Abstract
This report considers recent developments and ongoing debates around relational economic geography, and a growing body of work that has focused on economic practices as a means better to understand production processes and economic development. In particular, it examines the critical reaction to relational thinking within the subdiscipline, and the nature of the debate about the degree to which relational work is – and needs to be – regarded as distinct from more traditional approaches to economic geography. It then considers how relational economic geography has become inflected towards an epistemological and methodological focus on practice. It argues that this engagement with economic practices provides the basis to respond to some of the limitations identified with earlier work, and opens up fruitful new potential for theorizing the nature of agency in the space economy.
I Introduction
[There are] distinctive tendencies in economic geography to take new directions that differ from traditional approaches. I would refer to this as a trend toward a relational approach in economic geography. This should not be viewed as another turn, because it does not try to isolate those aspects of human life which are inseparable (see also Hudson 2004). It integrates economic and social, cultural, institutional and political aspects of human agency. (Bathelt and Gluckler, 2003, cited in Bathelt, 2006: 224)
Bathelt (2006) contended that a relational approach does not represent a turn because it is synthetic and integrative, but this point is arguable, and he later described relational work as ‘a distinct turning point’ (p. 226). (Sunley, 2008: 3)
II Questioning relational economic geography
I want to begin by picking up from where Harald Bathelt ended in his 2006 report, and assess the critical and sometimes spirited debate that has surrounded relational economic geography over the last seven years. In order to do this, I need quickly to revisit where the debate had reached and the assessment presented. This will be familiar to many readers, but I think four key arguments are worth identifying.
First, Bathelt (2006) identifies the impetus for a broad relational shift as rooted in an epistemological dissatisfaction with ‘traditional approaches’ to the discipline. These traditional approaches have tended ‘to focus on regions and other spatial representations as if they are actors’. The issue is that ‘the real agents – those people who act and interact in firms and other organizations to produce economic value – are often neglected (Swyngedouw, 1997; Maskell, 2001)’ (p. 224). Furthermore, Bathelt argues that traditional approaches often use spatial variables related to distance as explanatory factors to understand location decisions and spatial distributions of economic activities (cf. Barnes, 2001; Sheppard, 2000). The problem in short is that traditional economic geography has tended to neglect the role of agents who actively shape their environment (Bathelt, 2006: 224). Second, therefore, interest in relationality within the subdiscipline is thus a response to these limitations, and in particular a dissatisfaction with the conceptualization of agency in economic activity. Bathelt follows Yeung (2005) in seeing relational economic geography as encompassing ‘a strong actor perspective’ that ‘integrates economic, social, cultural and political influences on economic action’ and ‘rejects deterministic interpretations related to spatial categories’ in favour of a differently inflected emphasis on ‘a spatial perspective in analyses’ of the space economy (Bathelt, 2006: 230; cf. Yeung, 2005).
The third key argument is that the relational turn is advantageous insofar as it enables a more sophisticated understanding of ‘the consequences of global interdependencies and their relation to processes of local concentration and specialization’ (Bathelt, 2006: 224). It does this by adopting a different epistemological stance on agency and structure that draws on a wider debate in human geography and the social sciences. Foregrounded in Bathelt’s summarization are postmodern/structuralist insights from the cultural turn, but also arguments within institutional theory that reflect long-standing debates in modernist social science and, subsequently, critical realism around structure and agency (cf., for example, Giddens, 1984; Sayer, 2000). Relational economic geography thus suggests that economic agents act ‘according to particular economic and non-economic goals and strategies, as well as feelings and emotions’ (Bathelt, 2006: 229; see also Ettlinger, 2003; Massey, 2004). The results of this intentional action encompass intended and unplanned spatial outcomes which, in turn, impact the next round of decisions (Bathelt and Gluckler, 2003; Werlen, 1995).
Fourth and finally, in both epistemological and methodological terms, the argument is that economic geography needs to integrate a micro-perspective on human action that emphasizes its contextual, path-dependent and contingent nature. The implication is that ‘general spatial laws of economic action do not exist’ although simultaneously ‘the relational perspective does not rule out macro-theoretical considerations because human agency is not independent from the conditions of the capitalist system’ (Bathelt, 2006: 230). In this assessment, Bathelt suggests that relational economic geography invokes a broadly critical realist development of Giddensian and institutionalist structure-agency: of the macro-structures of the capitalist system which are transferred to the individual level through institutions in a process of downward causation. At the same time, there is also a process of upward causation (Hodgson, 2003). This process describes how micro-practices are translated into broader institutional arrangements which affect the macro-level.
