Capital’s welfare dependency: Market failure, stalled regeneration and state subsidy in Glasgow and Edinburgh

Comparing case studies of long-term, large-scale urban regeneration projects in Glasgow and Edinburgh, Scotland, this paper brings together two addendums to the rent gap model in the shape of the ‘reputational gap’ and the ‘state subsidy gap’. These neologisms are mobilised to clarify the risk-laden centrality of the state’s role in both the formation and potential closure of rent gaps in large-scale areas of disinvestment and devalorisation. Whilst such projects often appear as expressions of capital’s state-mediated extractive power over the built environment, we consider them as examples of capitalist failure or fragility – for even with striking levels of public subsidy to address ‘market failure’ the land and property market has not been reinvigorated according to plan. This highlights the need, we argue, for further critical scrutiny of failed or stalled urban regeneration projects as a means of foregrounding the instability, rather than the omnipotence, of contemporary urban capitalism.


Introduction
Critical studies of large-scale neoliberal urban development projects (UDPs) can often appear as fait accomplis, underscoring neoliberal urbanism's ability to achieve its primary goal: extracting value from the city. Ferguson (2010: 166) notes how the literature often concludes with a reiteration of critical agendas and a rather cursory nod to resistance: 'neoliberalism is bad for poor people, therefore we must oppose it'. To be clear, such critical scholarship is vital given the ideological dominance of media, state and market discourse in favour of predatory forms of neoliberal urbanism -yet it risks reifying the power of unstable capitalist urbanisation. The global financial crisis of 2007 to 2008 and the series of urban crises that have racked capitalism since the early 1970s make this clear (Harvey, 2012), whilst the impact of COVID-19 has only exacerbated economic uncertainties and contradictions (Brenner, 2020). The ongoing 'failure of capitalism' (Kliman, 2011), marked by a secular decline of growth and profit rates (Streeck, 2017), means that impressive macroeconomic forecasting models amount to 'little more than hunches with numbers attached' (Kliman, 2009: 53). As such, any naı¨ve faith in the efficiency of markets must give way to closer scrutiny of 'market failure', and the ever-mutating state strategies deployed to regulate, reproduce and salvage capitalism from its own internal contradictions. Yet, with only a few exceptions (Adams et al., 2017;Jones and Ward, 2002;Kitchin et al., 2014;Wallace, 2015), scholarship on the failure, stalling or collapse of UDPs is sparse.
We seek to rectify this lacuna with a comparative study of the two largest regeneration projects in Scotland: Waterfront Edinburgh (WE) on the city's northern edge and Clyde Gateway (CG) in Glasgow's East End. The former was first planned in the early 2000s, whilst the latter was initiated in 2008. Both are decades-long projects addressing brownfield, vacant, derelict or contaminated land, attesting to the importance of land -and its remediation, privatisation and commodification -in contemporary urban accumulation strategies (Christophers, 2018). Both were significantly derailed and modified by the 2007 to 2008 global financial crises, and both now navigate the ongoing economic impacts of COVID-19. Drawing on long-term engagement with both projects (Gray, 2022;Gray and Mooney, 2011;Porter, 2015, 2017;Kallin 2018Kallin , 2021, we assert that there is much to learn from scrutinising the interruptions and suspensions that routinely beset urban regeneration projects. Several themes emerge from our research. First, we reiterate the enormous public cost and risk associated with state-led and state-financed UDPs. Second, we argue that these public costs and risks are a sine qua non for project initiation and reproduction, especially in large-scale 'economically risky neighbourhoods' with marginal locations, poor amenities, fragmented ownership patterns and large amounts of vacant, derelict and contaminated land (see Hackworth and Smith, 2001: 469). Third, the assembly, sale and remediation of public land by public authorities are essential if not sufficient for ostensibly private urban regeneration processes. Fourth, the spectres of market failure, state rescue, and capital's 1 'welfare dependency' are omnipresent in such projects and no mere exception. Finally, the public risks/ costs associated with large-scale UDPs cannot guarantee success, underscoring the inherent vulnerability of capitalist reproduction -even when buttressed by massive public subsidy. Indeed, the confluence of pumppriming public subsidy and failed regeneration often results in significant loss of public land and money with neighbourhoods regularly remaining in regeneration limbo for decades.
To develop our argument, we build on Smith's (1979Smith's ( , 1996 rent gap thesis by combining our notions of the 'reputational gap' (Kallin, 2017), where the discursive difference between a maligned present and utopian urban future justifies large-scale intervention, and the 'state subsidy gap' (Gray, 2022), where the material-economic difference between an unprofitable urban scenario and a potentially profitable one must necessarily be bridged by state subsidy. Crucially, bringing these mutually reinforcing iterations of rent gap theory together sharpens our understanding of the formation and closure of rent gaps as two distinct but inter-related processes over time (see Krijnen, 2018). The paper begins with a discussion of rent gap theory with specific reference to large-scale areas of urban devalorisation, before outlining our theoretical framework, our case study contexts in Glasgow and Edinburgh, and how the melding of reputational and state subsidy gap theories illuminates multiple processes of state intervention in the formation and (non)closure of rent gaps in each UDP. In conclusion, we argue for more scrutiny of regeneration failure or stalling, and the always risk-laden relationship between rent gap formation and closure.
