The Iron Triangle of Urban Entrepreneurialism: The Political Economy of Urban Corruption in Spain

Despite a cascade of corruption scandals over the last decade, the role of political corruption has yet to be accounted for in the scholarship on Spanish urban entrepreneurialism. This is an omission that extends to the broader literature on the topic as well. This paper looks at the theoretical causes of this neglect and explores ways of integrating corruption into the study of urban entrepreneurialism. To do so, it offers a systematic analysis of the role of clientelism in the political economy of Spanish interurban competition. Contrary to the existing literature, which at best explains away corruption as a logical outgrowth of intensified competition for real estate capital, this paper argues that corruption is rooted in a political structure that precedes the emergence of urban entrepreneurialism. This structure is identified as an “iron triangle”, a complex of colluding interests formed by the state, the real estate industry, and political parties.


Introduction
In 2018, Spanish Prime Minister Mariano Rajoy was ousted from power after a corruption sentence confirmed that his party, the conservative Partido Popular (PP), had been running a decades-long scheme to trade politically allocated resources for illicit campaign donations. To nobody's surprise, the country's powerful construction companies figured prominently amongst the donors-after all, the fall of Rajoy was only the culmination of a decade of relentless scandal around urban development. The corruption avalanche began with the arrest of high-ranking officials of the Marbella city council in 2006, a spectacular event that was followed by a myriad of local scandals across the country, involving politicians from across the political spectrum. Over the years, the ramifications of these scandals gradually worked their way up the scales of the state. From 2009, evidence emerged of criminal networks specialised in brokering clientelist transactions for PP's donors at the regional level. In 2013, serious allegations of corruption reached the highest echelons of the state with the leak of the "B arcenas papers", named after Rajoy's chief treasurer, Luis B arcenas. These handwritten documents detailed the existence of an entire parallel bookkeeping system at the very top of the party, one that managed the party's bribe streams on the national scale.
This wave of scandals has had a profound impact in the country's imaginary, sparking public debates about the role of clientelism in the Spanish economy, and creating a chain of associations between corruption and the wave of urban overdevelopment that crashed the Spanish economy in the wake of 2007. Paradoxically, these developments have remained remarkably absent from the critical scholarship that over the last decade has examined Spain's dramatic experience with urban entrepreneurialism-a form of urban governance geared towards spatial competition for private capital flows Coq-Huelva 2013;Delgado Jim enez 2011;L opez and Rodr ıguez 2010). Indeed, on the rare occasions that corruption is cited in this literature, it tends to be explained away as an "extra-legal ramification" of inter-urban competition; a logical outgrowth of intensified processes of competition for real estate capital. But what is more striking is that this neglect transcends the Spanish literature. As we shall see, despite decades of sprawling scholarship, the theme of corruption has been largely neglected by the main theorists of urban entrepreneurialism as well. All of this poses a series of questions. First, why has the scholarship on urban entrepreneurialism omitted the role of corruption despite recurring evidence of its role in the built environment? Second, how can the analysis of corruption be integrated into the theory of urban entrepreneurialism? Third, what does the Spanish experience offer, both theoretically and empirically, to our understanding of how clientelism shapes urban governance?
To address these issues, this article offers a systematic analysis of the role of clientelism in Spanish urban entrepreneurialism, arguing that corruption was in fact constitutive of its political economic dynamics. Far from being a mere offshoot of inter-urban competition, the clientelist nexus between the state managerial class and the real estate industry is one that precedes urban entrepreneurialism and that has a logic of its own. This political structure was cultivated under Franco's dictatorship in the mid 20 th century and was later adapted to the political dynamics of the liberal democracy that succeeded it in the late 1970s. Through it, corporations competed for political influence to secure their market success, while political parties competed with each other by raising illicit funds to boost their campaign resources. Political patrons were incentivised to inflate their client's earnings as they usually exacted a previously determined portion of their profits, typically in the form of a 3% kickback. These dynamics then developed a symbiotic relationship with processes of inter-urban competition at the turn of the millennium, magnifying processes of real estate speculation at a time when local boosterism was highly popular amongst the electorate. By then, these clientelist exchanges were organised into regional, and even national, graft structures that went beyond the local logic of inter-urban competition.
The first part of the paper focuses on theoretical questions. It explores the omission of corruption from the literature on urban entrepreneurialism, tracing this neglect to the state theories that underpin this scholarship. To correct this, the article turns to critical studies of the "global South", borrowing the concept of informality to bring corruption into the analysis of urban entrepreneurialism. The second part of the article moves into the Spanish case, identifying the existence of an "iron triangle of urban entrepreneurialism": a three-way complex of colluding interests, the vertices of which are the state apparatus, political parties, and the real estate industry. The analysis starts by fleshing out the role of clientelism and informality in Spanish state formation, all to contextualise the role of the "iron triangle" in the Spanish political economy. The paper then unpacks each one of the "sides" of the triangle, dissecting the links between its vertices: the institutional mechanisms of clientelism (state-real estate); the criminal nexus between parties and developers (real estate-parties); and the reproduction of this kleptocracy (parties-state).
Urban Entrepreneurialism, Corruption, and the State Theorising the Competitive City: Urban Entrepreneurialism and Corruption Over the last three decades, a broad literature has emerged around the concept of "urban entrepreneurialism", a concept that explains the intensification of interurban competition in the midst of the rescaling of North Atlantic political economies. 1 The central premise of this literature is that the crisis of postwar political economies ("Fordist Keynesian regimes"), and their subsequent transition into neoliberal forms ("Schumpeterian workfare regimes"), was a process that had its mirror image in the urban sphere (e.g. Brenner 2003Brenner , 2004Harvey 1989;Jessop 1997aJessop , 1997bPeck and Tickell 2002). Under postwar political economies, North Atlantic nation-states were equipped with the capacity to coordinate industrial policy and roll out redistributive measures to offset the territorial imbalances of uneven development. Back then, urban governments were mainly responsible for relaying nationally planned policies at the local level. But this "urban managerialism" lost its purpose after the 1970s: as Fordist industry ran out of steam, Keynesian redistributive mechanisms were dismantled, and neoliberal agendas were "upscaled" to supranational institutions, resulting in a partial loss of de jure sovereignty for nation-states. The combined result was that cities and regions came under growing pressure to raise revenue of their own accord by attracting private capital flows, resulting in a heightened exposure to territorial competition. To meet this challenge, there was a trend to devolve or "downscale" the organisation of growth strategies to competing sub-national units. By the late 1980s, these pressures had given shape to entrepreneurial patterns of urban governance.
