Territory, power and statecraft: understanding English devolution

ABSTRACT In recent decades, the devolution of power to subnational regional authorities has formed a key element of what has been termed the ‘unravelling’ or ‘unbundling’ of the state in many parts of the world. Even in the United Kingdom, with its distinctive global reputation as a power-hoarding majoritarian democracy, the devolution of powers to Scotland, Wales and Northern Ireland since 1998 can be located within this broader devolutionary dynamic. In recent years, this process has focused on ‘the English question’ and a reform agenda that claimed to offer a ‘devolution revolution’. This paper offers the first research-led analysis of the scope, scale and implications of these post-2015 reforms to English governance. It utilizes Jim Bulpitt’s statecraft approach to explore the changing nature of centre–periphery relationships within England. The main conclusion has been that a ‘rhetoric–reality gap’ currently exists and a ‘devolution revolution’ has not occurred.


INTRODUCTION
Over recent decades, many countries have witnessed the decentralization of state functions in the form of both political devolution and administrative deconcentration (Asthana, 2013). Hooghe et al. (2016), for example, assert that there has been a global trend towards decentralization. Of the 52 countries they examined, two-thirds have witnessed an increase in their levels of regional authority. Such restructuring is viewed by commentators as a response to a variety of pressures including managing distinct subnational identities and cultures (Tang & Huhe, 2016), mitigating economic diversity (Martin, Pike, Tyler, & Gardiner, 2016), relieving the political and bureaucratic burden associated with centralization (Organisation for Economic Co-operation and Development (OECD), 2015), and changing political views on the contribution of decentralization to achieving economic and social policies (Hambleton, 2015;Jessop, 2016). In recent years the UK has witnessed far-reaching devolutionary reforms to Scotland, Wales and Northern Ireland and the introduction of an elected Mayor and Assembly in London (Flinders, 2009;Sandford, 2016b). And yet the paradox of these devolutionary measures was that (1) they were introduced within the constitutional parameters of the Westminster Model; and (2) they were periphery focused in geographical terms with little obvious thought to how by far the largest constituent nation of the UK (i.e., England) might be included. In May 2015, a Conservative government was elected with manifesto commitments 'to devolve powers and budgets to boost local growth in England' (Conservative Party, 2015, p. 1). Since then the government has rapidly initiated a potentially far-reaching reform agenda, statutorily underpinned by the Cities and Local Government Devolution Act 2016.
The aim of this paper is to analyze whether the post-2015 reforms to English governance have transformed centre-periphery relationships, and whether the approach of the current Conservative government is markedly different to that employed by previous UK governments. It draws upon the analytical framework and historical lens developed by Jim Bulpitt in his Territory and Power in the United Kingdom (1983). Bulpitt adopted the controversial position that the study of UK territorial politics had overemphasized the role of ideas and ideology, to the detriment of a far more basic focus on the art of governing and practical politicswhat he called statecraft. Crucially, he sought to examine territorial relations through a historical analysis of territorial politics in the UK and from the perspective of central government. For Bulpitt, territorial politics has been characterized by the relationship between the 'centre' and 'periphery'. The centre is defined as a 'political administrative community of senior ministers and top civil servants' and 'the periphery, or country, was usually, from the centre's viewpoint, all other places' (p. 3). He makes a distinction between 'high politics' and 'low politics'. The former involve matters that are regarded as primarily the responsibility of the centre and might include, for example, macro-economic policy and tax-raising powers. The latter covers those residual matters that in normal circumstances could be left to interests on the periphery. Bulpitt argued that the centre's operational code is determined primarily by its desire for autonomy over high politics and the reciprocal granting of control to the periphery over low politics. He described the separation of powers between centre/periphery and high/low politics as a 'Dual Polity ' (p. 3). Much of what shapes territorial politics could be explained by challenges to the dual polity equilibrium and the response of the centre. His framework has great contemporary relevance in that it offers a historically grounded framework to capture the challenges and changes in centre-periphery relations in England.
The analysis presented in this paper draws upon the findings of three research projects undertaken between March 2015 and August 2016. The core conclusion is that English governance sits within a well-established framework of centre-periphery relationships. Claims regarding the 'revolutionary' nature of this agenda are not therefore sustainable. What is identified is the emergence of an increasingly complex institutional landscape across England, lack of public engagement or public understanding, and even a pushback from local actors as the implications of the government's agenda become clearer.
The findings in this paper draw attention to the importance of understanding the 'politics' shaping the trajectory of English devolution. The paper makes a distinct contribution to the political science and regional studies literature in three distinct ways. First, by exploring the codes, strategies and resources employed by key actors to shape English local governance, it provides clarity on the political imperatives underpinning current devolution reforms. Second, it offers an intriguing insight into how institutions actually work and what drives territorial reform in the UK. Third, it explores how political momentum and control ebbs and flows in the policy process and why this makes territorial reform in England so challenging. This contribution has broader international implications for the analysis of territorial politics, multilevel governance and executive politics.
