The ‘imaginary’ challenge of remaking subnational governance: regional identity and contested city-region-building in the UK

ABSTRACT With the structuring of subnational governance driven primarily by economic goals, an issue that has become increasingly overlooked is that of identity. Drawing on interviews with stakeholders from the Sheffield City Region, the paper builds on Jones and Woods’ framework of 2013 of ‘material’ and ‘imagined’ coherence, demonstrating the ‘imaginary’ challenge of remaking subnational governance in the context of rescaling from regions to city-regions. It shows that historical regional identities can persist even in the absence of associated material components of governance, and that rescaling can create asymmetries between material and imagined coherence, resulting in competing imaginaries that hinder the new subnational arrangements.


INTRODUCTION
A well-established scholarly debate exists in urban and regional studies about the restructuring of economic and political spaces to accommodate the demands of an increasingly networked global economy (Brenner, 1998;MacKinnon, 2011;Swyngedouw, 1997). In contrast to the emphasis placed on 'regionalism' during the spatial Keynesian era as a stable subnational policy platform for delivering public services (Storper, 1997), the focus since has been on finding a form of metropolitan governance that prioritizes and best serves functionality, competitiveness and innovative capacity (Brenner, 2009;Davoudi & Brooks, 2020;Jessop, 2000). Fed by neoliberal rationality (Brenner & Theodore, 2002), this has given rise to a series of alternative and largely experimental efforts to dismantle and reconceptualize conventional regions into more narrowly defined relational spaces (Harrison, 2007;Massey, 2011).
While such decentralization experiments have been premised largely on economic arguments, many have noted the importance of identity in developing effective subnational governance arrangementsone that has nevertheless become increasingly overlooked (Rodríguez-Pose & Sandall, 2008;Van Houtum & Lagendijk, 2001). Some argue that subnational governance structures should reflect the history and identity of communities and thus match community boundaries (e.g., Hooghe et al., 2020). This raises important questions in the context of recent governance rescaling initiatives that have seen historical regions dismantled into a mosaic of city-regions, such as top-down devolution-led city-regionalization in the UK. A major assumption of such governance reforms is that the urban institutions, actors and citizens that occupy these new scalar arrangements will cohere, collaborate and coordinate to bring the proposed workings of a new growth model into effect. Yet, this overlooks the importance of established 'practices', 'relationships' and, importantly, 'identities' that reside in places and that impact, ultimately, the delivery of a new approach. Given that regions have been highlighted as sources of identity (Paasi, 2001;, there are questions as to whether new spatial imaginaries, such as city-regions, will supplant historically and culturally rooted regional identities and foster collaboration at the new level. This paper contributes to debates on the link between subnational governance, spatial imaginaries, identity and politics (Calzada, 2015;Davoudi & Brooks, 2020;Herrschel, 2002;Hincks et al., 2017;Keating, 2014;Paasi, 2009;Vallbé et al., 2018;Zimmerbauer & Paasi, 2013), through conversations with policymakers and entrepreneurs. Drawing on in-depth interviews with stakeholders from the Sheffield City Region (SCR) in the UK, and building on Jones and Woods' (2013) framework of 'imagined' and 'material' coherence, the paper explores the challenges of remaking subnational governance spaces. The research question informing this paper is: Why do some spatial imaginaries become institutionalized and accepted while others encounter resistance? We focus on city-region-building in the UK, which in England saw the regional governance tier in the form of regional development agencies (RDAs) abolished by 2012 and replaced by local enterprise partnerships (LEPs) and combined authorities (CAs) via devolution at the city-regional level.
The findings demonstrate the 'imaginary' challenge of remaking subnational governance, in our case study exemplified by the asymmetry between imagined and material coherence that sees the SCR imaginary, and by extension the city-regional economic development narrative, rejected by some local authorities. We show that a strong regional identity endures within the hinterland and is used strategically by local elites to legitimize demands for alternative scalar Devolution Deals. The ensuing local tensions and resistance to the city-regional arrangements stymie the prospects for SCR to become socially embedded. While 'imaginary', the challenge is real in its consequences, hindering collaboration and participation in city-regional activities.
Our findings contribute to the broader literature on devolution, rescaling and subnational governance, drawing attention to the importance of identity, community imaginaries and imaginary coherence in crafting subnational governance spaces. They highlight the need for devolution to move beyond economic-centric arguments and political preference. We argue that subnational governance arrangements need to exist at a scale that provides imagined coherence and aligns with bottom-up community identities. While the challenges surrounding regional identity in places such as SCR may be seen by policymakers as simple frictions to work through, the extent to which imaginaries become accepted, and the effectiveness of subnational governance, is contingent on subnational arrangements providing a coherent identity to local communities. We discuss the implications for city-regionalism and subnational governance more broadly.

LITERATURE REVIEW
'Identity' in regional economic development The shift from 'old' to 'new' regionalism has transformed identity discourses and changed the way regions are conceptualized. While endorsed by some as the embodiment of people, cultures and traditions associated with territoriality and 'historical depth' (Paasi, 2009;Raagmaa, 2002;Vainikka, 2012;Zimmerbauer & Paasi, 2013), regional identities are also promoted as fluid and adaptive to suit external markets and political projects (Allen & Cochrane, 2007). Yet, despite the force of globalization processes that continue to shape the way we think about regions, cultural differences (Paasi, 2001) and boundaries (Geschiere & Meyer, 1998) continue to matter.
Regions are discursive constructs that are constantly produced, shaped, reshaped and removed by policymakers in the processes of institutionalization and de-institutionalization (Zimmerbauer & Paasi, 2013). Enforced by the state in the form of subnational policy and institutional restructuring, a process of region-building to construct 'closer economic, political, security and socio-cultural linkages between states and societies that are geographically proximate' (Börzel, 2012, p. 255) assists in redefining the norms of centre-regional and intra-regional relationships (Brigevich, 2018). This links to literature on the role of 'spatial imaginaries' used by urban and policy elites to influence political and public opinion of highly selective readings of the form regions should take to promote subnational economic development (Davoudi & Brooks, 2020;Hincks et al., 2017;Hoole & Hincks, 2020). However, these ignore the presence of 'actually existing economies' and pre-existing ideas and imaginaries residing in place that are historically and culturally rooted.
