Levelling-up beyond the metropolis: is the UK government’s preferred governance model appropriate?

ABSTRACT We consider whether the UK government’s levelling-up governance model of Combined Authorities and metro-mayors is the most appropriate solution for beyond the metropolis. We draw on case study research from the Mersey Dee area between North East Wales and North West England. The paper addresses three propositions. First, that the underlying distinction between agglomeration-driven and place-based policy centres on assumptions regarding the homogenous and heterogenous character of place. The paper shows how, in the UK context, the city-region concept has evolved as an agglomeration-driven territorial construct with practical limitations. Proposition two focuses on how a distinctive character of place reflects its particular mix of firms and their resulting combined processes of agglomeration. Proposition three suggests that this mix of firms presents choices for the appropriate design of institutions locally and regionally. Finally, the paper illustrates why the present agglomeration-driven framework is a barrier to enabling levelling-up. Progressing ‘levelling up’ involves recognition that heterogenous local governance contexts are shaped by history, culture, and geography, where the success of place-based policies is not aided by the top-down imposition of governance models.

The distinctiveness of this case study centres on exploring about how and why firms, as an essential component of the local economy, relate to place and other firms to understand how the localness of an economy might be interpreted.Other case studies either might not address the local firm economy (Beel et al., 2020;Harrison, 2017;Harrison & Heley, 2015) or focus their interest on a particular cluster of companies, rather than on the mix of firms within the area (Nathan & Overman, 2013;Phelps et al., 2003;Pinch & Henry, 1999) In addition, the Levelling-Up White Paper (HMG, 2022) does not even consider the diversity of the local firm economy in its analysis at all.
The analysis of this paper centres on the following three propositions that derive from the case study research (Hildreth, 2021).These address the foundational character of place, the nature of its particular mix of firms and resulting processes of agglomeration, and what this means for the appropriateness of design of institutions: (1) That a key distinction between agglomeration-driven and place-based policy and institutional frameworks lies in their respective underpinning assumptions about the homogenous and heterogenous character of place.
(2) That a distinctive character of place lies in its particular mix of types of firms and resulting combination of processes of agglomeration, within and beyond the local economy.
(3) That the particular mix of firms in the local economy presents choices for the operational design of institutions across different spatial scales.
Each of these are underpinned by the key principles of place-based approaches.Firstly, that regional geographic context is critical, since local diversity is shaped by distinctive geographical, historical, cultural, social and institutional settings.Secondly, that investigating knowledge embedded in firms, people and institutions in place is essential for effective policy development, since it is unpredictable and not known in advance.
In following these principles, we argue that progressing 'levelling-up' involves recognition that heterogenous local and regional governance contexts are shaped by history, culture and geography, where the success of place-based policies is not aided by the topdown imposition of governance models (see also Fai & Tomlinson, 2019;Pugalis & Bentley, 2014).
(1) Key differences between agglomeration-driven and place-based policy and institutional frameworks Proposition one arose from examining the applicability of two contrasting representations of place and economy used to describe the Mersey Dee, as summarised in Table 1: as a 'city-region' (Mersey Dee Alliance (MDA), 2017) and as a 'locality' (Jones, 2019;Welsh Government, 2008).In the UK policy context, these in turn have become associated respectively with both the 'agglomeration-driven' and 'place-based' development approaches.These two models reflect different philosophical understandings of market mechanisms, the appropriate role of the state and the contributions of communities to local and regional development.These in turn lead to different epistemological observations regarding the nature of agglomeration, the significance of history, the nature of knowledge, and as a result, different conclusions about the design and role of institutions in local and regional development.Underlying differences between agglomeration-driven and place-based approaches centre on contrasting conceptual positions regarding the homogenous and heterogeneous character of place (Barca, 2011;Garretsen, McCann, Martin, & Tyler, 2013 ;Haughton et al., 2014;Overman 2014;Pinch & Henry, 1999).Table 1 summarises the key differences between these two approaches.