1 Critiques of the relational ‘turn’
In the last seven years, however, the uptake and development of relational economic geography has been uneven, diverse and the subject of critique. Regarding the latter, the reaction of some has been to contest the utility of the so-called `relational turn'. While these critiques have emerged from a number of different schools of thought within the subdiscipline, and are most certainly entwined in ongoing interventions about the impact of the cultural turn on economic geography (cf. Overman, 2004), they are often limited and more implicit than explicit engagements embedded within empirical papers or contributions on other topics (Peck, 2005). Within the constraints of this report, it is therefore most useful to summarize one of the main interventions that captures much of the critical reaction to date – Peter Sunley’s (2008) contribution in Economic Geography.
Sunley develops what I would suggest are at least five major lines of critique of relational economic geography. The first is that there is no singular relational approach to economic geography, but rather that this so-called ‘relational turn’ represented an elision of many different arguments and theories. He is critical of the ill-definition of the key concept of relation, suggesting that there is no clear differentiation between the concepts of relation, relationship and relational. Furthermore, he argues that the notion of ‘relational’ has ‘shifted to a high level of abstraction’ and ‘has lost much of its content and analytical grip’. In other words, much of the relational economic geography to which Bathelt refers lacks a clear understanding of what relationality is, and hence struggles to provide any overarching conceptual justification for the explanatory power of relational approaches. Sunley thus sees relational economic geography as an eclectic mix of work, ranging from focusing on firms and individuals, to networks and theories of the social construction of knowledge (Sunley, 2008: 19). In that sense, relational economic geography does not represent a coherent turn nor a programmatic agenda for the subdiscipline.
Second, Sunley questions the shift towards micro-scale research with a focus on individual actors and agents, and he doubts the capacity of relational approaches to develop generalizations. He argues that in reacting so strongly against identifying patterns and structures in economic landscapes – and in rejecting both analytical and substantive boundaries – relational economic geography has in fact lost sight of many of the valuable insights of institutionalist and critical realist approaches. His view is that it fails to offer ‘analytical models that prioritize causes’ and identify ‘causal mechanisms’. Consequently, it struggles ‘to discriminate between alternative economic theories and interpretations and has failed to deliver on its promise to provide mid-range proximate theories (Sunley, 2008: 3). It is important to note that this critical point falls in line with others insofar as relational economic geography has been seen to ignore the insights of a sizeable institutionalist literature (Peck, 2005). Sunley therefore is particularly wary of what he sees as a ‘separation’ of personal relations from ‘other regularities in behaviour’ as well as ‘the norms that regulate and shape them’ (Sunley, 2008: 4).
The third problem identified with relational approaches is that they give insufficient attention to processes of emergence and fail to recognize that different scales and layers of economic entities and institutions interact and co-evolve (Sunley, 2008). This seems largely aimed at the segment of the relational literature concerned with networks of various forms (e.g. Coe and Bunnell, 2003; Dicken et al., 2001; Yeung, 2003, 2005). The criticism is that in focusing on static networks or mapping relationships at a point in time, the significance of the dynamism of actors and entities is lost. Undoubtedly, drawing on the burgeoning literature around evolutionary economic geography (cf. Coe, 2010), the problem for Sunley is that much relational work tends ‘to dissolve scalar units into long chains of networks and spatial relations’ in a way that ‘misses the significance of emergence and overlooks some of the ways in which it affects the development of economic and social relations’ (Sunley, 2008: 15). The concept of scale may be problematic, but relational economic geography’s concern with networks throws the baby out with the bath water.
Fourth, one of the unexpected ‘conundrums’ that relational economic geography faces is that, despite claiming to be centrally concerned with the exercise of social and economic power in social relations, its actual appreciations of economic power and social relations are narrow and partial. For Sunley, this is essentially because it lacks the capacity to understand what used to be called ‘structural’ power, or macro-level power. He identifies the root of this problem as lying in relational approaches having become ‘increasingly preoccupied with micro-scale processes and ambiguous and imprecise networks’ (Sunley, 2008: 19).
Finally, there is the criticism that much relational economic geography lacks methodological rigour and consistency; it is, in other words, sloppy social science or, at best, it falls short of consistent agreement and implementation of methods that produce meaningful data on the space economy. This is in part due to the diversity and lack of clear definition of relations, but it is also grounded in a more fundamental epistemological argument about the need for generalization. For Sunley, the methods generally favoured by those relational contributors informed by poststructuralist ideas are incapable of producing useful mid-range concepts or meta-theories.