Importantly, spatial imaginaries and (regional) identity are entwined. As 'deeply held, collective understandings of socio-spatial relations' (Davoudi, 2018, p. 101), spatial imaginaries bring places into existence by fostering a shared sense of identity. The alignment of spatial imaginaries employed in subnational governance and place identity is therefore essential for the effectiveness of subnational governance. Davoudi and Brooks (2020) note the gulf between 'technocratic imaginaries' and the 'imaginaries of communities' in the context of city-regionalism, highlighting that spatial imaginaries also need to become socially embedded, namely to provide a sense of community and belonging. Similarly, Keating and Wilson (2014) refer to the emergence of 'regions without regionalism'. It is often these already established spatial scales and identities that are used to challenge new growth models, depending on whether central and local interpretations of the 'region' agree or conflict. Where there is resistance, this can be driven by uncertainty and a feared loss of identity of old organizations. Calzada (2018Calzada ( , 2019 demonstrates how resistance can lead to grassroots secessionist movements demanding alternative subnational governance and policy arrangements and the 'right to decide'. Therefore, 'identity' discourse is important in the economic growth or decline of regions (Semian & Chromý, 2014;Zimmerbauer & Paasi, 2013). Regional identity is defined as a type of 'collective identity' and 'key element' in the formation of regions (Brigevich, 2018). A distinction can be made between the regional identities pursued by regional elites (Semian & Chromý, 2014) and those of ordinary citizens (Paasi, 2001). The former can be far removed from the awareness (cognitive) and emotional attachment (affective) citizens have for a place (Keating, 1998). Furthermore, regional identities are strongly associated with history and 'images of the past' that are hugely influential on the patriotism and emotional attachments residents develop towards places (Šerý & Šimáček, 2012). Šerý and Šimáček (2012) also remind us of the importance of regional traditions for shaping regional character, and Raagmaa (2002) of the significance of the name of a region. This relates to 'regional consciousness' that is about how regions are imagined, understood and spoken about within communities (Vainikka, 2012).
Thus, Paasi (2011) further distinguishes between two types of regional identity, namely 'institutional structures' and 'regional consciousness'. This relates to Jones and Woods' (2013, p. 36) distinction between material coherence, namely 'the social, economic and political structures and practices that are uniquely configured around a place', and imagined coherence, which denotes a 'sense of identity with the place and with each other, such that they constitute a perceived community' in a way that fosters collective action. These provide a valuable framework for research on devolved regional geographies. The alignment of material and imagined coherence is important not only for national-regional cohesion but also for intra-regional alignment, as a region's success is heavily contingent on the extent to which the 'idea' of a region is shared: (1) amongst regional stakeholders for promoting common interests, shared visions, and mature and trusting relationships; and (2) between regional stakeholders and citizens for promoting a sense of belonging and regional pride, and ensuring their engagement with the region. As Semian andChromý (2014, citing Deas &Ward, 2000) emphasize, 'success depends on the interlacing of the regional authorities' development visions, their 'imposed' identity of the region, and the community's sense of belonging … on the willingness of inhabitants to participate in regional development and associated activities' (p. 265).
Identity is particularly important in the context of devolution and subnational governance rescaling whereby new spatial imaginaries are reified through institutional structures at a new scale. To foster collaboration, the new imaginaries need to become socially embedded and create a shared sense of identity. Yet there is a question as to whether they can replace old imaginaries, especially where the latter are linked to strong historically and culturally rooted identities. We examine these issues in the context of governance rescaling from regions to city-regions.

Subnational governance, scale and identity
The recent decades have seen a move away from centralized forms of governance in a global drive towards devolution (Calzada, 2017). This saw powers, authority and resources transferred from national to subnational scales, with governance tiers emerging at the regional, cityregional and metropolitan levels (Rodríguez-Pose, 2008). The search for the right scalar fix resulted in experimentation, with subnational governance rescaled from metropolitan areas to regions and from regions to city-regions. The latter have become the latest international trend in defining metropolitan spaces and the favoured scale for concentrating subnational economic activity and policy (Calzada, 2015;Moisio & Jonas, 2017). While the cityregion concept remains fuzzy, most definitions refer to an urban core linked by functional ties to its surrounding hinterland (Neuman & Hull, 2009). With city-regions regarded as regional engines of growth, much of the rationale behind city-regional governance is rooted in economic-and city-centric growth (Organisation for Economic Co-operation and Development (OECD), 2006), reflecting a shift in the devolution narrative from a bottom-up identity-based to an economic-centric discourse (Rodríguez-Pose, 2008).
This, however, is problematic as it overlooks the importance of historical and cultural identity rooted in place in providing social cohesion and fostering collaboration in economic development across places. While rescaling subnational governance remakes the material (i.e., institutional structures) at a new scale around new administrative boundaries, this is not automatically followed by the creation of new subnational identities, as previous imaginaries can persist and continue to manifest at the community level. As Paasi (2001, p. 138) notes 'regional consciousness has no necessary relations to administrative lines drawn by governments'. Thus, even when a region ceases to have an official status linked to formal structures and policy, it may continue to dominate the imaginations of local communities (Zimmerbauer & Paasi, 2013), and remain in existence for as long as people believe in and make claim to it (Paasi, 2003;Vainikka, 2012). This has key implications for subnational governance rescaling. As Healey (2009, p. 834) emphasizes regarding city-regions: to have significant effects and to endure through changes in intellectual fashions and political attention, the idea of the 'place' of a 'city-region' has to become embedded in key relations and imaginations within place itself. It has to act as a critical identity shaping force, mobilizing attention locally when neglecting externally. Only then, will a cityregion concept have significant effects in generating and maintaining synergies and resistances which will produce distinct place qualities.
We focus on these issues in the context of the rescaling of subnational governance from regions to city-regions in England, UK, where the subnational governance space has seen frequent remaking through rescaling, oscillating between regionalism and localism over the past few decades. In England, city-regions became the subnational arena of the moment following the election of a Conservative-led Coalition government in May 2010, which saw the demise of the regional governance tier and the establishment of LEPs and CAs as new local governance vehicles. Part of a new 'devolution' agenda promising the transfer of powers and resources from central to local government, this was in line with a broader shift towards new localism that saw increasing responsibility for managing economic change placed on local officials, business elites, The 'imaginary' challenge of remaking subnational governance: regional identity and contested city-region-building in the UK and voluntary sector representatives. The rescaling led to the formation of 39 LEPs in 2011. On a remit to create jobs and boost economic growth via resources delivered through City Deals and Growth Deals, LEPs were believed to better reflect functional economic geographies and local identities (Ayres & Stafford, 2014;Lowndes & Gardner, 2016;Marlow, 2019). Then, between 2011 and 2018, 10 CAsformed on the condition of electing a metro mayorwere created alongside LEPs for participating authorities to receive additional powers and funding from central government via asymmetric Devolution Deals.