Urban devalorisation and the necessity of state intervention Smith's (1979: 545) rent gap thesis, which refers to 'the disparity between the potential ground rent level and the actual ground rent capitalised under the present land use', remains a powerful heuristic for understanding processes of gentrification. Above all else, it implies 'an economic gap between actual and potential land values in a given location' (Smith, 1987: 463), the formation and closure of which is dependent on 'collective social action' involving numerous actors including builders, developers, landlords, banks, government agencies, estate agents, consumers and others (Smith, 1979: 545) as well as a multitude of social, economic, cultural, political, legal, discursive and environmental determinations (Clark, 1995;Krijnen, 2018;Smith, 1996). Indeed, gentrification in the OECD countries, especially in large-scale UDPs, has long been recognised as primarily state-led, state-based and statefinanced (Hackworth and Smith, 2001) and typically underpinned by well-established processes entailing the 'socialization of cost and risk and privatization of the possible benefits' (Swyngedouw et al., 2002: 552). Ghertner (2014Ghertner ( : 1554 challenges an assumed presupposition that reinvestment necessarily follows disinvestment and that 'land from which lower classes are displaced finds a ''higher and better use''. Yet, as numerous scholars have reiterated, rent gap theory was never conceived as a predictive explanation for gentrification in all its manifestations (Clark, 1988;Slater, 2017;Smith, 1987). Indeed, recent work has focused more on how rent gaps can be generated from relatively stable or even rising rent levels and how potential values can be maximised through housing financialisation (Aalbers, 2019;Christophers, 2022;Risager, 2022), platform capitalism and digital technologies (Fields, 2019;Wachsmuth and Weisler, 2018), the repeal of rent controls (Fields and Uffer, 2016;Krijnen, 2018;Teresa, 2019) or the use of debt-led shared equity schemes in newbuild 'affordable' housing (Kallin, 2020). The devalorisation-revalorisation relationship, then, is not always pivotal to rent gap creation and closure. Neither does an open rent gap guarantee closure. Profitability requires active state intervention with no guarantee of success, and potential failure is always intimately yoked to the profit logic that drives urban regeneration.
In an important intervention, Krijnen (2018: 441) argues that rent gap formation and closure should be viewed as distinct processes with no necessary corollary. Disaggregating these moments encourages us to scrutinise how rent gaps are formed, or not, in practice, and what particular conditions and forces inhibit or enhance the possibility of their closure. The classical notion of the rent gap as 'the spatial product of the complementary processes of valorisation and devalorisation' (Smith, 1986: 29) remains central, if not sufficient, to rent gap formation and closure, as our case studies reveal. But this complementarity is never removed from wider contradictions and the uneven development inherent to capitalist relations (as the original rent gap theory made clear) and should never be thought of as a relationship that always 'succeeds'. This cautionary note is especially important in the CG and Waterfront Edinburgh project areas. Both have long been characterised by poor amenities, fragmented patterns of land ownership, vacant, derelict and contaminated land, and recent histories of failed or incomplete redevelopment. In such conditions, state intervention, via public subsidy and boosterist discourse, is especially important in forming and potentially closing the rent gap. The concepts of the 'reputational gap' and the 'state subsidy gap' are deployed here precisely to grasp how statecraftinvolving policy, planning, financing, policing and discourse -aims to facilitate profitable possibilities for developers. We develop these concepts by way of an intriguing 'analytic puzzle' within gentrification and rent gap research (see also Gray, 2022).
