Policy transfer and convergence within the UK: the ...

© The Policy Press, 2012 • ISSN 0305 5736

193

Policy transfer and convergence within the UK: the case of local government performance improvement regimes Sandra Nutley,1 James Downe, Steve Martin and Clive Grace

Delivered by Ingenta to: University of Edinburgh IP : 129.215.149.92 On: Thu, 10 May 2012 08:18:36 Copyright The Policy Press

Analysis of the development and implementation of local government performance improvement regimes in England, Scotland and Wales over the last decade reveals congruence in policy goals but divergence in policy implementation and outcomes.The governments in all three countries had a common aim of improving local government performance. However, differences in ideology, the nature of central–local government relations, the numbers of councils and a political imperative for newly devolved administrations to be seen to pursue ‘home-grown’ solutions limited policy learning between different parts of the United Kingdom (UK) and in the case of Wales fuelled determined policy avoidance by policy makers.

Introduction The creation of devolved administrations in Northern Ireland, Scotland and Wales represents ‘a constitutional experiment of enormous significance’ (Shortridge, 2009: 143). It also provides new subnational arenas in which to study policy transfer and convergence. The devolved governments in Scotland and Wales have wide-ranging responsibilities for health, education, agriculture and other local public services, and this has opened up the possibility for policy divergence from the United Kingdom (UK) government and new kinds of policy transfer at subnational levels. (The stopstart nature of devolution in Northern Ireland, and in particular the suspension of the Assembly between 2002 and 2007, make it a less useful directly comparable case.) Recent studies have mostly examined high-profile policies in which there has been significant divergence post devolution, such as tuition fees in higher education and personal care for older people (Devolution and Constitutional Change, 2006; Cairney, 2009a; Keating et al, 2009). However, much policy still appears to be similar across the UK despite devolution (Cairney et al, 2010). In the decade following devolution, the UK, Scottish and Welsh Assembly governments all introduced performance improvement regimes for local government based on ‘whole authority assessments’: Comprehensive Performance Assessment (CPA) in England, Best Value Audit (BVA) in Scotland and the Wales Programme for Improvement (WPI).This article analyses the extent of convergence and divergence in these regimes, the factors that account for their similarities and differences and the extent to which their development and implementation was informed by policy transfer and learning between the three countries.The first section introduces the main concepts and the methods used to explore the regimes.The following three sections discuss respectively the extent of policy convergence/divergence, the determinants that have influenced convergence/divergence and the mechanisms of policy transfer and learning. Each section provides an overview of the findings from existing studies Key words: policy transfer • policy convergence • local government • performance frameworks

Policy & Politics vol 40 no 2 • 193-209 (2012) • http://dx.doi.org/10.1332/147084411X581880

194

Sandra Nutley et al

followed by an account of our findings. The final section presents our conclusions and discusses their implications for future research.


Concepts and methods Policy transfer and convergence have been researched at a number of different levels. Many studies have focused on policy transfer between nation states (Dolowitz and Marsh, 2000; Cook, 2008; Dwyer and Ellison, 2009), but there is also a growing literature on transfer and learning among subnational governments, particularly in federal systems (Kern, 2010). Subnational studies have considered transfer and convergence at regional and state levels, and between local authorities (Wolman and Page, 2002; Kern, 2010).While papers have tended to focus on policy diffusion within the American federalist system (eg Berry and Berry, 1999; Karch, 2007), there are studies of policy transfer and diffusion in other federal systems (eg Gow, 1992; Gilardi and Füglister, 2008). As highlighted in the introduction to this article, the creation of devolved administrations in the UK has also resulted in a growing literature on the extent of subnational policy transfer and divergence. Drawing together the main findings to emerge from this burgeoning literature is not straightforward as both the empirical and theoretical development of the field has been hampered by inconsistent and overlapping usage of concepts such as ‘policy transfer’, ‘policy convergence’, ‘policy learning’, ‘policy diffusion’ and ‘policy isomorphism’ (Knill, 2005). While some see ‘policy transfer’ as the umbrella concept that encompasses a wide array of subcomponents (Pemberton, 2009), others draw sharper distinctions between these concepts (Knill, 2005). In this article, we follow Dolowitz and Marsh (1996, 2000) in defining policy transfer as the process by which knowledge about policies, administrative arrangements and institutions in one political system is used in another. Policy convergence refers to whether there is increasing similarity over time in the policies adopted by two or more governments or administrative systems.Thus, policy transfer describes the mechanisms that can lead to policy similarity or dissimiliarity, whereas policy convergence is more directly concerned with the effects of a process of policy change. In discussing policy transfer, we also refer to learning and lesson drawing, by which we mean a more systematic process whereby policy makers study developments in other systems in order to evaluate their potential applicability in the home system (Rose, 1993, 2004). This need not necessarily result in action or change since the policy or programme in question may be deemed inappropriate for the home system – a process that Rose (1993) calls ‘negative lesson drawing’. Policy transfer and convergence/divergence are dynamic processes that need to be studied over time. So we undertook a detailed analysis of the development and implementation of the local government performance improvement regimes in England, Scotland and Wales over the period 2002–09. The importance of these regimes – central governments in all three countries rely on local councils to deliver a wide range of policies and public services efficiently and effectively – and the fact that the regimes have been subject to both convergent and divergent pressures made them appropriate case study subjects. We gathered evidence from three sources: a detailed analysis of policy documents, legislation and statutory and non-statutory guidance; a series of in-depth semi-structured interviews with 45 senior policy makers Policy & Politics vol 40 no 2 • 193-209 (2012) • http://dx.doi.org/10.1332/147084411X581880