However, primacy was given to economic growth targets and to cities, despite early criticism of city-centric economic growth, governance, and reductionism (Morgan, 2007). Thus, many cast doubt over the legitimacy of city-regionalization, deemed by some more as political fiat rather than bottom-up localism (Jones, 2013). Healey (2009, p. 833), for example, questioned 'who is doing the "summoning up" of the idea of a city-region, for what purposes, and in what institutional arenas, with what legitimacy and accountability?' Similarly, Harrison (2010) noted the underemphasis on how city-regions are constructed politically, highlighting a rather 'compromised city-regionalism'centrally orchestrated and constrained by political-administrative boundaries. Rees and Lord (2013) refer to this process as 'making space'a process of inserting the institutional architecture of the new localism within the multi-scalar hierarchy of governance, on the one hand, and as the territorial legitimization of (city-regional) space itself, on the other. They argue that LEPs were, to a great extent, politically reified based on political convenience reflected in the central government's preference for scale, agglomeration of critical assets, and the continuation of pre-existing partnerships. Likewise, many CAs seeking alternative arrangements to those prescribed by the central government were unsuccessful in securing Devolution Deals (Ayres et al., 2018). Indeed, many LEPs were superimposed over pre-existing arrangements, 'unable to escape the existing territorial mosaic of political-administrative units' (Harrison, 2010, p. 71). The influence of historical partnership legacies therefore cannot be ignored (Ayres & Stafford, 2014), as '[t]he rescaling of state space never entails the creation of a "blank slate" on which totally new scalar arrangements could be established' (Brenner, 2009, p. 134).
While scholars have highlighted the challenges of subnational governance in various contexts (e.g., Ayres et al., 2018;Calzada, 2018;Davoudi & Brooks, 2020;Gherhes et al., 2020;Keating & Wilson, 2014), the issue of identity remains less understood. As territorial identity is not confined within existing politico-administrative boundaries (Lackowska & Mikuła, 2018), this has seen demands for independence, autonomy, and even secession in many places across the world. Prominent examples include regional-nations with a strong cultural and historical identities such as Catalonia, the Basque Country, Scotland and the Flemish Region (Calzada, 2019; Hooghe et al., 2020; Vallbé et al., 2018). Recent studies point at the importance of identity in the context of city-regionalization, highlighting significant differences in territorial identities between citizens living in core cities and those living the suburban municipalities and peripheral areas (Lidström & Schaap, 2018). In Poland, for example, Lackowska and Mikuła (2018) observed stronger city-regional identities among those living in suburban areas, with a lower regional identity in core cities. Similarly, in Switzerland, Kübler (2018) found that citizens living in the suburbs have stronger intermunicipal attachments than citizens living in the core city, even in the absence of city-regional institutions, arguing that 'the functional integration of city-regions leads to a rescaling of citizens' territorial identities' (p. 64). Similar patterns were observed across Sweden regarding the intermunicipal orientation and identity (Lidström, 2018). However, in other cases such as the Barcelona metropolitan area and the Greater Stuttgart city-region, those living in the core cities of Barcelona and Stuttgart have stronger metropolitan and city-regional identity, respectively, than those living in the peripheral and surrounding areas (Vallbé et al., 2018;Walter-Rogg, 2018). Therefore, varying degrees of material and imagined coherence exist across places, but we are yet to understand the roots and implications for economic development.
In England particularly, Richards and Smith (2015) emphasize how state-driven devolution has been guided by economic rationale that bears limited or no resemblance to shared political, cultural or social identities that people relate to. In a case study of the Greater Cambridge Greater Peterborough (GCGP) LEPwound up in 2018 -Marlow (2019, p. 140) highlights that 'these long-standing dysfunctional features of England's subnational growth approach are "alive and well"', alluding to identity as part of unresolved tensions inherent in city-regional arrangements. While LEPs and CAs provide 'material identity' in their subregions, they also aim to construct new 'imagined identities' but as highlighted earlier, regional identities can persist. Therefore, there are questions as to whether this 'hard' approach to city-regionalization will ever filter to the 'soft' everyday and unite local identities to foster regional development (Beel et al., 2016), especially as 'few localities share Greater Manchester's economic coherence and cultural identity or its historic strengths in collaborative working across constituent boroughs' (Lowndes & Gardner, 2016, p. 365). Rescaling fragmented historical regions into a mosaic of subregions and city-regions, yet here is a question as to whether these have supplanted historically rooted regional identities.
While it has been shown that citizens develop multiscalar identities (Lackowska & Mikuła, 2018;Walter-Rogg, 2018), our understanding of the variation in how these manifest across regional and national contexts, and of the implications for subnational governance, remains limited (Lidström & Schaap, 2018). Therefore, this paper aims to answer the important question of why some spatial imaginaries become institutionalized and accepted while others encounter resistance. We build on Jones and Woods' (2013) framework of material and imagined coherence, examining the interplay between identity, scale and subnational governance in the context of rescaling from regions to city-regions in the UK, specifically in a city-region that is part of a wider region with strong historical and cultural roots.

EMPIRICAL FOCUS AND METHODOLOGY
The empirical focus of this paper is SCR, supported by the SCR LEP and SCR CA. Centred around the core city of Sheffield, the LEP and CA span four local authority areas that make up South Yorkshire (Figure 1), a metropolitan county located in Yorkshirean historic regional county in the north of England with a population of 5.3 million. SCR resembles a traditionally monocentric city-region, with Sheffield as the economic and employment core (Table 1). However, other major towns such as Doncaster are emerging alongside as key growth poles (One North-East, 2009). Therefore, while drawing on SCR-wide insights, our focus is on the local authorities of Sheffield and Doncaster as the two economic growth drivers in SCR.