If, following Smith, devalorisation is a necessary if not sufficient condition for revalorisation, Slater (2017) asks, why does gentrification rarely occur in the most devalorised parts of a city, but often in less devalorised areas that do not necessarily appear to offer the greatest potential for profit? He cites Hammel (1999bHammel ( : 1290, who provides one possible explanation: Inner-city areas have many sites with a potential for development that could return high levels of rent. That development never occurs, however, because the perception of an impoverished neighbourhood prevents large amounts of capital from being applied to the land. Slater (2017) suggests that the role of territorial stigmatisation in distorting land valuation, defaming the local neighbourhood and potentially thwarting inward investment may provide an important clue to this analytic puzzle. Returning to Krijnen's distinction between rent gap formation and closure, however, we would argue that territorial stigmatisation on its own is more likely to contribute to the formation rather than the closure of rent gaps by encouraging disinvestment and devalorisation of land values through discourses of denigration and decline. Territorial stigmatisation (which can be considered as a component of the negative side of the reputational gap) is often crucial to the creation of rent gaps, but it cannot provide a sufficient answer in itself to the question of rent gap closure -which is capital's ultimate objective. Hammel (1999a) clarifies this by making it clear that land valuation for highest and best use operates only as an ideal. In practice, 'land value is based on the sale of surrounding comparable parcels' (Hammel, 1999b(Hammel, : 1290. Capitalised land values are determined by the 'immediate or neighbourhood context' at the local level (Hammel, 1999b(Hammel, : 1291. As Hammel notes, this argument is compatible with Smith's theory, which emphasises the importance of those relations of propinquity that Hammel considers vital for potential rent gap closure: 'the ground rent that can be appropriated at a given site depends not only on the level of investment on the site itself but on the physical and economic conditions of surrounding structures and wider local investment trends' (Smith, 1996: 190). The rent gap, Smith (1996: 189-190) continues, is at its optimum when '[f]rom one block to the next' there is a sharp economic line in the landscape between 'different economic worlds' of existing market and profitable opportunity. It is irrational, he contends, for any real estate investor to commit large amounts of capital to the maintenance of a pristine building stock amid neighbourhood deterioration and devalorisation since any localised economic benefits are soon dissipated in a sea of depressed neighbourhood-wide ground rent values (Smith, 1996: 90-91). This reflects classic issues in land revaluation: the problem of attempting to create 'islands of renewal in seas of decay' (Berry, 1985) as opposed to overcoming 'islands of decay in seas of renewal' (Wyly and Hammel, 1999). The latter is a far more feasible option for developers since the former requires significant levels of state intervention to even initiate the process. This is where the state subsidy gap comes in.
It would be harsh to describe the CG and Waterfront Edinburgh UDPs as 'seas of decay', not least because that would risk reproducing the logic of territorial stigmatisation. Yet, the problem of largely contiguous urban devalorisation unmoored from adjacent higher land values is a central issue for both projects. In such 'economically risky neighbourhoods' (Hackworth and Smith, 2001: 469) -that is, risky for investors and developers -the combined theories of the reputational gap and state subsidy gap lend themselves to a deeper understanding of the formation and closure of rent gaps, disclosing how active state intervention is a sine qua non for the initiation, reproduction and potential completion of large-scale UDPs.

The reputational gap and the state subsidy gap
If territorial stigmatisation represses land values, potential rent gap closure only becomes meaningful where the denigrated present is juxtaposed against visions of a shining future. Enter the reputational gap, a concept aiming to capture the moment at which the gulf between reality (as it is represented) and potential (as it is imagined) is wide enough that it leads not only to rent gap formation but legitimises processes of state economic intervention that assist rent gap closure (Gray and Mooney, 2011;Kallin and Slater, 2014). 'Potential' is always imaginary and discursive, where the possibility of more profit is prefigured, crucially, by the idea of more profit, with all the symbolic trappings of wealth and 'progress' that undergird this imaginary (Kallin, 2017(Kallin, , 2021. The juxtaposition of 'what could be', held in tension with representations of 'what is', serves to increase the likelihood of intervention in the urban landscape. In this way, 'stigmatised territories are frequently denigrated in relation to the ''potential'' of the land they occupy' (Sisson, 2021: 664-665).
The reputational gap does not contest the economic logic of the rent gap, but highlights that (dis)investment decisions are rarely divorced from moralistic judgements that inscribe (and deny) a 'value' which is never only economic. As different histories of slum clearance, displacement and 'development' show, systems of valuation are often steeped in class disgust and racial discrimination, where 'narratives often become indistinguishable from the basic empirical identities of buildings, neighbourhoods, and entire cities' (Weber, 2002: 524). Recalling our discussion of Hammel's (1999aHammel's ( , 1999b work above, the reputational gap can exist in places where rent gap closure remains hypothetical, for it is easier to open but harder to close rent gaps in more 'risky', peripheral, or stigmatised neighbourhoods. This is where the state becomes crucial; in neighbourhoods where 'something must be done', the reputational gap takes on significant power, denigrating areas in order to lower land values for predatory capital while generating utopian images so that the idea of profitable urban renewal can be formed into some kind of reality. Representation, however, has limited power to concretely derisk economic investment in historically marginalised and devalued areas. Enter the state subsidy gap: the economic gap that the state must necessarily fill in order to de-risk development and pump-prime the conditions for potentially profitable private investment, especially in economically risky neighbourhoods. The state subsidy gap functions in both the formation and closure of rent gaps. In the first place, state intervention (or its conspicuous absence), abets devalorisation and the formation of rent gaps through processes of state retrenchment, disinvestment, abandonment, revanchism, stigmatisation and state exceptionalism. 2 In the second place, the state intervenes economically to address market failure and devalorisation, supporting revalorisation and the potential closure of rent gaps by de-risking the conditions for profitable investment. The state bridges the subsidy gap by such expensive and unprofitable initiatives as cheap land and property sales, land and property giveaways, complex processes of land assembly (often involving inflated land purchases from private landowners), the remediation of vacant and derelict land, tax breaks, expenditure on policing/security, and processes of eminent domain/compulsory purchase (see Gray, 2022). Combining reputational and state subsidy gap theories here aids understanding of the temporally extended, riskladen relationship between rent gap formation and closure, often occurring over decades without real success -even in developers' terms. While the reputational gap has an essential role in the formation of rent gaps, urban development ultimately requires concrete economic investment: this is where the state subsidy gap is imperative. Understanding the relation between these processes is vital for comprehending variegated strategies of state intervention over the longue dure´e of urban development processes. Before putting these terms to work together, we provide a brief overview of our UDP case studies.