2 Relationality and its critics: ships that pass in the night…
The critiques developed by Sunley and others have begun to elicit a limited response from those working within relational economic geography (e.g. Hassink and Klaerding, 2009; Jones, 2009; MacKinnon, 2011), and volumes could be written on some aspects of this debate (e.g. structure and agency). However, I want to argue, in considering emerging responses, that this debate has to some extent lost sight of the original criticisms of traditional approaches which provoked the relational shift in the first place. In fact, I would suggest that much of the critique levelled at relational economic geography has become absorbed in concerns that are in many respects peripheral rather than central to the original ontological and epistemological impetus for the relational shift.
However, this is not to say that criticisms of relational approaches do not have considerable validity and raise important questions that those continuing to work in relational economic geography need to address. By way of a brief summary – and not an exhaustive list – it is clear that definitions of what a relation is, how that is demarcated as a phenomenon, and understandings of how relations are important in shaping economic activity or outcomes require much further elaboration and development. Furthermore, the loose usage of relationality as an umbrella term for a range of quite distinct theoretical and empirical work is a significant limitation. It is also true that examples can be found of relational work that has under-theorized the nature of agency as well as the role of organizations and institutions; equally there is force to the criticism that some relational economic geography has over-focused on mapping relational networks at the expense of understanding the wider context of economic actors or the emergence of economic outcomes.
Yet, beyond this, I would argue that the character of the recent debate about relationality has (unfortunately) been diverted from the key issues that prompted the emergence of this shift in the first place, and to some considerable extent also has been dominated by a rather polarized assessment of the cultural turn and the ‘post-’ critique of modernist epistemologies (cf. Jones and Murphy, 2011). It has developed too much along an either/or trajectory that assumes: (1) that traditional and relational approaches to economic geography are irreconcilable; (2) that micro-level theorization cannot produce rigorous or effective theoretical macro-level generalization; (3) that relational economic geography has not yet developed a coherent and systematic theoretical framework and methodological approach and thus ‘the project’ has floundered; and (4) that the relational turn has eroded the capacity of economic geography to generate replicable or policy-relevant theoretical insight. In relation to the first two, recent interventions have gone to some length to contest these assumptions in arguing both that there is considerable common ground between relational and other approaches to economic geography, and also that mid-range theories are entirely possible to develop (Jones and Murphy, 2011). With respect to the latter two, interest in relational approaches has strengthened over the last seven years, with many seeking to develop more directly replicable mid-range theories. I will consider in particular how economic practice represents a key element of this mid-range theorization.
In this respect, the critiques identified in Sunley’s (2008) intervention are far from terminal to the ongoing relational shift, and do not address the fundamental factors behind the relational shift that Bathelt identified in his earlier report. Foremost is the sidelining of the fundamental dissatisfaction with attributing economic development to spatial constructs such as regions, and to over-stabilized structuralist concepts that are inadequate for conceptualizing where agency resides. Contra to Sunley’s argument, part of the impetus toward relationality was a recognition that physical places such as regions or industrial clusters are not the containers of ‘path dependence’, and increasingly so (Hess, 2004; Jones, 2008). While at times relational economic geographers may well have been too focused on individual practices and networks of relations at the expense of institutions, organizations and wider context, these limitations do not in themselves negate the remaining problem with ‘traditional’ approaches. Rather than deserving a final valedictory report, therefore, it is perhaps that relational economic geography has not successfully reconciled some of its own foundational questions about the spatial nature of economic agency and how entities that endure in time and space across places are shaped and develop to produce economic outcomes (again within, across and beyond places).
Thus, while the recent critique is undoubtedly an important intervention – and identifies very significant limitations with much of the so-called ‘relational’ economic geographical work to date – it largely avoids the key axes of debate. These are the questions about spatiality and economic agency and structure. In that sense, in providing a review of the state of the debate about relational economic geography, I would argue that recent critiques have identified important weaknesses and inconsistencies in how a loose relational shift has been implemented, but they have done little to address the foundational ontological and epistemological drivers behind it. Conversely, and acknowledging the undoubted value of the critiques, little work within relational approaches has responded directly to the critical challenges that Sunley and others identify around, for example, refining definitions of what an economic relation is or how methodological rigour might be achieved. Relational economic geography and its critiques are thus, to date, largely ships that pass in the night.