Policy transfer and convergence within the UK

195

and practitioners from central government, audit bodies, improvement agencies and local government, all of whom had been closely involved in the development or implementation of the performance regimes in their countries (see Table 1); and a workshop with a wider group of stakeholders from all the three countries to validate our emerging findings. Interviewees were asked a range of questions about the local government improvement regimes in their country: Table 1: Interviewees’ organisations and rolesa Organisations England (E1–E15) Audit Commission

Rolesb


Chief Executive, Former Chief Executive, Former Head of District Audit Cabinet Office Former special adviser to the Prime Minister on health and local government policy Department for Communities and Local Former Minister of State for Local Government Government Department for Constitutional Affairs Former Parliamentary Under-Secretary Improvement & Development Agency Executive Director, Former Executive Director Local authorities Six chief executives Local Government Association Former Chief Executive Scotland (S1–S16) Accounts Commission Secretary to the Commission, Member Audit Scotland Auditor General, Deputy Auditor General, Director of Public Reporting, Assistant Director of Public Reporting Convention of Scottish Local Authorities Strategic Director Improvement Service Chief Executive Local authorities Three chief executives, Assistant Chief Executive Scottish Government Best Value Policy Manager, Former Head of Justice Department, Former Minister of Finance, Head of Best Value and Performance Wales (W1–W14) Audit Commission in Wales District Auditor Local authorities Three chief executives Wales Audit Office Performance specialist Welsh Assembly Government Minister for Social Justice and Local Government, Former Senior Special Adviser to the First Minister, Director of Department for Constitutional Affairs, Equality and Communication, Head of Local Government Policy, Head of Local Government Strategy & Performance, Former Head of Local Government Policy Welsh Local Government Association Chief Executive, Performance and Improvement Adviser, Director of Governance and Improvement Total number of interviewees 45 Notes: All interviews took place during 2007/08. Roles are not listed in interviewee reference number order in order to protect the identity of those cited in this article.

a

b


196

Sandra Nutley et al

• how they had been developed; • what factors shaped their design; • the extent to which policy makers and inspectors learned from elsewhere in the UK and beyond; • what (if anything) inhibited or facilitated policy transfer; • what they believed the outcomes of the regimes in their countries had been. Interviews were taped and transcribed.We were granted excellent access to all of the main actors in each country and this enabled us to build up a very rich picture of the performance frameworks. To guard against the danger that these elite interviewees exaggerated their own roles in the design and development of the frameworks, and thus played down the extent to which they transferred policies from elsewhere, their responses were ‘triangulated’ against evidence from the documentary analysis and the workshop.


Extent of convergence and divergence Heichel et al (2005) draw together the findings from 74 studies of cross-national policy convergence.They report that nearly half (33) identified convergence, just 15 found no evidence of convergence or divergence. The other 26 studies identified convergence in some but not all of the policy dimensions that they examined. At subnational level, Kern’s (2010: 13) review of policy transfer within federal systems finds an even greater tendency towards convergence, because ‘policy transfer and learning within federal systems is easier to accomplish than policy transfer and learning beyond national borders’. However, on some issues, policy transfer results in convergence clusters, where pioneering states become centres of regional clusters, which can lead to policy polarisation (divergence) between different groupings of subnational governments. Studies within the UK provide examples of policy divergence in areas where, prior to devolution, policy had been similar, such as tuition fees in higher education (Keating, 2005).There are also examples where subtle differences in policy direction that existed prior to devolution have subsequently become more marked, such as the differing approaches to pupil testing (Cairney et al, 2010). There are yet other examples where initial policy divergence has been followed by reconvergence, such as the comprehensive smoking ban in public places (Asare et al, 2009; Cairney, 2009b). In cross- and subnational studies, researchers have found convergence on some dimensions of a policy but divergence on other aspects. For example, Hall (1993) argues that convergence in policy instruments and tools is more likely than convergence in policy paradigms.That is, the same policy instruments and tools might be used by two countries to achieve different goals. Conversely, Radaelli’s (2005) study of the diffusion of regulatory impact assessment finds that policy ideas and goals travel more readily than policy instruments and tools. In the UK, there are examples of convergence in terms of broad policy objectives but important differences in implementation and outcomes. For example, England and Scotland both introduced Anti-Social Behaviour Orders (ASBOs) but the way in which the policy was implemented differed – far fewer ASBOs were issued per capita in Scotland than in England (McGarvey and Cairney, 2008). It was therefore important that our research examined the extent of Policy & Politics vol 40 no 2 • 193-209 (2012) • http://dx.doi.org/10.1332/147084411X581880


197

convergence/divergence in local government performance improvement regimes not only in terms of broad policy choices but also in the ways in which they were implemented and their outcomes. We anticipated that the extent of convergence/ divergence would differ across these categories. Data about the three regimes were analysed using five categories based on an adaptation of Bennett’s (1991) classification of the policy dimensions on which there might be convergence:


• goals – a coming together of intent to deal with commonly defined policy problems; • instruments – the statutory basis of the policy; • design style – the way in which policy responses are formulated (for example whether processes of policy formulation are consensual or conflictual); • content and tools – the administrative rules and implementation practices that embody the policy; • outcomes – the effects of the policy. The results are summarised in Table 2 and are discussed below.