Urban context(s)
By the 1970s, following exceptional levels of industrial decline, capital flight and comprehensive demolition without redevelopment, Glasgow, and its East End in particular, suffered wholesale urban abandonment and devalorisation, characterised by vast swathes of derelict contaminated land. The Glasgow Eastern Area Renewal (GEAR) project (1976)(1977)(1978)(1979)(1980)(1981)(1982)(1983)(1984)(1985)(1986)(1987), a proto-neoliberal public-private urban regeneration project, was established to tackle 'the most striking example of metropolitan decline in the United Kingdom' (Wannop and Leclerc, 1987: 70). Notably, for this discussion, GEAR failed to pump-prime significant private investment and resolve urban decline despite substantial public funding (Gray, 2015: 208-211). With another period of sustained disinvestment came deepening urban decline and territorial stigmatisation, opening up a reputational gap -a new urban frontier (Gray and Mooney, 2011) -that the 2014 Commonwealth Games (CWGs 2014) and CG projects (and accompanying state intervention) promised to finally close in the late 2000s.
Though Edinburgh is better known as a prosperous city of finance and tourism, its industrial decline and reputational defamation parallel Glasgow's East End in the Northern Docklands, which is dominated to the west by the former Granton Gasworks (the largest in Scotland by the mid-1960s, mothballed in 1987) and to the east by the former Granton Harbour (closed to commercial traffic in 1974). By the late 1990s, Granton contained over half of the city's vacant land (City of Edinburgh Council Development Committee, 1997). Waterfront Edinburgh Limited (WEL) referenced the reputational gap explicitly, calling Granton a 'cut-off, polluted, contaminated and forgotten part of Edinburgh' in one sentence before promising a 'world-class destination' in the next (Llewelyn-Davies, 2000: 2).
Following devolution from the UK in 1999, the Scottish Government (then the Scottish Executive) largely followed UK 'urban renaissance' policy, emphasising private sector involvement and vesting powers of coordination into bodies that defied categorisation as either 'public' or 'private' (Jones and Evans, 2006). The 'private' companies set up to lead both projects typified this era. Waterfront Edinburgh Limited (WEL) was a partnership between the City of Edinburgh Council (CEC) and Scottish Enterprise (themselves a non-departmental body of the Scottish Government) whilst CG is a partnership between Glasgow City Council (GCC), South Lanarkshire Council and Scottish Enterprise. Both receive(d) funding from the Scottish Government, with each company constituted to access grants unavailable to local authorities. Whilst this arrangement has remained largely unchanged in Glasgow East, in Granton, WEL has since been superseded by Granton Waterfront (GW). Interestingly, this is a project but not a company, pursued by a state-led partnership without private partners (led by CEC, with National Galleries Scotland, National Museums Scotland, The Scottish Government, The Scottish Futures Trust -infrastructure-financing arm of the former -and Edinburgh College). There are significant differences between each area, but in each case the state acts as both cheerleader and financier, bearing substantial risks and costs underpinning potential regeneration.
Urban regeneration in Glasgow's East End was given a major boost when the city won the bid to host the 2014 Commonwealth Games in 2007, dovetailing with the longer-term, ongoing, CG project . The CWGs 2014 was led by a separate partnership between the Scottish Government, GCC, Commonwealth Games Scotland (sporting body) and Glasgow 2014 Ltd (Organising Committee). Whilst they relied on existing venues, major new developments -the Emirates Sports Arena and the Athletes' Games Village -were constructed in Glasgow East with CG preparing the land and infrastructure. In terms of scale, CG, 'Scotland's biggest and most ambitious regeneration programme', is the larger of the two projects, covering 840 hectares and containing an estimated 340 hectares of vacant and derelict land as of 2008. The land is a patchwork of Council-owned land and private property, with CG given the role of land assembly and disposal so that regeneration can proceed under its aegis. Primary objectives are the creation of 21,000 new jobs, 10,000 new housing units, a population increase of 20,000, 400,000 square metres of business space, and the pump-priming of £1.5 billion of private sector investment in the project area. These plans were immediately imperilled by the 2007 to 2008 global financial crisis, though the threat was mitigated to some extent by previously contracted infrastructure for the CWGs 2014.