In the incredibly prescient article that popularised the concept, David Harvey (1989) identified three defining characteristics of this "urban entrepreneurialism": (1) a move away from service provision toward place-marketing and local boosterism, in order to make areas attractive for investment; (2) a displacement of rational planning for speculative urban development, often accompanied by local governments offering to absorb the risks of investors; and (3) the replacement of top-down forms of government for network-based, public-private partnerships.
The urban entrepreneurialism literature has successfully captured the basic dynamics of inter-urban competition inaugurated by "globalisation", inspiring a vast number of empirical case studies. But there has remained a notable gap in The Iron Triangle of Urban Entrepreneurialism this literature: the issue of corruption. In the broadest of terms, corruption denotes an abuse of entrusted power, though what this specifically means in a given context is bound to be shaped by struggles over what constitutes acceptable practice and governance (Doshi andRanganathan 2019a, 2019b). Here we will be focusing on "material" and "territorial" forms of corruption, that is, socially determined abuses over the distribution of resources and space. These forms of corruption typically require breaches in public mandates to the benefit of private actors, whether these breaches are administrative (distortion of routine bureaucratic procedures) or political (distortion of policy-making procedures) (Heywood 2007).
None of the theorists of urban entrepreneurialism cited above have dedicated much attention to the role of fraud and clientelism in urban governance. This is striking considering that corruption dynamics frequently underpin processes of accumulation through the built environment, not least due to the great dependence of urban development on politically allocated resources (e.g. permits, planning decisions, public contracts). A notable example is that of the Brazilian construction firm Odebrecht, which recently admitted before US authorities to having systematically paid $788 million in bribes across three continents (BBC 2019). It appears that the company even dedicated a fully functioning business unit to its bribing operations, as if political influence were another factor of production. As we shall see, the omission of corruption from Anglo-Saxon theories of urban entrepreneurialism extends to the literature on Spain as well, which is all the more striking considering the recent abundance of urban corruption scandals there.
The question, then, is why has the scholarship on urban entrepreneurialism neglected the role of corruption in spatial processes of capital accumulation?
The Limits to Functionalism: Urban Entrepreneurialism and the State A close inspection reveals that the gap around corruption in the urban entrepreneurialism scholarship is based on the different conceptualisations of the state underpinning this scholarship, which are imbued by varying degrees of functionalism. For example, much of Harvey's work relies on a highly structuralist theoretical system of his own creation: the switching circuits of capital (Harvey , 1982(Harvey :412-445, 2003. According to this system, urbanisation under capitalism has a counter-cyclical relationship to industrial overaccumulation. Whenever profitability thins in the manufacturing sector, what Harvey calls the primary circuit of capital, capital switches its investment to the built environment, the secondary circuit, which kicks in like an emergency power supply that keeps accumulation running. 2 In this schema, in which capital is represented as a selfpropelling force, the state merely performs a supportive role. The state either finances capital's pre-ordained switch, or enforces it, by removing sources of friction (e.g. unwelcome regulations, social unrest) or opening new markets for accumulation (e.g. privatisations, imperial expeditions). Harvey's theory of urban entrepreneurialism should be seen in light of this system. As patterns of capital accumulation morphed in the wake of the industrial crisis of the 1970s to give a greater salience to the built environment, urban entrepreneurialism emerged as a distinct form of statecraft to support this process.
The theory of capital-switching sees the state through a functionalist prism: i.e. state policy reflects what capital needs. This assumes that the state always acts as a manager of the collective interests of capital, and as such, renders invisible the ways in which formal policy channels are distorted by special interests, corruption included. At the heart of the problem is that neither the state nor capital are monolithic entities. The state managerial class is always traversed by conflicting interests, both in a strategic sense (i.e. differing views of how to manage the state) and in a material sense (i.e. differing forms of class reproduction for different types of state officials). Likewise, the bourgeoisie is split into fractions (e.g. landowners vs. developers, small businesses vs. transnational corporations, etc.) and their political interests only rarely align. The implications of this complexity for any theory of the state are two-fold. First, the idea of a state acting in unison on behalf of the collective interests of capital is highly implausible. To be sure, capitalist states do custody the reproduction of capital accumulation-at the very least to preserve their own political continuity-however, they never do so in the abstract, but rather, by attending concrete political dilemmas that always require them to make strategic choices and prioritise certain interests over others. Second, we should not take the impartiality of capitalist states at face value. Though capitalist states rule through formally impartial channels, the fact that state institutions are always traversed by political contradictions means that state officials will be pressured to bend, circumvent, or subvert formal rules in order to further their own goals. This requires us to take seriously the fact that, beneath an outward image of state autonomy, there are groups furthering their agendas through informal, clientelist means.
Neo-Gramscian state perspectives are better positioned to grasp this complexity. Bob Jessop's extensive writings have repositioned the concept of urban entrepreneurialism upon a wholly different theoretical platform: his own "strategic-relational" theory of the state (Jessop 1997a(Jessop , 1997b. 3 Jessop represents the state as a malleable battlefield, one being constantly reshaped by political conflict, such as the clash between different fractions of capital for the representation of their interests. For Jessop, the capacity of states to manage the reproduction of capital cannot be taken for granted-on the contrary, this is precisely what needs explaining. To be operational, states need to find ways to suspend (or at least, contain) the conflicts that traverse them. This requires two types of political efforts: state projects, to unite the state managerial class around a coherent institutional apparatus; and hegemonic projects, to resolve conflicts between particular interests and the general interest (Jessop 1990:9, 161). The strategic-relational approach has the potential to open up room to account for the politics of corruption. For example, Jessop insists that hegemonic projects will continue to privilege the "economic-corporate" needs of a dominant class or class fraction so long as this does not threaten the continuity of the hegemonic order.
