Neoliberalism and the geographies of marketization

1 Why marketization?

Neoliberalism is conceived as the insertion or installation of markets as the underlying institution or mechanism for organizing society. For example, David Harvey (2005: 3) characterizes neoliberalism as the ascendance of a ‘market ethic’ in which ‘contractual relations in the marketplace’ are deemed the most efficient and most ethical means to organize society. This installation of markets as the organizing force in society entails a necessary role for the state. As Tickell and Peck (2003: 167–8) point out: ‘More often than not, the practice of neoliberalism has little to do with laissez-faire deregulation…but instead is associated with the extensive deconstruction and reconstruction of institutions, often in the name of or in the image of “markets”’. This means that neoliberalization is better understood as the ‘mobilization of state power in the contradictory extension and reproduction of market(-like) rule’ (2003: 166, emphasis in original). Consequently, neoliberalization has led to ‘a tangled web of state-regulated oligopolies, profit-orientated enclaves and pseudo markets’ (2003: 167). While several geographers have taken up the task of untangling this mess, as we will outline below, it is surprising that so few have sought to grapple conceptually with the diversity and variety of the processes, outcomes, forces and agents at work in this ‘mobilization of state power’ to extend markets, or marketization. In focusing on marketization – or, more accurately, marketizations – it is our intent to explore analytically the how and why of this ‘state mobilization’.

Before we discuss existing geographical literatures on neoliberalism and marketization, however, it is pertinent to define ‘marketization’. In broad terms, and drawing on the work of Peck and Tickell (2002) and Tickell and Peck (2003), it involves the introduction of markets and market forces into the state, primarily into its functioning, authority and legitimation – issues we will come back to below. Likewise, Hendrikse and Sidaway (2010: 2039, emphasis in original) argue that marketization is about the reconfiguring of the relationship between state and market ‘so that they become more thoroughly intermeshed’ (see also Aalbers, 2013). According to Peck and Tickell (2002: 388), ‘state power was mobilized behind marketization and deregulation projects, aimed particularly at the central institutions of the Keynesian-welfarist settlement’. In follow-up, Tickell and Peck (2003: 169, 180) describe marketization as both ‘primitive’ and ‘amoral’, reflecting the deregulatory focus of the roll-back phase of neoliberalization (i.e. 1980s–1990s). When it comes to the later, roll-out phase (i.e. post mid-1990s) they align marketization more closely with new forms of public service delivery such as public-private partnerships (discussed below).

What are missing in these accounts, however, are the specifics and details of these marketization processes (i.e. the what, why, how, who and where). As Wendy Larner (2003: 509) puts it, ‘The differences between processes such as deregulation, privatisation, and marketization were rarely discussed [in the literature], and their effects on people and places all too predictable’. A useful illustration of this is provided by Weller and O’Neill (2014: 117), who argue that the forces of marketization actually ‘bind multiple levels of government into the maintenance of public service provision’, meaning that the insertion of markets is complex – to say the least – and that it does not entail a withdrawal of the state, or even a lessening of its influence, since the state ‘remains in control and accountable’ (cf. Springer, 2014).

2 Marketization in human geography

A number of scholars within and outwith human geography have provided more detailed accounts of marketization, which we turn to next. Here we highlight two main strands of thought in human geography: (a) the concern with marketization in the ‘neoliberal natures’ literature; and (b) the concern with marketization in relation to the literature on public-private partnerships.

Some of the more detailed work on marketization comes from geographers working on the neoliberalization of nature; for want of space, all we can we do here is briefly highlight some key reviews of this literature and debates in this area. First, Castree (2008: 142) defines marketization as ‘the assignment of prices to phenomena that were previously shielded from market exchange or for various reasons unpriced’. Hence, marketization necessarily entails previously non-market phenomena, whether protected or simply unpriced, which are then transformed by the insertion or creation of markets. Second, in her critique of Castree, Bakker (2009: 1781, 1784) emphasizes the increasing discussion of ‘marketized forms of nature’ without defining what this marketization entails, apart from representing a particular ‘tactic’ of neoliberalization. Bakker (2010) builds on this work, however, in her later discussion of ‘neoliberal natures’ in this very journal. She reviews how different scholars have dealt with marketization, which she defines as where ‘markets determine resource allocation and pricing’ (2010: 723); she highlights work on water markets, fishing quotas, human organ markets, emissions trading and ecosystem services. While Bakker highlights the diverse ‘targets’ of marketization, she does not analyze the diversity in marketization itself. Third, Castree (2010: 1728) returns to marketization in a later review essay where he defines it as ‘rendering alienable and exchangeable things that might not previously have been subject to a market calculus’ – which would include public goods. He distinguishes it from the use of market proxies – as does Bakker (2010) – which he defines as ‘making remaining state services more market-like in their operation’. Here Castree reiterates the notion that marketization involves a transformation from non-market to market calculus (e.g. Castree, 2008). With both Castree and Bakker the distinction between ‘marketization’ and ‘market proxies’ seems unnecessary since both involve the introduction of a market calculation – whether simulated or not – into human activity. We, however, do think that there is some need to differentiate between types of market, as discussed below.