Policy goals and instruments The evidence demonstrated that policy makers in all three countries shared common policy goals and used congruent policy instruments. In all three cases, whole authority assessments were introduced through a combination of legislation and statutory and non-statutory guidance. All three regimes had their origins in the Best Value duty of continuous improvement, which was introduced throughout the UK in the late 1990s by the New Labour government as a quid pro quo for a moratorium on the compulsory competitive tendering of local government services (Martin, 2000). All three frameworks used whole authority assessments to evaluate a council’s overall performance because policy makers in all three countries believed that it was ‘corporate capacity’ – reflected in clarity of purpose, robust performance management systems, efficient use of resources and effective partnership working – that determined a local authority’s ability to achieve sustained improvement in the services it provided (Martin et al, 2010). And all three performance frameworks were overseen by the principal audit body – the Audit Commission in England, Audit Scotland (working on behalf of the Accounts Commission in Scotland) and the Wales Audit Office (Downe et al, 2008). The evidence therefore pointed to striking commonality in the goals and instruments used in the three countries.


198

Sandra Nutley et al

Table 2: Local government performance improvement regimes: areas of convergence and divergence England (CPA) Improve performance by strengthening corporate capacity Policy Primary legislation instruments Statutory and nonstatutory guidance Assessments conducted by Audit Commission Policy design style Designed largely by the Audit Commission and imposed on local authorities Policy content Mode of operation External challenge, ‘naming and shaming’

Scotland (BVA) Improve performance by strengthening corporate capacity Primary legislation Statutory and nonstatutory guidance Assessments conducted by Audit Scotland Designed by task force, which drew on existing voluntary practice

Wales (WPI) Improve performance by strengthening corporate capacity Primary legislation Statutory and nonstatutory guidance Assessments conducted by Wales Audit Office Designed by tripartite group with powerful local government influence

External challenge, education and persuasion

Local ownership of need to improve, collaborative

Assessment criteria and methodology

Uniform implementation of rules-based scoring system reflecting national priorities Annual assessments of upper-tier authorities Published performance scores and narrative report Externally imposed by ministers on advice of Audit Commission

Best Value principles tailored to local priorities

Joint risk assessment tailored to local circumstances

Five-year rolling programme of all authorities Published narrative report

Annual assessments of all authorities

As last resort imposed by ministers on advice of Accounts Commission

As last resort imposed by ministers on advice of Wales Audit Office

Initial challenges to CPA followed by widespread acceptance Credited with positive impact but concerns about susceptibility to gaming High-level ministerial support but opposition from other political parties

Initial concerns but broad acceptance among most councils

Widely welcomed by councils. Smooth implementation process Perception that impact might be positive but lack of evidence to support this Concerns about lack of transparency and need for more ‘teeth’


Policy goals

Frequency

Reporting

Intervention

Policy outcomes Process success

Programmatic success

Political success

Credited with positive impact on corporate capacity of authorities Widespread political support – seen as legitimate by all key stakeholders

Confidential joint risk assessment



199


Policy design style and content In contrast to the policy goals and instruments, there were marked differences in policy design style and content. Whereas CPAs were imposed ‘top down’ by the UK government and the Audit Commission with relatively little discussion or consultation with local authorities, both the WPI and BVAs were developed through consensual processes in which local government representatives played a prominent role.The assumption in England was that local government required a robust external challenge. CPAs were therefore designed to provide a ‘shock to the system’ and published performance ratings were used to ‘name and shame’ councils that were judged to be underperforming.Those that continued to languish at the bottom of the performance league table faced the prospect of centrally orchestrated intervention.The Scottish framework also included external reporting and challenge, but there were no performance scores.The aim was to raise councils’ awareness of their weaknesses, rather than to make a public example of them. Meanwhile, Welsh ministers adopted the most localist stance. They argued that while external assessment could help to highlight problems, it was local officers and councillors who had responsibility for addressing poor performance. So assessments made under the WPI were not published and there were no explicit sanctions for poor performance. There were also differences in the nature and frequency of the assessments. CPAs used a universal rules-based scoring system.The Scottish and Welsh frameworks were less prescriptive and gave far more scope for local priorities and self-assessment. In England and Wales, local authorities underwent annual assessments, whereas each Scottish council was subject to just one BVA between 2003 and 2009.

Policy outcomes The policy outcomes associated with the three regimes also differed. It is difficult to gauge the success of CPAs, BVAs and the WPI using statistical analysis because of the lack of comparable performance data across the three countries (Andrews and Martin, 2010). However, our interview data provided clear and consistent evidence about their perceived effectiveness.We assessed actors’ perceptions of outcomes using a framework derived from Marsh and Sharman (2009), which differentiates between: • process success (the smoothness of policy introduction and implementation); • programmatic success (the effectiveness, efficiency and resilience of a policy); • political success (the level of political support for and perceived legitimacy of a policy in the eyes of key stakeholders). In terms of ‘process success’, the WPI scored highest. Given the role that local government representatives played in its design, it was not surprising that this framework met with relatively little resistance ‘on the ground’. BVAs in Scotland encountered more design and implementation difficulties, particularly in the early stages, and some local councils complained that the judgement criteria were weighted in favour of larger councils that had the staff and other resources to be able to manage the assessment process effectively. Initially, CPAs met with considerable resistance,