The paper is structured as follows. The next section provides a brief descriptive account of the evolution of English devolution policy, which highlights the longterm absence of any clear constitutional vision. Then it outlines Bulpitt's 'statecraft approach' to understanding centre-periphery relationships and argues that it has much to offer a contemporary analysis of UK territorial politics. The third section sets out the paper's methodological approach, while the fourth section presents the empirical analysis, reflecting the core themes in Bulpitt's thesis: governing codes, political resources and governing strategies. This supports the core conclusion that the 'devolution revolution' should actually be interpreted as a continuation of the dual polity equilibrium that has characterized UK territorial politics for at least half a century. The final section reflects upon the broader empirical and theoretical implications of this argument.

CONTEXT
England has been a landscape of almost permanent administrative reconfiguration and rescaling during the second half of the 20th century (e.g., Banks, 1971;Garside & Hebbert, 1989;Mawson & Bradbury, 2006). Devolution in Scotland, Wales and Northern Ireland was associated with identity and democratic representation, and the devolution debate in those areas has been founded upon the creation of strong subnational institutions with democratic elections, significant policy responsibilities and accepted status within the unwritten UK constitution. This bottom-up public pressure has been almost entirely absent within England. Over at least the past 50 years, UK governments of various colours have sought to deconcentrate powers and functions to regional or local tiers of government. The Labour governments of the 1960s established regional economic and planning councils for nine regions in England. These developed a considerable intelligence capacity but lacked political and economic clout, eventually being abolished in 1979 (Hogwood & Keating, 1982). Under the Margaret Thatcher Conservative governments, private-sector, geographically focused urban development corporations were used to catalyze economic development in deprived urban areas. Spatial scales shifted again in 1992-94 with the creation of nine government offices for the English regions to coordinate bids for European Union funding along with numerous strands of development funding from central government.
The Tony Blair and Gordon Brown New Labour governments sought to draw together the constitutional/territorial dimension of the 'English question', but in practice their policy was equivocal in its effect. Deputy Prime Minister John Prescott drove the establishment of regional development agencies (RDAs), alongside expanded government offices for the regions, and 'regional chambers' bringing together local government leaders and regional stakeholders. The ultimate aim was the introduction of elected assemblies in the regions of England. However, a first referendum, in the North East, saw an elected body rejected by a large margin in November 2004, and the policy was quietly abandoned. In 2007, the 'regional chambers' were replaced with 'local authority leaders' boards' (Ayres & Stafford, 2009), and government policy began to shift towards city-regions and functional economic areas (FEAs) (Ward, Deas, Haughton, & Hincks, 2015). Thus, successive governance 'solutions' within England have been imposed by the centre, and have been driven purely by an economic/technocratic outlook (Tomaney, 2016). Devolution in Scotland, Wales and Northern Ireland constituted a break with the practice of 'power-hoarding majoritarianism' by the UK central government, but this remains in place with regard to England (Rees & Lord, 2013;Richards & Smith, 2015). Since 2010, local areas have enjoyed greater discretion over the formation of supra-local governance and partnership arrangements (Pugalis & Townsend, 2013;Waite, MacLennan, & O'Sullivan, 2013), but this has not been accompanied by the transfer of additional power.
The current English devolution policy initially hinted at a move away from this impasse. It evolved during George Osborne's term as Chancellor of the Exchequer (May 2010-July 2016). It eschewed a constitutional approach: the new policy cast devolution more as a functionally efficient means to achieve agreed policy outcomes. This served both to challenge local areas to articulate their desired outcomes instead of entitlements to 'powers', and to reassure central government participants that they would retain some control when power was devolved (Political Studies Association (PSA), 2016). The early focus was upon 'city deals', negotiated between the government and local representatives of FEAs, twinned with competitive bid-based funding of specific development and regeneration projects. But doubts persisted about the degree to which these shifted power and responsibility (Pike & O'Brien, 2015).
In response to growing stakeholder and think-tank support (e.g., City Growth Commission, 2014;Cox, Henderson, & Raikes, 2014;Morrin & Blond, 2014), city-deals were overtaken by a series of 'devolution deals'. These began with the Greater Manchester Agreement in November 2014. They have consisted mostly of pan-urban 'combined authorities' with directly elected mayors managing a spread of new powers and programmes (Gains, 2016;Pike, Coombes, O'Brien, & Tomaney, 2016).
To date, as Table 1 illustrates, a number of other devolution deals have been agreed between central government and specific English localities, with six 'city-region' mayors expected to be elected in May 2017. However, central-local relations surrounding the process have frequently been fraught, and a number of deals have collapsed after being agreed. A deal for the large rural area of East Anglia was split in two (Cambridgeshire/Peterborough and Norfolk/ Suffolk); several small district councils in Norfolk rejected the successor deal, and the government withdrew support. Similar events played out in Greater Lincolnshire, driven in large part by the requirement for directly elected mayorsan unofficial deal breaker for the government. One participant in a successful deal area described the mayoralties as 'a price worth paying' (Lord Peter Smith; House of Lords Hansard, 22 June 2015, c1412). Areas that sought alternative governance arrangements for devolution arrangements simply found that no deal was forthcoming. The then Chancellor, George Osborne, stated that 'I will not impose this model on anyone: but neither will I settle for less' (Osborne, 2015).