In the UK context, the Combined Authority and metro-mayor model has been conceptually associated with the agglomeration-driven model.This institutional approach emerged, particularly post-2010, in England from a growing interest in the functional economy.This resulted in a programme of City Deals and the implementation of Combined Authorities, and in England the election of metro mayors, as the preferred devolution model.The Levelling-up White Paper (HMG, 2022) most recently emphasised continuity in extending this approach universally, with a slightly amended form within County Councils, and most recently eliminating Local Enterprise Partnerships (LEPs): The preferred model of devolution is one with a directly elected Leader covering a welldefined economic geography with a clear and direct mandate, strong accountability and the convening power to make change happen.(HMG, 2022, p. 135) As this suggests, this city-region approach is consistent with a top-down territorial form of governance and leadership (Cheshire et al., 2014, pp. 39-44;Pike et al., 2017, p. 125).The approach was influenced by a study across 122 European functional urban regions by Cheshire and Magrini (2009), 'where the level of decision-making is a good fit with a city's economic footprint this is associated with better economic performance' (BIS, 2010, p. 16).The underlying logic to this approach is rooted in the neo-classical economics literature on public finance and fiscal federalism (Oates, 1972;Tiebout, 1957) to interpret whether variations in levels of public expenditure might provide economic dividends (Cheshire et al., 2014;Cheshire & Magrini, 2009;Pike & Tomaney, 2009).OECD studies offered support by indicating that productivity benefits from city size were likely offset  Barca, 2011;Cheshire et al., 2014;Hildreth & Bailey, 2013.by any fragmentation of governance, as measured by the number of local municipalities within a metropolitan area.This is because such fragmentation might obstruct transport investments and effective land use planning across the metropolitan area, increasing congestion and reducing the city's attractiveness to people and firms.Furthermore, it was argued that the establishment of a metropolitan (or city-region) body might mitigate such adverse effects by half (Ahrend et al., 2011;OECD, 2015).
There have been a number of challenges to whether this city-region model is in practice the most appropriate devolution approach for England.For Haughton et al. (2014), it is about 'agglomeration boosterism', that builds on existing successful local economies.For Waite and Morgan (2019, p. 783).'metrophilia', is a 'pervasive and uncritical embrace of city-centric perspectives in spatial planning'.Meanwhile, Hoole and Hincks (2020, p. 3) consider that the city-region model reflects 'a geo-political strategy actively employed by the national state to serve their own interests, in conversation with local considerations'.In addition, there have been concerns that this city-region model is not anyway appropriate to medium-sized and smaller cities and rural areas, not least because of the likely pull of capital, employment, and people towards larger agglomerations from rural areas and weaker towns and cities (Beel et al., 2020;Harrison, 2017;Harrison & Heley, 2015).
In addition, the city-region model might be a better fit in a monocentric metropolitan context, such as for Greater Manchester, which more closely reflects the conditions described in Cheshire and Magrini (2009).In practice, mixed urban and rural areas beyond the metropolis, may be more likely to reflect intra-urban and relational polycentric characteristics combining a mix of spaced-out centres, of which no one dominates and may play different roles and where economic activity is distributed rather than concentrating (Davoudi, 2003;Hildreth, 2021;Parr, 2004;Shaw, 2023) (see section 3.1).
An alternative place-based analysis could explore a localities framework.This approach also recognises that people live their lives functionally across places for different reasons e.g. in terms of working, shopping, education, housing, and accessing culture.It has applications where places combine both 'material'having institutional structures that hold a place together, and 'imagined'a shared sense of identity, leading to common patterns of behaviour and geographical reference pointscharacteristics (Jones, 2019;Jones & Wood, 2013).Unlike the territorial construct of city-region model, the locality approach combines three characteristics of placeterritorial, relative, and relationalas summarised in Table 2.
The localities approach values having political structures to enable shared action across space.But the territorial is situated alongside the relative and relational qualities of place to emphasise a place's diverse, distinctive, and porous character, where it is recognised that places might have fuzzy boundaries.As a result, it is essential to combine relational focus both 'within' and 'beyond' the place to create institutional coherence (Haughton & Allmendinger, 2017;Jones, 2019;Jones & Wood, 2013;Massey, 2011).Table 3 summarises key characteristics that differentiate the city-region and locality models.In particular, the city-region model is homogenous (in relation to city-size), with its institutional focus topdown on scale, administrative and market barriers, and structure.By comparison, the locality model is heterogenous with its institutional focus bottom-up relative and relational characteristics underpinned by building trust through place-based leadership.
(2) Proposition 2 -Distinctiveness of place in its mix of firms and processes of agglomeration Proposition two is that a distinctive character of place lies in its particular mix of types of firms and resulting combination of processes of agglomeration that apply within and beyond the local economy.This proposition arises from exploring how the localness of an economy might be understood from investigating how its firms relate to place, other firms, and institutions.Whilst a place might contain a particular clustering of firms, it is also possible that a diversity of firms might be identified that respond to different processes of agglomeration for their location in space.To consider a broad range of companies, three models of firm networking described by Gordon and McCann (2000), as 'pure agglomeration', 'industrial complex' and 'social network', are applied to the case study in this paper.In so doing, this opens access to a wide range of relevant and more recent literature about firms from different perspectives that cross the placebased and space-neutral agglomeration-driven policy divide, which is central to this case study analysis.