200

Sandra Nutley et al

especially from councils that were judged to be performing badly. However, within two to three years, English local government had largely acquiesced. The growing acceptance over time of CPAs and BVAs reflected their ‘programmatic success’. Most of our interviewees in England and Scotland believed that the regimes achieved their overall objective of encouraging improvements in corporate capacity, although there were residual concerns about the burdens the regimes placed on local authorities. In contrast, most interviewees in Wales were uncertain about the effectiveness of the WPI and there was widespread dissatisfaction with what was seen as a lack of ‘hard’ evidence of improvement by Welsh councils. As for ‘political success’, CPAs enjoyed the confidence of ministers for several years, but failed to secure support from across the political spectrum. Opposition parties continued to argue for its abolition and its successor regime (known as Comprehensive Area Assessment) was scrapped by the incoming coalition government in 2010. Initially, the WPI was popular with both ministers and local authorities, although not with auditors. However, over time, all parties expressed concerns about its effectiveness. BVAs achieved the most consistent political support among our interviewees and were strongly endorsed by a wide-ranging independent review of scrutiny arrangements in Scotland (Crerar, 2007). Overall, the analysis of the three regimes confirmed our expectations that the extent of convergence/divergence would vary across the different dimensions of the policy. Following on from the common origins of the three regimes in the duty of BestValue, we found congruence at the level of broad policy goals but divergence with regard to their detailed implementation and their perceived success.The importance of the differences in implementation should not be underestimated. For many interviewees, they meant that the regimes ‘felt’ very different. As an interviewee from Scotland noted, “although many aspects of public service audit and inspection in England and Scotland look as if they are converging, the overall feel of these regimes … is different” (interviewee S1 – senior manager, Audit Scotland).

Determinants of convergence and divergence Previous studies have identified at least four main reasons why policies might converge: • concurrent pressure, where policies converge because nations or states facing similar economic, social and technological challenges arrive independently at similar solutions; • direct coercion, where a policy is imposed on one country or state by a higher authority; • indirect coercion, where functional interdependence creates spillovers that lead to policy convergence; • policy learning, where nations or states learn and adopt lessons from elsewhere (Dolowitz and Marsh, 1996). Although in theory it is possible to distinguish between these categories, in practice policy convergence tends to result from combinations of them.The evidence suggests that geographical proximity, shared social and economic characteristics, similarities in institutional arrangements and strong cultural ties all help to facilitate policy transfer Policy & Politics vol 40 no 2 • 193-209 (2012) • http://dx.doi.org/10.1332/147084411X581880



201

and convergence (eg Dolowitz and Marsh, 1996; Knill, 2005; Lenschow et al, 2005; Hood, 2007). In the case of the UK, the factors influencing convergence have included most of these factors, although direct coercion is uncommon due to a reasonably clear division of competence between the UK government and the devolved administrations and the lack of systems to provide for common standards (Adams and Schmuecker, 2006). In general, England has been identified as the innovator in public service provision and this has led to pressures for Scotland and Wales to follow its lead (Cairney et al, 2010).This is especially true where a policy attracts media attention and where public opinion supports the need for policy uniformity; the UK government and the devolved administrations share a commitment to measuring their success in performance terms and they are sensitive to ongoing funding and performance comparisons (Jeffery, 2009). England undoubtedly has the greatest policy development capacity and this has long encouraged the borrowing of policies (policy learning) by Scotland and Wales. Prior to devolution, the role of the Scottish, Welsh and Northern Ireland Offices was often to implement policies conceived in London with only relatively minor modifications (Keating et al, 2003). Set against these pressures for convergence, there are other factors that have encouraged divergence. While England, Scotland and Wales share many similar social and economic problems, there are institutional, demographic and cultural differences that affect perceptions of these problems and how they might be tackled. The devolution settlement is highly permissive of divergence and the existence of distinct policy communities at devolved levels on some policy issues (such as school education) has enabled the devolved administrations to formulate and sustain divergent policies (Raffe, 2006).There are also political pressures for divergence as the case for devolution rested to a large extent on the need for distinctive policies that addressed the particular needs of Scotland, Wales and Northern Ireland (Bradbury, 2003). This seems to have been particularly marked in Welsh politics where researchers have found a well-established presumption in favour of policy distance from England (Adams and Schmuecker, 2006), but in both Scotland and Wales divergence has sometimes occurred as a result of a reluctance to be seen to be following an ‘English model’ (Raffe, 2006). There is some evidence that the capacity for divergence has increased over time; as policy development capacities and self-confidence in the devolved administrations have grown, they have become less susceptible to convergence pressures (Cairney et al, 2010). In the light of this literature, we anticipated that the main determinants of the congruence identified in the regimes would be similarities in institutional arrangements, shared ideologies about the effectiveness of performance auditing, and policy learning as a result of limited policy development capacity in the devolved administrations. As the regimes have not attracted much media attention, we did not anticipate that media or public opinion would be a significant influence. The main determinants of divergence were anticipated to be differences in central–local relations and scale of local government (ie the numbers of local authorities), combined with the political imperative to do things differently in Scotland and Wales. With regard to congruence in policy goals, the ongoing dominance of the UK government during the early days of devolution was important. In 1999, Wales was still governed in terms of legislative policy goals by England. Scotland had a greater Policy & Politics vol 40 no 2 • 193-209 (2012) • http://dx.doi.org/10.1332/147084411X581880