Following a City Deal in 2012 and Growth Deal in 2014, SCR became the second city-region after Greater Manchester to reach a Devolution Deal with central government in 2015. Conditional on electing a metro mayor in May 2017, this would give SCR an extra £30 million per annum over 30 years to boost growth. However, due to heightened tensions between SCR stakeholders in 2016including disagreement over the station location of the High Speed 2 (HS2) rail line in the region that 'exposed antagonistic relations and the perseverance of local politics and diverging interests within the SCR' (Hoole & Hincks, 2020, p. 1597)a mayor was not elected until 2018. The region then entered a period of further turmoil between 2017 and 2019 when Doncaster and Barnsley began campaigning for devolution to Yorkshire. Moreover, while the SCR 'brand' aims to provide a cohesive imaginary and civic identity, it is up against a strong Yorkshire regional identity that is commemorated yearly on Yorkshire Day. This issue has been highlighted elsewhere by Marlow (2019) who showed that the GCGP LEP had 'retained strong local identity of two distinctive and different medium-size cities with large rural hinterlands' (p. 143). SCR's contested monocentric status and the pre-existing regional identity provide an appropriate setting for examining rescaling and the extent to which the new SCR imaginary has been successful in galvanizing local actors and fostering participation in economic development activities.
To examine this, we adopted a qualitative methodological approach. In total, 34 in-depth interviews were conducted between November 2015 and December 2018 with six different stakeholder groups: local authority officials (11), entrepreneurs (12), SCR LEP representatives (six), chamber of commerce representatives (two), SCR CA representatives (two) and one civic organization representative (Table 2). Most interviews were conducted in Sheffield and Doncaster, with a small number of interviews from Rotherham, Barnsley and Bassetlaw to gain a wider perspective. The interviews were typically 60-90 minutes in length, primarily face to face, and recorded with interviewee consent. Interview transcripts were thematically coded based on prior theoretical assumptions and novel ideas or contradictions that emerged from the data. This followed an iterative process of reasoning involving 'deductive' and 'inductive' techniques until 'circularity' within the research process was reached (Bryman, 2012). The codes were grouped based on similarity and revised and refined through constant comparison with the data and key literature, yielding the final themes presented in the next section.

FINDINGS
The findings are presented in relation to three overarching themes which together highlight an asymmetry between material and imagined coherence in SCR and the persistence of a regional consciousness. First, SCR is characterized by a weak imagined coherence that fosters local disengagement and perceptions of inter-place rivalry. Second, the importance attributed to core cities in Devolution Deals creates a fragile material coherence, generating local governance tensions. Third, this has seen the resurgence of a regional identity discourse at the local level, prompting some local authorities to use regional identity strategically to push for a regional Devolution Deal.
A weak imagined coherence fosters local disengagement and inter-place rivalry perceptions The interviews highlighted that, rather than providing a new identity for local communities to coalesce around, the SCR brand is in fact a source of tensions at the local level. Criticism levied towards SCR for being an 'artificial construct' (INT17-LA) that local communities do not identify with revealed a weak imagined coherence. As a stakeholder explained, 'Just the term "city-region" … not everyone wants to live in cities. … Not a lot of people live in cities, actually' (INT4-LA), highlighting that the city-centric identity does not resonate with hinterland communities. Moreover, the name of the city-region itself is problematic for some local communities and a hindrance to local engagement, as the core city-centric identity fails to represent all localities: What's it called? Sheffield, not South Yorkshire. … What does the Sheffield LEP represent? It might just be an awkward typo, but it still carries on through.
(INT22-ENT) If the name of the city-region was the Barnsley, Doncaster, Rotherham and Sheffield City Region … it might be different. (INT12-LEP) The 'imaginary' challenge of remaking subnational governance: regional identity and contested city-region-building in the UK If you look at devolution … it just looks like Sheffield getting bigger.
This creates the perception that Sheffield as a core city accrues the majority of economic benefits at the expense of city-regional peripheries, that 'Sheffield has obviously got much more money going into it' (INT20-ENT), and that members 'will be competing for some of the business' (INT23-ENT). The issue appears to have been exacerbated by the rescaling of the Yorkshire Forward RDA to the SCR LEP and CA which has seen economic development initiatives and funding 'centralised' in the core city: [Funding] should be given to each individual town … because if it's done through Sheffield, Sheffield keeps it all and Rotherham, Doncaster, other towns like that will see hardly any of it because they'll focus on particular projects that are 'by chance' based around Sheffield and the growth of Sheffield as a city.
Such concerns were also expressed by entrepreneurs located in the hinterland in relation to business support provision. There is a shared view among many entrepreneurs that businesses located within the city-regional periphery are disadvantaged by the city-regional setting, which they argue sees funding and support benefiting Sheffield-based businesses disproportionately: The concern would be that somebody like Sheffield, that perhaps is bigger, vibrant, more figures, more turnover, more people, would get more resources than perhaps a one-man-band in Doncaster. Furthermore, the weak city-regional identity is evident in levels of engagement, or rather disengagement, between local communities. For example, many businesses within the periphery are reluctant to engage with the LEP to access business support, retaining a preference for local support provision: I would never speak of the Sheffield regional bit as the first point of contact. … I would just think of the likes of Business Doncaster. … I almost feel as though Doncaster and Sheffield are rivals, and any other town, place is a rival. … There's always been this funny relationship between Sheffield and Doncaster so that's why I'd probably try to avoid it.  This reluctance to engage with the city-region is also reflected in the low local response rates to economic surveys distributed under the SCR brand:  This clash between local identities and the SCR imaginary is particularly accentuated in Doncaster's case. While the formation of the SCR LEP and CA saw the town caught in the SCR narrative, an apparent apathy and even opposition to SCR is seeing the town increasingly disengaged from the city-regional scene: Therefore, city-regionalization efforts are not supported by a widely shared city-regional identity. Rather than civic cohesion and collaboration, the weak imagined coherence has led to local disengagement, fostering perceptions of inter-place rivalry. City-regional peripheries, such as Doncaster, that do not identify with the cityregion have 'started to forge [their] own identities as well' (INT13-LEP).