The clustering of 'pure agglomeration' firms is essentially urban in character.Firm entry into the local economy is relatively easy through property rental or purchase (depending on the price) to join a market characterised by many competitors.Within an agglomeration-driven framework, firms' benefit from matching, sharing, and learning opportunities created within the urban space around them to derive agglomeration economies of scale.These firms operate with a loose set of firm and institutional relations and draw on knowledge external to the firm and internal to urban space from the diverse public and private organisations sharing that space (Duranton & Puga, 2004;Overman et al., 2009).
In contrast, 'industrial complex' firms are usually part of a multi-national enterprise (MNE) structure.They may coincidently collocate with other industrial complex firms to maximise internal economies of scale by minimising location factor and distance transaction costs.These firms make significant long-term capital investments in machinery, equipment, infrastructure, and real estate to locate.Firms' relationships are internalised and integrated vertically within the firm's global production networks (GPNs) (i.e. of firms, institutions, and markets) across different spatial scales from local to global (Coe et al., 2004;Parr, 2002).The embeddedness of firms with 'social network' characteristics in place, meanwhile, is not related to internal or external economies of agglomeration, but rather whether or not they have needs and loyalties which keep them anchored in the local economy.Here the key factor that underpins these firms' relationships with other firms and institutions is trust.But it should be stressed that these ties are not territorial.Rather, they may operate across varied geographies that are independent of bounded space.As a result, social network firms focus of on developing a collaborative approach that enables trust though shared support, practice, and accountability (Amin & Thrift, 1994).
As a result, the three models contrast in processes of agglomeration and their knowledge flows and relationships.The pure agglomeration firm is most closely associated with the city-region model of territorially bounded space.The industrial complex firm may have a complex relation with place from the diversity of its GPN firm and institutional and market relationships, which will be more relational and relative than territorial.For the social network firm, trust may not be territorially spatial.Rather, their connections to place and relations will operate across varied relational and relative geographies independent of bounded space.
(3) Proposition 3the local mix of firms presents choices for the operational design of institutions.
Proposition 3 builds on proposition 2 by suggesting that the particular mix of firms in the local economy presents choices for the operational design of governance institutions.This is because agglomeration processes, knowledge flows and relationships for each type of firm suggest contrasting relationships with place.Given that the pure agglomeration firm is most closely associated with the city-region model, the institutional response will be centred on territorial initiatives to build agglomeration economies and reduce market and administrative barriers.For industrial complex firms, the local institutional challenge is that MNE firms are likely to centre their focus on competitiveness and retention of internal control of company knowledge with a global integrated plant and supplier structure.For the local response to MNEs, there may be opportunities for 'strategic coupling' between territorial networks and MNE GPNs, through processes of value creation, enhancement, and capture (Coe et al., 2004).Yet these are unlikely to occur through territorial bounding, but through effective multi-level relational networks operating often across regional geographies (Bailey et al., 2015).For social network firms, where the emphasis is placed on trust, institutional relations may be sought within the specific geographical, historical, institutional, and cultural contexts of place that address principles of place leadership and institutional thickness.As a consequence, rather than assuming, as within the government's preferred agglomeration-driven institutional model, the centralising agglomeration of firms, it may be more appropriate to be open to the possibility of diversity of firm types with their contrasting agglomeration processes, knowledge flows and relationships.Each of the three types explored in this paper in turn present different institutional challenges.For the pure agglomeration firms, it may require a focus on the role of urban centres.For industrial complex firms, it may entail enabling strategic coupling with GPNs, by engaging relationally within multi-level and spatial networks.For social network firms, the capacity to cultivate trust is important, from being responsive to knowledge from firms and the labour market in shaping what is most needed going forward.In combination, these suggest qualities of place-leadership that are not primarily about structure.Rather, they require the place-leadership skills of boundary spanning, being context sensitive and referent power through which capacity to build trust and work across multi-level geographical and organisational boundaries are critical qualities (Beer et al., 2019).