However, in the last part of this report, I argue that more recent work has inflected relational economic geography in a direction that both in part responds to some of the identified weaknesses of the first phase of relational work and also develops a research agenda that transcends some of the polarized nature of the debate about relationality to date.
III Relational economic geography 2.0: the growing emphasis on practice
Much of the recent literature that might fit into Bathelt’s (2006) categorization of relational economic geography has been characterized by a different epistemological emphasis than earlier work. While this further development is itself diverse, it is possible to identify a growing conceptual concern for practice as a means better to capture and specify the nature of agency that is caught up in the relations between economic actors. This is not to argue that other strands of relational economic geography have not also developed further – notably work on global production networks (e.g. Cumbers et al., 2008; Nadvi, 2008) and the broad ‘German school’ of relational economic geography (Bathelt and Gluckler, 2011; Berndt and Boeckler, 2009, 2011; Glückler, 2006; Grabher et al., 2008; Ibert, 2010). However, even in these areas there is evidence of relational economic geographical thinking engaging with conceptions of economic practice as a means to theorize the nature of relations.
The roots of economic geographical concern with practice are both long-standing and diverse, but Jim Murphy and I argued in a recent review that a diverse body of work over the last 5–10 years marks this growing epistemological and methodological interest in practice among economic geographers (Jones and Murphy, 2011). We identified four threads within this work, centring, respectively, around: (1) institutional theories and approaches; (2) governmentality; (3) diverse livelihoods and everyday economic practices; and (4) communitarian and relational approaches. It is, of course, this final thread that is the concern here, but the others are worth identifying because what might be termed relational economic geography overlaps in many areas with these other literatures.
Within recent relational approaches, and refining further the arguments of the 2011 paper, there are several developments in practice-oriented thinking within relational economic geography. First, relational economic geographers have increasingly taken intra- and inter-firm practices as their central concern. A key focus has been to identify, interpret and explain the dynamic nature of the relationships that shape production and retailing activities (Palmer and O’Kane, 2007; Wood and Reynolds, 2012), finance (Buttle, 2008; Clark et al., 2009; Hall, 2006), knowledge transfer (Larner and Laurie, 2010; Vallance, 2011), learning and innovation processes (Hall, 2007, 2009; Ibert, 2007, 2010), value chains and global production networks, and industrial clusters (James, 2007; Murphy and Schindler, 2011). Although there is a diversity of empirical and theoretical foci, relationally-inclined scholars have recently often viewed practices as everyday relational processes that constitute economic action and hold communities or firms together within, and in relation to, particular geographic contexts (cf. Jones and Murphy, 2011). Much of this work has drawn upon the interdisciplinary ‘communities-of-practice’ (CoP) literature (Amin and Roberts, 2008). For example, relational concerns with practices have examined tacit knowledge held within industrial, value-chain or intra-firm communities, knowledge that is often only realized in the ‘doing’ of business (Faulconbridge, 2006, 2008; Gertler, 2008; Hall, 2008). In this view, learning within firms, clusters and industries is driven by more than simply the aggregation of individual sources of human capital; it is instead the product of collectively legitimated (everyday) social practices wherein and through which knowledge is embedded (Jones and Murphy, 2011).
A second aspect of the recent practice-oriented relational literature is concerned with the sociospatial dynamics of industrial organization. Earlier relational work identified how networking or relational practices act as ‘ordering mechanisms’ through which power is articulated and globalized economic activities are organized (Amin, 2002; Dicken et al., 2001), and this has been developed in the more recent literature through a concern with how industries and firms too are organized through common social practices (e.g. Certomà, 2011; Murphy, 2006; Pain, 2008). These are practices that legitimate, control and coordinate business activities, and that can help to create the relational proximity (and in some cases trust) needed for firms to act at a distance in a globalized economy (Faulconbridge et al., 2009; Glückler, 2006). Examples include work that has examined the relational practices that generate trust or mistrust in the venture capital industry (Jones and Search, 2009; Wray, 2012a, 2012b), the way in which project-based work is undertaken in creative and cultural industries (Watson, 2012; Watson and Ward, 2013) and how relational practices structure technology clusters (Ahlqvist, 2013; Wang et al., 2010). A further development is work that considers the practices that surround temporary sites of production, knowledge exchange and learning in the global economy, such as trade fairs and exhibitions (Bathelt and Spigel, 2012).
