this kind of terminology would have sounded 40 years ago, when the public service had its own terms for most of these roles and activities. ‘Citizens and businesses expect the same levels of access and personalization from public services as they receive from leading private sector organizations such as Amazon and Tesco. They expect to be able to access information from multiple locations and in ways that suit them rather than the providers’ (H.M.Government, 2010, p9).
This love affair with a ‘businesslike approach’ may, or may not, have come to an end with the business-induced global financial crisis of 2008-9 (it is too early to be sure – and we should note that the UK government’s praise for supermarket management - above - post dates the crisis). Yet even if it has been badly dented, this way of thinking has now penetrated so many public sector organizations, and has been promoted to an entire generation of public service managers, so its effects will surely not disappear overnight. Furthermore, it is not only that business thinking has been ingrained in minds, business practices have also been embedded in standard operating procedures, including budgeting and accountancy rules (Newberry and Pallot, 2005) and personnel regulations:

‘In almost all [EU] Member States the differences between civil servants and other public employees and civil servants and private sector employees have become fewer. Still, a few member states stick to the traditional patterns and maintain at least some traditional features. Other Member States have almost entirely aligned the working conditions within the public service and between the public- and private sector’ (Demmke and Moilanen, 2010, p139)

Another development that has gathered pace over the past quarter century has been the evolution of critical, large-scale public IT systems which are essentially supplied, maintained and run by for-profit transnational corporations:

‘[G]overnments no longer run their own IT functions. To get their systems built, developed, and managed they increasingly rely on the global IT industry, specifically the giant ‘systems integrator’ companies such as Electronic Data systems (EDS), IBM-Accenture, Cap-Gemini-Ernst and Young, Lockheed Martin and their like’ (Dunleavy et al, 2006a, p1)

In short, business-like practices are now embedded in the warp and weft of many (but not all) public services in many (but not all) countries. Whether this is something to be celebrated, cautiously welcomed, or deeply regretted remains an on-going debate.
One further aspect of government-as-business thinking should also be mentioned. Just as public servants have been enjoined to think of themselves as service providers, working to a business plan, so governments have frequently redefined citizens who use public services as consumers. Again, this has been controversial, at least in academia, and many words have been spent delineating the differences between ‘citizens’, ‘clients’, and ‘consumers’. One research project showed how, in the UK in the decade after New Labour came to power in 1997, the government had advanced the concept of a ‘critical citizen-consumer’ as a key figure in a number of its sectoral policies. Evidence that such a paragon actually existed - at least in large numbers - was not, however, forthcoming. Surveys seem to indicate that the public can be quite discriminating between different circumstances and contexts, and do not particularly want to be treated as actively-choosing, information seeking rational decision-makers in many of their interactions with public authorities (Clarke et al, 2007).
Deepening the answer – 3 – limited convergence on NPM
Some influential voices have claimed that there has been convergence, and that that convergence has been towards the NPM model. Here are just three examples of that – the first a leading American professor, the second an equally well-known Australian and the third a more recent academic overview:

‘The movement has been so striking because of the number of nations that have taken up the reform agenda in such a short time and because of how similar their basic strategies have been’ (Kettl, 2005, p1)