In hindsight, the local stakeholders admitted that not enough effort has been dedicated to obtaining civic buyin and building the SCR brand in the public perception: There's quite a lot of apathy anyway and … lack of engagement from the public on Sheffield City Region deal because I don't think people will see it as touching their life particularly. And I don't think we've done a great job as individual authorities and collectively through the City Region to engage local people with it. I mean, it's difficult enough as officers to understand how it all works and to be able to explain the benefits and to accept that there are overall benefits and it is overall a good thing for us, without being able to sell it to the public. SCR therefore lacks the city-regional identity that other places have developed prior to the institutionalization of city-regional space (e.g., Kübler, 2018;Lackowska & Mikuła, 2018;Lidström, 2018). It displays a weak cityregional identity in the peripheries, similar to other European city-regions (Vallbé et al., 2018;Walter-Rogg, 2018), and even direct oppositions to the core city-centric narrative, all of which indicates a weak imagined coherence. It exemplifies how this manifests at the local level through disengagement, stymieing city-regional governance. This illustrates the importance of pre-existing bottom-up territorial identities which cannot be 'manufactured' through top-down institution building, as city-regional institutions (i.e., the material) in themselves are insufficient to foster citizen buy-in and the development of city-regional identities. The result is a hinterland that feels disconnected from the city-regional imaginary. As an interviewee emphasized, in SCR 'there's still almost a divide that needs to be broken  INT17  LA  INT18  SCR CA  INT19  ENT  INT20  ENT  INT21  ENT  INT22  ENT  INT23  ENT  INT24  ENT   INT25  ENT  INT26  ENT  INT27  ENT  INT28  ENT  INT29  ENT  INT30  LA  INT31  SCR LEP  INT32  LA   INT33  SCR LEP  INT34 LA Note: a All participants remained anonymous. LA, local authority representatives; ENT, local entrepreneurs; SCR LEP, Sheffield City Region local enterprise partnership representatives; CC, chamber of commerce representatives; CA, Sheffield City Region combined authority representatives; and CO, civic organization.
The 'imaginary' challenge of remaking subnational governance: regional identity and contested city-region-building in the UK before it can be fully embraced' (INT25-ENT). Unlike other established city-regions, such as Greater Manchester, SCR does not benefit from 'a more cohesive civic identity' (INT4-LA). Instead, some localities have retained a strong regional identity: 'Yorkshire has more of an identity than Sheffield City Region … so people are going to relate more to Yorkshire' (INT17-LA). This indicates the existence and persistence of multiscalar identities (Lackowska & Mikuła, 2018) within the city-regional periphery that connect to the wider region, the consequences of which are unpacked in the next sections.
A fragile material coherence: core-periphery polarization and local governance tensions Further to the weak imagined coherence, a key issue is that the SCR governance setting polarizes economic development interests, leading to local governance tensions as opposed to fostering collaboration between localities. Two interviewees emphasized: While an agreement was reached on the surface to form the SCR LEP and CA to negotiate a Devolution Deal, in reality the new arrangements are marred with tensions. Far from creating a blank slate (Brenner, 2009), the SCR LEP and CA have inherited many of the issues that permeated the area throughout its political history: There's always been a reluctance of the coalfields to work with Sheffield. Sheffield's always really been seen as the bad brother, and Barnsley, Doncaster and Rotherham are the coalfield areas, but the problem is they don't work together either. There doesn't seem to be a regional perception that the only way to get the region going is there's got to be integration and interconnection between the local authorities.

(INT24-ENT)
This is also evident in the political past of some localities which attempted to 'break away' economically and seize opportunities beyond South Yorkshire: Fast forward to the present day, these tensions persist, with the relationship between the four districts of South Yorkshire still 'a bit choppy' (INT18-CA). Thus, a further challenge uncovered by the interviews is the fragility of the SCR's material coherence. Beyond public perception issues, the city-regional arrangements fail to provide a constructive framework for collaborative economic development activities, with deep-seated tensions surfacing at the local level. One of the most contentious issues seeding these tensions is the core city-centric focus, with some interviewees describing SCR as 'Sheffield-centric' (INT28-ENT), as 'too insular within Sheffield' (INT27-ENT). Also recounted by an interviewee was how 'the general feeling outside of Sheffield is that a lot of these things are Sheffield running roughshod over everybody else' (INT21-ENT), with another emphasizing: Sheffield naturally sees itself in any city-region as the centre of gravity and there's a lack of humbleness and humility perhaps not from any one individual but just the challenge when you're looking at devolution from a city-region concept rather than a region concept. It's lack of equity … what would be perceived.
Indeed, the peripheral character of the towns surrounding Sheffield is regarded as a weakness that hinders the realization of their growth potential. This is particularly acute in Doncaster's case as the town attempts to assert its growth potential and ambitions. Highlighting distance from the core as an economic disadvantage, two officials stated: Doncaster is a bit on the edge of the City Region … further away from Sheffield, which means it doesn't have quite the same impact on Doncaster as it does, say, somewhere like Rotherham in terms of drawing away economic growth.

(INT33-LEP)
Probably a key weakness is the fact that Doncaster is overshadowed by Sheffield and also Leeds … so quite often it may be overlooked in terms of where the government may want to invest in schemes or external businesses may want to come and invest.
Importantly, this is fostered by the national-level devolution agenda which emphasizes core cities as economically superior powerhouses. As an official emphasized, 'Cities is where we see the growth. That's been the narrative. Towns understandably feel left behind' (INT12-LEP). Indeed, city-regional peripheries like Doncaster which is emerging as 'the second biggest growth pole' (INT30-LA), feel particularly disadvantaged by these arrangements: There are some tensions around the role of core cities versus mid-sized cities like Doncaster. … There's an assumption in many quarters at national level and regionally that, if you consolidate all your assets and support to the core city, by definition everybody else benefits, but city-regions are different. Sheffield City Region is polycentric. Doncaster increasingly is punching its weight in terms of the 160 Cristian Gherhes et al.
growth it's supporting in the Sheffield City Region, so do government adequately recognise the role of mid-sized cities as well as core cities? Because we're always going to be up against it … but we've got our own large economy with our own distinctive offer. We've got assets that many core cities can't compete with. … It shouldn't be polarised and it shouldn't be seen as one thing or the other. (INT32-LA) Critically, this not only fuels an unfavourable public perception that further weakens the city-regional imaginary but erodes the material coherence of the LEP and CA. Described by an interviewee as a 'political struggle' (INT34-LA), the consequences of this polarization create local tensions, with SCR membership deemed by some as a hindrance to local economic development. For example, local stakeholders in Doncaster highlighted the challenges of realizing the economic opportunities afforded by its geographical location for the town's benefit in the current setting: In terms of commuter flows, we look to the West and we look to the East, so Wakefield, Leeds, but also North Lincolnshire, Humberside. We've got strong economic connections there. So yeah, we are in the SCR camp and obviously working as hard as we can to make that a success, but we've also got major economic ties and great potential … [and] we need to be nurturing those global trade flows beyond the SCR. … That's a real big challenge for Doncaster. (INT32-LA) Thus, while in other places bottom-up demands call for the formation of city-regional institutions as new territorial governance 'material' to give those places the coherence reflected in citizens' territorial identity (Kübler, 2018), in SCR, city-regional institutions are seen as a hindrance to local economic opportunities and a further marginalization of local voice. The case study thus illustrates the challenges and consequences of top-down imposed technocratic scalar fixes that are disconnected from local realities. The fragility of the material coherence is also visible in the economic geography of the city-region, specifically the interconnectednessor lack thereofof city-regional localities. This is seen to stymie inter-local economic practices and to limit the economic development potential of the peripheries: I can't get a train from Barnsley to Doncaster or Rotherham. It doesn't even exist. They closed the lines down in the 60s … so I can't even improve the line speed because it doesn't even exist. … That's an impediment to local growth, because actually, if you could then connect up with some of those core cities, that would make a huge difference for learning potential, for growth, aspiration, et cetera.