Yet, paradoxically, neither Jessop nor his followers (e.g. Brenner 2003Brenner , 2004) have given much attention to corruption, which raises the question of what has prevented them from doing so. In his books on state theory, Jessop has dedicated brief discussions to the politics of clientelism, which he defines as "the exchange of political support in return for favourable allocation of politically mediated resources" (see Jessop 1990Jessop :162, 2015. In these discussions, he explains away clientelism as an exceptional and inherently unsustainable mode of governance. To support this claim, Jessop relapses on a functionalist form of reasoning: because oligarchic capture may weaken the state's transformative capacity (state project failure) and/or result in a hegemonic crisis (hegemonic project failure), "[n]either situation can be said to favour the reproduction of capital in general, and it is therefore in the political interests of capital to press for the transformation of the state" (Jessop 1990:162). As we shall see, this logical deduction is contradicted by historical evidence: long-term capitalist growth appears to be perfectly compatible with clientelist modes of governance.
A further problem lies in Jessop's representation of the state managerial class, to which he seems to afford a degree of functional coherence that he refuses to other classes. Jessop operates under the supposition that it is in the interest of state managers to safeguard the formal structures of the state, sometimes describing them as guardians of state unity in periods of hegemonic instability (Jessop 1990:9). But this neglects how the state managerial class is composed of different fractions-career bureaucrats, politically appointed administrators, elected politicians-and these have different dynamics of class reproduction. By extension, it also neglects how the different elements within the state managerial class often resort to informal means to ensure their own reproduction.
The next step, then, is to make adjustments to the strategic-relational prism so that it can cast light on the ways in which state power is exercised beyond the structures of legality. Only in this way will we be able to render visible the role of corruption in the governance of urban capitalism.

The Informal Dimensions of State Power: Corruption and the State
Corruption and clientelism have long been framed as antithetical to a proper functioning of the market. Mainstream corruption discourses are shaped by modernisation theory, which sees corruption as a symptom of incomplete capitalist development (for an excellent review, see Doshi andRanganathan 2019a, 2019b). During the Cold War, the source of corruption was identified in the legacy of pre-modern patrimonialism, which resulted in a blurred separation between the public and private realms. This discourse was reformulated in the 1990s, in the context of the neoliberal push for market-driven reforms in "developing economies". From then on, the source of incomplete development became an insufficient degree of market discipline, which left unrestrained the supposedly inherent "rent-seeking behaviour" of government officials. As is often the case, the techniques of domination tested out on the "South" were soon redeployed against the popular classes of the "North". At the turn of the 2010s, European elites navigated the Eurozone crisis by scapegoating the countries of the continent's "periphery"-mainly the European South-which were said to be dragging the whole continent down due to the immaturity of their social formations (Dooley 2018). Articulated in terms of cultural dysfunctionality, the immaturity discourse was deployed to instil market discipline on the profligate and corrupt societies of the European south.
Against this, most critical scholars (rightly) shifted the blame onto the Eurozone's institutional architecture, which in the decade preceding the crisis devastated the industrial base of the South and locked Mediterranean countries in a cycle of debt-led growth (Charnock et al. 2014:79-81). The problem is that the dynamic of this debate has encouraged many authors to downplay the impact of corruption in the countries of the European South, Spain included. The point should not be to deny the material existence of these abuses, as this only obscures the realities of class domination, but to reframe our approach to corruption in a more constructive way. To do so, we have much to learn from the critical scholarship on the so-called "global South". The reaction against mainstream corruption discourses by critical analysts of the global South has not been to downplay the existence of corruption, but to craft a repertoire of theoretical strategies to integrate the study of corruption into critical scholarship.
For our purposes, we must highlight three of these insights (see Doshi and Ranganathan 2019b). First, this literature has highlighted how the culturalist/racialised character of mainstream corruption discourses has served to overlook the existence of corruption in "developed" countries. To bring this back to Europe, rather than explaining away the role of corruption in the periphery, one could simply point out the evidence of analogous practices in the core. For example, only last year, the Austrian government fell after the leader of the junior coalition party (the farright FP € O) was filmed trading public contracts and privatisations to a Russian construction company that was offering financial and media support for his party (Opratko 2019). Second, some scholars have pointed to the lack of evidence sustaining the claim that corruption hampers market success. A frequently cited example is the accelerated growth of the countries of the Asia-Pacific region, where statesanctioned forms of "crony capitalism" were pivotal in their developmental strategies (Kang 2002). Third, there is a broad consensus that corruption cannot be analysed, let alone addressed, without careful reflection on histories of state formation (Doshi and Ranganathan 2019b). The role of corruption in a social formation cannot be theorised in the abstract, because corruption is a complex phenomenon that refracts social relations in multiple ways. As such, it needs to be understood in relation to the specific historical and political dynamics of a given social formation.
To build these insights into a theoretical framework, the notion "informality" provides a useful conceptual tool (Haid and Hilbrandt 2019;Roy 2005). Informality highlights how power is exercised in ways that go beyond the formal realm of the law, either because these are extra-legal outright, or because they involve a flexible deployment of existing laws and regulations. Importantly, the concept highlights the mutual constitution of formality and informality-the informal is always a "state of exception" created by the formal actions of the state. Despite its broadness, the concept is helpful insofar as it casts light on the interstices of the public-private divide, where a growing amount of decisions are made, even in the "global North". In the sphere of real estate, for example, the concept has been deployed to describe the unplanned outgrowth of urban areas, the usurpation of land by developers with the acquiescence of the state, or the creation of para-statal vehicles to fast track urban development, which is a typical feature of urban entrepreneurialism (see Doshi and Ranganathan 2019b). At a methodological level, the concept forces us to not take for granted the formal legal structures of the state, and as such, helps us challenge "the myth of formality in the global North" (Jaffe and Koster 2019).