‘There are various ideas of what is involved in public management reforms. However, as the process has continued there has been convergence as to what is involved in the reforms’ (Hughes, 2003, p51)
‘no doubt NPM represented a global trend that affected public sector decision making worldwide’ (Gualmini, 2008, p75)
This is an interpretation from which I beg to differ. I can understand why, looking at the surface policy commitments in certain countries at certain times, observers might arrive at the conclusion that convergence was taking place. But I would argue that either a wider (geographic) and/or a deeper (implementation-focused) analysis would lead in a very different direction.
There are at least three reasons why many governments chose to resist or only reluctantly or selectively embrace the trend to NPM. We could call these ‘reach’ ‘grip’ and ‘motivation’. Some national governments have limited reach, in the sense that their constitutional and structural position does not permit them to carry out wide-ranging reforms right across the public sector. They cannot ‘do’ blanket NPM even if they want to. There are powerful sub-national authorities with their own powers and ideas, and it is unlikely that, at any given moment, all these will be of the same ideological orientation or policy preference as their central governments. Germany and the USA are prominent examples (and, indeed, in such countries it may well be local governments that are the leading standard-bearers for PMR). So in these countries, unlike, say France or the UK, it is hard for a central government, however enthusiastic, to transform the whole of the public sector. In fact, internationally, comprehensive, synchronized public sector reform is rare almost to vanishing point.
Governments also vary in terms of their grip – that is how firmly and effectively they are able to transmit their demands and instructions down their own vertical hierarchies. In some cases transient or minority governments face powerful public sector unions or strong professional groupings, and are unable to carry through radical change. Some recent Belgian federal governments would fall into this category, as would some Nordic coalition governments. In other cases - such as the Thatcher and Blair administrations in the UK - a single party government with a substantial majority and disciplined party support lasts for many years and is able to face down union and professional resistance.
Finally, there is the question of motivation. Some national governments have been effusively enthusiastic about NPM ideas, others have been cautiously welcoming and others still have actively resisted them. Thus one could number the Reagan and G.W.Bush presidencies, and the Thatcher premiership, as strongly pro-NPM, while many of the Nordic governments since 1990 have been cautiously and selectively welcoming of some NPM instruments - for example by accepting performance management, but resisting the widespread application of market mechanisms within basic public services (Pollitt and Bouckaert, 2011). At the other end of the scale the Mitterand Presidency in France and the EU Commission under Santer were regimes that were distinctly critical of ‘Anglo-Saxon’ ideas. Subsequently the Kinnock Reforms (EU Commission) and the Sarkozy administration in France proved somewhat more receptive to certain NPM concepts, but in each case the ‘conversion’ was both muted and heavily ‘translated’ into the local institutional circumstances (Bezes, 2010; Pollitt and Bouckaert, 2011).
If we take all three reasons together - reach, grip and motivation - one can see that New Zealand 1984-93 and the UK 1979-2010 were rather exceptional. They were among a quite small group of countries where reach, grip and motivation all permitted, or encouraged, a general movement in an NPM, business-like direction. They were not, as some commentators believed, prototypes of a wave that would sweep across the globe. They were certainly the origins of a wave, but one which encountered a series of very different sea shelf and coastal formations, so that in some places it came ashore like a tsunami while in others it was reduced to a mere murmur, and hardly noticed, or was even excluded by cultural and institutional seawalls. One can hardly call NPM a global phenomenon when, for many years, it had such limited effects on central governments in states as important as France, Germany and Japan. Yet in 1992 Osborne and Gaebler had famously written that:

‘If the rise of entrepreneurial government is an inevitable shift rather than a temporary fad…one would expect it to happen in other nations [than the USA] as well. And to a startling degree, it has. A similar process of transformation is underway throughout the developed world’ (Osborne and Gaebler, 1992, p328)

Since then, whatever plausibility this theory of converging transformation may once have had has evaporated. We have accumulated a small mountain of scholarly studies showing how NPM has been adapted, translated, diluted, superseded or downright rejected by a large number of countries (e.g. Barzelay and Gallego, 2010a, b; Bezes, 2010; Borki and Van Berkel, 2007; Christensen, 2009, Christensen and Lægreid, 2001; Goldfinch and Wallis, 2010; Kickert, 1997; Kickert 2008; Kuhlmann, 2010; Ongaro, 2009; Pollitt, 2006, 2007a; Schroter, 2010; Smullen, 2010). A major recent study of civil service reforms in the 27 EU states arrived at the conclusion that:

‘different historical traditions and cultures as well as HR systems have a considerable impact on public management moderisation paths and on the outcomes of HR reforms’ (Demmke and Moilanen, 2010, p243)