Another area of human geography where neoliberalism and marketization are closely associated is in the work of scholars on public-private partnerships (PPPs). While not all these scholars necessarily link PPPs to neoliberalism or neoliberalization, their research outlines several important issues. The likes of Clark and Root (1999), Leyshon and Thrift (2007), Hillier and van Wezemael (2008), Musson (2008) and Cumbers (2012), for example, focus on the UK’s Private Finance Initiative (PFI) as a geographically-specific mechanism for transforming the relationship between state and markets. PFI is a capital financing scheme first introduced in the UK by John Major’s Conservative government (1992–97), but which was really taken up by the later New Labour government (1997–2010) as a key way to resolve the public investment deficit built up since Thatcher-era capital spending cuts (Kitson and Michie, 2000). There are several ways to frame the changes induced by PFI. First, Clark and Root (1999) emphasize the link between PFI and forms of (much needed) urban and regional capital investment. Second, Leyshon and Thirft (2007) argue that ‘PFIs are highly attractive to large institutional investors’ since they enable them to manage their assets and liabilities over long time periods. Third, Hillier and van Wezemael (2008: 176 ) argue that PFI is a Deleuzian assemblage of practices and actors in which finance ‘becomes the goal of action and thus surpasses its role as a means’. Fourth, Musson (2008) highlights the transformation in thinking within the state itself as government departments adopted ideas from ‘new public management’ (NPM), especially the idea that the state should be a facilitator rather than provider of services. Finally, Cumbers (2012) identifies PFI and contracting out as specific forms of privatization underpinned by neoliberal argument about minimal state involvement in the economy. While these geographers offer valuable insights, what is missing, analytically at least, are comparative perspectives of marketization processes, since PFI is a highly-specific form tied to the British political economy.

Other geographers and planners have focused on PPPs more generally, looking further afield than the UK. An important overview of marketization’s impact on cities, for example, is provided by Graham and Marvin (2001) in their book Splintering Urbanism. Others provide detailed accounts of PPPs in international development (Miraftab, 2004), public service delivery (e.g. Hearne, 2006, 2011), and infrastructure planning (e.g. Siemiatycki, 2010, 2011, 2012; Haughton and McManus, 2012). First, Graham and Marvin (2001: 150) analyze the transformation of ‘networked’ infrastructure through a process of unbundling that, quite deliberately, opens it up to ‘marketability’ (their term). They highlight six models of infrastructure unbundling in their book (pp. 153–61), which cuts across different countries and scales but is driven by private sector preferences for standalone and dis-integrated projects (i.e. excludable and rivalrous – and therefore lacking the quality of public goods). At the same time, the state retains a key role as creator of markets through public contract tendering, concessions and price regulation. Second, a number of scholars have studied the expansion of PPPs around the world. Miraftab (2004), for example, explicitly argues that PPPs are a ‘Trojan horse’ in development, enabling the insertion of markets in the provision of public goods like infrastructure. She argues that much of the wider social science literature on PPPs is mechanistic in that it deals with forms and terms of contracts – plus their benefits and problems – rather than questions of equity and justice (e.g. Kwak et al., 2009). The work of scholars like Hearne (2006, 2011), Siemiatycki (2011, 2012) and Haughton and McManus (2012) provides a corrective here, in that they look specifically at PPPs as drivers of fragmenting governance and uneven development. For example, Hearne (2011) discusses the conflict between the financial and social objectives of PPPs in Ireland, resulting from new contractual arrangements that lock-in public service provision to neoliberal forms of governance. Siemiatycki (2011) illustrates that infrastructure unbundling (see Graham and Marvin, 2001) has uneven spatial rationales and effects; that is, some unbundled infrastructure is more attractive to private capital (e.g. road over rail infrastructure), while other forms are problematized by the subsequent geographical fragmentation of governance and institutional jurisdiction (e.g. regional public transit networks). Finally, Haughton and McManus (2012) argue that PPPs are a clear example of ‘hybrid neoliberalism’ in that a belief in (unregulated) markets plays out in varied and diverse ways in different places and at different scales (also Peck and Tickell, 2002; Castree, 2006).

This brief discussion of PPPs in the human geography literature should illustrate that there are a number of issues we need to unpack in order to theorize the diversity of marketization processes. To start with, the current literature is dominated by an analytical and empirical focus on infrastructure – as such, there has been a more limited engagement with the marketization of public service delivery, another side of the entanglement of the state and markets (see Weller and O’Neill, 2014). This does not mean that there is no work in this area, but what has been written raises further questions. First, a number of geographers, planners and others have studied contracting out (or outsourcing). For example, Warner and Bel (2008: 723) have argued that contracting out is a ‘form of privatization because private firms get residual gains’; this perspective is largely based on a theoretical concern with examining the benefits of public monopoly versus market competition espoused by public choice theory. In earlier and related work, Warner and Hefetz (2002) contrast the benefits of markets with those of ‘regionalism’, by which they mean the economies of scale derived from the regional and urban organization of public services. In more recent work, Warner and Clifton (2014) highlight the way that privatization is frequently a national-scale process (especially in the Europe Union), and suggest that marketization processes might be constituted at more local scales. However, they conclude by arguing that ‘marketisation of policy is complex and multi-layered’ (p. 57), reiterating the arguments of Weller and O’Neill (2014).

Indeed as far back as the 1990s, Doogan (1997) suggested that marketization processes play out at the local scale. He argues that the marketization of public services is inherently tied to a (scalar) separation of (local) service needs from the system of their (local and national) provision. In particular, this results from the privatization and decentralization of previously large, centralized bureaucracies (e.g. nationalized industries, national utilities, etc.). Foreshadowing Graham and Marvin (2001), Doogan (1997) highlights the fragmentation of public service delivery, especially in labour markets, as services are outsourced. Furthermore, Doogan (1997) illustrates the extent to which such marketization is underpinned by contract limitations and constraints, rather than market price competition. While these studies provide useful, spatially-sensitive analyses of the restructuring of policy and service delivery, especially outsourcing, they are mainly based on a dichotomy between market and state, rather than an analysis of the entangling of state and markets that we see as characteristic of marketization processes.

3 Marketization in other social science literature

We now turn to the broader literature on marketization in the social sciences. We think that human geography can benefit from an examination of other literatures on marketization, especially when it comes to three critical issues: (a) discourses of efficiency; (b) administrative restructuring; and (c) forms of accountability.