(  Importantly, these issues have persisted throughout the years. In fact, they have become more visible with new projects requiring horizontal collaboration between cityregional actors. This was the case of the HS2 rail line intended to increase rail capacity in the UK and better link the North with the South: Therefore, in contrast to what has been attributed to the success of Greater Manchester as 'a combination of cohesive local intergovernmental relations and a culture of collaborative working' (Haughton et al., 2016, p. 420), the evidence suggests that SCR relations are lacking the maturity and trust to overcome long-established, parochial interests, contributing to a fragile material coherence.
Using regional identity strategically: the push for 'One Yorkshire' In the context of a weak imagined coherence, since 2017 SCR has been involved in a campaign for Yorkshirewide devolution. This came about when Doncaster and Barnsley pledged their allegiance, together with 16 other local authorities across the region and the Mayor of SCR, to devolution for Yorkshire. Meanwhile, Sheffield and Rotherham continued in their efforts to push a SCR Devolution Deal over the line. While in part fed by diverging local identities and resistance to city-centric policies, as outlined in the previous sections, also relevant to this change of direction was a resurgence of 'regionalism' following the 2014 Scottish independence referendum that brought to the fore the possibility for alternative devolution arrangements as a solution to London-centrism. This desire to explore alternative approaches to growth was reinforced by the 2016 Brexit vote: Indeed, arguments in favour of a return to regionalism were presented that drew on a post-EU Referendum discourse, with claims that the One Yorkshire vision is an attempt to recognize local histories and identities and reconnect with local communities. This was emphasized by one interviewee's pledge that 'what's really interesting about One Yorkshire is it's of the place' (INT1-LA) and another who described how: The 'imaginary' challenge of remaking subnational governance: regional identity and contested city-region-building in the UK A really important point … missed in devolution, is the connection with people … we need to do something that is for and with people … add to it the post-Brexit world about scale and brand and identity and the need to articulate a voice, and you've got the genesis of the One Yorkshire proposition. (INT5-LA) While gaining new momentum in 2017, the idea of devolution to Yorkshire was prominent in the early 2000s when the former deputy Prime Minister John Prescott led an unsuccessful campaign for an elected Yorkshire Assembly. In recent years, the movement found new life through the creation of the Yorkshire Devolution Movement and Yorkshire Party in 2012 and 2014, respectively. Sustaining through different institutional fashions, the main argument for devolution to Yorkshire has remained consistent, described by Giovannini (2016) as the 'historical, cultural, economic and territorial distinctiveness' (p.591) of the region linking to its 'shared identity' (INT11-LA) that is an 'in-built feeling' (INT16-CO), and in strong contrast to the lack of any affiliation residents have with SCR: 'People still associate themselves with Yorkshire. They're Yorkshire born and bred. Sheffield City Regionpeople don't know what it is' (INT8-CC). Therefore, while in other places the functional integration of city-regions led to a rescaling of citizens' territorial identities (Kübler, 2018), this has not been the case in SCR where the formation of the CA and LEP has not automatically rescaled citizen's territorial identity. In the city-regional peripheries, this appears to remain fixed to the region. Relating to this, one interviewee's claim resonated with Rees and Lord's (2013) concept of 'making space' that denotes the way that city-regions were formed at the intersection of 'viable political-territorial city-regional geographies' and the 'abstract constructions on which they were based' (p. 691): if you bring this down to a people scale, because we're dealing with people who live in places ultimately, I think they struggle to understand what the Sheffield City Region is, because really it's just an economic geography that's been drawn up to fit around the other bits.
It was thus the region's strong identity that One Yorkshire supporters used to garner wider support from the public, as indicated, for example, by the inclusion of the Yorkshire rose emblem in all their campaign material. Alongside the claim to identity, however, interviewees also spoke favourably of the economic advantages devolution to Yorkshire would bring to the region. As evidenced below, this argument is based on the premise that the region's strong identity and 'established global brand' (INT11-LA) that had 'international significance' (INT7-LA) would boost its competitiveness: We've got our eyes on what we see as the bigger benefits, which would come through the agglomeration of having a Yorkshire area … Yorkshire has got international significance … so we see it as something that is a marketable position … as opposed to something like a South Yorkshire brand.  This claim is supported by an independent economic study carried out to support the One Yorkshire case that findsas one interviewee recallsthat 'One Yorkshire Devolution would or could rather boost the economy of Yorkshire by as much as £30 billion a year' (INT16-CO). For places like Doncaster, the push for One Yorkshire is also an attempt to take advantage of economic opportunities within the region, beyond SCR boundaries: 'in economic terms, Leeds is where … a lot of the regional growth is coming from' (INT4-LA). This is against a backdrop of SCR's relatively weak economic position in comparison to other core cities (OECD, 2020). 1 It was partly for this reason that it was claimed Sheffield was not supporting devolution to Yorkshire since it would 'rather be a big fish in a little pond than a little fish in a big pond' (INT16-CO). Moreover, some interviewees were of the view that working together at a larger scale would strengthen their national position and influence. This was especially the case for supporters in SCR that had witnessed the cityregion's weakening position and a growing uncertainty over the future of the existing arrangement. Once one of the front runners of devolution, since 2016 SCR has been falling behind other city-regions in its devolution journey due to the delay in electing a metro mayor and signing the SCR Devolution Deal. In this context, interviewees in Doncaster suggested that devolution to Yorkshire would increase their influence over government for obtaining the kind of Devolution Deal that was wanted locallyor at least in the peripheriesas well as enhance their competitiveness relative to other areas bidding for devolution funds and powers: [For] allowing those conversations with central government, scale helps. That's why we're favouring the Yorkshire side of things … it does give you the kind of massing and scale to be able to have a discussion and a conversation with central government around strategic things. (INT7-LA) If we're working together with democratic legitimacy over 5.3 million people, which is the size of Scotland, then there's a better chance of influencing and lobbying and getting the deal that we need. (INT4-LA) Thus, just as other places with strong city-regional orientations are politically 'out of place' in the absence of cityregional arrangements (Kübler, 2018), peripheral localities in SCR and Yorkshire more widely with a strong regional identity see themselves as politically 'out of place' in the absence of a regional governance tier that is perceived to provide a better scalar alternative. Importantly, by 162 Cristian Gherhes et al.