NPM, of course, is no longer ‘new’. Indeed, it has been referred to as ‘middle aged’ (Hood and Peters, 2004) or even dead’ (Dunleavy et al, 2006b). Just as some earlier writers hailed a convergence towards NPM, some more recent commentators have proclaimed a ‘progression of paradigms’ with traditional bureaucracy leading to an NPM phase, which in turn gave way to a post NPM phase variously characterized as networks, joined-up government and, most pervasively, ‘governance’ (Pollitt and Hupe, 2011). ‘The need for effective governance of…interactions, and therefore for new forms of public management such as network steering will increase…’ (Kickert et al, 1997, p190). Or ‘it has been argued here that the NPG [New Public Governance] has become the dominant regime of public policy implementation and public services delivery…’ (Osborne et al, 2010, p414).
However, it seems that ‘the problems which exist with persistent arguments for a NPM convergence also exist for the successor “post-NPM” myth’ (Goldfinch and Wallis, 2010, p1100). Once more, certain trends - real enough in certain periods in certain countries - have been amplified, simplified and generalized so that they can be presented as a common international direction. Once again, the precise empirical domain and technical content of these trends is often left vague while large claims are nevertheless advanced for their relevance (for early and more recent examples of this approach, see, respectively Kooiman, 1993 and Osborne, 2010). Until we can clearly distinguish what is governance and what is not, and until we can count the number and density of networks in different countries and sectors, and how these have changed over time, we should exercise caution in our claims for convergence. Even more challenging, it would be desirable to be able to link ‘governance’ and ‘networks’ to outcomes, and to establish whether and why these approaches produced better results than traditional bureaucracy or NPM. Very little of the ‘governance’ literature seems to meet these criteria. Much of it polarizes between high level, abstract, rather disembedded process models and very local and specific case studies. There is a big gap in the middle.
Deepening the answer - 4 - talk, decisions, practices, results
Those who conduct empirical field research into public management reforms quite often find that ‘the reform’ (whatever the particular example may be) changes steadily the further one moves from the original debate to the government white paper or ministerial speech, and then through local implementation and on to the final results. It is challenging for governments to keep a tight hold on reforms as they percolate out from the ministry to the further reaches of operational units and sub national levels of government. Indeed, it is not unusual for reforms simply to run out of steam and fail to reach the frontline at all. Back in 1994 Savoie published a widely-read comparison of the Thatcher, Reagan and Mulroney administrations, in which he pointed out that many of the Mulroney government’s public management polices had failed to carry through implementation to achieve their proclaimed objectives (Savoie, 1994). Very common, also, is the situation in which a new initiative makes a brief impact but, after a few years, all trace of it has disappeared. A few years ago colleagues and I rang round European public service quality initiatives that had been presented at international quality conferences two and four years previously. In an uncomfortably high proportion of cases our respondents said they knew nothing about the initiative we were talking about, or that it had finished now.
Table 1 helps us to understand why the spread and impacts of NPM (or any other fashionable model) may sometimes be exaggerated. Basically, it is quicker and easier to research the headlines of talk and decision than to go out into the field and look in detail at operational practices and final outcomes. Thus, for example, a quick survey of official documentation shows that executive agencies in the UK, Sweden, Finland and the Netherlands all have performance indicator systems. This could be seen as an example of convergence, with a strong NPM flavour (performance measurement and results-oriented management are central planks in the NPM model). However, detailed fieldwork reveals, that these indicators are used in very different ways and with different consequences in the four countries (Pollitt et al, 2004). Similarly the Guamini quotation at the start of the section on convergence (above) is based on a survey of government legislation, so it refers only to our ‘decision’ stage, not to practice or results. Before leaving this point we should note that the Talk-Decision-Practice-Results framework has several implications for comparative analysis. Inter alia it suggests that we should try to compare like with like (decisions with decisions, or results with results). Comparing (say) talk and decisions in country A with practice in country B is potentially misleading (and unfair).





More and more people are talking and writing about a particular idea (e.g. contracting out)

Quick and cheap. Monitoring what people are talking and writing about is fairly straightforward.


The authorities (governments, public boards etc) publicly decide to adopt a particular reform

Again, quick and cheap. The public decisions of the authorities can usually be located quite quickly (on the Net, often without leaving one’s desk)


Public sector organizations incorporate the reform into their daily operational practices

Probably requires expensive and time-consuming fieldwork. This needs both funding and access.


The results (outcomes) of the activities of public agencies change as a result of the reform

Final outcomes are frequently difficult (and expensive) to measure. Even more frequently there is an attribution problem, i.e. one cannot be sure how much of the measured change in outcomes can be attributed to the reform itself, as opposed to other factors.


(After Pollitt, 2002)