First, there is an important set of literatures dealing with how marketization processes are bound up with diverse discourses of distributive and allocative efficiency and flexibility of markets (cf. the inefficient and bureaucratic state). A number of scholars argue that the shift of policy delivery – whether infrastructure or public services – to private sector providers is premised on an assumption that the private sector is more efficient (e.g. Kay and Thompson, 1986; Vickers and Yarrow, 1991; Farnsworth and Holden, 2006). On the one hand, these discourses can play out internally, promoting the insertion of markets into organizational units and policy functions. On the other hand, they can play out externally, as general discourses about the general efficiencies engendered by the private sector management of policy delivery. In the first instance, Musson (2008) and Flinders (2005) highlight how government organizational restructuring has been driven by ‘new public management’ (NPM), an ‘internal’ discourse centered on denigrating government delivery of services in comparison with visions of efficient private sector delivery. This discourse (a) legitimates a rethinking of the state as market facilitator rather than service provider, thereby promoting the delegation of delivery functions to the private sector; and (b) legitimates the introduction of ‘internal’ markets into organizational units and structures (discussed below). The second instance raises broader political questions around the relationship between market and state actors. For example, Sally Ruane (2010: 530) notes a ‘dual objective of technical and ideological education’ within British policy discourses around public-private partnerships (PPPs) like PFI. In particular, the word ‘partnership’ takes on the role as signifier of political legitimation, emphasizing an important difference between marketization (‘good’) and privatization (‘bad’). Two other scholars, Graeme Hodge and Carsten Greve, make a similar point in their analysis of PPPs as a ‘language game’ (Hodge and Greve, 2007, 2010). In their reading of PPPs they note the ‘governance risks’ associated with private sector involvement in policy delivery, especially in light of different national (and negative) experiences with privatizations in the 1980s – a key characteristic of roll-back neoliberalism (Peck and Tickell, 2002). Hodge and Greve (2010) argue that PPPs, especially the notion of partnership, helps policy-makers and politicians to avoid terms like privatization and outsourcing. In this sense, marketization processes are deeply political, not only in their discursive framing but also in the way that they represent the (intentional or unintentional) continuing attempts to ‘lock-in’ policy to market-based instruments (Hodge and Greve, 2007).

Second, marketization processes are not only about the creation of new markets in policy delivery; as noted above, they also entail the insertion of market principles in non-market areas of life, especially the public realm. This means more than the dominance of discourses of efficiency, however. It involves the organizational restructuring of public administrative structures, practices and subjectivities, as outlined by Crouch (2011) and Peters (2012). The blurring of the state and markets is not limited to public sector restructuring, however; it also entails the spreading of market principles through particular imaginaries of citizenship. In his discussion of marketization, Whitfield (2006a, 2006b), for example, outlines the transformation of managerial practices as they become a technocratic assessment of universal, as opposed to contextual or particular, criteria which are then embedded in specific targets (e.g. school or hospital performance). These targets represent attempts by the state to create competition and new incentive structures through ‘internal markets’ that will reward performance rather than job status, encouraging all service providers from managers to frontline staff to ‘up-their-game’ (Bel et al., 2010, 2013; Crouch, 2011). As Whitfield (2010a) points out, such pseudo-markets, or market proxies, are also meant to ensure that service providers have ‘correct’ information about the service demands of users – now conceived of as ‘consumers’ – so that they can respond effectively and flexibly to users and their changing demands. The insertion of these market principles assumes or requires, however, the corresponding transformation of the identities and subjectivities of both service providers and service users. This leads John Clarke (2004) to argue that there is a growing emphasis on individual self-sufficiency within (marketized) policy delivery, especially when it comes to imagining the users of the welfare services. However, according to Clarke (2004), this construction of citizens as welfare ‘consumers’ is more of an (neoliberal) imaginary of the reformers themselves than a reflection of the demands of citizens. Moreover, whether this transformation involves markets or market proxies, the effect is the same. As Aldred (2008: 45) highlights, for example, it leads to the: ‘contractualisation of relationships between different organisational units within a range of public services … [which] helps to turn the services exchanged into commodities even if they are not sold on the open market’.

In the end, this restructuring of administrative structures, practices and subjectivities is bound up with the entangling of the state and markets, and has been associated by Clifton et al. (2006) with the (contradictory) emergence of a ‘regulatory’ state. Greater regulatory oversight is needed as services are outsourced to the private sector since the state now needs to correct market failures and ensure competition (also see Crouch, 2011).

Third, the final set of concerns relate to forms of accountability, whether financial or political. According to a number of scholars, financial pressures underpin a range of marketization processes. Taylor-Gooby et al. (2004) argue that the welfare state has been framed as a ‘burden’ on the economy. Within debates on neoliberalism this demonstrates how new forms of accountability center on competiveness as the financial backbone of a country’s public services (Larner, 2000). In a similar vein, Flyvbjerg et al. (2003) argue that public infrastructure spending came to represent a significant (political and financial) risk for governments, especially relating to cost overruns and revenue shortfalls from public investment. Generally, this transfer of risk from state to markets is a key issue when it comes to marketization processes – it is meant to provide financial accountability (e.g. transparency) and political accountability (e.g. an end to ‘pork-barrel’ politics). Asenova and Beck (2010) note that PPPs, for example, are premised on the transfer of financial risk to the private sector contractors, who are responsible for any increased cost. However, whether or not any financial risk is actually transferred is another question, as Loxley (2012) points out. Moreover, Loxley also highlights the contentious relationship between financial and political accountability, in that the latter has become dependent on the former through various forms of ‘balanced budget’ legislation and tax cuts. In this sense, if governments want to be politically accountable (i.e. fulfil their election promises), they are tied to forms of financial accountability that may curtail their room for manoeuvre (also see Whiteside, 2012). Financial and political accountability, therefore, can butt up against one another in a contradictory manner.