To sum up: in the same way that the theory of urban entrepreneurialism needs strategic relationism to highlight the political conflict involved in inter-urban competition, strategic relationism needs to integrate the informal dimensions of state power into its outlook if it is to ever account for the role of corruption in the governance of the built environment. The remainder of the article will illustrate this theoretical synthesis through a systematic analysis of the role of clientelism in Spanish urban entrepreneurialism.

The Iron Triangle of Urban Entrepreneurialism in Spain Urban Entrepreneurialism in Spain
Over the last decade, a number of studies have borrowed the insights of the urban entrepreneurialism literature to theorise the rise of inter-urban competition in Spain and its impact on Spanish capitalism more broadly ; Coq-Huelva 2013; Delgado Jim enez 2011; L opez and Rodr ıguez 2010). 4 For all their differences, a common narrative can be discerned in these studies. The story begins with a crisis in the primary circuit of capital: the exhaustion of the industrial-developmental project of Franco's dictatorship in the 1970s. With the collapse and dismantlement of large swathes of Spain's industrial base, the secondary circuit kicked in. Private investment relocated to real estate in the 1980s, a turn that was actively supported by the state through housing policy, land-use planning reforms, and financial regulations. The process coincided with Spain's transition into liberal democracy, which brought a dramatic rescaling of the state apparatus: Franco's unitary state devolved vast competences over the built environment to a quasi-federal structure, one composed of 17 regional governments and a pluriverse of municipal democracies. This state restructuring, coupled with Spain's growing exposure to European capital after joining the EEC in 1986, intensified processes of inter-urban and inter-regional competition for private capital flows, setting in motion processes of urban overproduction. The logical culmination of this process was the spectacular real estate bubble that crashed the Spanish economy in 2007-2013.
The notion of urban entrepreneurialism certainly captures well the boom in place-marketing of these years. Patterns of inter-urban competition as theorised by Harvey became first discernible from the early 1990s, with the most paradigmatic cases being hosting of the Olympic Games in Barcelona, or the opening of a Guggenheim Museum in post-industrial Bilbao. Yet, it was at the turn of the millennium when urban entrepreneurialism came into full swing and its logic began to filter down to even the smallest municipalities. The legislation of a new national land-use planning framework in 1997-1998 (the so-called "build-anywhere law') jumpstarted a property boom that was then turbocharged with Eurozone finance in the early 2000s. The result was a "prodigious decade" for Spanish residential capitalism: nine million property transactions; six million new homes; and 50% new job posts, the majority in the service and construction industries (L opez and Rodr ıguez 2010:250; von Zeschau 2011). In the face of these market opportunities, a "boomtown mentality" swept the country. Local governments began to build electoral coalitions around urban development projects to attract private investment into their areas. To put their cities "on the map", municipalities became eager to upgrade their position within "the spatial division of consumption" (Harvey 1989:9): by encouraging tourism, gentrification, the construction of spectacular buildings, or hosting cultural events. Since most property buyers sought beach-side properties along the Mediterranean Coast, the phenomenon was particularly acute in the eastern and southern seaboards of the country. The mayor of Villajoiosa, a small town on the coast of Alicante, put it this way: "Every town in our area is succeeding and everyone is growing rich. We cannot afford to stand still" (quoted in Burriel de Orueta 2008).
Like their Anglo-Saxon counterparts, however, studies of Spanish urban entrepreneurialism have neglected the role of corruption in spatial processes of capital accumulation, a curious omission given how corruption and urban development have become almost synonymous in the country's public imaginary after a decade of relentless scandal. A simple word search reveals that Charnock et al. (2014) do not award the term "corruption" a single mention in their book, whereas L opez and Rodr ıguez (2010) only afford it a handful of passing mentions in their 500-page study. And in the rare occasions that corruption is mentioned, it tends to be described as a logical outgrowth of urban entrepreneurialism: The fiscal weakness of municipalities and their dependence on the evolution of real estate markets has tended to subordinate their autonomy to that of corporate agents just as a large part of corporate profits have come to gravitate around local government decisions. The over-abundance of corruption cases in recent years is only the extralegal ramification of this mutual dependence. (L opez and Rodr ıguez 2010:354, emphasis added) Corruption, then, appears as a functional phenomenon, a symbiotic relationship between local states and developers in a context of intensified competition for real estate capital. 5 Though this partly explains why so many municipalities were willing to neglect checks and balances to attract property developers, it does not account for the full scale of political corruption in Spain. By now there is considerable evidence that local clientelism was coordinated at the regional and even national levels by criminal structures tied to the financing of political parties. In 2009, it emerged that a criminal network had been brokering local clientelist transactions between corporations and the conservative Partido Popular (PP) across the Valencian region. Subsequent investigations showed that this particular criminal organisation (the "G€ urtel network") had also penetrated other regions (Madrid, Galicia), where other networks of this sort existed as well (Jim enez and Villoria 2012). After 2013, The Iron Triangle of Urban Entrepreneurialism evidence from the B arcenas papers suggested that what had previously appeared as local bribing scandals were in fact part of a broader graft system fused onto the party's national financial apparatus (Gallero 2016). Put differently, what initially seemed like local disconnected affairs were in fact tied together by a broader system of illicit party finance that went beyond the logic of inter-urban competition.
As with the main theorists of urban entrepreneurialism, the neglect of corruption in these studies is symptomatic of a difficulty to accommodate it in the model of state-capital relations built into the theory of capital switching. In these studies, the Spanish state's efforts to channel investment towards the built environment from the 1980s appear as a way of managing the needs of capital as a whole. Yet, often forgotten in this literature is that attempts to stimulate the real estate industry in Spain from the 1980s onwards emerged not only to reinvigorate capital accumulation in the abstract, but also out of concrete efforts to tackle a mounting housing deficit, curtail house-price increases, and find employment for an unruly proletariat (Pedro 2001:273-299). In other words, if state managers opened the floodgates of investment on the built environment, they did so to address political priorities in a strategic manner. In a similar vein, the Spanish ruling class' tendency to engage in corruption cannot be understood through the mechanics of capital switching, but rather as the product of a historical clientelist structure, one that compels political parties to compete with each other by raising campaign finance through bribery, as it simultaneously pressures corporations to buy political influence in order to compete over state-allocated resources.