Thus, central preferences were decisive in determining the structures through which locally devolved powers were to be exercised. This was allied to opaque and largely secret deal negotiations, also the subject of considerable critique Prosser, Renwick, Giovannini, Sandford, & Flinders, 2017;Sandford, 2016b). Participants may have perceived this too as a price worth paying to obtain new powers, but it stymied the process in a number of localities. Research has also indicated that the scope of devolution was firmly limited in practice by government priorities (Communities and Local Government Committee, 2016;PSA, 2016;Sandford, 2016b). At the time of writing, the prognosis that devolution deals provide a successful formula for devolving power in England is, therefore, under challenge.
The new mayors will take office in the context of significant cuts to local public spending (Innes & Tetlow, 2015; National Audit Office (NAO), 2016), and the departure of Osborne, architect of the 'Northern Powerhouse' agenda that provided the backdrop to many deals, from the government in July 2016. This makes understanding the forces underpinning developments in English governance even more timely, especially in relation to the territorial dynamics and constitutional forces that may have (intentionally or unintentionally) been unleashed. As Tony Wright has argued, constitutional reforms in the UK are generally not explosive or revolutionary but generally occur through the gradual widening of initial 'cracks' in policy areas: 'It is a "crack" that provides space for "wedges" that extend [what] has already been achieved' (Wright, 2004, p. 870). The question arises whether, after 50 years of imposition of territorial 'solutions' on England by central government, the latest generation of devolution policy can be a 'crack' that leads to more enduring shifts in governing relationships between English localities and the UK government. Scotland, Wales and Northern Ireland have seen substantial change in their relations with the UK centre, culminating in their recognition in statute as permanent institutions (a significant, if purely declaratory, development). By contrast, analysis of central-local relationships in England has often been crowded out by updates on the rapid developments in devolution deals (for a detailed review, see Sandford, 2016a). In order to move beyond description, it is necessary to place recent events within a conceptual and theoretical framework that is historically attuned to the British political tradition. Thus, the next section outlines Bulpitt's statecraft theory, which is capable of exploring the embedded politics of devolution in the sense of the codes, strategies and resources through which new centre-periphery relationships are negotiated.

THEORY
This section outlines the core elements of Bulpitt's framework and explains how it can be utilized in relation to contemporary developments. It also situates Bulpitt within the broader regional studies literature on English devolution to demonstrate both its complementarity and capacity to advance our understanding.
Bulpitt's framework Bulpitt's framework identifies options in relation to governing codes, political resources and governing strategies, which can generally be used to identify one of four broad models of centre-periphery relations. Bulpitt ascribes analytical primacy to 'the Court', a small number of individuals that form the decision-making political elite. As rational actors, they will be primarily focused on 'the art of winning elections and achieving some necessary degree of governing competence in office' (Bulpitt, 1986, p. 21): that is, statecraft. Winning elections and displacing blame are the primary drivers of any decision-making calculation and in this regard 'governing codes' play a crucial role. Governing codes form a set of relatively coherent principles or rules underlying policies and policy-related behaviour and five were identified (Table 2).
On resources: he suggested that a centre aspiring to dominance would seek a hegemonic unionist culture, a constitution that obstructed periphery articulation, a bureaucracy with strong territorial penetration, a mass party system controlled by central politicians, the economic resources necessary to dispel territorial deprivation and time to devote to peripheral politics. (Bradbury, 2006, p. 567) These resources are elaborated further in Table 3. If one assumes that the centre is aiming for maximum possible control of the periphery, Bulpitt suggests that 'it will need to possess most (just how many is not clear) of [these] resources' (Bulpitt, 1983, p. 63).
In terms of implementing and sustaining a territorial code, the national government could adopt a 'hands off'  laissez-faire strategy, which implies 'an indifference to control over economic and social responsibilities' (Bradbury, 2006, p. 567). It might adopt a 'hands on' promotional strategy that necessitates central intervention, but which can be costly. Finally, an arm's-length regulatory strategy involves governing at a distance and may have contractual characteristics (Table 4). A key issue to note is that different sections of the same central government department often possess different strategic intentions that can change over time (Ayres & Stafford, 2014). 'Moreover, the centre may adopt different strategies in different peripheries, and the intentions as listed may be designed to produce different results' (Bulpitt, 1983, p. 63). This mixture of codes, resources and strategies led Bulpitt to identify four broad models of centre-periphery relations ( Table 5).