1 Diversity of markets

Marketization processes are characterized by new forms of entanglement between state and markets. We particularly emphasize the plurality of markets since these can take many forms. In his work on ‘Polanyian economic geographies’, for example, Jamie Peck (2013: 1555) argues that Karl Polanyi’s notion of the economy as ‘instituted process’ was a ‘multilogical one’ providing the means for ‘understanding variegated economic formation’. In this sense, the market economy is multiple, involving a variety of instituted market processes; that is, there is not just one form of market economy.

From the literature it is possible to identify at least three types of market – or market proxy – that are entangled with the state. As Tickell and Peck (2003: 167) suggest, the introduction of markets is not simple or clean, but involves ‘a tangled web of state-regulated oligopolies, profit-orientated enclaves and pseudo markets’. This highlights the fact that markets supplement – rather than supplant – the state in the delivery of public goods (e.g. public services and infrastructure). As both Castree (2008) and Bakker (2010) highlight, marketization entails the market pricing of things like public goods (e.g. hospital cleaning, roads, electricity, etc.) that were not previously priced or were ‘shielded from market exchange’ (Castree, 2008: 142). This does not, evidently, mean that markets replace the state; markets can be instituted by the state and are dependent upon the state for their operation, as we outline below. Moreover, it can also involve the adoption of market(-like) proxies and principles in the delivery and management of public goods, as outlined by scholars who focus on market discourses like new public management (NPM). These discourses legitimate the insertion and installation of markets and market proxies within the state. In this sense, markets are not simply an imposition on the state, they are also very much integrated within the state; thus markets are not instituted as an alternative or replacement for the state.

We can see some of the diverse geographies of marketization by looking at the forms of markets that have arisen in the delivery of public services and infrastructure. First, new forms of external service contracting (i.e. outsourcing or contracting out) created by the state represent competitive markets for the delivery of (quasi-) public services like waste collection, street lighting, cleaning and road maintenance (Farnsworth and Holden, 2006). What distinguishes this type of market from others we discuss below is the possibility for price competition between private sector providers (Bel et al., 2013); in turn, this means that it is a market supposedly driven by cost, innovation and efficiency. One example of this market is the outsourcing of welfare provision in places like the UK. Programs like the Flexible New Deal (FND), for example, have been contracted out to private sector (and some voluntary sector) providers (Atkins, 2010). FND commodifies ‘job training’ and ‘job placements’ in an attempt to reduce unemployment and increase employability. This commodification enables competition, which plays out geographically in different ways. On the one hand, markets are promoted as both a means of local control and of local responsibility; as such, outsourcing has been described as a form of ‘selective decentralization’ by Clarke (2004: 36). On the other hand, Clarke also argues that outsourcing involves the establishment of new ‘agencies and agents as the proxies of state power’ rather than replacement of the central state by localized, competitive markets.

Second, new forms of contractual (market) arrangement have been created by the state to enable the private sector to build, operate and manage public services and infrastructure. While couched in the language of ‘public-private partnership’ (see Hodge and Greve, 2010), these contractual arrangements represent another form of market, one that is distinct from outsourcing (above). Unlike outsourcing, these markets include both medium-term (i.e. 20–30 years) and long-term (i.e. 100 years) concessions, franchises and leases, which cover the delivery and running of services and infrastructure (Whitfield, 2006a; Bel et al., 2013). These contractual arrangements are distinct as markets in that they do not involve price competition as such because they involve long-term contracts with strict bidding specifications (Hearne, 2011). More specifically, the bidding process limits price competition because only large consortia of businesses have the necessary capacity to both bid for contracts and then run any franchise or concession afterwards (Siemiatycki, 2012).

Contractual markets are distinct, geographically-speaking, because they are dependent upon the spatial and scalar ‘unbundling’ – or fragmentation – of existing public services and infrastructure as theorized by Graham and Marvin (2001). Consequently, new forms of (state-market) governance are tied into this contractual delivery and running of services and infrastructure, something both Hearne (2011) and Siemiatycki (2011) highlight. The state has to coordinate and facilitate the activities of diverse social actors across a significant time period and across multiple scales (see Leyshon and Thrift, 2007). It goes without saying that this governance differs significantly between countries. For example, in much of Europe the national state has led to the delivery of large-scale road and rail transport PPPs (Albalate, 2014). Conversely, in North America and Australia, the sub-national state (e.g. provincial or state) has been responsible for providing major transport facilities, and this has led to different forms of long-term contractual arrangement like asset monetization (Whitfield, 2010a; Haughton and McManus, 2012). We come back to these differences below. These contractual markets are distinct because they lack certain forms of competition, at least when it comes to pricing. While there may be competition in terms of bidding for franchises or concessions – as with William Baumol’s notion of ‘contestability’ (see Kay and Thompson, 1986) – the length of these contracts then militates against any benefits of continued price pressures on productivity, performance or value for money.

Third, the state has created new markets as the result of the privatization of state-run ‘network monopolies’ like electricity, natural gas, telecommunications and water, as well as transport infrastructure and public transit (see Crouch, 2011). As Graham and Marvin (2001: 78) point out, these network monopolies ‘were seen to require a significant capital outlay, recouped over a long period of time’ and ‘represented a form of “embedded” or “sunk” capital that realistically could not be dismantled or moved’. This meant that their privatization could not be accompanied by a concurrent expansion of competition; moreover, private sector actors could not be tied into long-term contracts without leaving significant and unresolvable uncertainties (e.g. potential rising prices or falling demand, unexpected financial crises, political changes, etc.). Instead, new privatized monopolies were created, which necessitates the creation of new regulatory agencies to curb (too much) monopoly power – this helps us to distinguish between regulated markets as an example of marketization and other forms of privatizations (Kay and Thompson, 1986). Examples of these regulatory agencies from the UK include OFCOM (telecommunications), OFWAT (water and sewage), and OFGEM (natural gas and electricity) (Parker, 2009). Across Europe, Clifton et al. (2006) show that many of these regulatory agencies were introduced prior to the ‘privatization’ of utilities, illustrating how markets are instituted by the state, supporting the claim that neoliberalism cannot be equated with deregulation (Peck and Tickell, 2002).