fragmenting regional identities and attempting to remake and anchor new identities in regional urban cores, the topdown city-regional agenda in England appears to have reignited regional consciousness within hinterlands, fuelling opposition to city-regionalism and demands for rescaling. Despite the rejection of a One Yorkshire devolution proposal by central government in early 2019, followed promptly by the signing of a SCR Devolution Deal, support for One Yorkshire continues from within SCR. This can be interpreted as a bottom-up response to the perceived marginalization experienced by places outside of core cities which have also become known in the literature as places that are 'left behind', that 'don't matter' (Davoudi, 2019;Rodríguez-Pose, 2018), while demonstrating the ongoing importance of identity in regional economic development regardless of top-down preferences for scale.

DISCUSSION AND CONCLUSIONS
The findings show that, in SCR, devolution-led cityregionalization is not supported by a widely shared cityregional identity and a strong governance framework that fosters collaboration. We show that the formation of SCR failed to seek and obtain the legitimacy of local communities, with some localities feeling trapped in a city-regional narrative to which they do not feel they belong. A weak imagined coherence fosters disengagement and perceptions of inter-place rivalry, with some hinterland communities feeling disconnected from the city-regional narrative and contesting the SCR imaginary. The city-centric SCR brand is a barrier to constructing a shared city-regional identity, being inconsistent with the community imaginary and perceived to disadvantage city-regional peripheries. Moreover, the material coherence provided by the LEP and CA is fragile, being weakened by a city-centric growth discourse perceived to foster core-periphery polarization. The ensuing local governance tensions further weaken the precarious SCR imaginary, hindering collaboration and participation in cityregional affairs. Instead, a regional Yorkshire identity endures in the hinterland, reflecting the persistence of regional consciousness. These dynamics led to the resurgence of regional identity discourse locally, prompting some local authorities to leverage regional identity to push forward the One Yorkshire agendaa region-wide Devolution Deal proposition constructed around the argument that the wider region provides superior economic development opportunities for local communities (i.e., beyond SCR borders) and better reflects local identities. The dominance of economic arguments in favour of regional devolution highlights that identity is, however, secondary to the economic discourse (Hoole & Hincks, 2020;Rodríguez-Pose & Sandall, 2008), being used strategically by local elites to legitimize demands for alternative governance structures. Thus, One Yorkshire is partly a response to city-centric policies that are seen to marginalize city-regional peripheries. Support for One Yorkshire can be seen as an attempt to challenge an existing centrally led devolution model (Hoole & Hincks, 2020) defined in mainly 'functional' terms towards adopting a more 'bottom-up', place-relevant approach based on identity and 'political' regionalism.
While SCR has recently managed to secure a Devolution Deal, the underlying challenges highlighted in this paper cannot be overlooked. Beyond the One Yorkshire devolution proposition supported by the Yorkshire Party, identity-driven political claims have seen the rise of political parties claiming for regional independence, like the Northern Independence Partysimilar to other places in Europe and around the world where territorial identity and the demand for self-rule have been mobilized to influence subnational governance (Calzada, 2019;Hooghe et al., 2020). Given the historical roots of such tensions and continued opposition to the city-regional narrative, the bottom-up demand for regional devolution is rather a 'deferred problem' (Hoole & Hincks, 2020, p. 16), with issues likely to persist and re-emerge in the future. Meaningful devolution therefore needs to transcend an obsession with scale and economic centricity (Gherhes et al., 2020) and seek to better align with bottom-up community imaginaries. Otherwise, it risks fuelling identitybased politics, destabilizing subnational arrangements as opposed to fostering local collaboration.
Therefore, our paper demonstrates the 'imaginary' challenge of remaking subnational governance based on political convenience and the top-down imposition of centrally preferred governance scales (Rees & Lord, 2013). It shows that historically and culturally rooted regional identities cannot be simply remoulded around relatively recent technocratic conceptualizations of city-regions as functional economic geographies. Building on Jones and Woods (2013) framework, we demonstrate that politically driven rescaling can create asymmetries between material and imagined coherence, and show how competing imaginaries can hinder subnational arrangements. Specifically, historical (regional) identities can persist and continue to provide a sense of belonging to local communities, outside of new administrative boundaries imagined by political elites. In our case study, a city-centric narrative perceived as exclusionary exacerbates the 'imaginary' challenge, generating local tensions and resistance. In the UK, the challenge is arguably compounded by the constant remaking of subnational governance, which has not allowed sufficient time for new imaginaries to become established and accepted. Indeed, institution-building and public acceptance of new institutions is a slow and drawn-out process (Harding, 2020). In some cases, such as Manchester City Regionhailed as an exemplar of city-regionalismefforts to build city-regional institutions span decades (Deas, 2014). This, however, is not the case in city-regions such as SCR which, as relatively new propositions, are not supported by similar histories of collaborative working and image-making efforts that are essential to legitimizing new spatial imaginaries (Lowndes & Gardner, 2016;Van Houtum & Lagendijk, 2001). Old, or already lived, imaginaries can thus be perceived to provide better governance The 'imaginary' challenge of remaking subnational governance: regional identity and contested city-region-building in the UK alternatives. Given that city-regionalism is in a fledgling state, the 'imaginary' challenge may be merely an effect of evolutionary tensions that such places undergo in establishing their identity.