202

Sandra Nutley et al

degree of autonomy but was still heavily influenced by the Scottish government’s links to Westminster via the Scottish (now Scotland) Office. In addition, all three countries were influenced by New Public Management ideologies (Clarke and Newman, 1997) and the associated emphasis on managing public services through performance auditing. However, when it came to the detail about how public services should be encouraged to improve, Scotland and Wales did not buy into England’s top-down imposition of targets, testing, league tables and star ratings. Instead, the emphasis was on collaboration, persuasion and shared ownership of the improvement agenda (Downe et al, 2010). As a result, the greater policy development capacity of England did not lead to as much policy learning from the English regime as might have been anticipated (we discuss this further in the next section). In relation to the determinants of divergence, differences in central–local relations were important. Scotland and Wales increasingly adopted distinctive approaches to central–local relations and the regulation of local public services (Bennett et al, 2002; Laffin, 2004; Gallagher et al, 2007).They explicitly sought to work in ‘partnership’ with the local governments in their countries (Martin and Webb, 2009) and this created distinct local government policy communities in Scotland and Wales, which not only felt able to do things differently but also thought that it was imperative that they do so. In part, the small number of local authorities in Scotland (32) and Wales (22) compared with England (150 single-tier and county councils) was seen as an important reason for avoiding an English-style, top-down performance improvement regime. Interviewees in Scotland argued that there was little point in compiling performance league tables consisting of a small number of local authorities of widely varying sizes and serving very different kinds of areas.The WPI was said by interviewees to reflect the statutory requirement placed on the Welsh Assembly to work in partnership with local councils. In addition, the Welsh Assembly’s limited administrative capacity left civil servants dependent on local government for policy input (Jeffery, 2006a). As a result, Welsh local authorities were able to wield considerable influence (Laffin, 2004) and auditors found themselves in a weaker position than their counterparts in England or Scotland. As we anticipated, there was also a strong political imperative to do things differently, particularly in Wales. Here policy makers seized on the WPI as an early opportunity to put in place a distinctive approach consistent with the-then First Minister’s boast of the ‘clear red water’ that was opening up between policies conceived in Cardiff and Westminster. The subtext was that this should help to build support for the fledging Welsh Assembly, which was still uncertain of popular support in the wake of the wafer-thin majority that voted in favour of devolution.There was also an element of ‘pay-back’ for local government’s support for the creation of a new assembly. The anticipated pressures for convergence (similar institutions, ideologies and policy transfer) were influencing factors but their effects were outweighed by pressures for divergence (central–local relations, the scale of local government and a political imperative to do things differently). Performance management ideologies limited divergence at the level of overall goals, but important differences in the ‘theories of improvement’ adopted by each country led to divergence in implementation (Downe et al, 2010). The process was, however, a dynamic one and the balance between the pressures for convergence/divergence shifted over time. As we discuss in the concluding section, towards the end of the period of our study, there were signs of Policy & Politics vol 40 no 2 • 193-209 (2012) • http://dx.doi.org/10.1332/147084411X581880


203

increased self-confidence in the devolved administrations. But rather than fuelling divergence, as suggested by Cairney et al (2010), this made policy makers in Scotland and Wales more receptive to lessons that could be learned from England and elsewhere.


Policy transfer mechanisms Previous studies have identified that wholesale borrowing of a policy is rare. In practice, policy makers tend to copy some but not all aspects of a programme or to create hybrids, which combine elements of policies from several countries. Studying familiar problems in unfamiliar settings may also act as a source of inspiration, helping to generate novel solutions to problems faced at home (Dolowitz and Marsh, 1996, 2000; Rose, 2004). In terms of the direction of policy transfer within a federal or devolved government context, this can occur from the centre to the periphery, from the periphery to the centre, and around the periphery (Cairney et al, 2010). Previous studies have suggested that in the UK, centre–periphery policy transfer – whereby policies formulated in England are adopted in Northern Ireland, Scotland and Wales – is the dominant form of interaction (Keating, 2007). Periphery-to-centre transfer is considered to be rare and there seem to be few channels and opportunities for periphery-toperiphery transfer. Studies of policy transfer have identified the main actors involved as elected officials, political parties, bureaucrats/civil servants, pressure groups, policy entrepreneurs/experts, consultants, think tanks and transnational or supranational institutions (Dolowitz and Marsh, 2000). In the UK, the main channels of policy transfer are frequently identified as civil servants, professional networks and political parties (Jeffery, 2006b). Taking these factors into account, we anticipated that the direction of any policy transfer was likely to be from England to Scotland and Wales, but that this was more likely to take the form of inspiration and adaptation than direct copying. We also anticipated that civil servants, professional networks and political parties would be the main channels for any lesson drawing and policy transfer. We found examples of centre-to-periphery learning in our study, but this was more evident in Scotland than in Wales. Scottish interviewees reported that BVAs had been influenced directly by experience in England. Audit Scotland recruited staff from England and this enabled it to draw on lessons from the experience of undertaking CPAs. One interviewee explained that there was “a very deliberate focus on looking at the way the existing audit mechanisms could be adapted … rather than looking at starting afresh … we aimed to learn from both CPA and WPI … CPA was probably more influential” (interviewee S7 – senior manager, Audit Scotland).The process was to adopt and adapt what were seen as the ‘good bits’ of CPA, such as the corporate assessment process, while avoiding the aspects that the policy community in Scotland did not like:“the rules-based assessment and the application of particular labels for how well a council was performing” (interviewee S7 – senior manager, Audit Scotland). The minutes of meetings between Welsh civil servants, auditors and local government representatives record that they conducted an analysis of CPAs in England. However, our interviewees played down its importance. Some emphasised that they were committed from the outset to developing a ‘home-grown’ approach and one interviewee told us that they had used a “fairly limited evidence base to inform the Policy & Politics vol 40 no 2 • 193-209 (2012) • http://dx.doi.org/10.1332/147084411X581880