In summary, there are at least three different forms of market instituted by the state, all of which necessitate theoretical unpacking. These can be distinguished by differences in competition and pricing. We can identify at least two different forms of competition in these markets. First, there is competition in markets; that is, between competing producers or service providers (e.g. cleaning services). Second, there is competition for markets; that is, for long-term contracts or franchises (e.g. privatized railway franchises). We can also identify three forms of pricing in these markets. First, there are markets with competitive prices where there are competing providers (e.g. outsourced service delivery); second, there are markets with contractual pricing that involve bidding for franchises or concessions and long-term contractual arrangements (e.g. PPPs); and, third, there are markets with regulated pricing (e.g. privatized utilities). Next we look at the different rationalities, instruments and actors underpinning these different market mechanisms.

2 Diversity of market rationalities

We can distinguish between marketization processes on the basis of epistemological claims about the benefits of markets, and the attendant strategies deemed necessary to achieve those claims – what we call ‘rationalities’ here. These rationalities form part of Foucault’s (2008) notion of ‘governmentality’, or the art of governing. Neoliberal governmentality, according to Foucault, does not entail deregulation; rather, it necessitates the production of freedom through ‘the establishment of limitations, controls, forms of coercion, and obligations’ (2008: 64). Where rationalities come into play is when we can identify specific strategies underpinned by particular legitimating discourses. These rationalities, moreover, can be highly diverse, meaning that it makes little sense to talk of market-centered rationality – or marketization – in the singular. We can identify a number of rationalities that constitute the diversity of marketization processes – these include: (a) the rationality of efficiency and inefficiency; (b) the rationality of value for money; and (c) the rationality of responsibility.

First, rationalities of efficiency and inefficiency and value for money (VFM) are related to one another, although distinct in that the former reflect a rationality of market actors and the latter a rationality of the state. On the one hand, they are similar in that they legitimate market competition through the framing of markets as efficient allocators and distributors of societal resources, especially when compared with the state. On the other hand, they are distinct in that VFM is bound up with questions of (prudent) public spending. In relation to the former, a rationality of efficiency is implicitly also a rationality of inefficiency; therefore, it makes more sense to talk of an (in)efficiency rationality. Underpinning this rationality, according to the likes of Flinders (2005), Musson (2008), Crouch (2011) and others, is the assumption that markets are (naturally) efficient and the state is (naturally) inefficient. This is treated as a given because the former involves competitive (or consumer) pressure, which creates incentives to both reduce costs and meet demand. Of course, this rationality is contested with, for example, there being ample evidence that private providers of public services are no more efficient than the state, as Warner and Clifton (2014) argue in their work. Despite being contested, however, this has not stopped the continued outsourcing of public services on the basis of this rationality. It is, therefore, important to ask why.

While Peters (2012) makes the argument that assumptions about (in)efficiency simply boil down to labour costs – that is, labour costs can be reduced when services are delivered by private providers, thereby reducing direct public expenditures (e.g. wages, benefits, etc.), if not indirect public expenditures (e.g. low-wage subsidies, tax credits, etc.) – others stress different implications. According to Crouch (2011: 24–5), (in)efficiency means that the allocation of resources should reflect consumer choice, and this is best achieved through market mechanisms. As such, it legitimates a range of practices and strategies. As Clarke (2004) argues, this rationality produces an ‘abstract individual’ as the underlying focus for the organization, practices and knowledge of public service delivery. The ‘public’ – in public service or public good – is thereby transformed from a collective entity (i.e. society) into an individual entity (i.e. consumer). These rationalities play out in specific ways, notably when it comes to the restructuring of state organizations in light of theories of new public management (NPM). According to Flinders (2005), Musson (2008) and Peters (2012), NPM facilitates new rationalities and strategies of individual performance, individual pay bargaining and limited, temporary employment. These implications for work practices and conditions are evident in the outsourcing of public services. It is interesting to note, at least in Doogan’s (1997) work on the UK, the contradictory effects of the fragmentation of the labour market engendered by the outsourcing of local service delivery to meet individual consumer demands. Although it is meant to create efficiencies from competition between small, local providers, it actually leads to the subsequent amalgamation and concentration of providers into large, multinational firms (e.g. Serco, Capita) as the latter take over public service delivery (see also Crouch, 2011). Thus such rationalities not only (re)configure state actors, but also market actors.

Second, according to Rutherford (2003), market mechanisms are also equated with value for money (VFM) when it comes to state spending. He goes on to argue that in instituting VFM as the overriding goal in policy delivery, the state places ‘targets, competitions and cost’ above other objectives, which leads to ‘comparisons with the private sector’ making state actors more ‘receptive to private sector participation’ (2003: 45). This VFM rationality is tied to particular forms of governance, especially those involving new (long-term) contractual arrangements like public-private partnerships (PPPs). VFM rationality naturalizes private sector involvement in public service and infrastructure delivery. Consequently, VFM legitimates the new forms of market that are centered on contractual arrangements, often long-term ones, rather than market competition. This VFM rationality is evident across a diverse number of contractual forms. Hodge and Greve (2010) illustrate this diversity in PPPs by outlining the difference between joint ventures, long-term infrastructure contracts (LTIC), policy networks, community development, and urban renewal projects. These PPPs can vary quite considerably between countries; for example, LTIC are public-financed in the USA and private-financed in the UK (2010). What they share in common is that they represent a ‘language game’ in which new forms of assessment – e.g. ‘whole-life project costs, risk transfers and risk-adjusted discount rates’ – come to replace ‘cheapness’ as the criteria of VFM (2010: S12; also Hodge and Greve, 2007). This temporal displacement of cost over the life-cycle of the project – i.e. ‘buy now, pay later’ (Flinders, 2005) – constitutes the rationality behind infrastructure PPPs, legitimating them as a reduction in the immediate fiscal impact of investment on other public spending objectives (Hillier and van Wezemael, 2008). Moreover, the length of the contractual arrangements, while reflecting life-cycle cost calculations, binds the immediate cost of physical assets with the ongoing cost of its operation, thereby reframing contractual terms as competition for a market, if not in a market. However, in his research on Ireland, Hearne (2011) points out that these sorts of contractual arrangement (e.g. PPPs) frequently include revenue guarantees for PPP operations to ensure that private operators are not impacted by changing service levels (e.g. rising or falling user demand) over the lifetime of the contract. In this sense, VFM does not reflect reducing the actual costs of PPP projects per se, but rather the accounting and accountable costs of projects to the state at the start of the contract.