On Edinburgh's waterfront, recent attempts at urban regeneration have been more convoluted, involving overlapping visions, shifting landowners, and a confusing mixture of labels. Edinburgh Waterfront refers to a wide stretch of coastal land that, since the early 2000s, has framed disparate development projects at Leith Docks, Ocean Terminal, Western Harbour and Granton. The Waterfront Edinburgh project was launched in 2000, concerning only the land at Granton. The original plan covered some 140 hectares of land, split ownership between WEL (57 hectares), Forth Ports at Granton Harbour (39 hectares) and National Grid on the former gasworks site (44 hectares). With land remediation and infrastructure acting as bait for private investment, WEL sought to create 'one of the finest development opportunities in Europe' (Waterfront Edinburgh Limited, 2002: 1), but they were never vested with the power to coordinate the other landowners (Kallin, 2021). Unrealised plans include: multiple luxury hotels, a World Trade Centre and even an artificial island with a beach. The original 'masterplan' was interrupted to such a degree that the project either side of the 2007 to 2008 crisis looks markedly different. Since 2020, after a decade of stalled development, the GW project inherited WEL's and National Grid's land. To simplify matters, the case study is presented here as three separate sites, mapping onto the original division of landowners, referred to here as the WEL landholdings, the Granton Gasworks site and Granton Harbour.

The state to the rescue: Glasgow and Edinburgh
Here we provide an overview of specific sites within the two UDPs to show the dynamic relationship between the reputational gap, the state subsidy gap, and the rent gap. In both cases decline and stigma have been juxtaposed against growth and success (the future) and there is substantial evidence of state subsidy. Crucially, if the opening and closure of rent gaps are temporally separated (sometimes by decades), state intervention is similarly dispersed over time. The symbolic devalorisation that accompanies the opening of the reputational gap might be generations in the making, even if it is only when discourses of a 'better' (alternative) future begin circulating that we can conceive of this as an intentional gap. Once stigmatisation and city boosterism occur in tension, then the reputational gap becomes clear. But this tension is never enough to materially transform neighbourhoods. This is where the state subsidy gap comes into play.

Glasgow: Clyde Gateway
A reputational gap at the CWGs 2014 Athletes' Village site in Dalmarnock, and the East End more generally, was generated by decades of neglect and disinvestment by local and national authorities (Gray and Porter, 2017), and by virulent forms of territorial stigmatisation from politicians and sections of the media, most recently objectifying the area with Conservative Party-led 'broken society' narratives: 'Glasgow's Guantanamo', 'a hideous social experiment', 'a ghetto ringed by some of the saddest statistics in Britain' (see Gray and Mooney, 2011: 12-16). CWGs 2014, however, provided the possibility of closing this reputational gap (explicitly referenced in regeneration agency literature), with the Games and CG projects promising largescale state intervention to close the yawning state subsidy gap between unviable and profitable investment opportunities (see Gray, 2022). Processes of rent gap closure by the local and national state on the Games Village site have included land assembly, remediation and site preparation to the sum of more than £30 m of public subsidy (Gray and Porter, 2015). The public cost of remediation alone on the 35-hectare site was £7.7 m, with funding from GCC via the Scottish Government's Vacant and Derelict Land Fund. Despite the onerous public cost of land expropriation, assembly and remediation, City Legacy Consortium, the public-private partnership who won the bid to construct the Athletes' Village, were gifted the entire site at 'nil cost' in order to avoid any problems the consortium might face raising private finance (Stewart, 2009).
Meanwhile, several other derelict, vacant and abandoned land parcels adjacent to the site were purchased from developers by GCC at a total cost of approximately £30 million. To take the most notorious example of these 'dodgy land deals' (Gray and Porter, 2015), Charles Price bought property from GCC along the muchdenigrated Springfield Road area in 2005 to 2006, shortly preceding the announcement that Glasgow would host the Commonwealth Games, for approximately £8 m, before selling it back to GCC in 2008 for £20 m inclusive of VAT (Gray, 2022). Notably, no compulsory purchase orders (CPO) were ordered to expropriate developers such as Price, despite legal provision for such exceptional measures in Section 42 of the Commonwealth Games Bill, yet a blanket CPO was placed on all housing and retail properties on the Dalmarnock Village site in August 2010 (Gray and Porter, 2015).