The central autonomy model (M1) seeks to produce a structure of territorial politics in which the centre is prepared to allow considerable 'operational autonomy to peripheral governments and political organizations, so long as they do not challenge its autonomy over matters of 'High Politics' (p. 65). This assumes a capacity on behalf of 'the Court' to insulate the centre from peripheral penetration on matters of importance. The capital city bargaining model (M2) suggests more 'interference in the centre's affairs but often in a cooperative fashion and generally only on issues of interest to the periphery' (Bradbury, 2006, p. 568). The central authority model (M3) describes a situation whereby there may be some periphery support for the centre's mandate but there is a consistent need for continuous central assertion in the periphery. Finally, the coercive power model (M4) involves systematic central intervention and use of threats. However, this model is resource intensive, unstable and not democratically desirable.
The nature of the specific relationship or model owed less, according to Bulpitt, to ideology and more to pragmatic politicsor statecraftand the need to adopt an efficient and proportional relationship. This was demonstrated through a historical methodology that identified several stages of centre-periphery relations in the UK. Using this framework Bulpitt provided 'the long view' of the UK's territorial politics and identified 'the central autonomy model' (Table 5, M1) as the dominant model of centre-periphery relations for at least three centuries. British territorial management was therefore a history of continuity rather than change based around a 'dual polity' whereby the national government preferred to concentrate on 'high politics'. Analyses of Scotland, Wales and Northern Ireland through a Bulpittian framework have concluded that the central autonomy model remains dominant there despite the strength of devolved powers (Bradbury, 2006;Convery, 2014). Our Bulpittian analysis continues in this historical and qualitative vein to look at the political imperatives shaping the most recent chapter in English territorial governance.
Bulpitt's contribution to the regional studies literature on English devolution Our analysis contributes to ongoing and pertinent debates in the regional studies literature. Below, we illustrate this by drawing on a number of recent studies that have particular significance for Bulpitt. These studies are relevant because they deal directly with territorial management and two key facets central to the current devolution deal debate, i.e., drives to boost economic development and attempts to reduce spatial disparities. They also highlight the importance of historical context, the path dependent nature of territorial management and the centrality of statecraft in managing centre-periphery relations. The following discussion explores the connection between our Bulpittian analysis and this body of work in the regional studies literature. Pike et al. (2016), for example, recognize the importance of historical context and institutional legacies in shaping pressures for territorial change and continuity. In their recent analysis of the abolition of RDAs in England they explore how actors in particular political and economic contexts shape and negotiate transitions to new territorial arrangements. Likewise, Rees and Lord (2013) suggest that the creation of local enterprise partnerships (LEPs) in city-regional areas in the early 2010s in England was 'as much about the realpolitik of finding workable governance arrangements and consonant political objectives as they were about uncritically following economic logic' (p. 684). Like Bulpitt, they recognize the pivotal role of critical actors at the centre and how politics serve to shape territorial reform over a more evidence-based approach. Their  Territory, power and statecraft: understanding English devolution qualitative analysis, based on the views of key insiders, offers an illuminating account of English territorial reform from the perspective of those closest to the process. Pike, Rodriguez-Pose, and Tomaney (2017, p. 47) 'trace the evolution of thinking about local and regional development in order to situate current debates in their sometimes neglected historical context'. While Bulpitt takes a historical political stance steeped in the British tradition, Pike et al. explore territorial rescaling in a global context by drawing on international literature on macroeconomic factors, such as global financial flows and agglomeration economies, and the roles of creative individuals and networks in shaping place and territory. They find that, despite calls for a common theoretical language on local and regional development, there are 'highly variegated international practices and experiences' (p. 47). Martin et al. (2016) examine English governance by examining the persistence of spatial inequalities and the challenges of rebalancing the economy. They do so from a spatial economic perspective. Like Pike et al. (2017) and Bulpitt, they also take a historical perspective, examining economic growth data from 1871 to conclude that spatial economic imbalances, most notably between the north and south of England, are deeply rooted in the highly 'centralized nature of the national political economy' (Martin et al., 2016, p. 342). They conclude that 'only a bold and radical change in that political economybased on a devolution and decentralization of economic, financial and political poweris called for' (p. 342). While it is too early to pass judgement on the eventual impacts of the current devolution deals, our Bulpittian analysis offers a valuable insight into whether the Conservative government's devolution plans represent early evidence of the radical change that Martin et al. call for in their recent work.
The issue about entrenched UK government positions and the path-dependent nature of English territorial management is also identified by Harrison (2007) and Etherington and Jones (2016). Harrison (2007) explores the inability of spatial reforms in England to deal with the problems of rebalancing the UK economy, inequality and spatial injustice. He concludes that 'the spatial and scalar reorganisation of the state from one scale to anotherin this case from the region to the city-regionmerely distracts attention away from the state's inability to manage the capitalist tendency for uneven development' (p. 328). In a similar vein, Etherington and Jones (2016) explore the failure of more recent spatial reforms around the devolution deals and localism to deal with spatial inequalities. Like Harrison, they conclude that 'failure arises because of the primacy of a neoliberal-dominated strategy orientated towards the market' (p. 371). Interestingly, they draw on notions of governance and metagovernance failure as important to understanding both the limitations to and contradictions of devolution and city-region building in England. They argue that: 'Metagovernancethe "government of governance" through "over-seeing, steering, and coordinating governance arrangements" (Bell and Hindmoor, 2009: 11)has received minimal detailed attention in urban and regional studies' (p. 373).