Clientelism and Informality in Spanish State Formation
As indicated earlier, any analysis of corruption in any given political economy must stem out of a historical analysis of state formation. This section will do so while teasing out the role of the informal in the governance of the built environment, an analysis that will prove necessary to understand the dynamics of urban corruption in contemporary Spain.
Clientelism in modern Spanish politics has a long history, but the contemporary symbiosis between the state managerial class and the real estate industry traces its origins to the fascist dictatorship of Francisco Franco . At the time, clientelist exchanges were a common practice across the dictatorship's opaque bureaucracy, as state officials frequently traded bribes for vital administrative procedures (Cardes ın 2016:286). The realm of construction and real estate was one of the most permeated by clientelist exchanges. To enter the market, developers had to acquire a politically allocated license, and once in, the "friendliest" were rewarded with public construction contracts, tailored urban development plans, or even the lease of prisoners to work on construction sites. In this environment, construction firms were compelled to develop strong personal and professional ties with the dictatorship's elites in order to prevent being outflanked by betterconnected competitors. As such, clientelism became the informal means through which select firms could secure commercial success in an illiberal state.
The informal nexus between state and real estate expanded dramatically after 1959, when the regime embarked in a national-developmentalist project that inaugurated a generalised "state of exception" in urbanistic affairs. The urgency to expand the country's industrial base required the construction of as many working-class dwellings as possible, with little regard for quality or surrounding urban infrastructure (Pollard 2009). To support a rapidly expanding construction industry, the government did not hesitate to engage in frequent breaches of legality. Indeed, the central state got into the habit of overruling local planning considerations so that housing could stretch into poorly connected industrial parks, to the point that cities began to sprawl in oil-spill shapes that did not match official planning projections (Naredo 2010;Neuman 2010:110-116).
This pattern of urbanisation spurred a powerful backlash from the late 1960s, when grassroots urban activists (the movimiento vecinal or "neighbourhood movement") began to demand better living conditions for working-class urban dwellers. Despite the movement's ties with the outlawed Partido Comunista, the movement became so strong that government officials were forced to settle planning decisions with its leaders in informal negotiations. In Madrid, this effectively established an alternative clientelist system to represent the interests of the urban poor (Neuman 2010:113-116). The ultimate political demand of this movement was the devolution and democratisation of the local state, so that self-governed communities could ensure their access to public services, infrastructure, and culture (Mart ınez i Muntada 2011).
After the end of the dictatorship, these demands were engraved in the Constitution of 1978, which set in motion a broader rescaling of the state apparatus. Vast competences over land-use planning and housing were transferred to a new quasi-federal structure of 17 regional governments. However, most power over the built environment was devolved to the municipalities, to be ruled by democratically elected governments (Coq-Huelva 2013:1220-1223). In addition, from 1985 local councils were granted executive powers over their local savings banks (cajas de ahorro), credit institutions that specialised in the financing of real estate. This extension of devolution acted as a hegemonic project to appease cultural minorities and working class activists, but also as a state project to unify the state apparatus under the command of the then dominant Partido Socialista, as it replaced large swathes of the old fascist administration with elected officials (Cardes ın 2016:292-296).
As the vigilance of the neighbourhood movement fizzled out over time, the new political elite engaged in power grabs that deformed the new municipal democracies into elective dictatorships. Formally, the Constitution equipped the local state with a permanent bureaucracy to supervise the new local governments, but these bureaucratic positions were often left vacant and their roles passed onto temporary clerical workers that were employed by the local civil service (Garc ıa Quesada et al. 2013:624-627). This was a fragrant conflict of interest, as it rendered the supervisors dependent upon their supervisees for their salary and working conditions. Moreover, to further remove themselves from the purview of administrative law, some local executives began to outsource public management functions to private external bodies, all under the neoliberal pretence of lifting the dead hand of the state. The most resounding case is that of Marbella, where between 1991 and 2006 the city council created a parallel administration of private companies and foundations that was filled with handpicked friends and relatives until it became larger than the official bureaucracy itself (Garc ıa Quesada et al. 2013:628).
In this context, the corruption of the Franco years was never really uprooted, it merely adapted to the new shape of the state. The new liberal democracy devolved decision-making power to a myriad of regional and local nodes, and created a political elite under permanent electoral competition for the control of these nodes. Yet, in the absence of effective checks and balances, the new political structure spawned new forms of corruption: old patterns of clientelism in the real estate market were now joined by an additional form of clientelism in the electoral arena, according to which political parties were compelled to exact bribe monies in order to expand their campaign capacities-or else risk being overrun by their rivals. Put differently, now state officials not only sought bribes to enrich themselves, but also to reproduce their class power over the state.
These clientelist dynamics, thus, cannot be seen as a functional outgrowth of urban entrepreneurialism-on the contrary, they preceded it and facilitated its emergence. By the time the property bubble took off at the turn of the millennium, Spain had already in place a historically grown clientelist structure mediating between state and capital, one with a logic of its own. This informal structure can be imagined as an "iron triangle", in the American sense of a "complex" of colluding interests. The three vertices of this triangle are: (1) the state apparatus; (2) the construction industry; and (3) political parties, all joined together by different forms of informality, from the illicit use of public power for private gain outright, to the bending of existing regulations and exploitation of legal grey areas. The following sections will explain the workings of this informal political structure by examining each one of the "sides" of this triangle in turn.

First Side: The Institutional Mechanisms of Clientelism
The studies of the Spanish crash alluded to earlier have only acknowledged the existence of the first side of the triangle, the clientelist connection between the state (vertex 1) and the construction industry (vertex 2). However, these studies have not provided much detail about how the mechanics of these exchanges worked. It is thus worth doing a quick survey of the patterns of political corruption around Spanish urban development. Most can be categorised according to a triad of clientelist transactions: (1) urbanistic; (2) contractual; and (3) financial.