Methodology
Our study of the Mersey Dee is organised around four elements.First, in order to understand what makes a place, we consider the appropriateness of the two functional representations of place introduced in Section 1 above, used to describe the Mersey Dee: as a 'locality' and a 'city region'.This is explored by examining the Mersey Dee's economic relationships both internal and external to the area, as reflected in an analysis of its labour markets and firm-to-firm and institutional relationships, together with a mix of key policy documents.It is also informed through an analysis of the Mersey Dee's character and identity from an analysis of the historical interaction of its different centres from foundations in Roman times to becoming the functionally connected economy it is today.
Second, we seek to investigate the contribution of economic institutions to place.This is in order to focus on how institutions, through their formal and informal behaviours, contribute to realising the economic potential of the Mersey Dee, and in multi-level terms as a part of North Wales into North West England, that encompasses the crossborder economy.This part draws on a review of relevant policy documents and board papers that have been important in influencing the Mersey Dee, from across North Wales and NW England, together with semi-structured institutional interviews across the public and private sectors.
Third, we seek to investigate the role of firms in place, by uncovering knowledge about how firms relate to place through their relations with other firms and institutions, both locally and outwardly, as well as reasons for their location in the area today.We draw on the framework of firms models originally compared by Gordon and McCann (2000) and an approach to build evidence on a firm-by-firm basis in place described by Markusen (1996Markusen ( , 1994)), across sectors and different ownerships.We identify, approach, and interview a selection of Mersey Dee firms that have horizontal and vertical firm-to-firm and institutional relations that are both local to and external from the Mersey Dee.
The study includes results from 46 firm interviews, reflecting the primary industrial sectors found within the Mersey Dee area, including aerospace, automotive, chemicals, food, nuclear, renewables, financial services, ICT, and tourism. 1 A structured process to select, arrange, conduct, and transcribe interviews was followed in each of the five local authority areas.A semi-structured interview format was applied, to both enable consistency between interviews, whilst allowing additional probing questions to draw further elucidation.Interviews were held face-to-face on company site's, with Managing Directors or other senior managers, lasting between one to two hours.These were conducted primarily between 2014 and 2016.
To enable 'an active process of self-reflection' (Cochrane, 1998, p. 130), continuous efforts were made to test interviewee and other research data in interactive ways over the period 2015-2019.These included: seeking feedback from interviewed firms on interview results, giving presentations to different institutional audiences and receiving feedback; presenting reports and sharing evidence with the MDA Board and member local authority committees; holding small roundtable discussions with companies in the area regarding the results; providing evidence to a Welsh Government review of funded research and innovation in Wales.
Fourth, we draw together findings with regard to the interaction of place with firms and institutions, particularly comparing outcomes with the two representations of place of the city region and locality, within their respective agglomeration-driven and place-based frameworks.

Results
(1) What kind of place?
The Mersey Dee is a functionally connected economy, with different spatial and agglomeration characteristics to those of northern English city-regions or Cardiff and Swansea in Wales.Today the MDA includes the local authorities of Flintshire and Wrexham in Wales and the Wirral and Cheshire West and Chester in England but included Denbighshire at the time of research interviews (see Figure 1). 2 History and geography have combined to give the Mersey Dee complementary polycentric qualities that functionally connect its core residential and industrial centres and cross-border labour market.It is characterised by having two small cities, towns, and a rural character, with functional relationships into its city neighbour Liverpool.The area is also characteristically distinctive with its concentration of manufacturing employment on large industrial estates.As a result, much of its employment is neither urban in character, nor having been dispersed from urban areas.Labour market connections between Chester and Deeside and Ellesmere Port are particularly strong with Wrexham also linked by travel to work movements into the rest of Mersey Dee.Wrexham also forms a more localised labour market.There is also a nodal role for Birkenhead within the Wirral.This combination of factors has resulted in 'functional polycentricity'from the spatial organisation of firms, as much as 'morphological polycentricity'by the distribution of urban settlements of different sizes (Hall & Pain, 2006).Thus, the Mersey Dee most closely reflects distributed patterns of employment, as in 'the tendency of economic activity to cluster in several interacting' centres (Davoudi, 2003;Özkul & Hildreth, 2016), with functional relationships shaped by complementary industrial and settlement patterns between villages, towns, a city and large industrial estates.This pattern enables opportunities for the spread of economic sectors and diversification of the local economy.For workers, it offers choices of where to live and work.It is possible to live in any of the centres in the Mersey Dee and travel to another with less congestion and land costs than would be the case in a metropolitan environment (Özkul & Hildreth, 2016).Such an approach is recognised in the National Development Framework of Wales, where places with more than one centre have 'scope to distribute houses, jobs and prosperity according to the needs of the region' (Welsh Government, 2008, p. 5, 49).