Bulpitt's focus on statecraft equates to this description of metagovernance. In his account, the Court seeks to manage the separation of powers between centre/periphery and high/low politics by creating what it views as the right environment or arena for doing business. Below we argue in detail that the evidence from English devolution deals backs up Bulpitt's perspective: but we also argue that aspects of the evidence point towards a shift in the UK government's approach to metagovernance. Our detailed analysis of the UK government's codes, resources and strategies complements the valuable work of Etherington and Jones (2016)  These questions facilitate the presentation of the research findings and underpin the core argument: that the post-2015 devolution reforms should be understood as a continuation of 'the central-autonomy model' (Table 5, M1).

METHODS
The evidence presented in this paper has been collected through three research projects conducted between March 2015 and August 2016. Key data and findings from across the three datasets were triangulated to ensure the validity and reliability of the findings presented. This involved verifying core observations using data emanating from all three projects. The first project (P1)  The second project (P2) was an Economic and Social Research Council (ESRC)-funded project that focused on devolution to two English regions (South Yorkshire and Hampshire/Isle of Wight). This project received face-to-face comments from 37 expert witnesses including officials, policy-makers, special advisers, former ministers and serving politicians from local and central government. It adopted a comparative case design whereby one 'pure' assembly (Assembly North) consisted of 32 members of the public; and one 'mixed' assembly (Assembly South) consisted of 23 members of the public and six politicians (for a full methodological statement and analysis, see Crick Centre, 2016). The insights from P1 and P2 are used to inform and substantiate the analysis and argument offered in this paper. The final project (P3) tracked the devolution deals that were announced (and often amended) during 2015 and 2016 and also analyzed a vast set of primary governmental reports, ministerial statements, parliamentary papers, plus 'grey literature' from think tanks, consultants, professional peak organizations etc. This project was based within the House of Commons Library and ensured that the other projects were informed by the very latest and up-to-date information.

RESEARCH FINDINGS
This section draws upon the data and findings collected in the three research programmes discussed in the methods section, and structures them around the statecraft framework that was the focus of the theory section in order to delineate the dominant model of centre-periphery relationship and therefore the extent of change.

RQ1. What territorial management code has shaped English devolution under the current Conservative government?
Findings indicate that the governing code is one of 'central control of local governments' (Table 2, C3) but with elements of 'local elite assimilation' (C2) and weak 'citizen mobilization' (C5). C3 is framed around imposing legal frameworks, policy objectives or creating new incentives. In this context the devolution deals are really little more than a new form of top-down control whereby new incentives are offered (primarily financial or related to policy discretion) within 'the shadow of hierarchy'. Thus, the text of the deals will govern the central-local relationship, and stakeholders at the focus group (P1) were in agreement that central government was dictating how this 'text' was to be developed. While senior Department for Communities and Local Government (DCLG) and HM Treasury officials outlined a 'desire to devolve as much as possible to local areas' they also recognized that power and control was to be 'granted by central government and not taken from it' (P1). A senior official from a think tank suggested that the guidelines for developing deals 'were purposefully vague and light touch to provide the government with the wriggle room required to seek the outcome they wanted' (P1).
Lack of formal or written guidance affords repeated opportunities for the enforcement of the centre's norms, and this has become more systematic in more recent deals. For instance, the West Midlands deal, signed in November 2015, states that it will 'develop a business case for an innovative pilot to support those who are hardest to help. The business case should set out the evidence to support the proposed pilot, cost and benefits and robust evaluation plans … ' (HM Treasury, 2015b, p. 11). A number of 'readiness conditions' are stipulated for the West Midlands to take on 19+ skills funding, including the completion of a joint 'area review process' and finalizing of risk-sharing arrangements. More recently, the three March 2016 deals include a provision for the authority: [to] develop a full implementation plan, covering each policy agreed in this Deal, to be completed ahead of implementation. This plan will include the timing and proposed approach for monitoring and evaluation of each policy and must be approved by the DCLG Accounting Officer. (HM Treasury, 2016, p. 20) Other elements of this governing code involve a degree of 'local elite assimilation' (C2), largely through appointments to the LEPs that have spearheaded centre-periphery negotiations on the part of localities (NAO, 2016;Pike et al., 2016). Deal negotiations were generally conducted in secret between LEP representatives, council leaders and the government. The publication of the final deal was the first point at which most local councillors, stakeholders and members of the public knew about the agreement (Blunkett, Flinders, & Prosser, 2016). Indeed, this view is confirmed by Kenealy (2016, p. 1) who claims that in the Manchester deal, for example, 'key decisions were taken quickly and by a small number of key officials'. The vast majority of interviewees referred to a 'streamlined and closed process' (P1, local government official) whereby central and local actors could exert strong leadership (Bentley, Pugalis, & Shutt, 2016). Interestingly, the majority of respondents in P1 research expressed their support for closed discussions and process-light arrangements. As an HM Treasury official indicated, 'something is lost if you try to standardize process too much. In many instances informality has afforded stakeholders the right environment to broker deals for their areas' (P1). Members of the public wereunsurprisinglyfar less impressed with a dominant governing code that seemed to operate through a set of informal, secretive, elite-to-elite relationships. This was a core finding of P2 where members of the citizens' assemblies overwhelmingly voted in favour of greater public participation in order to legitimate the devolutionary processes. More specifically, when able to decide and then vote upon the key priorities for the new devolved governance arrangements 'public engagement and transparency' emerged as a leading issue (Crick Centre, 2016, p. 33). Furthermore the analysis of subsequent rejections of 'deals' by local councilssuch as North Somerset, Gateshead, Erewash, Norwich, Breckland, Great Yarmouth etc.suggests that a concern regarding public engagement or 'the missing link in the devolution debate'  has eroded confidence in the reforms (Hammond, 2016). An official from a leading think tank opined: 'there is a very real chance that devolution deals will fall apart at the implementation stage without the commitment from a broader range of stakeholders' (PI). And yetparadoxicallya weak strain of 'citizen mobilization' (C5) runs through the government's plans, as a commitment to accept a directly elected mayor was generally a government requirement for the granting of a deal (HM Treasury, 2015a, p. 1).