204

Sandra Nutley et al

design of WPI. [We] did not do any research across other countries” (interviewee W6 – senior official, Welsh Local Government Association). In contrast to Scotland, there is no evidence that Welsh policy makers asked the question: ‘Under what circumstances and to what extent would a programme now in effect elsewhere also work here?’ (Rose, 1991: 4). Rather, decisions were driven by the desire to be seen to take a distinctive approach – and in particular one that was different from England. An interviewee told us: “there is a huge amount of prejudice [against English policy instruments] … it seems to be particularly strong within the Welsh Assembly Government. I’m sorry to say that it is easier in Wales to introduce an idea fromVenezuela than London” (interviewee W4 – local authority chief executive). Although the auditors and their senior colleagues in the Audit Commission’s headquarters in London fought a rearguard action to insert some elements of CPA into the new Welsh regime, it soon became clear that the odds were stacked against them. There was no appetite for lesson drawing based on a dispassionate assessment of the appropriateness to Wales of a CPA-type approach. Rather, this was a case of what might be called ‘policy avoidance’, driven by the political imperative to be seen to be taking a different approach to its neighbour. There was little evidence of policy transfer and learning from the periphery to the centre. Given that CPA was introduced before the other two regimes, it is perhaps not surprising that there was little evidence of policy learning in its development. Most of the English policy makers and practitioners we interviewed knew very little about BVAs or the WPI, and many expressed surprise at the idea that they might have anything to learn from these regimes. One workshop participant, for example, told us: “it hadn’t occurred to me to find out about what was happening in Wales and Scotland until today” (public official from England). This reflected a belief that the performance of public services in the devolved territories – in Wales in particular – lagged behind those in England, although this was linked largely to the widely publicised failure to reduce hospital waiting times, rather than any analysis of local authority performance. There was only limited evidence of periphery-to-periphery policy transfer. The representative bodies of local government in Scotland and Wales had exchanged information. Civil servants from the two countries also reported that they met on a fairly regular basis, but there was little evidence that they had influenced each other’s approaches to the development of their local government performance frameworks. The development of the local government performance improvement regimes in two of the three countries (England and Scotland) was determined largely by officials and civil servants. CPAs were designed by the Audit Commission with some input from civil servants. In Scotland, the BVA architecture was shaped by auditors and civil servants in consultation with representatives of local government. In Wales, the principal architect of the WPI was a special ministerial adviser to the First Minister, and although the detailed design involved auditors, local government officials and civil servants, the process was more politicised than in the other two countries.Think tanks, policy entrepreneurs and transnational networks had almost no influence in the development of the regimes, although others have found these to be influential in cross-national policy transfer (Stone, 2000, 2004). Nor was there evidence from our research to support Cairney et al’s (2010) suggestion that political parties in the UK act as powerful forces for policy transfer and convergence. Rather, divisions Policy & Politics vol 40 no 2 • 193-209 (2012) • http://dx.doi.org/10.1332/147084411X581880


205

within the Labour Party acted as a force for divergence as ‘Old Labour’ politicians in Wales deliberately dissociated themselves from the Blairite reform agenda being pursued in England.


Conclusions Our analysis of the development of local government performance regimes in England, Scotland and Wales shows that, faced with similar pressures and in pursuit of the same broad objectives, policy makers in all three countries put their faith in formal methods of audit and assessment. However, the apparently similar regimes that they developed were implemented in quite different ways. As a result, they had a very different ‘feel’ to each other and were perceived as having different outcomes in terms of rates of improvement.The reasons for the divergence in implementation can be traced to differences in central–local government relations and contrasting views about the effectiveness of top-down targets and performance monitoring.The number of authorities in each country was also important. Our study found mixed evidence of policy transfer and policy learning in the decade immediately following devolution. Policy makers in England seemed to be uninformed and unconcerned about developments elsewhere.There was no attempt on their part to learn lessons from BVA or the WPI. Some policy makers in Scotland and Wales had a high level of what might be called ‘policy awareness’ and tracked developments in England. However, they responded in different ways. The Scottish auditors who designed BVAs studied CPAs and were unabashed about adopting and adapting elements of the English framework.When Scottish local authorities became aware of the more transparent rules-based system embodied in CPAs, they lobbied successfully for greater transparency in the second round of BVAs. In 2009, CPAs were replaced by a new framework known as Comprehensive Area Assessment (CAA), which scrutinised the performance of local partnerships rather than focusing entirely on local councils.The second round of BVAs in Scotland, which also started in 2009, followed suit by implementing new ‘Shared Risk Assessments’, which attempted to give a holistic assessment of outcomes in each local area. In 2002/03, the key actors in Wales rejected CPA in principle and pursued a path of determined policy avoidance. But five years on they were more willing to look at and learn from Scotland and England. In 2009, the Welsh Assembly Government passed a new local government measure – Local Government (Wales) Measure 2009 – which gave ministers powers to intervene directly in failing authorities and required the auditors to publish performance data on each council in Wales, as was already happening in England and Scotland. Interviewees believed that this reflected increased confidence on the part of Welsh actors. As the devolved institutions had become more established, there was less of a need to be seen to be different from England and they became more open to learning lessons from their nearest neighbour. As one explained: “if we were starting afresh today, we would probably try to learn from England” (interviewee W7 – senior official, Welsh Local Government Association). Over time, the Welsh Assembly Government has also shown increasing interest in policies from Scotland, notably the duty on local partners to cooperate, which is enshrined in Scottish law, and there are now fairly frequent contacts between senior officials.