Urbanistic clientelism revolves around the manipulation of the market potential of the land. The peculiarities of Spanish urban planning grant local governments a great degree of discretion in the distribution of the economic power of land. When classifying land as fit for urban development, local governments estimate the worth of the re-zoned land assets as if these had already been fully developed and then fix their value by law (Fern andez 2005;Marcinkoski 2016:68-72). As per the "build-anywhere" law of 1997-1998, the procedure to calculate this value was the so-called "residual method", according to which the costs of production are subtracted from an estimate of the potential market value of the final realestate asset, using average house prices as a benchmark. As property prices boomed in the 2000s, the price differentials reaped from these re-zoning procedures became immense. Fern andez (2005) has estimated that re-zoning operations would prompt land asset prices to escalate, on average, between 2000% and 6500%. This political mechanism of asset-price inflation sustained land (and thus, house) prices high during the years of the bubble, acting as a remarkable instrument for urban entrepreneurialism. Unsurprisingly, this procedure also became a formidable breeding ground of clientelism. Numerous corruption cases involve politicians using these powers informally, to reward friends and family (or even themselves) with a re-zoning of their land assets, or to give developers a tipoff of an impending reclassification so that they could purchase rural lands ahead of time.
Clientelism in land-use planning also revolves around the capture of regulatory procedures to facilitate the approval of construction projects. During the bubble, this often came attached to the opaque practice of "urbanistic agreements" (convenios urban ısticos)-perhaps the most evident instance of informality being built into Spanish urban governance. Largely unregulated until 2007, these procedures allowed local governments to tailor urban plan amendments to the interests of specific developers. In exchange, local governments would demand that developers included some sort of facility to the benefit of their electorates: e.g. a local care home or a municipal sports centre (Garc ıa Quesada et al. 2013:623). Negotiations of these agreements would unfold in a bilateral fashion and without any minimum requirements for transparency. The content was usually withheld from the public and sometimes from the local state itself.
In addition, urbanistic clientelism also involves the deliberate neglect of environmental controls, generally by overlooking the ecological impact of urban development projects. Rather than exploiting a legal grey area, the kind of informality involved in these transactions typically involved bending the letter of the law. For example, whereas regulations forced local governments to commission reports on the environmental impact of urban developments, these same regulations did not specify that they had to take heed of those reports (Ord oñez 2016). The proliferation of golf courses in provinces at risk of desertification is a good indication of where environmental assessments have been wilfully ignored.
Contractual clientelism involves distortions in the legal procedures to allocate public contracts. The marks of contractual clientelism are most evident in the acceptance of additional expenditures in public contracts (sobrecostes). It is not uncommon for adjudicating authorities to assume the unforeseen costs of their contractors, a practice that can inflate the original value of a project several times over. The most resounding examples are found in large-scale infrastructure contracts, almost invariably to the benefit of Spain's largest construction conglomerates. A notorious case is Valencia's City of Arts and Sciences, a complex of futuristic buildings built that became the flagship of the city's entrepreneurial project. Though the original budget was €382m, it ended up costing the Valencian regional government €1.28bn (Serra 2015).
Another mechanism of contractual clientelism is the rigging of design contests. This refers to the ways in which supposedly competitive contests for the adjudication of public contracts are manipulated to favour specific companies. This distortion typically involves the informal manipulation of technical reports to arbitrarily inflate the score of a design project. An ongoing case investigates the role of the Aragonese regional government in the rigging of a design contest to adjudicate the development of an industrial park to a large conglomerate called Acciona. Email evidence suggests that the contest's technical director had orders from his superiors to use discretionary criteria available to him to "readjust" the score of Acciona's offer (L opez-Fonseca and Montaño 2014). The Aragonese government would later pay the company €150m in unforeseen costs, a sum that allegedly financed the kickbacks for the officials involved in the rigging of the contest.
Finally, financial clientelism relies on corporate malpractice in the banking sector, generally through the politically controlled savings banks (cajas de ahorro). In the 1980s local councils were allowed to handpick a large section of the executive boards of their local cajas, and in the 1990s some regional governments began to stretch their power into these bodies as well. According to the estimates of an IMF (2012:8) report, by 2009, out of 43 savings banks at least 16 granted political representatives a 40-50% share of voting powers in their corporate governments. It was not uncommon for property developers benefitting from certain urban planning decisions to then finance their building operations with loans from a local caja controlled by their same political patrons or their party friends (Coq-Huelva 2013:1223-1226. Perhaps unsurprisingly, these credit institutions also acted as reckless financiers of local and regional strategies of urban entrepreneurialism during the bubble, resulting in "a delirious cycle of development proposals, municipal approvals, and easy credit lubricating this urban transformation" (Marcinkoski 2016:72). The Bank of Spain would describe Valencia's Caja de Ahorros del Mediterr aneo (CAM) as "the worst of the worst" (Biot 2012). This caja would accompany its regional government in all sorts of extravagant projects (including the aforementioned City of Arts and Sciences). This was rooted in informal complicity: CAM was in the habit of loaning credit at very low interest (sometimes 0%) to its very own board members, a malpractice that the regional government, the official regulator, neglected to flag.

Second Side: The Party-Developer Nexus
Having examined the institutional mechanisms of clientelism, the next step is to explore the connection between the construction industry (vertex 2) and political parties (vertex 3), an informal nexus that organised clientelist exchanges above and beyond the local logic of urban entrepreneurialism. Over the last decade, the majority of political corruption cases have involved illicit payments not just to the public officials providing the political favours but also to their political parties. The informal nature of these financial transactions did not preclude their systematisation. Political parties farmed out their bribe-collecting activities to specialised criminal networks, ran by a mixture of high-ranking party members and businesspeople. The main purpose of these informal networks was to broker between parties and developers, collecting tolls on the channels of influence that they facilitated, usually 3% of the profits of the rewarded company-i.e. a "kickback". These bribe monies were then split between the criminal networks brokering the transaction, the political officials making the necessary decisions, and the financial hierarchies of their political parties.