Two other features of the process resonate with Bulpitt's characterization of centre-periphery relations in the UK. The first relates to Bulpitt's focus on 'Court politics' and the small number of key individuals that form the decision-making political elite. In relation to post-2015 English devolution, the 'Court' was very small with the policy being personally associated with and driven by one person: George Osborne (Ayres, 2017). This might explain the manner in which several Whitehall departmentse.g., the departments for Education and Skills and Work and Pensionsseemed reluctant to engage in devolutionary discussions and why negotiations over specific policy areas (e.g., housing, pre-16 education etc.) often ended abruptly. This variation in Whitehall commitment to policies was a key element of Bulpitt's analysis as he emphasized how different departments may have very different strategic intentions, and that these may ebb and flow over time (Pike & O'Brien, 2015).
A second issue relates to 'blame games' and 'blame displacement', which Bulpitt saw as a key factor in central government's selection of a governing code. In the context of far-reaching public sector budget cuts a number of interviewees (P1) and assembly members (P2) viewed the deals as an attempt to devolve responsibility for making unpopular cuts to public services. Likewise, a DCLG official outlined 'the two acid tests for deal proposals are that local areas can do it cheaper and better than the centre' (P1). This raises the question of why local authorities would risk becoming a lightning rod for centrally imposed public cutbacks in return for relatively modest financial and policy incentives. The answer can be found in Wright's (2004) 'cracks and wedges' thesis. Local authority leaders have consistently suggested that the importance of the deals lays not in their current initial content but in their potential to evolve into a quite different relationship in the future. For example, a recent Localis and Grant Thornton (2015, p. 16) report a local authority official described the devolution deal process as 'the most positive negotiating experience I've ever had. The government genuinely tried to do this in partnership. The civil service is on board … '. This view was echoed by a number of local government respondents during interviews (P1 and P2) and the focus group (P1). English localities were therefore adopting a pragmatic position based on the realpolitik that (1) local authorities are under increasing fiscal pressure, (2) 'this was the only deal on the table' (local government official, P1) and (3) the constitutional balance of power and resources lies firmly in the hands of the centre.
RQ2. What political resources are being invoked to promote the governing code?
Evidence suggests that all the resources identified by Bulpitt (Table 3) are still available to the centre in England (though they have weakened considerably in recent years in relation to Wales, Northern Ireland and especially Scotland). However, they are not being used as assiduously as in previous eras; and this has opened the door for a weakening of the central autonomy approach. The resources are examined in turn.
The existence of a 'hegemonic unionist culture' (R1) seems to have waned (Mitchell, 2002). Political disputes in England are mostly refracted through traditional party politics and cleavages of class and economics, not through territorial claims. But a 'hegemonic culture' could also be interpreted as an acceptance by local areas of the government's narrative of the role and purpose of English devolution. This lies behind the assumption that local areas seeking devolution must 'give the confidence to devolve' (Localis & Grant Thornton, 2015, p. 32), by demonstrating how they can deliver improved outcomes within the framework of the government's approach. A local government official (P1) described 'building relationships and assurances with central government as critical to securing a deal'. Bulpitt's (1983, p. 63) second requirement is a constitution which 'does not positively assist the articulation of peripheral interests' (Table 3, R2). There are no formal mechanisms of intergovernmental relations within England, and no representative apparatus such as an upper house of parliament. Local authorities compete with members of parliament to articulate local interests and their influence is dependent on the quality of their relationship with Whitehall (Wills, 2016). Thus, the central autonomy model continues to dominate on paper. But there is evidence that the devolution deal process has constituted a limited break with this orthodoxy. Local participants have reported that it is the first genuine attempt to transfer power that they had been involved in (Cox & Hunter, 2015). Central and local government interviewees described 'relationships as far more collaborative and less adversarial than in the past' (central government official, P1), and there is evidence of some localities pushing back against central demands (see below). A senior official from the Greater Manchester Combined Authority described their relationship with the centre as based on 'mutual understanding and shared goals' (P1), although it was clear from the focus group that not all localities shared this positive experience (Henderson, 2015).