206

Sandra Nutley et al

It is important not to overemphasise these more recent signs of convergence or to exaggerate the appetite for policy learning. Overall, the evidence points to three local government performance regimes that were still developing largely in parallel even in 2009. While their architects were paying more attention to developments elsewhere in the UK, they continued to be driven primarily by developments within countries rather than by systematic learning between them. The UK government’s decision in 2010 to abolish the CAA regime and the Audit Commission at the same time as Wales was implementing a ‘harder-edged’ version of the WPI (laid out in the Local Government [Wales] Measure 2009) reinforces this point. These findings have a number of implications for future research on policy transfer and convergence. They confirm the importance of studying policy developments in detail, over time, taking account of contextual factors and drawing on a range of different actors’ perspectives. They highlight the value of enquiring into different aspects of policy convergence/divergence, from goals through to outcomes. And they suggest that following devolution, the UK does provide a source of valuable empirical data for studying policy transfer and convergence.They are also consistent with the wider literature, which has highlighted policy areas where there have been pressures to adopt similar policy goals and instruments but significant differences in implementation. Our findings also echo previous research, which has pointed to the limitations on lesson drawing and policy transfer (Rose, 1993; Evans and Davies, 1999; Marsh and Sharman, 2009). Our findings highlight the importance of policy avoidance as a reason for a lack of policy transfer between governments even in apparently favourable conditions such as those offered by the UK. In the specific case of the UK, policy avoidance was shaped by an overriding political objective to demonstrate the need for and efficacy of devolution in both Scotland and Wales.These findings suggest that policy avoidance is most likely to occur in devolved governments that have been recently created and/or endowed with new competences that were formerly the remit of central government. It is also possible that policy avoidance is most marked, and policy transfer least likely, when a policy area connects with a distinctive, meta-policy platform, such as the rejection in Scotland and Wales of competition as a means of improving public services. Future research might usefully consider whether this is the case and whether our findings are applicable more widely. Finally, while we have made an initial attempt to compare the perceived success of the three local government performance improvement regimes, it is clear that there is far more to be done in this area.As Marsh and Sharman (2009) have suggested, scholars interested in policy transfer should give more attention to policy success and failure. This is not a straightforward proposition because of the complexity of issues involved in assessing the multidimensional nature of success (Marsh and McConnell, 2010). However, the differences that we found in perceptions of the process, programmatic and political success of the three regimes suggest that future studies would do well to take a multidimensional approach to assessing outcomes.



207

Note 1 Corresponding author.


Acknowledgements This article draws on research funded by the Economic and Social Research Council (ESRC) Public Services Programme award number 166-25-0034. The authors wish to acknowledge the helpful comments received from two anonymous reviewers. References Adams, J. and Schmuecker, K. (eds) (2006) Executive summary, devolution in practice 2006: Public policy differences within the UK, London: IPPR. Andrews, R. and Martin, S.J. (2010) ‘Regional variations in public service outcomes: the impact of policy divergence in England, Scotland and Wales’, Regional Studies, 44 (8): 919–34. Asare, B., Cairney, P. and Studlar, D. (2009) ‘Federalism and multilevel governance in tobacco policy: the European Union, the United Kingdom and the devolved UK institutions’, Journal of Public Policy, 29 (1): 79–102. Bennett, C.J. (1991) ’What is policy convergence and what causes it?’, British Journal of Political Science, 21 (2): 215–33. Bennett, M., Fairley, J. and McActeer, M. (2002) Devolution in Scotland:The impact on local government,York: Joseph Rowntree Foundation. Berry, F. and Berry,W. (1999) ‘Innovation and diffusion models in policy research’, in P. Sabatier (ed) Theories of the policy process, Boulder, CO and Oxford:Westview Press. Bradbury, J. (2003) ‘The political dynamics of sub-state regionalisation: a neofunctionalist perspective and the case of devolution in the UK’, British Journal of Politics and International Relations, 5 (4): 543–75. Cairney, P. (2009a) ‘Has devolution changed the “British policy style”?’, British Politics, 3: 350–72. Cairney, P. (2009b) ‘The role of ideas in policy transfer: the case of UK smoking bans since devolution’, Journal of European Public Policy, 16 (3): 471–88. Cairney, P., Keating, M. and Hepburn, E. (2010) ‘Policy convergence, transfer and learning in the UK under devolution’, Paper presented at the Seminar on Policy Learning and Policy Transfer in Multilevel Systems, Edinburgh, 22–23 January. Clarke, J. and Newman, J. (1997) The managerial State, London: Sage Publications. Cook, I.R. (2008) ‘Mobilising urban policies: the policy transfer of US business improvement districts to England and Wales’, Urban Studies, 45 (4): 773–95. Crerar, L. (2007) The Crerar Review: The report of the Independent Review of Regulation, Audit, Inspection and Complaints handling of public services in Scotland, Edinburgh: Scottish Government. Devolution and Constitutional Change (2006) Final report of the Devolution and Constitutional Change Programme, Edinburgh: ESRC Devolution Programme, www. devolution.ac.uk/final_report.htm Dolowitz, D. and Marsh, D. (1996) ‘Who learns what from whom: a review of the policy transfer literature’, Political Studies, 44 (2): 343–57.