Deepening the answer - 5 - faith and evidence
In 1995 I published an article which argued that in many cases no serious evaluation had been attempted of many of the main NPM reforms and, further, that such evaluations as had been instituted had been unable to reach convincing conclusions on such basic issues as efficiency gains and changes in effectiveness (Pollitt, 1995). The title was ‘Justification by faith rather than works?’ Given that at the time I thought I was doing little more than stating the obvious and backing it up with an analysis of a few well-known studies, I was surprised at how widely the article was subsequently cited. At the time of writing, in mid 2011, a major EU Framework Project is investigating the impacts of NPM reforms across the EU ( The project is far from complete, but early findings based on a meta analysis of academic and practitioner documentation from most EU states is that the number of studies which convincingly connect data on changes in efficiency or effectiveness to specific management reforms is modest indeed (Pollitt and Dan, 2011). Evidently the 16 years in between the 1995 study and COCOPS in 2011 have seen relatively little change in this particular respect.
There are good reasons why convincing evaluations are so thin on the ground. These are both technical and motivational (Pollitt, 2009). Technically, reforms are often composed of many different, if related, interventions, and may well take place at the same time as other changes, so sorting out exactly which intervention or combination of interventions produced exactly which results can be a formidable undertaking. Furthermore, results often develop over time, so that an evaluation conducted at reform plus one year may yield a different picture from a later evaluation at reform plus three years. For example, a downsizing and refocusing reform may yield poor results after twelve months, when staff are still worried and resentful, and the positive results of the refocusing have yet to come to fruition. Given more time, however, staff will become accustomed to the new regime, and will have learned and honed the new procedures, giving a much better set of results overall. Ideally, also, an evaluation would commence well before the reform, so as to collect a full set of data on the ex ante situation, for later comparison with the ex post. In practice this is seldom achieved. Another desideratum of classic evaluation - a control group - is even rarer (although occasionally one may find ‘natural experiments’, such as when performance measurement regimes diverge in England, Wales and Scotland - Bevan, 2010).
Motivationally, politicians are not necessarily comfortable with the elongated timescales involved in most large scale administrative reforms. They may hope to gain kudos from the original reform announcement and launch, but how many of them expect to be around in, say, five years time when the full range of impacts will be much clearer? Further, if they are experienced in such matters they may well suspect that the results will include unexpected as well as expected effects, and that the overall impact will turn out to be less pure and less radical than the promises made at the time of the reform launch. The motivation to invest in systematic long term evaluations is therefore less than overwhelming. At the same time both politicians and management consultants are members of professions that must always have something new to say (Goldfinch and Wallis, 2010, pp1111-1112). Several scholars have traced how past failures can be forgotten in the bright new dawn of the next round of reform (Brunsson, 2006; Sundstrom, 2006). Others have shown how, in those countries where the barriers to reform are relatively low, politically-driven reforms can arrive helter-skelter, with little chance to learn much in depth from the previous reform before it is replaced by the latest idea (Pollitt, 2007b).
None of this is to say that experience and evidence is wholly ignored. The relative absence of high quality formal evaluations is not the same thing as an absence of learning. Learning does go on, but seldom in textbook fashion. Programmes deemed, however accurately or inaccurately, to have been failures become the things to be avoided in the next round of reform. Trial and error does seem to be a common learning mode, even if the trials are political trials and the errors are often media-amplified dramas.
Neither is the sparsity of formal evaluations unique to NPM-type reforms. If we look at networking, partnership and governance-type reforms we find a similar story. To begin with, definitions of networks and governance are multiple and slippery and methods for counting and measuring networks and network interactions are either primitive or wholly absent (Pollitt and Hupe, 2011). Only a few papers of the many papers using these concepts manage to conduct what might be regarded as an orthodox scientific analysis of outcomes (e.g. Meier and O’Toole, 2001). Studies of the outcomes of network governance are particularly thin on the ground. Apart from the famous Meier and O’Toole studies of Texas school districts (e.g. Meier and O’Toole, 2010), even the few worthy attempts that have been made contain measures of ‘outcomes’ which are both subjectively perceptual and heavily laden with process-like elements (such as ‘involvement of actors’ and whether innovative ideas are developed during the project - Klijn et al, 2010).
One kind of evaluation that has blossomed since the beginning of the 21st century is international comparisons of general aspects of government, such as ‘good governance’, ‘bureaucratic quality’, ‘transparency’ and so on (Pollitt, 2011). Comparing general features of governance in different states is complex and difficult. That has not stopped analysts producing large numbers of purportedly comparative indices of various aspects of government and governance (e.g. Accenture, 2009; Advisory Group on Reform of the Australian Government Administration, 2009; OECD, 2009). The World Bank’s ‘World Governance Indicators’ (WGIs) have probably been the best-known measures (Arndt, 2008), but one source lists more than 400 indicators (IADB, 2007). The UK has arguably the world’s most serially-reformed public sector, and it was notable that a 2008 Cabinet Office report gave great prominence to international comparisons, including a preface from the then Prime Minister in which he said he wanted ‘every element of our public services to be the best in the world’ (Cabinet Office, 2008, p7). Some well-informed commentators see this growth in international comparison as an almost irresistible force or, at the very least, one unlikely to be stemmed in the near future (Hood et al, 2008).
Unfortunately, however, the popularity of these league tables has made only a modest contribution towards helping us understand how effective particular types of management reform have been. This is because the measures tend to be highly aggregated, and these aggregate ‘results’ cannot usually be traced back to specific reforms. Sometimes, indeed, the aggregate is itself a somewhat incoherent jumble of inputs, processes, outputs, and outcomes, the overall meaning of which is very hard to discern (Pollitt, 2010).
Deepening the answer - 6 - cycling and alternation
In a long term study of administrative reforms in the US federal government, Light noted a strong tendency for oscillations and alternations (‘tides’) around certain basic trade-off issues (Light, 1997).