Third, the previous rationalities of (in)efficiency and VFM are also bound up with notions of responsibility and accountability. Whereas the centralized, national state is discursively constructed as inefficient and unresponsive to particular, contextually-specific and localized user needs, market-based instruments are framed as individually liberating and inherently responsive – that is, they respond to consumer demand (Whitfield, 2010b). Such discourses are constructed in and operate across the public, private and third sectors, promulgated by new business policy networks (e.g. European Services Forum), think tanks (e.g. Institute for Public Policy Research), and public-private agencies (e.g. Partnerships UK), according to Ruane (2010). One key area where responsibility and accountability come to the fore is in the framing of (financial) risk transfer. According to less critical literature on marketization, especially PPPs, the transfer of risk is one of the key benefits of PPPs; see Deloitte (2006), Kwak et al. (2009) and Bel et al. (2013), for example. This form of risk transfer is meant to motivate ‘a greater response to customer needs’ on the part of private sector providers (Bel et al., 2013: 305). Whether this is the case with all marketization processes is not clear, however, since some markets (e.g. regulated markets) have no need to be responsive to customers, only to the regulator. Generally, this rationality suggests that only certain social actors (e.g. private sector business) are – or, can be – responsible and accountable to the public. This responsibility and accountability motif plays out geographically in interesting ways. For Rutherford (2003: 42), it reflects the notion that markets ‘encourage greater local control of service provision’; for Aldred (2008: 51) it implies that ‘risk absorption is localised’ while ‘private capital and the national state attempt to devolve responsibility for service failure’; and for Flinders (2005: 230) it leads to an ‘increasingly diverse and fragmented state’ where ‘a clear and effective bond of accountability between the governors and the governed’ is lost. While this rationality no doubt plays out in different ways in different contexts and at different scales, it primarily enables the downloading of responsibility onto the local state actors, while the national state does not have ‘to bear directly the political costs’ (Peters, 2012: 221).

3 Diversity of market instruments and actors

It is useful to come back to Graham and Marvin (2001) when we consider the diverse instruments and actors in marketization processes. Echoing research in science and technology studies, they argue that cities are best understood as socio-technical systems – that is, a combination of social (e.g. public services) and technical (e.g. artefacts) infrastructures. They argue that the ‘modern infrastructural ideal’ – i.e. infrastructure monopolies, standardized technical networks and integrated cities – came under attack from the 1960s and 1970s as the result of economic and technological crises and restructuring (2001: 91–4). The integrated and networked geographies of urban centers – in the Global North at least – were undermined by the extension of neoliberal rationalities, which we outlined above. ² Here we highlight three instruments and actors constituting these marketization processes: (a) the geographical unbundling and the (contradictory) bundling of public service delivery and infrastructure development; (b) the financing instruments and forms of monetization at play; and (c) the state-market agencies created as part of the extension of markets into the delivery of public goods.

First, and drawing on the earlier work of Clark (1999) on ‘pension fund capitalism’, Graham and Marvin (2001: 97) emphasize that market-based instruments for delivering public goods are tied to specific geographical forms of delivery of those public goods. Although Graham and Marvin focus on infrastructure, their arguments can be extended to public services more generally. They highlight the geographical ‘splintering’ or unbundling of public goods as a necessary step towards the restructuring of their delivery; primarily this is because of the pressures from finance capital on the state to create attractive investments, which are bound up with certain kinds of tradable asset (i.e. those that can be made excludable). In this sense, financial attractiveness was and is increasingly tied to stand-alone (and often physical) assets rather than integrated networked assets; for example, a road rather than a public transit system, or a contract to deliver services in one place rather than across places or jurisdictions. This is a point supported by Musson (2008: 3) when he argues that ‘public bodies and private contractors tend to negotiate deals on a case-by-case basis’. As Siemiatycki (2011) and others note, this means that this splintering leads to significant uneven development as some places (e.g. cities) end up better served by newly unbundled assets or services than others. This is because there is existing demand in those locations (i.e. plenty of users) creating path dependence in the returns on those investments (i.e. more users in cities and therefore more potential for user fees and therefore more investment).