208

Sandra Nutley et al

Dolowitz, D. and Marsh, D. (2000) ‘Learning from abroad: the role of policy transfer in contemporary policy making’, Governance, 13 (1): 5–24. Downe, J., Grace, C., Martin, S.J. and Nutley, S.M. (2008) ‘BestValue audits in Scotland: winning without scoring?’, Public Money & Management, 28 (1): 77–84. Downe, J., Grace, C., Martin, S.J. and Nutley, S.M. (2010) ‘Theories of public service improvement: a comparative analysis of local performance assessment frameworks’, Public Management Review, 12 (5): 663–78. Dwyer, P. and Ellison, N. (2009) ‘“We nicked stuff from all over the place”: policy transfer or muddling through?’, Policy & Politics, 37 (3): 389–407. Evans, M. and Davies, J. (1999) ‘Understanding policy transfer: a multi-level, multidisciplinary perspective’, Public Administration, 77 (2): 361–85. Gallagher, J., Gibb, K. and Mills, C. (2007) Rethinking central–local relations in Scotland: Back to the future?, Hume Occasional Paper 70, Edinburgh:The Dave Hume Institute. Gilardi, F. and Füglister K. (2008) ‘Empirical modelling of policy diffusion in federal states: the dyadic approach’, Swiss Political Science Review, 14 (3): 413–50. Gow, J. (1992) ‘Diffusion of administrative innovations in Canadian public administrations’, Administration and Society, 23 (4): 430–55. Hall, P.A. (1993) ‘Policy paradigms, social learning and the state: the case of economic policy-making in Britain’, Comparative Politics, 25 (3): 275–96. Heichel, S., Pape, J. and Sommer, T (2005) ‘Is there convergence in convergence research? An overview of empirical studies on policy convergence’, Journal of European Public Policy, 12 (5): 817–40. Hood, C. (2007) ‘Public service management by numbers: why does it vary? Where has it come from? What are the gaps and the puzzles?’, Public Money & Management, 27 (2): 95–102. Jeffery, C. (2006a) ‘Devolution and local government’, Publius, 36: 57–73. Jeffery, C. (2006b) ‘Devolution and divergence: public attitudes and institutional logics’, in J. Adams and K. Schmuecker (eds) Devolution in practice 2006: Public policy differences within the UK, London: IPPR. Jeffery, C. (2009) ‘Devolution, public attitudes and social citizenship’, in S. Greer (ed) Devolution and social citizenship in the UK, Bristol: The Policy Press. Karch, A. (2007) Democratic laboratories: Policy diffusion among the American states, Ann Arbor, MI: University of Michigan Press. Keating, M. (2005) ‘Higher education in Scotland and England after devolution’, Regional and Federal Studies, 15 (4): 423–35. Keating, M. (ed) (2007) The government of Scotland: Public policy-making after devolution, Edinburgh: Edinburgh University Press. Keating, M., Cairney, P. and Hepburn, E. (2009) ‘Territorial policy communities and devolution in the UK’, Cambridge Journal of Regions, Economy and Society, 2: 51–66. Keating, M., Stevenson, L., Cairney, P. and Taylor, K. (2003) ‘Does devolution make a difference? Legislative output and policy divergence in Scotland’, Journal of Legislative Studies, 9 (3): 110–39. Kern, K. (2010) ‘Policy transfer and learning among regions: experiences from federal systems’, Paper prepared for the Seminar on Policy Learning and Policy Transfer in Multilevel Systems, Scottish Policy Innovation Forum and University of Aberdeen, Edinburgh, Royal Society of Edinburgh, 22–23 January.




209

Knill, C. (2005) ‘Introduction: cross-national policy convergence: concepts, approaches and explanatory factors’, Journal of European Public Policy, 12 (5): 764–74. Laffin, M. (2004) ‘Is regional centralism inevitable? The case of the Welsh Assembly’, Regional Studies, 38: 213–23. Lenschow,A., Liefferink, D. andVeenman, S. (2005) ‘When the birds sing: a framework for analysing domestic factors behind policy convergence’, Journal of European Public Policy, 12 (Oct): 797–816. McGarvey, N. and Cairney, P. (2008) Scottish politics, Basingstoke: Palgrave. Marsh, D. and McConnell, A. (2010) ‘Towards a framework for establishing policy success’, Public Administration, 88 (2): 564–83. Marsh, D. and Sharman, J.C. (2009) ‘Policy diffusion and policy transfer’, Policy Studies, 30 (3): 269–88. Martin, S.J. (2000) ‘Implementing BestValue: local public services in transition’, Public Administration, 78 (1): 209–27. Martin, S.J. and Webb,A. (2009) ‘Citizen-centred public services: contestability without consumerism’, Public Money & Management, 29 (2): 123–30. Martin, S.J., Downe, J., Grace, C. and Nutley, S.M. (2010) ‘Validity, utilization and evidence-based policy: the development and impact of performance improvement regimes in local public services’, Evaluation, 16 (1): 31–42. Pemberton, S. (2009) ‘Editorial’, Policy & Politics, 37 (3): 313–16. Radaelli, C.M. (2005) ‘Diffusion without convergence: how political context shapes the adoption of regulatory impact assessments’, Journal of European Public Policy, 12 (5): 924–43. Raffe, D. (2006) ‘Devolution and divergence in education policy’, in J. Adams and K. Schmuecker (eds) Devolution in practice 2006: Public policy differences within the UK, London: IPPR. Rose, R. (1991) ‘What is lesson-drawing?’, Journal of Public Policy, 11 (1): 3–30. Rose, R. (1993) Lesson-drawing in public policy: A guide to learning across time and space, Chatham, NJ: Chatham House. Rose, R. (2004) Learning from comparative public policy: A practical guide, London: Routledge. Shortridge, J. (2009) ‘Debate: ten years of devolution’, Public Money & Management, May, 142–3. Stone, D. (2000) ‘Non-governmental policy transfer: the strategies of independent policy institutes’, Governance, 13 (1): 45–62. Stone, D. (2004) ‘Transfer agents and global networks in the “transnationalisation” of policy’, Journal of European Public Policy, 11 (3): 545–66. Wolman, H. and Page, E. (2002) ‘Policy transfer among local governments’, Governance, 15 (4): 477–501.

Sandra Nutley ([email protected]), University of Edinburgh Business School, Edinburgh, UK James Downe, Steve Martin and Clive Grace Centre for Local & Regional Government Research, Cardiff Business School, UK