After the event, the Village accommodation was retrofitted and sold for private homeownership (300 units) and social housing (400 units). Initially, 1400 private homes were planned, with 1100 for private sale and 300 for social rent. Yet the 2007 to 2008 crisis shifted the ratio between private and social sector housing to more equal terms (Wainwright, 2014), with the state mediating market failure. A second residential phase planned by the Consortium for private homeownership (125 units) remains unrealised, leaving a significant part of the site vacant and overgrown. In 1999, the site comprised 1589 homes, almost exclusively social rented, in structurally-sound tenement buildings. As such the Dalmarnock regeneration process, including the longer term devalorisation process preceding the Games -routinely ignored by media and commentators -translates into a loss of nearly 1200 social housing homes and around 3000 people displaced (Gray, 2015).
Nearby, the 64-hectare Shawfield National Business District site, targeting the commercial office sector, is pivotal to meeting CG's ambitions on business space and jobs in the area. The site was once home to J&J Whites Chemical Works, which left behind a legacy of highly toxic soluble hexavalent chromium contamination (Walker, 2005), necessitating extensive remediation work. The reputational gap formed from a treacherous aggregation of corporate neglect, environmental degradation, vacancy and dereliction, territorial stigmatisation, and sustained disinvestment/under-development since the Works' closure in 1965. Such catastrophic devalorisation means that rent gap closure following a simple devalorisation-revalorisation dialectic is fraught with difficulty. The extent of state subsidy employed here became a sine qua non for merely potential rent gap closure. The first phase, involving land assembly, remediation, compulsory purchase and infrastructural improvements across 11 hectares, necessitated £27.8 m of public subsidy from European, national and local authority sources (Gray, 2022). A further £14.2 m of subsidy was committed to remediate 2.5 hectares on phase 2 of Shawfield, including new road infrastructure on the first phase of the Shawfield Regeneration Route. Despite this significant outlay, however, 53 hectares of heavily contaminated land over phases 2 and 3 still require remediation and infrastructural improvements, at a projected cost of £54.2 m. Thus, current estimated public costs of site remediation and preparation amount to £96.2 m (Clyde Gateway, 2019). To date, only one building -the Red Tree Magenta building -has been developed on phase 1 (for 169 workers), suggesting that significant public subsidy does not assure new development. In what amounts to a begging letter to the Scottish Government and GCC, CG have made it clear that capturing subsidy is both an explicit objective and a necessity if development is to occur at all: 'There can be no argument that without intervention by the public sector, of which Clyde Gateway is currently the main vehicle, market failure will remain' (Clyde Gateway, 2019). Bridging the gap from 'what is' to 'what could be' remains elusive: Shawfield National Business District remains far from completion, with considerable doubts attached to obtaining the estimated £50 m plus public subsidy still required.
Riverside Dalmarnock is an intriguing example of the political economy of land assembly/disposal in the Clyde Gateway area. The site was long dominated by Dalmarnock Power Station (built 1915, demolished 1980. It subsequently 'lay vacant and derelict, often used for illegal dumping and fly tipping' (Scottish Housing News, 2015), leading to profound forms of territorial stigmatisation -as replicated in adjacent sites throughout the East End. However, as we have argued, the reputational gap is at its most potent when stigmatisation and city boosterism occur in tension with active state intervention, and the coming of CWGs 2014 and CG projects opened an opportunity for potential rent gap closure via complex delivery of state subsidy for urban redevelopment.
After CWGs 2104, CG sold the 6.5-hectare site to Link Group for £5.7 m in April 2015 (along with another 2.4-hectares, worth another £1.69 m). Yet, the funding for this 'acquisition by Link' was provided by Glasgow City Council -one of the main partners in CG -via Scottish Government funding (Scottish Housing News, 2015). This followed considerable prior investment by CG, who paid Murray Estates £4.5 m in 2010 for the heavily contaminated site (concrete, steel and Japanese Knotweed), which was bought by Murray Estates for only £375,000 in 2005 (again, just before Glasgow was announced as host city of the Games). CG then spent a further £3 m of public money on land remediation and preparation before GCC bought the site on behalf of Link (The Herald, 2015). With total public spending on the land at £9,239,000, the loss on the deal amounted to £4,416,600 sans VAT: effectively, GCC bought the site from themselves, via Murray estates and CG, only to gift it gratis to a developer at a multi-million-pound loss (The Herald, 2015). As Lord Smith of Kelvin, former Chair of CG, explained: 'This deal epitomises what Clyde Gateway is all about. We have stepped in to deal with market failure' (Scottish Housing News, 2015).
After being gifted the site, Link Group, one of the biggest social landlords in Scotland, and Laurel Homes, a private housing developer, have built 562 homes on site: 206 private homes (Laurel Homes), and 356 'affordable' homes (Link Group). Despite the immense public subsidy that made the project feasible, the overall total is skewed towards private homeownership and intermediate (marketised) 'social housing': 344 homes, including private sale, new supply shared equity (NSSE) and mid-market rent (MMR); 218 homes for social rent. This reflects a general trend of social housing commodification across the CG project, where targets were never set for social rented housing (Gray, 2015: 226) and where 80% of all housing constructed so far has been designated for private sale or intermediate forms of tenure (Garcı´a-Lamarca and Gray, 2021).