Few of the interviewed firms were found to have pure agglomeration characteristics.Even those that did commonly retained strong personal place ties to the Wirral or to Chester.It is likely that a specific study of Mersey Dee urban centres -Chester, Wrexham, and Birkenheadwould identify more firms within this category.Nevertheless, it is unlikely, at least so far, that pure agglomeration firms are of such central importance in the local economy as firms with industrial complex and social network characteristics.
The industrial complex model applies to companies, primarily MNEs, that invest into the Mersey Dee.Their firm-to-firm relations were found to operate vertically within the international complex, exposing the Mersey Dee externally to the global economy and decisions made by strategic-decision makers at company headquarters (HQs).Firms' decisions to locate in the area were influenced by a combination of location factor and distance transaction costs, combined with a historical legacy of local public infrastructure investment.Today, these plants compete for re-investment with other company sites within their vertical company structures.As a result, they are exposed to company global production networks (of firms, institutions, and markets).For the Mersey Dee, the contribution of local labour markets may be of value, due to the spatial immobility and flexibility of skills.In addition, regional universities may be well placed as trusted partners in innovation.In this situation, the local economy is not envisioned as a bounded territory, but as an open network of relative and relational spaces operating within the broader network of firms local to global relations.Within this, a key aim for local and regional development networks is to enable processes of value creation and capture (Bailey et al., 2018) and in so doing, to enable conditions that create connections (termed as 'strategic coupling') between firms and their regions (Coe & Yeung, 2019).
Our findings also confirm academic observations that spatial patterns of employment and integration of firms' relationships within the local and regional economy reflect how production is organised internationally within the industrial complex. 3In the Mersey Dee, such differences might be associated with a MNE's history of local investment.Thus, evolved firms (firms investing into the Mersey Dee prior to 1980), with their longer histories of reinvestment and production reinvention were less likely to be peripheral within firms' production networks.Incoming firms (firms investing within the Mersey Dee post-1980) were more likely to peripheral in these networks.This was reflected in the breadth of firms' locally devolved responsibilities, delegated responsibilities for product innovation, scope to develop local and regional supply networks and the likelihood of recruiting graduates in addition to school leavers.Also, evolved firms are more likely to engage in trust-based relationships with universities.Nevertheless, there was a caveat that many of the industrial complex firms in the area are assembly plants and only a few have HQ roles.Table 4 provides a summary of differences we identify between firms investing in the area post-1980 and pre-1980.Overall, this shows how the pre-1980 firms were more likely to have broader local firm to firm and institutional relations, be given broader local responsibility (e.g. for local innovation) within their MNE complex, than the more recent arrivals.
A characteristic of the Mersey Dee is that local (or indigenous) firms commonly illustrate trust-based behaviours, through their relational ties with other firms.They were formed within the areaas a new start-up, by related diversification or from take-over by owners with local ties.Often these ties are very local with personal or identity commitments by their owners, for example, to Helsby, or Wrexham or to the Wirral.These may be family ties, or historical association to a particular location.For two firms it is due to immovable equipment.For others, it is because of the local availability of specialist skills.As a consequence, indigenous firms are largely immobile with their ties to place.A small minority of firms currently weighing up options of relocating are doing so with regard to a local alternative site.There is a place-dimension, with social network traits being local, but not urban.Such firms frequently approach innovation as a collaborative activity of responding to their customers' needs and challenges.They often seek loyal, and where possible, local suppliers.They might engage in institutional relations through business networks, collaborative partnerships with universities and valued personal engagement with local and national government, the latter particularly in Wales.They take responsibility for institutional relations and are prepared, when encouraged, to contribute to bottomup processes for building place-representation and enabling supportive interaction in the local economy.This is reflected, for example, in the call by Professor Keith Ridgeway, the Chair of Industry Wales 2020-2022, to develop an 'industrial commons' (Pisano & Shih, 2009) to foster the sharing of tacit knowledge and collaborative relations to drive innovation in products and processes in the local economy (Ridgeway, 2020).
(3) What kind of institutions?What emerges from this research is that the Mersey Dee is segmented in its firm economy, with contrasts between evolved, incoming, and indigenous firms in their patterns of industrial complex, social network, and more limited, pure agglomeration characteristics.In earlier decades, the Mersey Dee was centred on an industrial complex model of local economic development.But there is evidence that the quality of firm-to-institutional relations is becoming more complex as the firm economy diversities and the social network model becomes more important.