The question of a 'central bureaucratic machine' (R3) as a mechanism of governance is more nuanced. A recurring theme from interviewees in both central and local government was a concern that central government lacks the resources to negotiate and monitor increasing numbers of devolution deals (P1). This would hinder the functioning of a central autonomy model, but some local respondents also saw a lack of central resources as a threat to devolution: it could affect (1) the quality of central-local negotiations, (2) the ability to explore more innovative and creative possibilities, and what one local government respondent referred to as (3) 'parity of opportunity if not parity of outcomes to all areas' (P1). One view mooted by a number of interviewees was the need to strengthen the 'local bureaucratic machine', bringing stronger negotiating partners for government into being.
The resource of a 'mass party system' (R4) is not as strong as it may appear. The leading devolved areas are all Labour dominated, and a number of interviewees commented on their political alliance with a Conservative government. A think tank official (P1) stated that: 'Labour-led areas had seen the potential in devolution and were less inclined to play party politics.' Finally, 'managing the economy so that peripheral interests are protected from economic deprivation' (R5) and 'sufficient time to devote to peripheral politics' (R6) have not been utilized to great effect. The global financial crisis and ensuing austerity measures in the UK have in fact exacerbated concerns about deprivation, social justice and spatial inequalities in England.
The following section examines the governing strategies shaping centre-periphery relations.
RQ3. What governing strategies are being pursued to implement the governing code?
Research suggests that the governing code is mainly pursued via 'hands on' (Table 4, S2) intervention. There was a perception amongst interviewees (P1) and expert witnesses (P2) that Whitehall has its 'priority areas to secure a deal with, whether for political or economic motives' (academic interviewee, P2). Indeed, respondents referred to informal and off-the-record discussions between so-called Westminster 'big hitters' and localities with a view 'to knocking heads together to secure a deal' (central government official, P1).
But despite the enduring presence of many tools of control associated with central autonomy, data also suggest the beginnings of a shift in central-local relations. Some local government interviewees reported that government was 'desperate' to conclude deals (P1)contrasting with its claims that it did not intend to oblige any areas to do so. Regarding the negotiations themselves, the data suggest that the government did not hold all the cards. One local government interviewee reported that a local leader had been able to extract concessions from the government: 'she threatened to walk out. She literally had her coat on and was in the doorway, and they saidall right' (P1). Some participants at the focus group (P1) also reported having successfully 'pushed back' the government on a number of points where initially the government had refused to consider devolution and on firming up commitments to delivery within the deal texts. This suggests an inkling of a long-term relationship with more affinity to the 'capital city bargaining model' (M2), where the localities negotiate on more equal terms. More recently, sector representatives have suggested in the media that devolution is floundering amidst a loss of momentum in the centre: this hints again that any move towards this model would be dependent on central government intentions.
The mayoral role demonstrates the ambivalence of the centre's position. The newly elected mayors may be able to articulate, defend and satisfy their interests within the institutional complex of the centre using their democratic mandate. But the constraints of the deals may require them to take on the 'less gratifying functions' of government, with little leeway for independent action. The requirements for monitoring, evaluation, and implementation plans in the deals suggest that the centre intends to maintain hands-on control via output targets. How this translates into practice depends upon two main factors: whether central government will have the resources (or the will) to monitor in this way and how far elected mayors will be able to push back against such control. This brings us to reflect on the wider implications of the analysis of English regional devolution offered in this paper.

CONCLUSIONS
This paper has explored recent developments in the governance of England. It has achieved this by presenting the findings of three interrelated research projects using Bulpitt's statecraft approach. The main conclusion has been that a 'rhetoric-reality gap' currently exists and a 'devolution revolution' has not occurred. Bulpitt's 'central autonomy model' remains the dominant mode of central statecraft and to some extent the recent deals might be more accurately described not in the language of 'devolution' but as a 'new partnership' in centre-periphery relationships. This finding demonstrates the continuation and intensification of the transactional approach identified by Pike & O'Brien (2015) in their analysis of city deals. Indeed, this is exactly the interpretation promoted by the former Conservative Deputy Prime Minister, Lord Heseltine: When one talks of devolution it's not realistic to talk about freedom. This is a partnership concept. Central government are elected and they are entitled to have their manifestos implemented and it cannot be contemplated there is a sense of freedom at a local level which can actually frustrate the clear mandates upon which governments are elected. … I am sympathetic to the word 'partnership' rather than 'freedom' or 'devolution'. (Heseltine, 2015, p. 1) To some extent this conclusion is not surprising and has been noted elsewhere (Tomaney, 2016). Despite the hyperbole of politicians rhetoric-reality gaps are far from uncommon in polities around the world and the British political tradition is well known for being evolutionary, organic and conservative. The question is really then one of what makes these findings particularly distinctive or of relevance to political scientists and regional studies scholars around the world? Three brief responses help to tease out the broader relevance of the research presented in this paper. The first focuses on the issues of clarity; the second on how institutions work and the third on policy momentum and control. At a very basic level one of the most significant contributions of this paper has been to offer an accurate, coherent and up-to-date snapshot on the reform of English local governance. The importance of mapping the changing topography of this institutional landscape should not be understated given the rather ad-hoc, sporadic and confused rollout of this agenda to date. The key strength of Bulpitt's framework in this context lies in the manner in which it dissects the underlying strategies and resources of key actors. The descriptive element of this study has therefore identified the emergence of an increasingly complex and asymmetrical governance framework at the subnational level.