Edinburgh: Waterfront Edinburgh
Seeking to transform the so-called 'scarred and derelict' area of Granton into 'a desirable place in which to live and work' (Gracie, 2003: 119), £9 m was spent on decontaminating WEL landholdings, followed by investment in new roads and pressure for a light rail system (WEL, 2002). The commitment of £375 m in public funding for the North Edinburgh Tram Loop, though not directly linked to WEL's budget, was considered central to the project's success or failure (Waterfront Edinburgh Limited, 2003). Exemplifying the state subsidy gap, expenditure on public transport was expected to 'have a significant impact on the commercial viability and overall success' of the project, otherwise there was 'a clear risk that the area would remain isolated' (Tym & Partners, 2007). WEL's (state-funded) promoters drew the reputational gap wide open to justify closing the state subsidy gap: if 'Granton had the biggest concentration of social and employment issues in the city', the site was 'absolutely fantastic' if you could imagine it completely reconfigured (cited in Kallin, 2018: 46). WEL (2002) aimed to 'increase land values to the levels where the aspirations of the Master Plan [became] viable' through assembling land for developers to acquire without the risk of remediation, transport, or reputational reboot, thus attempting to rent gap closure. This was only partially successful. The largest building completed pre-crisis was Phase 1 of the 'Upper Strand', a £75 m joint venture between the Burrell Company and Places for People, the UKs largest residential landlord -a housing association that increasingly pursues for-profit development (a statutory-conforming minimum of 15% of this block was earmarked for 'affordable' status). Phase 1 stood alone for years, and the shifting fate of the site is still traceable. On the tallest block, conceived as a 'landmark' tower, the sea-facing wall is almost entirely devoid of windows, intended to adjoin a Phase 2 that was never built.
The 2007 to 2008 financial crisis hit WEL hard, with landholdings plummeting from £33 million to £14.5 million (Edinburgh Evening News, 2008). The company was wound down in 2014, at which point WELowned assets (including unsold land) were brought back 'in-house' to CEC. The dynamic of state intervention mitigating the consequences of market failure is once again clear: the 'private sector' vehicle of the local state, having failed, is bailed out by the local state. Yet the state subsidy gap (like both the reputational gap and the rent gap itself) is never fully closed. The original WEL landholdings have seen only piecemeal developments since, including several residential developments by Housing Associations, but these remain characterised by incoherent policy and planning.
The 45 (Christophers, 2020: 330). CEC intervened, purchasing the site for £9 m and initiating a second phase of state-led regeneration via GW in 2020. Covering 120 hectares of land, this picks up the pieces left by the original WEL project and by the private sector (National Grid). Both failed according to initial targets -the reputational and rent gaps remained openso the state intervenes again. GW is also included in The Edinburgh and South East Scotland City Region Deal, providing access to funding from both the UK and Scottish Governments for housing, infrastructure and 'sustainability' tied to economic growth. The council are pushing for 'full investment' through this route, but almost half of projected costs (£307.9 m of £689.1 m) are already covered by various state funding streams (City of Edinburgh Council, 2021). In late 2021, the UK Government also announced £16.5 m of funding to rehabilitate the remaining gasometer via their 'Levelling Up' policy and CEC have committed £196 m to 'accelerate the regeneration'. Time will tell if this pump-priming public investment will finally lure private capital to the waterfront.
In the mid-2000s, following the example of WEL and National Grid (but refusing to cooperate with either), Forth Ports, partnered with private developer Gregor Shore, developing two luxury blocks at Granton Harbour with a combined total of 255 apartments on reclaimed land. The blocks were valued at £163 million in early 2007 (gregor.co.uk), but the developer was bankrupt by October the following year, highlighting how 'risky' the entire project remained (and hence how necessary state subsidy remains). Forth Ports then declared that their land had 'no immediate development value' whatsoever (Bolger, 2011), banking it for the best part of a decade. In the interim -between grand plans, stasis, and new grand plans -the reputational gap opened wide again, with the ruins of Granton's failure inscribing a new layer of stigma (Kallin, 2021). Recently, development has been revived, with two Housing Associations acquiring land to construct mixed tenure housing, whilst the land acquired and sold on after the bankruptcy of Gregor Shore has been incorporated into a new private development branded 'Edinburgh Marina'. If it comes to fruition, this will incorporate 102 private 'luxury' apartments (£660,000 for a 2-bed flat; £1,000,000 for a penthouse) alongside 'luxury' retirement apartments (£534,000 for a 2bed flat; £1,077,000 for a penthouse), and a Hyatt Regency Spa Hotel -a vision with no qualms about its exclusivity (https://edinburgh-marina.com).