Today's governance challenge is not about choosing between the three different firm models of firms and their respective implications for institutional design as summarised in Table 5.This shows how the prevalence of different firm models in the context of place, pose contrasting institutional challenges.As a result, there may be within the Mersey Dee a concurrent focus on the contribution of towns and two small cities to the area's economy, responding to a historical dependence on inward investment, whilst enabling and integrating a stronger indigenous economy that exists alongside the existing MNEs.This is not just a local challenge for a mixed rural and urban area but rather requires effective collaboration within a multi-level relational framework of governance that extends locally to nationally, in which different processes of agglomeration coincide together in the context of place.

Discussion: implications for the governance of levelling-up agenda
This section considers broader implications for the governance of the levelling-up agenda, in the context of the three propositions outlined in Section 1.The significance of proposition 1 is that it illustrates how the Combined Authority and metro-mayor model has become conceptually underpinned by a space-neutral agglomeration driven framework.Thus, whilst Government makes the case that its approach to devolution is about empowering local leadership and decision-making based on recognised geographies (HMG, 2022, pp. 137-138), there are risks for a deal making approach if underpinned with a conceptual framework that emphasises territorial structure over trust and multilevel relationships.Firstly, it reflects a deal-making approach, whereby an area agrees a devolution deal in exchange for adopting the preferred Combined Authority and metro-mayor model, or an adapted leadership model in Counties.Secondly, it is a piecemeal approach.Whilst the ambition might be to eventually roll out this model as an institutional solution for the reform of local government across England, it is being addressed without any overview of how individual Combined Authority and County deals will collectively make the sub-national governance of England stronger and more effective when the different pieces of the jigsaw fit together relationally.Third, this is compounded by constant churn in institutions (such as the scrapping of Regional Development Agencies, their replacement with Local Enterprise Partnerships and the latter's proposed abolition in the 2023 budget).Fourth, it reinforces the culture of Central Government allocating funding through different funding pots through an ad-hoc, short-term and fragmented way.Fifth, it is driving policy towards further proposed Combined Authority and County governance structures that bare little relation to the original aspiration of aligning governance structures to the functional economy (Shaw, 2023).
In practice, a different philosophical approach could have been taken towards the governance of England's metropolitan areas.Within Western Europe, England is unusual in not combining metropolitan areas with regions.For example, in Germany, metropolitan areas are established within a federal (or regional) structure.Their purpose is to achieve effective cooperation across urban and regional areas within functional metropolitan areas.They have been built bottom-up, often with cooperation going back to the 1950s.There are wide differences in their models of governance, land area, urban and rural structure, and fields of responsibility.In Germany, the focus of metropolitan areas is not to be administrative units.Rather, they are platforms for regional cooperation, even to the point that their boundaries may overlap with more than one Federal State (Blätter, 2017;Council of Europe, 2018).
In summary, whilst the city-region agglomeration-driven model in England is focused on the appropriate bounding of functional territorial space, the place-based localities approach gives priority to enabling places to design their institutional framework to promote integration and coordination, as appropriate for a heterogenous, open, and globalised world.Underpinning this are three significant differences between the city-region and the localities models in relation to their approach towards governance institutions.First, that because of its underlying homogenous spatial assumptions, the agglomeration-driven model marginalises the role of history in processes of agglomeration of different places (Garretsen & Martin, 2010).Second, whilst the former approach centres on the territorial concept of place, the latter addresses its territorial, relative, and relational dimensions.Third, is that the city-region model downplays the significance of informal institutional factors in valuing devolution, given that their direct and indirect effects are difficult to measure (Cheshire et al., 2014).
In contrast, this paper identifies particular factors that shape the distinctive institutional context of the Mersey Dee.First, we posit that the area's institutional characteristics are shaped by historical timing and patterns of path dependency.Today's local industrial landscape was enabled by public and private investment in manufacturing and former ordinance sites in association with, or immediately after, the Second World War and by post-1970s inward investment policies.For example, the former is connected with aircraft manufacture at Broughton, the nuclear industry in West Cheshire and the Wrexham Industrial Estate.The latter can be seen in the Deeside Industrial Estate, on the former Steelworks site and with coordinated efforts to attract new and repeat inward investment.These, along with the development of the area's core settlements enables the complementary distribution pattern of employment described earlier.These in turn have enabled a collective institutional identity that underpins multilevel cross-border collaboration regionally between North Wales and North West England.