In light of our arguments about marketization, it is interesting to note that the unbundling – or splintering – of assets is accompanied by a (contradictory) bundling of the delivery instruments and actors. This is most evident in the discussion of PPPs (e.g. Rutherford, 2003; Hearne, 2006, 2011; Hillier and van Wezemael, 2008; Siemiatycki, 2010, 2011, 2012). While we characterize state-market partnerships as a particular instrument – or set of instruments – for the delivery of public services and infrastructure, they also entail a reconfiguration of state and market actors. There are a range of such instruments, all with their own peculiarities. Although we cannot go into too much detail here, others have done so already. For example, Siemiatycki (2010) outlines at least four PPP models: Design-Bid-Build (DBB), Design-Build (DB), Design-Build-Finance-Operate (DBFO), and Build-Own-Operate (BOO) (also Shaoul et al., 2006). ³ Building on Siemiatycki, we suggest that this range of instruments runs from least to most marketized. What really distinguishes such PPPs as an instrument of marketization, however, is that they entail a bundling of activities and actors in the delivery of public goods (Rutherford, 2003). This means that there is no single market mechanism at work, but rather there are multiple, overlapping markets involving their own state-market entanglements. For example, a number of scholars have identified labour market restructuring with various marketization processes (e.g. Doogan, 1997; Peters, 2012). As Doogan (1997) argues, local variation in outsourcing – itself an instrument of marketization – has meant that there is increasing geographical variability in labour markets resulting from the fragmentation of labour as public services are ‘packaged’ into discrete work contracts.

According to Siemiatycki (2010), the bundling of activities in infrastructure PPPs – which would also apply to services outsourcing – includes planning, delivery, financing, operations and management of facilities or services. They are usually bundled together in one contract (i.e. franchise, lease or concession) and contractual relationship. Contractors, in turn, usually comprise consortia of private sector actors, including financial investors (e.g. banks, pension funds), design and construction companies, and operations and management companies. Alongside the private sector actors, however, state actors play a key role in creating and facilitating the ‘market’ conditions – or, perhaps more accurately, ‘contract’ conditions – suitable for marketization. This is why Hillier and van Wezemael (2008) identify diverse actors in assemblages of PPPs; that is, assemblages are more than the private sector contractors. In their study of a PFI school, for example, Hillier and Wezemael identify: (1) an assemblage of multi-scalar state actors (e.g. Local Education Authority, central government ministry, etc.) seeking to enroll private finance; and (2) a PPP governance assemblage of multi-scalar state and market actors (e.g. local government, private sector partners, etc.). Ultimately, they show that PPPs are a hybrid network of state-market actors entangled in the use of specific market instruments like PFI. It is also why Miraftab (2004: 94) argues that ‘State decentralization arguments that lay the groundwork for advocating PPPs reveal contradictory expectations that through them the government both enables and regulates the market’. Generally, what differentiates market instruments and actors are the different market mechanisms underpinning them, as noted above, meaning that different instruments and actors are enabled and regulated quite differently by the state; that is, the state does not simply create the market, it creates or institutes a diversity of markets.

Second, the bundling of service delivery in marketization processes outlined above is distinct from privatization and involves diverse entanglements of capital and state in the delivery of public goods. There are a range of financial instruments and actors that can underpin marketization processes, for example. We want to highlight two here: (1) PPP financing and (2) asset monetization. With PPP financing, much of the literature discusses and critiques the supposed transfer of risk from the state to the market (e.g. Asenova and Beck, 2010) and the side-stepping of state budgetary spending limits (e.g. Ruane, 2010). Risk transfer is premised on a specific form of calculation. It is dependent on forms of life-cycle assessment in which the costs of various activities – such as planning, construction, maintenance and operations – can be specified in each PPP or service contract. Evidently this is difficult to calculate with any certainty over the lifetime of a long-term contract (e.g. 20–40 years). Hearne (2006) points out that this type of risk transfer is a key instrument for shifting responsibility from the state to private sector companies, for both the delivery and operations of the service or asset. Hearne (2011) goes on to highlight, however, that such life-cycle calculations do not necessarily lead to any government savings, or even transfer of risk. Part of the reason for this is that financial terms often include guarantees for contractors so that any change to expected financial returns (e.g. from falling usage, rising interest rates, etc.) is offset by (contractual) government payments (i.e. subsidies).

Several countries in the Global North initially used PPP financing to shift public investment off government budgets, including Greece, Portugal, Spain and the UK. As Siemiatycki (2010) highlights, the UK led the way when it came to PPPs, although other countries have a long history of these types of contractual arrangement (e.g. Spain). Similar pressures in the Global South are evident as (ideological or financial) unwillingness to make public investment has driven the transfer of public services to private providers. For example, Bond (2010) and Ahmed (2010) show how this happened in South Africa (water) and India (electricity) respectively. In the Global South the impact of foreign operators and investors on the delivery of public services is now a major issue; this entails a shift in emphasis from the interests of a broader public to those of private business. No matter whose interests are favored, however, it is increasingly evident that these sorts of long-term contractual arrangement have (and will continue to) significantly reduced the state’s room for manoeuvre when it comes to pursuing alternative policy instruments. As Siematycki (2010: 55) notes, long-term contractual arrangements, especially where they include restrictive financial clauses (e.g. noncompetition), will likely limit ‘government flexibility to respond to changing conditions and public feedback’. How this plays out geographically is of critical importance. On the one hand it is likely to limit the capacity of local government to plan for local demands (e.g. coordinating public transit, ending user fees, etc.), while on the other hand it is likely to further embed supranational (and global) regulatory regimes like EU competition regulations across jurisdictions (see Warner and Clifton, 2014).

There is, in light of the discussion above, a major threat of institutional lock-in to particular financial instruments, especially resulting from long-term contracts. An important example of this institutional lock-in is asset monetization, as outlined by Whitfield (2010a). This involves long-term leasing of existing public assets (e.g. highways, parking meters, buildings, etc.) to a private company for an upfront lump sum payment to the (usually sub-national) state; such monetization efforts largely depend on the ability to introduce or maximize user fees (e.g. tolls). Asset monetization is primarily associated with North America, where restrictions on PPPs still exist in some jurisdictions (Garvin, 2010). Examples include: the 99-year lease of Highway 407 in Greater Toronto for an upfront payment of $3.1 billion to the government of Ontario; the 98-year lease of the Chicago Skyways for an initial payment of $1.83 billion; and the 75-year lease of Chicago parking meters for $1.15 billion (Ashton et al., 2012; Albalate, 2014). When it comes to the Global South, asset monetization has been more closely related to various forms of land-grabbing through long-term land leases (Springer, 2011). Such examples suggest that asset monetization is far from a North American phenomenon alone. The differences between these geographies would also suggest that asset monetization is not simply a means to divest public assets and secure cash to pay off government deficits in return. As Springer (2011) argues, such forms of marketization in places like Cambodia constitute part of broader patronage networks or what he calls ‘klepto-neoliberalism’. Whatever the underpinning drivers, such marketization can be highly problematic in the long-term as any public assets are unlikely to be recoverable once divested, even though they nominally should return to public control at the end of the lease.