This leads into a related focus on how institutions work and what drives reform agendas. The work of George Tsebelis (Tsebelis, 2002) is particularly relevant in this context due to the manner in which it develops the concept of veto players within a legislative context and in relation to specific policy areas. A veto player isput simplya player in the political game with the capacity to prevent change from the status quo. The conceptual value of this approach to understanding how institutions work lies in the manner in which it reveals the existence of embedded but generally unequal or asymmetrical distributions of power within and between institutions. Any reform will therefore generally produce 'winners' and 'losers'. This explains the ebb and flow of reform demands and the role of crises in potentially jolting systems out of their pre-existing pathways. Bulpitt's analysis underlines the role of the national government as the key veto player with the capacity to facilitate or block changes to the status quo.
The almost embedded obstacle to reform is therefore an unwillingness amongst those political elites that currently benefit from the existing status quo to support reforms that are explicitly designed to shift the balance of power. In this context comparative political and institutional analyses have consistently highlighted the importance of two factors: crises and constitutional entrepreneurs. Crises open 'windows of opportunity' by potentially jolting institutions or constitutional configurations out of established pathways and thereby facilitating the introduction of new structures and relationships. The opening of a 'window of opportunity' is not on its own enough to guarantee that an opportunity for change will be taken. Established political elites are likely to seek to frame or control the crisis quickly in such a way that the reform parameters are reduced and the 'window' is quickly closed. In this context the role of an individual with a personal and explicit commitment to a specific reform agenda and who has the capacity to remove 'executive blockages' (i.e., a constitutional entrepreneur) is often vital. This account can be drawn upon to explain failures in the meta-governance of English devolution. This historical institutionalist account complement explanations of failure based on the predominance of a neo-liberal strategy put forward by Etherington and Jones (2016). This Bulpittian analysis also provides an in-depth account of the political drivers and constraints shaping governance reform in England that contributes to the international literature on understanding the context specific factors shaping territorial reform globally (Pike et al., 2017).
In this instance, a 'window of opportunity' was opened by a constitutional entrepreneur in the form of the Chancellor of the Exchequer, George Osborne, who was very clearly personally driving the agenda (Hambleton, 2015;Ward et al., 2015). In many ways Osborne's role was to remove the 'executive blockages' that other members of his government placed in the way of the reform agenda. The obvious risk of having a policy attached to a constitutional entrepreneur is that if the individual loses their capacity to shape and control the emerging meta-governance, as Osborne did when he was not offered a place in Theresa May's Conservative government from July 2016 onwards, then the policy agenda may well wither and the 'window of opportunity' close.
This brings us to a third and final point that highlights the broader significance of recent developments in the UKthe issue of policy momentum and control. The British constitution is well known for evolving according to what has become known as the 'cracks and wedges thesis' (Wright, 2004). This suggests that change generally occurs through the imposition of relatively minor concessions in relation to rules, procedures or relationships which then provides a wedge that can be used to expand and develop the initial concession into a more significant change over time. The relevance of this argument for this paper is that it may well be far too soon to really understand the implications of recent reforms to English governance. Elected mayors have the potential to act as 'wedges', pushing for greater local purchase on policy through the 'cracks' afforded by the devolution deals. The interim Mayor for Greater Manchester, Tony Lloyd, has already described the devolution deal to Greater Manchester as 'transformational' and over time this may be true. To make this suggestion is not to undermine the core argument of this paper regarding the maintenance of a clear 'central autonomy model'. And yet looking to the future there is clearly a strong chance that if the economic, social and democratic benefits of 'devolution deals' begin to be realized then the centre-periphery model will shift, implying that the government has ceded some control. Stronger English localities with visible 'metro mayors' and greater governance capacity will, so the theoretical logic and international experience suggests, move into a position whereby they can work with government in partnership to broker arrangements that suit their local circumstances. That is, the 'centre-periphery model' may evolve more towards Bulpitt's 'capital city bargaining model' and the UK's continued existence as an international exemplar of a powerhoarding majoritarian polity may itself have to be reassessed.

DISCLOSURE STATEMENT
No potential conflict of interest was reported by the authors.