Similar cycling can be seen the history of UK reform. In 1988 the Thatcher government launched a major restructuring programme to carve out large blocks of operational activity and place them in executive agencies that would be ‘at arm’s length’ from ministerial departments (Oliver, 2003; Pollitt et al, 2004). During the ensuing decade more than 130 new agencies were set up, and at the peak they collectively employed more than 70% of UK civil servants. In 2002 an official review by the Blair government concluded that this multiplication of agencies had brought improvements, but that it had increased the distance between ministries and operational management and led to a loss of policy co-ordination (Office of Public Services Reform, 2002). A number of agencies were then merged with each other or drawn back into departments or at least subject to tighter ministerial control (Talbot and Johnson, 2007). The 1980s impulse to decentralization had been replaced with a desire for more central control and co-ordination.
Within the 13 year life span of the Blair/Brown New Labour administration (1997-2010) central government’s approach to performance measurement swung from moderate to intense and back again, so that by the end the government was officially distancing itself from its former period of intense central targetry, and proclaiming the need for professional service deliverers to have greater freedom to take their own decisions and set their own targets (Cabinet Office, 2008). The coalition government that succeeded them took this rhetoric further, denouncing their predecessor’s performance measurement regime as inhibiting of local innovation, and declaring a ‘bonfire of targets’. Soon, however, new frameworks of ‘milestones’ and ‘standards’ appeared, and operational managers could be forgiven for thinking that in practice they still had a great deal of upwards performance reporting to do.
Whilst the metaphor of cycling does serve well to describe certain movements in public management reform, it should not be interpreted too literally. When a system cycles between centralisation and decentralization that does not mean that it returns to exactly the same point that it had reached some years before. If there is an alternation between an urgent priority for increased efficiency, and then, later, an equally strong emphasis on accountability and transparency, that does not mean that in the later part of the alternation efficiency sinks back to where it was before the efficiency drive. Thus, for example, the drive for efficiency in phase one may well raise efficiency, but it does not have to fall back (or certainly not the whole way back) during the succeeding phase two focus on accountability and transparency. Thus, over time, even if attention swings between two (or more) prime goals, the levels to which they are achieved continues to increase over time (Pollitt, 2008, pp56-59). Transparency in 2011 is generally far greater than it was twenty or thirty years ago, even if there are some fallings back, and some deliberate attempts to reverse to the trend (Roberts, 2006).
A pattern there is, but what is perhaps just as significant as the pattern itself is what it is not. Thus, despite the claims of some observers, there is not a clear pattern, sweeping through the world in a succession of paradigmatic stages – NPM, networks, governance, or anything of that nature. Neither is there a stable pattern of evidence-based progress: most states do not appear to have launched carefully-evaluated reform programmes, then modified them in the light of the findings, then moved on to a higher level. Indeed, careful, formal evaluation has been the exception rather than the rule, and where it has occurred it has usually been inconclusive.
On the other hand a purely fashion-driven model does not fit very well either. Fashions there have been (and continue to be), but in each case the fashion is less than universal, with important states resisting the latest version of modernity, or advancing their own alternatives. And there is evidence of some empirical learning, even if it is patchy and discontinuous. New Zealand and the UK both retreated from the extreme fragmentation induced by the NPM reforms of the 1980s and 1990s. The US changed its organizational procedures for handling emergencies after Hurricane Katrina. None of these were fashion moves – on the contrary, they represented learning from unsatisfactory experiences. Formal evaluations may not have been the driving force in these cases, and such evaluations as there were may have fallen a long way short of the models advanced in evaluation textbooks, but there were some, and there is no reason to assume that they played no part at all in the shifts of direction.
What we see, therefore, is a story of fashion-influenced trial and error, with some strong currents but no global tides. It is forward-looking story, in which more attention has often been paid to the demands of today and the promise of tomorrow than to the experience of yesterday (Pollitt, 2007b; Pollitt, 2008). It is also, for the most part, the story of a series of processes, rather than a story of discrete events. PMR almost always takes time, and the precise nature of a given reform almost always evolves somewhat as it goes along. That is why scholars need to study the process over time rather than relying too heavily on the statements and claims made at the launch of each reform.
It is important to bear in mind the limitations of the particular level of analysis at which this paper is pitched. This has been a slightly stratospheric overview, seeking and finding large, but rather crude patterns over a considerable period of time. It may well be that an analysis conducted at a more detailed level might uncover different patterns, in the same way as an overview of an anthill may show chaos, but a tighter focus on, say, soldier ants, will yield a much clearer pattern to their activities. There is, for instance, a good case for research at the level of specific tools. Work done on (inter alia) TQM and performance related pay have already revealed suggestive regularities (Joss and Kogan, 1995; Perry et al, 2009; Zbaracki, 1998). What such studies tend to show is not that TQM or PRP always have the same effect (far from it) but rather that it may be possible to identify certain contextual features which significantly raise or lower the probabilities of successful implementation. Thus we can begin to develop patterns of tools-plus-contexts, which may help us understand the kinds of circumstances in which particular categories of result are more or less likely. In some cases such an analysis may indicate that a tool is fairly robust, in the sense that it usually produces at least some results across quite a wide variety of contexts. In other cases, however, a tool may depend on so many unusual or difficult conditions being in place that it is likely to work well only in rare and highly favourable conditions (PRP may fall in this category).
This has been an exercise in peering through an awful lot of trees to try to discern the shape of the wood. It is obviously possible to make a different selection of key elements than the six set out at the beginning of this paper. I would argue that these six are, however, fairly firm and well-evidenced. They are an important part of the big picture, if not all of it. With three decades of hindsight the pattern we see is one of an enterprise which has been both uncertain and imprecise. Significant successes have been recorded, but have been balanced by many failures. Overblown and subsequently disappointed claims for prospective achievements have not infrequently earned the mistrust of both citizens and public service staff. A new industry has grown, but the areas of useful knowledge and technique which it has assembled are offset by the continued presence of oversimplification, insensitivity to context and history and, in a few cases, downright snakeoil marketing. Public management reform may have become a policy sector in its own right, but its theoretical underpinnings appear to be disjointed, its evidence base is patchy and its political identity remains as yet unsure.
Lessons for the World Bank?
[This is the section I have added to the original academic paper]
I would suggest a number of ‘lessons’ which could be drawn from the foregoing analysis:

  1. Big models, such as NPM or ‘good governance’ or ‘partnership working’, often do not take one very far. The art of reform lies in their adaptation (often very extensive) to fit local contexts. And anyway, these models are seldom entirely well-defined or consistent in themselves. Applying the big models or even standardized techniques (benchmarking, business process re-engineering, lean) in a formulaic, tick-box manner can be highly counterproductive.

  2. To take the first point further, the whole idea that there is one model or set of principles that can or should be applied everywhere is suspect. As many scholars and some practitioners have been observing for decades, there is no ‘one best way’. The whole exercise of reform should begin with a careful diagnosis of the local situation, not with the proclamation of a model (or technique) which is to be applied, top down. ‘No prescription without careful diagnosis’ is not a bad motto for reformers.

  3. Another, related point is that task differences really do matter. A market-type mechanism may work quite well when applied to refuse collection but not when applied to hospital care. Sectoral and task differences are important, and reformers should be wary of situations where their advisory team lacks substantial expertise in the particular tasks and activities that are the targets for reform.

  4. PMR is always political as well as managerial/organizational. Any prescription or diagnosis which does not take into account the ‘way politics works around here’ is inadequate and incomplete. Some kernel of active support from among the political elite is usually indispensable.

  5. PMR is usually saturated with vested interests, including those of the consultants/advisors, and the existing public service staff. To conceptualise it as a purely technical exercise would be naïve.

  6. Successful PMR is frequently an iterative exercise, over considerable periods of time. Reformers must adapt and also take advantage of ‘windows of opportunity’. This implies a locally knowledgable presence over time, not a one-shot ‘quick fix’ by visiting consultants.

  7. It does work sometimes! But, as indicated at the outset, humility is not a bad starting point.


Accenture (2008) An international comparison of the United Kingdom’s public administration, London, Accenture
Advisory Group on Reform of Australian Government Administration (2009) The reform of Australian government administration: building the world’s best public service, Canberra, Commonwealth of Australia, October
Andrews, M. (2010) ‘Good government means different things in different countries’, Governance 23:1, pp7-35
Arndt, C. (2008) ‘The politics of governance ratings’, International Public Management Journal, 11:3, pp275-297
Barber, J. (2007) Instruction to deliver: Tony Blair, public services and the challenge of achieving targets, London, Politico’s
Barzelay, M. (2001) The new public management: improving research and policy dialogue, Berkeley, University of California Press
Barzelay, M. and Gallego, R. (2010a) ‘The comparative historical analyses of public management policy cycles in France, Italy and Spain: symposium introduction’, Governance 23:2, pp297-307
Barzelay, M. and Gallego, R. (2010b) ‘The comparative historical analyses of public management policy cycles in France, Italy and Spain: symposium conclusion’, Governance 23:2, pp209-223
Bevan, Gwyn (2010) ‘Impact of devolution of health care in the UK: provider challenge in England and provider capture in Wales, Scotland and Northern Ireland? Journal of Health Services Research and Policy, 15 (2). pp. 1-2
Bezes, P. (2010) ‘Path dependent and path-breaking changes in the French administrative system: the weight of legacy explanations’ pp158-173 in Painter, M. and Peters, G.B. (eds.) Tradition and public administration, Basingstoke, Palgrave/Macmillan
Brunsson, N. (2006) Mechanisms of hope: maintaining the dream of the rational organization, Copenhagen, Copenhagen Business School Press
Cabinet Office (2008) Excellence and fairness: achieving world class public services, London, Cabinet Office
Christensen, J. (2009) ‘Danish public management reforms before and after NPM’, pp279-299 in J. Goldfinch and S.Wallis (eds.) International Handbook of Public Management Reform, Cheltenham, Edward Elgar
Christensen, T. and Laegreid, P. (eds.) (2001) New Public Management: the transformation of ideas and practice, Aldershot, Ashgate
Clarke, J.; Newman, J.: Smith, N.; Vidler, E. and Westmarland, L. (2007) Creating citizen-consumers: changing publics and changing public services, London, Sage
Dean, M. (1999) Governmentality: power and rule in modern society, London, Sage
Demmke, C. and Moilanen, T. (2010) Civil services in the EU of 27: reform outcomes and the future of the civil service, Frankfurt, Peter Lang
Dunleavy, P.; Margetts, H.; Bastow, S. and Tinkler, J. (2006a) Digital era governance: IT corporations, the state and e-government', Oxford, Oxford University Press
Dunleavy, P.; Margetts, H.; Bastow, S. and Tinkler, J. (2006b) ‘New Public Management