However, even in well-established city-regions with a long tradition of city-regional cooperation, such as Greater Stuttgart which has experienced metropolitan governance for decades, surrounding areas have not developed the same strong city-regional identity and attachment as those living in the core city (Walter-Rogg, 2018). This raises questions over whether identity formation is merely a question of time. In fact, Chojnicki (1993) argues in the context of regionalism that the existence of a common regional identity should be a prerequisite for the institutionalization of a region. This common identity fosters place attachment and springs a sense of community (Lackowska & Mikuła, 2018), which are critical to fostering cohesion and civic engagement (Kübler, 2018). Importantly, the rescaling of governance space is not automatically followed by a rescaling of territorial identity. Thus, Chojnicki's (1993) argument can be extended to any spatial scale, namely that the institutionalization of a spatial scale ought to be supported by the pre-existence of a community identity that confers imagined coherence to that scale. To paraphrase Davoudi and Brooks (2020), a spatial imaginary can only become socially embedded when a significant number of people imagine themselves as belonging to that community, regardless of the scale involved, be it regional, city-regional, or metropolitan.
The implications extend beyond SCR and the English context. Our case study highlights the consequences of top-down 'manufactured' scalar fixes that clash with community imaginaries in the context of city-regionalism. However, asymmetries between material and imagined coherence can manifest in different ways across places. For example, they can be more extreme, such as demands for territorial autonomy or independence as seen in places across Europe and around the world (Calzada, 2018;Rodríguez-Pose & Sandall, 2008) or, as our case study illustrates, 'softer' demands for devolution at a scale that is coherent with the community imaginary. Thus, not all demands for devolution are driven by secessionist aspirations. Unlike places such as Catalonia, Basque Country, Corsica, and Scotland, recognized widely as regionalnations, localities in SCR and Yorkshire are not advocating for independence but for a more constructive and inclusive subnational governance framework. These peripheral localities want a voice and for that to be recognized in governance structures, as current arrangements are perceived to marginalize that voice. Our case study thus provides an additional perspective of a region that is not as prominent as those previously studiedwhat we might consider ordinarythereby illustrating that the issue of territorial identity is pervasive and can influence the effectiveness of governance at different scales, and that tensions are likely to arise where the material and the imagined are not mutually reinforcing.
While geographically focused on SCR, our paper yields key implications for the devolution agenda in the UK and beyond. First, it challenges the way top-down city-regionalism has been imposed across England, particularly in regions that similarly are made up of localities with distinct identities (Lemprière & Lowndes, 2019). As Van Houtum and Lagendijk (2001, p. 765) emphasize, 'proximity is not a cause for interaction', and therefore bundling localities around core cities based on spatial proximity and economic functionality rationales does not provide a solid foundation for collaborative local governance. In regions with strong historical and cultural roots, politically driven city-regionalism can create what Deas and Giordano (2003) refer to as competing geographies of identity. The dominance of core cities can be perceived as encroaching on local identities, with hinterlands subsumed to an overpowering core. This stifles collaboration between the urban core(s) and their hinterlands, which is essential to the functioning of a city-region (Rodríguez-Pose, 2008). However, where regional identity is fragmented and strong subregional or local identities exist, establishing arrangements at the regional scale can be equally challenging (Lemprière & Lowndes, 2019). While the challenges surrounding regional identity may be seen by policymakers as simple frictions to work through, the extent to which imaginaries become accepted, and the effectiveness of subnational governance, is contingent on subnational arrangements providing a coherent identity to local communities (Davoudi & Brooks, 2020). For example, efforts to establish city-regional arrangements in places where strong city-regional or intermunicipal identities already exist (e.g., Kübler, 2018;Lackowska & Mikuła, 2018), are unlikely to encounter the level of opposition seen in SCR. If imposed, spatial imaginaries can create a disjuncture with bottom-up community identities which goes against the constructive delivery of these economic and social imaginaries of place.
There is therefore a need to move beyond economiccentric devolution. At a time of heightened challenges relating to Brexit and the ongoing Covid-19 crisis, and a pledge from central government to connect local recovery with 'levelling up', more than ever before our findings emphasize the importance of taking local and regional identity discourse in local economic development policy seriously. If further reformsas anticipated in the recently promised publication of a landmark Levelling Up White Paper in late 2021are to be about 'strengthening community and local leadership, restoring pride in place, and improving quality of life in ways that are not just about the economy' (HM Government, 2021, p. 30), they will need to foster collaboration across places and better reflect bottom-up identities and preferences for scale. This will help enhance subnational governance capacity and capability for improving the effectiveness of local areas to respond.
Our case study contributes to the broader literature on devolution, subnational governance, scale and identity (Calzada, 2017;Davoudi & Brooks, 2020;Keating & Wilson, 2014;Vallbé et al., 2018), demonstrating the importance of identity and community imaginaries in crafting subnational governance spaces. It highlights that 164 Cristian Gherhes et al. imagined communities can exist and persist even in the absence of associated material components of governance. While the material can be endlessly remade, imaginaries lag behind and can create resistance to new arrangements that do not align with community identities. Moreover, as material and imagined coherence can exist at different scales, asymmetries can exist between subnational governance arrangements and the territorial identities that spring community imaginaries. These risks creating, (re)surfacing or exacerbating local tensions, thus undermining rather than facilitating local economic development, and can become a source of local tensions and legitimize identity-based political claims for territorial autonomy. Given the diverse geographies of city-regions and other types of subnational and multi-scalar governance arrangements, as is exemplified within SCR but the consequences of which are relevant more broadly in different contexts, nationally and internationally, there is scope for future research to expand on our study and examine these dynamics in other contexts and at different scales. Similar identity-related issues are apparent in other places across the world like the Basque Country, Catalonia, and Galicia in Spain (Calzada, 2019;Vallbé et al., 2018), the Greater Stuttgart city-region in Germany (Walter-Rogg, 2018), the Flemish Region of Belgium, Quebec (Canada), as well in places across Bolivia, Colombia, Ecuador, Nicaragua, Panama and Venezuela, the latter of which have seen increased mobilization of territorial identity and demand for self-rule among indigenous communities (Hooghe et al., 2020). Future studies can explore the interplay between scale, governance, and identity within and between multi-scalar arrangements. Particularly, polycentric city-regions, polycentric urban regions, non-core city-regions, and places that are part of more than one subnational arrangement (e.g., places that are part of two cityregions) provide interesting case studies for additional insights into the dynamics between identity and subnational governance. Finally, given ongoing identity-fuelled territorial tensions across the world, future studies can examine how other places have sought to build materially and imaginarily coherent subnational governance spaces and the strategies and approaches employed.