The foremost example of a political organisation corroded by these practices is the conservative Partido Popular (PP). Some of the earliest evidence surfaced in 1992 (the Naseiro affair, named after the party's chief treasurer at the time). The scandal was unearthed by intercepted phone conversations between politicians, who were openly discussing the bribes that the party was receiving in exchange for public decisions (Gallero 2016:137-148). The conversations also suggested the existence of internal power struggles to control these illicit revenue streams, as well as the awareness of the party's national superstructure of the existence of these practices. Everything points to the continuity, extension, and systematisation of these practices until their resurfacing in the scandals of the 2010s. Police reports openly speak of an uninterrupted continuum between the practices of the Naseiro scandal and the more recent B arcenas affair, which, as addressed in the opening, revealed the existence of an entire informal bookkeeping system to manage the party's bribe streams on a national scale (Gallero 2016:145).
Over the course of the last decade, successive police operations have uncovered the existence of several criminal networks acting as intermediaries between PP and its illicit donors. Perhaps the most infamous is the so-called "G€ urtel network", a criminal organisation formed by businessmen and high-ranking party members that operated in the Valencia, Madrid, and Galicia regions. The network cultivated relationships with party members in government positions and used their own companies as a front to disguise the party's election campaign activities with false invoices (El Mundo 2010). In exchange, the network demanded political favours, either for their own companies or for the companies that had contracted their brokering services, from which they would exact part of their profits.
In the midst of the "Taula affair", a more recent scandal relating to the finances of the Valencian branch of the party, the leader of one such intermediary network admitted in court to having brokered hundreds of clientelist transactions, claims later corroborated by 12 different "donors" (EFE 2016). During his testimony, he claimed to have exacted 2-3% on the profits of the companies rewarded in every transaction, part of which went straight into the party's coffers. Interestingly, he detailed how bribe revenues were then distributed by the party to five different slush funds in the Valencian region alone: one slush fund stored the party's local branch in the city of Valencia, the region's capital; three would have corresponded to the party's provincial branches in the regions; and one final fund was managed by the party's regional structure, rising above the others and feeding off from them (Zafra 2016). Witnesses to the Taula affair also claimed in court that the Valencian regional slush fund in turn fed into the one managed by B arcenas at the national level, suggesting the existence of a bribe-collection hierarchy fused onto the financial apparatus of the party, from the bottom to the top. This would corroborate the claim made by B arcenas himself that "PP ran a slush fund in every region and every province for electoral expenses" (quoted in Martialay 2014). 6 Though evidence of similar practices in other parties is not lacking, it remains unclear to what extent other parties developed informal systems of criminal finance as large and as comprehensive as that of PP. The closest examples are perhaps the scandals surrounding the Catalan nationalist party Converg encia i Uni o, an organisation accused of the systematic exaction of 3% in kickbacks from public contractors in Catalonia. Though the bribe-farming systems of Converg encia bear remarkable similarities to that of PP, one striking difference is that in this case part of this graft would have also gone to relatives of the party leader, like a mafia family (eldiario.es 2017). In the case of Spain's main social-democratic party, Partido Socialista (PSOE), despite having been involved in all too familiar clientelist transactions at the local and regional level during the years of the bubble, evidence of a system of criminal finance as widespread as that of PP has yet to emerge. To be sure, there is plenty of evidence that the party was engaged in practices identical to those of PP during the 1990s. For example, the following headline story opened the newspaper El Pa ıs in 1991: On Wednesday, the Minister for Public Works, Josep Borrell, convened the CEOs of the largest construction companies in his office to urge them, "on behalf of the Prime Minister and myself", to stop making illicit payments to political parties in exchange for construction contracts ... Borrell was adamant and insisted that if any of the companies gathered there received illicit proposals "in the name of PSOE" [his party] they should communicate this to him directly ... Sources from the construction sector admit that the payment of bribes, between 2% and 4% of the value of the project, is a common practice in the allocation of public contracts. This corruption extended to every level of the state administration: national, regional, and municipal ... In the sector, frequent comments are made about the differential costs of graft across local and regional governments. (Miss e 1991) Reminded about the headline in a recent televised interview, Borrell himself made the following observation: "... it seems evident that the more a party holds power at the local level, the more these phenomena proliferate, sometimes not as isolated cases, but as organised party structures involving many people at once" (see Evole 2016).

Third Side: The Reproduction of the Kleptocracy
As addressed earlier, part of these bribe monies went to their brokers as a fee, part went to the politicians pushing levers of the state to favour the briber, and part went into the coffers of the politician's political party (hence the convenience of the 3% figure). In the case of PP, these bribe moneys went to a local slush fund and were then distributed up the party's territorial scales, nourishing a hierarchy of provincial, regional, and national slush funds. The B arcenas papers, the parallel bookkeeping system designed to manage the national slush fund, indicates how these illicit cash pools were put to use: informal bonuses to the leadership-ensuring the collusion of the upper echelons at large-and systematic campaign spending fraud (Mercado et al. 2013). This last outflow seems to have been the main purpose of the entire structure, a form of electoral "doping" designed to reproduce the party's access to the nodes of state power by informal means. This form of political accumulation constitutes the third side of the iron triangle, the one that connects political parties (vertex 2) to the state (vertex 3).
Campaign spending fraud is made fairly easy by a number of loopholes in the Spanish party finance system. In theory, Spanish political parties ought to finance themselves through a system of public subsidies proportional to the degree of representation that they achieve (Jim enez and Villoria 2012:127-129). Established in 1977 with the aim of supporting the country's nascent liberal democracy, this system left a number of legal ambiguities when it came to private donations that legislators have been reluctant to address ever since. The first regulations were introduced in the mid-1980s, when private contributions were capped at a certain amount per donor each year, and state contractors were banned from making contributions due to conflict of interest (Jim enez and Villoria 2012:127). However, the practice of anonymous donations persisted, allowing parties to simply split large donations into several unidentified contributions.
Moreover, although the official accounts that parties report reflect that around 80-95% of their finances come from public subsidies, there are plenty of "back doors" to conceal much larger sources of informal funds. The finances of political parties fall under the purview of the Court of Audit (Tribunal de Cuentas), an institution equipped with scarcely dissuasive sanctions and self-defeating legal technicalities. Perhaps the most striking is that a political party is not legally regarded as a single financial unit comprising all its different territorial organisations. This means that the accounts that parties submit to the Court do not automatically include party foundations, associated companies, or local party branches. This legal multiplicity has allowed them to easily evade the purview of the state by simply moving funds around (Jim enez and Villoria 2012).