With both these financing instruments, the primary actors are often sub-national governments bound by both balanced budget legislation and increasing responsibility for the delivery of public services. In the USA, for example, a number of states have no regulations enabling the use of PPPs (Garvin, 2010), which means that asset monetization represents a useful instrument for redeploying public funding from one activity to another. As financing instruments, both PPPs and asset monetization create incentives for private companies to sell assets onwards, leading to an active secondary market in infrastructure assets. The work of Leyshon and Thrift (2007), Orr (2007) and Clark et al. (2011) outlines a range of issues with these secondary markets, one being that lines of responsibility for particular assets become increasingly blurred as assets are purchased and resold. This also illustrates how financing is itself geographically constituted through the establishment of ‘large-scale diversification across sectors and geographies to mitigate the consequences of political risk [e.g. nationalization] at the portfolio level’ (Orr, 2007: 8). While there are commonalities between PPPs and asset monetization, they are distinguished by the spatial and material limits to asset monetization, in that certain types of asset – but not usually services – can be easily monetized, including roads (e.g. tolls) and parking space (e.g. parking fees), while others cannot (e.g. railways and transit) (Siemiatycki, 2011). In contrast, PPP financing can be used across a range of public services.

Third, it is important to examine the emergence of new state-market actors like delivery agencies, regulators and so on – all such actors are tied into the extension of particular forms of market. Many new state-market actors first emerged with privatization in the 1980s and early 1990s, especially as a result of the need to create new regulatory agencies (Clifton et al., 2006). While privatization was popular in certain countries (e.g. the UK) during this period, Cumbers (2012) argues that it frequently resulted from new forms of market liberalization driven by new supra-national regulatory regimes like the European Union (EU). Interestingly, marketization processes, at one and the same time, involve market instruments – like PPPs or outsourcing – to unbundle assets at one scale, such as local government services, and other instruments to integrate markets at another scale, such as the EU’s internal market. In this sense, the private financing and monetization of public goods necessarily entails, on the one hand, the commodification of public services, their provision and their delivery, and, on the other hand, the regulation of these commodification practices by bringing them within the purview of the state as market-maker. The entanglement of state and markets, in this sense, does not mean that the state no longer has any say over how the market ‘delivers the goods’, as it were; rather, and like Weller and O’Neill (2014: 117) suggest, ‘contracted out services remain tightly regulated, governed by benchmarks and performance standards’. Even where there is no formal state (re)regulation (e.g. new regulatory agencies), marketization processes do not imply a lack of ‘state-like’ oversight. As Aldred (2008: 41) highlights in her work on the UK National Health Service (NHS), national policy is often ‘entrusted not to the state itself, but to a part-privatised quango [quasi-autonomous non-governmental organization]’. These quangos are not privatized entities, nor are they necessarily state entities; they represent a new state-market actor. However, they remain proxies of the state, whether or not market instruments and actors constitute them.

A number of scholars suggest that the emergence of new state-market actors – ranging from regulatory agencies through to quangos – illustrates the expansion of a ‘regulatory state’, especially in places like the EU (e.g. Majone, 1994). For example, Clifton et al. (2006: 740) argue that new regulatory agencies led ‘to changes in state functions focusing on correcting market failures and promoting competition via rule-making rather than budgetary allocations’. They highlight how the insertion of markets necessarily involves reregulation, not deregulation – that is, it involves the further entanglement of the state and markets, not their disentangling. This reregulation is associated with roll-out forms of neoliberalism, according to Peck and Tickell (2002), and is most evident when it comes to what Graham and Marvin (2001: 78) call ‘network monopolies’, discussed above. Examples range from airports through to various utilities like telecommunications and electricity generation (Kessides, 2005; Parker, 2009). However, creating markets from these network monopolies turns them into ‘privatized monopolies’, since cost barriers preclude multiple, competing providers (e.g. more than one railway network, electricity grid, and so on). Markets in these monopoly sectors can only be instituted through the establishment of regulatory mechanisms and agencies.

Obviously, the state is not hollowed out as a result; rather, it underpins this particular form of regulated market in which new regulatory actors set rules on competition, pricing, profit, etc. For example, Parker (2009) highlights a number of such agencies in the UK, including: OFCOM (telecommunications), OFWAT (water and sewage) and OFGEM (gas and electricity). It is important to remember, and unpack geographically, that not all regulators operate in the same way. There is considerable diversity around the world resulting from inherited institutional differences. Parker and Kirkpatrick (2005) provide a useful illustration of how prices can be regulated in different ways. For example, the UK has price controls and pricing formulae (e.g. regulating train fares), while the USA has rate of return regulations designed to ‘cover all legitimate operating and capital costs while providing the firm with a “fair” rate of return on the capital employed’ (2005: 245). These different pricing instruments entail different market incentives; supposedly the former (i.e. price controls) provide incentives for efficiency improvements, while the latter do not because it does not matter how profits are made (i.e. efficient profits are the same as inefficient profits). When it comes to privatized monopolies, the creation of a regulated market – as opposed to price or contract markets – is necessary because of the specific social and technical costs associated with entry into that market; namely, and as Wrigley (1992) pointed out, this involves high sunk costs that limit the number of competitors in a market or bidders for a market.