Second, we suggest that such history may reinforce patterns of institutional conduct that may be inclusive or exclusionary.Evidence supports the inclusive characteristics of the MDA institutional project.There are important cultural differences in working across national borders because of language, culture, as well as economic and social differences.However, within and beyond the Mersey Dee there exists a strong culture of collaboration towards shared aims.Earlier partnership efforts to attract and retain inward investment has left a long-term legacy of collaboration.This is reinforced by the collective memory of recovery from economic crises across North East Wales in the late 1970s.These provide both an institutional memory of what can go wrong and an incentive towards pulling together in crises, reinforcing a culture of shared norms and values.
Third, because of its distinctive geographical position, it should be stressed that the Mersey Dee regionally provides a strategic bridge to connect regional institutions across North Wales into North West England.As such it demonstrates characteristics of a multi-level space of governance that is set up for a specific purpose or role (Bache et al., 2016;Hooghe & Marks, 2003).
Fourth, we illustrate the centrality of institutional strengths that are typically not associated with the city-region model.These comprise the capacity to replicate institutions, sharing knowledge, as well as trust in a shared purpose and adaptability.These derive from informal institutional characteristics of common geography, industrial and settlement history, cultural identity, and institutional settings.Geography has given the area an advantage of being central to industrial connections and shared trade routes across North Wales into North West England, encouraging multi-scalar collaboration.Industrial and settlement history has gifted the area its polycentric distributed pattern of employment and fostered a sense of cross-border identity.Institutional settings have been shaped out of shared experiences and responding to previous challenges.These experiences are important as the area adapts to a post-Covid-19 pandemic and Brexit (Bailey et al., 2021;Bailey et al., 2023) and faces transformation challenges around Industry 4.0 and Net Zero (De Propris & Bailey, 2021).Formal organisational structures are important and by contrast reflect a comparative weakness in the context of public sector austerity in England and Wales.But whereas new funding and stronger structures can be replicated, the informal institutional characteristics which underpin the Mersey Dee are much more difficult and slower to reproduce.

Conclusions and further implications for policy
In summary, there is a need to move on from the present piecemeal emphasis in government policy on individual City or County Deals built around an inflexible framework, to more strategic thinking about how the English sub-national economic geography might work together (for example within a national spatial plansee Wales as an example), with much more flexible options locally.Indeed, evidence suggests that place-based policies are more effective when organised for functional, rather than administrative areas, particularly when they combine 'material' and 'imagined' relative characteristics (see Section 1).This is to ensure there are linkages between cities, towns, and their common commuting zones, as well as between urban and rural areas.And, for such a framework to be effective, scale matters, centred on functional areas.This is, we argue, not properly recognised in the current governance approach underpinning the levelling-up agenda.
Despite the place rhetoric apparent in government pronouncements, there is in reality little evidence of a place-based approach in practice.Key elements that make up a placebased methodology seem to be missing.It should be stressed that effective implementation of place-based policies is underpinned by uncovering knowledge embedded in firms, people, and institutions in place is essential for effective policy development.Such knowledge about the regional economy and its market dynamics is generated and exchanged in a local context.But it is also unpredictable, and not known prior to investigation.It therefore needs to be uncovered by bottom-up participatory processes that engage both local/regional actors and the State to build consensus (Barca et al., 2012;Barca, 2011;Hildreth & Bailey, 2014).This is challenging to do well, as such a process is developed from a strong culture of trust.Here it helps if the geography of regions (and urban areas) has a clear functional basis.In addition, successful adaptation requires strong formal and informal institutions: formal, in that they are designed to be clearly accountable multi-level structures, and informal, that they enable building a culture of trust.
A top-down imposed deal-making approach does little to facilitate this.Rather than emphasising structure (as seen in the current approach), it would be better to focus on how to build a long-term culture of trust, effective place-leadership (Beer et al., 2019;Sotarauta & Beer, 2017), and a multi-level process of working.Indeed, the importance of trust came up repeatedly in our firm interviews.There were many examples amongst indigenous firms, in how they operated with their customers and in their relations with universities.The technologically more advanced MNEs in particular raised it as an issue (such as Unilever's partnership with Liverpool University).Building trust seems more relevant in this example beyond the metropolis than agglomeration economies.