Notably, this development takes place against the wishes of CEC, which took the developer to court in 2020, claiming that planning permission granted in 2003 was not fit for purpose (Matchett, 2020). If the Granton Harbour development seems autonomous from state subsidy, it should be understood that its viability relies on the 'sea of [infrastructural] renewal' on adjacent land -underpinned by extensive state fundingwhich provides crucial positive externalities surrounding this 'island' of private sector renewal. Moreover, this case study also corroborates arguments that Housing Association developments became increasingly attractive for developers and financier's post-2007 to 2008 crisis, since they offered stable investment, albeit with lower returns, underwritten by guaranteed state support (Wainwright and Manville, 2017). This suggests that any prominence of housing association properties in regeneration projects may well signify private sector failure and state rescue. In this respect, Granton Harbour echoes the Games Athletes Village and Riverside Dalmarnock on the CG site, where private sector housing market failure has paradoxically opened up the possibility for more social housing provision, albeit heavily mediated by the market.

Conclusion
Despite the entrepreneurial myths of free market ideology, large-scale UDPs are typically massively underpinned by state intervention, which opens the reputational gap (through disinvestment, territorial stigmatisation and the propagation of new visions for old places) and seeks to close the state subsidy gap (via substantial pump-priming funding, socialisation of risk and financial aversion of market failure). Rather than publicly funded social democratic programmes of social provision (public housing, public transport and other forms of public goods), such projects are the real 'subsidy junkies' of contemporary urbanism -where market failure is endemic and state intervention is tacitly admitted as an essential correction of market failure, even if free market ideology cannot bear to admit it. Drawing on a key argument in Harvey's (1989) classic paper charting the shift from managerialism to entrepreneurialism, Ward and Wood (2021) argue that analyses of risk absorption by the public sector, and public sector funding of privatised infrastructural provision, have been downplayed in urban regeneration research. We agree. We also believe that these investigations should start from an interrogation of market failure as no mere anomaly but as an inherent feature of urban regeneration and urban capitalism.
By combining the 'reputational gap' (Kallin 2017(Kallin , 2021 and the 'state subsidy gap' (Gray, 2022), this paper underscores the centrality of active and multi-modal state intervention in both the formation and potential closure of rent gaps over the longue dure´e, describing the complex forms of state intervention (and non-intervention) over time that are a sine qua non for urban development to proceed -never mind reach completion -in large-scale areas of urban devalorisation. Attempts at rent gap closure are extremely context dependent (Risager, 2022), however our two case studies corroborate our argument that the reputational gap and state subsidy gap can be productively mobilised to comprehend rent gap processes in cities with quite different socioeconomic contexts, yet with very similar problems in addressing specific large-scale 'economically risky neighbourhoods'. The combination of our respective neologisms helps illustrate how the state intervenes at multiple levels to address the fraught relationship between profane urban reality and urban promise, contrasting the rhetoric of urban entrepreneurialism with the banal realities of capital's subsidy dependence. Such command of subsidy could be seen as a manifestation of power by capital, mediated by the state. However, we contend that it is better seen as an index of weakness, expressing the failure of capitalist production and a secular decline in growth and profit rates since the 1970s (Harvey, 2012;Kliman, 2011). In particular, our two-fold theoretical framework here illustrates how the durational formation and closure of rent gaps is fraught with risk, and how reliance on state intervention across these processes over time (often decades) is a vulnerability rather than a strength of capital. For without such state mediation, multiple regeneration projects would fail to launch, never mind deliver. Moreover, substantial state intervention never guarantees 'successful' urban regeneration projects, as our comparative study shows, and when it does -on developers' terms -private gain is typically inextricably dependent on public pain (Swyngedouw et al., 2002).
Some results from our comparative study suggest that the failure of urban capitalism may produce potentially socially beneficial outcomes -as witnessed in the use of the social housing sector to fill the void left by fleeing (or never-existing) private capital investment. This is evidenced in the CWGs 2014 Athletes' Village following the financial crisis of 2007 to 2008, and in Granton, which has latterly incorporated a more proactive strategy of affordable housebuilding after two decades of market failure. Yet, these isolated incidences were apparently compelled by a lowering of market ambitions rather than structurally transformative progressive policy regimes. And since speculative finance capital is increasingly mobilised within UK social housing (Kallin, 2020;Wainright and Manville, 2017;Wijburg and Waldron, 2020), such potentially liberalreformist developments should be examined with careful scrutiny as another instance whereby the apparent autonomy of local state urban development is interpellated by private capital whilst being underwritten by state subsidy. In conclusion, we hope that urban scholars will embrace a research programme investigating regeneration failure as a means of establishing the instability of capitalism as a reproductive schema. This critically engaged strategy -underpinned here by our dual rent gap theoretical framework -offers a deeper scrutiny of capitalism's vulnerabilities and dependencies, exposing its weak points and insecurities in ways intended to assist the critical inquiries and struggles of those with most to benefit from overcoming neoliberal urbanism's gross inequities.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.