The easiest way to mask funds in this manner is through party foundations or shell companies. Conveniently, different rules apply to party foundations with regard to private donations. Not only is there a higher cap on contributions, but foundations are also exempt from the rule that bans public contractors from donating to political parties as well as from the "cooling-off" period, designed to avoid the arrival of delayed payments derived from clientelist transactions (Jim enez and Villoria 2012:130-131). In the 1980s, it was common for donors to commission generic reports to a shell company that would then funnel the payment into the party's coffers. This fraudulent practice was at the epicentre of the two main political finance scandals of the 1990s: the Naseiro affair (1992), and the Filesa affair (1991)(1992)(1993)(1994)(1995)(1996)(1997), which struck, respectively, the conservative PP and the social-democratic PSOE (Heywood 2007). A more recent example is the "Palau affair" afflicting the Catalan nationalist party CiU, in which companies awarded with public contracts by the Catalan government made large contributions to the party's foundation that were undisclosed to the Court of Audit (eldiario.es 2017). However, as explained earlier, shell companies, rather than party foundations, were the preferred mode of disguise in PP's more recent scandals.
The exact extent to which these practices were used during the years of the bubble is a question that is bound to remain unanswered. But one indicator should capture our imagination: the vast amount of €500 notes in circulation in these years. This denomination, with little practical everyday use for most people, was standard in money laundering and bribing operations. Spain gathered 26% of the Eurozone's total at the peak of the bubble in 2006 (L azaro 2016). Even though this figure reflects broader illicit transactions as well, it nevertheless points to the intrinsic connection between "dirty money" and urban development.
During the 2000s, many local candidates ran electoral campaigns boosted by the bribes of developers, something that reinforced conflicts of interest in urban planning and created patterns of competition between parties over criminal sources of finance. This created a perverse structure of incentives, according to which entire party structures derived illicit electoral finance from inflating the profits of the construction industry by informal means. In turn, parties derived a competitive edge from this criminal political economy, as it enabled higher campaign spending. In the process, the local urban planning office (concejal ıa de urbanismo) became a highly strategic institution for these processes of political accumulation. This issue was openly addressed in a recent televised interview involving several formerly high-ranking political figures (see Evole 2016). Discussing the problem of widespread political corruption in urban development, Josep Piqu e, a man who occupied several cabinet positions at the turn of the millennium before joining the ranks of the construction industry, made the following remark: "In local councils, where there is a strong tradition of forming coalition governments, the junior partner tends to always demand..." "The urban planning office"-the sentence is finished by Esperanza Aguirre, head of the Madrid region between 2003 and 2012. Piqu e concludes sarcastically: "What a coincidence."

Conclusion
The literature of "urban entrepreneurialism" has successfully captured the dynamics of inter-urban competition that have arisen in the wake of neoliberal globalisation; however, it has proved unable to grasp the role of corruption in the governance of the built environment. This is due to this concept's emphasis on the role of the state in managing the reproduction of the capitalist economy as a whole, which inflects a functionalist bent onto the different theories of the state that underpin it. These end up exaggerating the coherence of the state apparatus, taking at face value the capitalist state's hegemonic self-representation as an impartial arbiter of competing interests. Not only is the capitalist class ridden with factionalism, as the strategic-relational versions of urban entrepreneurialism already demonstrate, but so is the state managerial class as well. This means that state managers have class interests of their own that should be factored into the analysis, since their abuse of entrusted power is often not just about personal gain, but about ensuring their class reproduction as well.
The difficulty to integrate this complexity into the study of urban entrepreneurialism is patent in the literature on Spain, where a number of sophisticated studies have explored this phenomenon in the country but have been unable to factor corruption into their accounts despite overwhelming evidence. Tellingly, on the few occasions that corruption is mentioned, it is simply assumed to be a functional expression of inter-urban competition-i.e. it helped local states attract investment.
As the article has argued, the study of urban governance can address these issues by drawing methodological insights from studies of the "global South", which have developed a repertoire of theoretical strategies to account for corruption in critical ways. From this toolkit, the concept of "informality" is particularly useful, as it casts light on the manifold ways in which state power is exercised by bending, twisting and circumventing formal legal rules. Factoring informality into a strategic-relational approach to the state-one that not only addresses intercapitalist competition for state influence, but also the dynamics of class reproduction of state managers-is a good start to tease out how corruption often embeds itself into the governance of the built environment. However, this must always be grounded in a careful historicisation of state formation in the political economy in question, since apparently similar corruption practices can mean a variety of things and fulfil divergent social roles across different contexts.
These theoretical points have been showcased with an analysis of the role of clientelism in Spanish urban entrepreneurialism. Contrary to much of the literature, this article has shown the existence of a clientelist structure that, informally but systematically, organised the colluding interests of state, political parties, and the real estate industry-what has been referred to as an "iron triangle of real estate". The analysis has traced the origins and evolution of this informal structure to the shifting forms of the Spanish state, which have in turn shaped different patterns of class reproduction for state managers. Through it, corporations competed for political influence to secure their market success, while political parties competed with each other by raising illicit funds to boost their campaign resources, vital to access or revalidate state power. Since political patrons exacted a portion of the earnings secured for their clients (a 3% kickback or "commission fee"), they were incentivised to inflate the profits of the real estate industry, as this increased the bribes that then swelled their electoral war chests.
The analysis has shown that these clientelist dynamics cannot be reduced to a functional expression of inter-urban competition. Not only did the "iron triangle" precede the take-off of urban entrepreneurialism, but it also displayed a logic of its own, one that transcended the local sphere, as clientelist networks were organised at the regional and even national scales. Nevertheless, the workings of this informal political structure magnified the reckless local boosterism of these years, inflating the property bubble to new heights. Corruption, thus, was integral to the real estate boom that imploded the Spanish economy in the wake of the great recession.