In summary, we argue that place-based approaches in the UK haveparadoxicallybeen effectively constrained into a one-size-fits-all top-down approach from the government under its levelling-up agenda with a focus largely on agglomeration economies underpinning a Combined Authority devolution model.This line of thinking has been pushed by prominent think-tanks which have used an agglomeration economies argument to justify a focus on cities.The danger is twofold: firstly, that agglomeration economies is anyway overstated, or more complex in relation to positive and negative externalities, even in the context of cities (Coffey et al., 2023), and secondly that this risks failing beyond the metropolis (with implications for levelling up) given the complexity of firm networking relationshipsas seen in the Mersey Dee examplewhere a focus on agglomeration economies may be unsuited to the reality of the local economy and firm networking.Rather a genuinely place-based approach requires space for alternative forms of institutional set up, and the need for more bottom-up forms of policy development to support different forms of firm networking, especially in the context of challenges around Net Zero and Industry 4.0 (Bailey & Tomlinson, 2022).Indeed, our findings reinforce a case for the place-based principle that institutions both shape and are shaped by economic geography and history.We suggest that the rooting of economic activity into local institutional fabric is important to realise local economic potential.It also points to how the institutional fabric is also built bottom-up and not just top-down, with strengths layered and combining through different spatial levels of relationships.Our work illustrates the multi-level character of institutional relations that stretches well beyond the Mersey Dee itself.It also indicates that informal institutional characteristics that are shaped relationally over long periods of time are an important foundation for formal institutional effectiveness.Here the idea of institutional thickness is helpful.Amin ad Thrift (1994) point to it having four underlying characteristics.First, the presence in a locality of many institutions of different kinds.Second, a high level of interaction between these institutions.Third, this interaction enables coalition and collective representation.Fourth, this leads in turn to a culture of shared norms and values.What we identify is that it is not the existence of many institutions that is important, but rather how the fostering of shared tacit knowledge and collaborative resources enables processes to drive innovation and value creation in the local economy (Barca et al., 2012) and value capture (Bailey et al., 2021).
Evidence suggests the dominant, ongoing agglomeration-driven model remains a barrier to addressing the heterogeneous nature of place.Progress towards 'levelling up' is more likely to be achieved, not by continuing top down, short term, and fragmented initiatives centred around 'deals', but rather through a long-term approach to subnational economic policy, where the centre values differences in connections between and across places with firms and institutions.Indeed, progressing 'levelling up' involves recognition that heterogenous local governance contexts are shaped by history, culture, and geography, where the success of place-based policies is not aided by the top-down imposition of governance models.The Mersey Dee case study illustrates that the way forward does not lie primarily in structures or piecemeal initiatives.Instead, these should be founded on central government working with, and incentivising regional and local actors to build a long-term culture of trust, effective leadership and strategic multilevel working focussed on the long-term realisation of local economic potential.

Notes
1.An explanation of the research methods applied in this research may be found in 'Chapter 4 -Researching a place-based framework' in Hildreth (2021, pp. 79-116) Wirral also represents Liverpool City Region.3.For example, Markusen's (1996) distinction between satellite (shallow local firm relationships) and hub-and-spoke firm (more locally integrated) relationships, Massey's (1995) differences in the spatial division of labour between company plants depending on how control over processes is devolved in different locations within the company complex and Yeung (2020) about centrality or peripherality within global production networks.

Figure 1 .
Figure 1.Map and location of the Mersey Dee.Source: This figure is reproduced with the permission of the Mersey Dee Alliance.

Table 1 .
Agglomeration-driven and place-based development frameworks.

Table 2 .
Three characteristics of locality space.

Table 3 .
Two institutional models: city-region and locality.
RelativeNoabout scales and administrative/market barriers Yeshistory, independence, and inter-dependence, likely enabling 'imagined' and 'material' characteristics Relational Nofocus internally only Yescombines within and out there Key words Structure (leadership) Trust (leadership)

Table 4 .
Differences between evolved and incoming industrial complex firms regarding their embeddedness in place.
, which is an open access document available in UCL Discovery at: https://discovery.ucl.ac.uk/id/eprint/10130797/ 2. The MDA grew out of a recognition of shared economic, social, and interests across the area covering West Cheshire, Wirral and North East Wales, building on a strong foundation of joint working from the 1990s, with a partnership formed in 2007.Today, in addition to the local authorities of Flintshire and Wrexham in Wales, and Cheshire West and Chester in England, the MDA's membership also includes Chester and Wrexham Glyndŵr universities, Coleg Cambria, Transport for Wales and the Welsh Government.