is dead – long live digital-era governance’, Journal of Public Administration Research and Theory, 16:3, pp467-494

Goldfinch, S. and Wallis, J. (2010) ‘Two myths of convergence in public management reform’, Public Administration 88:4, pp1099-1115
Gore, A. (1997) Businesslike government: lessons learned from America’s best companies, Washington DC, National Performance Review
Gualmini, E. (2008) ‘Restructuring Weberian bureaucracy: comparing managerial reforms in Europe and the United States’, Public Administration, 86:1, pp75-94
H.M.Government (2010) Government ICT strategy: smarter, cheaper, greener, Cabinet Office, London, HMSO
Hood, C. (1976) The limits of administration, Chichester, Wiley
Hood, C. and Jackson, M. (1991) Administrative argument, Aldershot, Dartmouth
Hood, C. and Peters, G.B. (2004) ‘The middle aging of New Public Management: into the age of paradox?’ Journal of Public Administration Research and Theory, 14:3, pp267-282
Hood, C.; Dixon, R. and Beeston, C. (2008) ‘Rating the rankings: assessing international rankings of public service performance’ International Public Management Journal, 11:3, pp298-358
Hughes, O. (2003) Public management and administration: an introduction (3rd ed.) Basingstoke, Palgrave/Macmillan
Inter-American Development Bank (IADB) (2007) Datagob: governance indicator database, ( accessed 01-09-08)
James, O. (2003) The executive agency revolution in Whitehall: public interest versus bureau-shaping perspectives, Basingstoke, Palgrave/Macmillan
Joss, R. and Kogan, M. (1995) Advancing quality: Total Quality Management in the National Health Service, Buckingham, Open University Press
Kettl, D. (2005) (2nd ed.) The global public management revolution, Washington DC, The Brookings Institution
Kickert, W.; Klijn, E-H. And Koppenjan, J. (eds.) (1997) Managing complex networks: strategies for the public sector, London, Sage
Kickert, W. (ed.) (2008) The study of public management in Europe and the US: a comparative analysis of national distinctiveness, Abingdon, Routledge
Klijn, E-H.; Steijn, B. and Edelenbos, J. (2010) ‘The impact of network management on outcomes in governance networks’, Public Administration, 88:4, pp1063-1082
Kooiman, J. (2003) Modern governance: new government-society interactions, London, Sage
Kuhlmann, S. (2010) ‘New Public Management for the ‘classical continental European administration: modernization at the local level in Germany, France and Italy’, Public Administration 88:4, pp1116-1130
Light, P. (1997) The tides of reform: making government work 1945-1995, New Haven, Yale University Press
Meier, K. and O’Toole, L. (2001) ‘Managerial strategies and behavior in networks: a model with evidence from US public education’, Journal of Public Administration Research and Theory, 11:3, pp271-293
Meier, K. and O’Toole, L. (2010) ‘Beware of managers not bearing gifts: how management capacity augments the impact of managerial networking’, Public Administration, 88:4, pp1025-1044
Mulgan, G. (2009) The art of public strategy: mobilizing power and knowledge fro the common good, Oxford, Oxford University Press
National Audit Office (2006) Central government’s use of consultants, HC 128, Session 2006-2007, London, The Stationary Office
Newberry, S. and Pallot J. (2005) 'A wolf in sheep's clothing? Wider consequences of the financial management system of the New Zealand central government', Financial Accountability and Management, vol.21:3, pp. 263-277
OECD (2009) Government at a glance 2009, Paris, OECD
Office of Public Services Reform (2002) Better government services: executive agencies in the 21st century. London, Cabinet Office (
Ongaro, E. (2009) Public management reform and modernization: trajectories of administrative change in Italy, France, Greece, Portugal and Spain, Cheltenham and Northampton, MA., Edward Elgar
Osborne, D. and Gaebler, T. (1992) Re-inventing government: how the entrepreneurial spirit is transforming the public sector, Reading, Mass., Addison Wesley
Osborne, S. (ed.) (2010) The New Public Governance: emerging perspectives on the theory and practice of public governance, London and New York, Routledge/Taylor and Francis

Perry, J.; Engbers, T. and Jun, S. (2009) ‘Back to the future? Performance-related pay’, Public Administration Review 69:1, pp39-51

Pollitt, C. (1990) Managerialism and the public services: the Anglo-American experience, Oxford, Blackwell
Pollitt, C. (1995) ‘Justification by works or by faith? Evaluating the New Public Management’, Evaluation, 1:2, October, pp. 133–154.
Pollitt, C. (2002) ‘Clarifying convergence: striking similarities and durable differences in public management reform’ Public Management Review 4:1, pp471-492
Pollitt, C. (2006) ‘Performance management in practice: a comparative study of executive agencies’, Journal of Public Administration Research and Theory 16:1, pp25-44
Pollitt, C. (2007a) ‘Convergence or divergence: what has been happening in Europe?’ pp10-25 in C. Pollitt, S. Van Thiel and V. Homburg (eds.) (2007) New Public Management in Europe: adaptations and alternatives, Basingstoke, Palgrave/Macmillan
Pollitt, C. (2007b) ‘New Labour’s re-disorganization: hyper-modernism and the costs of reform – a cautionary tale’, Public Management Review, 9:4, pp529-543
Pollitt, C. (2008) Time, policy, management: governing with the past, Oxford, Oxford University Press
Pollitt, C. (2009) ‘Structural change and public service performance’, Public Money and Management, 29:5, pp285-291
Pollitt, C. (2010) ‘Simply the best? The international benchmarking of reform and good governance’, pp91-113 in J.Pierre and P.Ingraham (eds.) Comparative administrative change and reform: lessons learned, (Festschrift for Guy B.Peters), Montreal and Kingston, McGill-Queens University Press.
Pollitt, C. (2011) ‘Moderation in all things: international comparisons of governance quality’, Financial Accountability and Management (in press)
Pollitt, C.; Talbot, C.; Caulfield, J. and Smullen, A. (2004) Agencies: how governments do things through semi-autonomous organizations, Basingstoke, Palgrave/Macmillan
Pollitt, C. and Op de Beeck, L. (2010) Training top civil servants: a comparative analysis, Leuven, Instituut voor de Ovreheid (downloadable at www.public
Pollitt, C. and Bouckaert, G. (2011) (3rd ed.) Public management reform: a comparative analysis: NPM, governance and the Neo-Weberian State, Oxford, Oxford University Press
Pollitt, C. and Dan, S. (2011) The impacts of the New Public Management in Europe: a meta-analysis, paper to be presented at the European group for Public Administration Conference, Bucharest, September
Pollitt, C. and Hupe, P. (2011) ‘Talking about governance: the role of magic concepts’ Public Management Review, 13:5, pp1-18
Roberts, A. (2006) Blacked out: government secrecy in the information age, Cambridge, Cambridge University Press
Sahlin-Andersson, K. and Engwall, L. (eds.) (2002) The expansion of management knowledge: carriers, flows and sources, Stanford, California, Stanford Business Books
Saint Martin, D. (2005) ‘Management consultancy’, pp671-694 in E. Ferlie, Lynn Jnr., L. and Pollitt, C. (eds.) The Oxford handbook of public management, Oxford, Oxford University Press
Savoie, D. (1994) Thatcher, Reagan, Mulroney: in search of a new bureaucracy, Toronto, University of Toronto Press
Schröter, E. (2007) ‘Reforming the machinery of government: the case of the German federal bureaucracy’, pp251-271 in R.Koch and J.Dixon (eds.) Public governance and leadership, Wiesbaden, Universitäts-Verlag

Simon, H. (1946) ‘The proverbs of administration’, Public Administration Review 6, pp53-67

Smullen, A. (2010) Translating agency reforms: rhetoric and culture in comparative perspective, Basingstoke, Palgrave/Macmillan
Sundström, G. (2006) ‘Management by results: its origin and development in the case of the Swedish state’ International Public Management Journal 9:4, pp399-427
Talbot, C. and Johnson, C. (2007) ‘Seasonal cycles in public management: disaggregation and re-aggregation’, Public Money and Management, 27:1, pp53-60
Zbaracki, M. (1998) ‘The rhetoric and reality of Total Quality Management’, Administrative Science Quarterly, 43, pp602-636

Share with your friends:

The database is protected by copyright © 2020
send message

    Main page