Third, the body of work concerned with alternative economies has also become increasingly concerned with economic practices of various actors in shaping the nature of the economy. Drawing on broader relational approaches, recent research has focused on the governing practices that shape the development of industries and regional economies well as their alternate forms. Examples would include work that has examined the development of Islamic finance and banking within the global financial system (Pollard and Samers, 2013) or how governing practices are caught up in regional and urban growth strategies and outcomes (Suau-Sanchez et al., 2012). At the scale of individuals and groups, economic geographers concerned with diverse livelihoods have also become increasingly concerned to theorize economic practices within a relational framing to understand, for example, household economies and the relationship of everyday practices to wider economic development (Slocum, 2007; Smith and Stenning, 2006; Stenning et al., 2010)
Finally, and perhaps most recently, the engagement between relational economic geography and practice has proven very fruitful in a body of work that is most concerned with interrogating the relationship between performativity and economic relations (Berndt and Boeckler, 2009). The broad German school of relational economic geography has in this sense been most active, clearly evident in work concerned with markets and finance. Drawing from related work on performativity in socio-economics (e.g. Mackenzie, 2007; Mackenzie et al., 2007), Berndt and Boeckler (2009), for example, provide a wide-ranging review and theoretical arguments for the development of a distinctly geographical understanding of how practice/performance shapes the nature of market development and market behaviour. Other work has applied this understanding to a wide range of markets such as those caught up in transnational commodity chains or the internal markets within TNCs (Berndt and Boeckler, 2011).
Overall, these recent practice-oriented strands within relational economic geography encapsulate in part a response to criticisms levelled at earlier relational work. Berndt and Boeckler (2009), for example, provide a set of mid-level conceptualizations of markets understood through the lens of relational performativity that they argue are applicable to a variety of market formations (e.g. financial markets, commodities, supply chains). Equally, contributions such as that of Vorley et al. (2012) seek to move beyond static conceptions of networks and to unpack the nature of relationality between economic actors within those networks. In this way, recent relational economic geography has taken up the challenge of refining scope for developing generalizable concepts and theories of economic agency without an over-emphasis on micro-practices and individual actors. Some of this work is also engaging actively with more traditional economic geography – certainly institutional and critical realist approaches – but does not set them in opposition to poststructuralist grounded work such as that using actor-network theory. There are also signs that there is a response to Sheppard’s challenge to relational economic geography (as with other subdisciplinary areas) to reinvigorate its political economic purpose (cf. Sheppard, 2011).
However, while an epistemological shift toward practice is helpful in these respects, it has also been argued that this practice-oriented work needs a clearer epistemological and methodological framing in order effectively to respond to some of the identified weaknesses in existing relational work. To do this, practice-oriented relational economic geography still faces methodological and empirical challenges. Three are of particular relevance (Jones and Murphy, 2011): (1) it must clearly demarcate the boundaries of particular practices such that they are discernible as factors constituting or driving larger-order socio-economic phenomena; (2) it must be able to identify those practices that have a significant impact on socio-economic outcomes; and (3) a practice-oriented epistemology means little unless it enables economic geographers to make generalizations about the space economy.
IV Conclusion
This ‘practice shift’ as applied to relational and other strands of economic geography does not necessarily warrant labelling as a ‘turn’; it is not intended as a characterization of a coherent ‘school’, draws on diverse theories of practice, and also has long antecedents within the social sciences that precede debates about relational economic geography (Jones and Murphy, 2011). What it does provide scope for, however, is the capacity to move the rather polarized terms of the debate about relational economic geography forward to a more fruitful engagement which accounts for the limitations of traditional approaches that remain unaddressed, as well as with the identified weaknesses in relational thinking to date. The critical response to the so-called relational turn has thus provided a useful impetus to re-evaluate what may have been at times a narrow epistemological and methodological focus. Yet, if it is grounded in the proposition that one cannot understand durable economic forms and their implications for evolutionary trajectories through the study of networks and the micro-social economic world, then this objection is misguided. The issue that relational economic geographers need to address is the degree to which relational approaches have succeeded in doing precisely this, not whether it is worth doing (Jones and Murphy, 2011). And the conceptual differences between relational and other approaches are arguably less than many have argued (Hassink and Klaerding, 2009). In that respect, the identifiable shift towards practice offers the opportunity better to conceptualize economic agency through mid-range theories that do not rely on either the unsophisticated conceptions of agency or the spatial categories that prompted the original dissatisfaction with the ‘traditional approaches’. Relational economic geography has thus certainly moved on from this situation since the last report, and the challenge now is to move beyond the polarized terms of the recent debate. Part of that requires a recognition that ‘traditional’ approaches to economic geography have themselves also become more sophisticated in addressing the limitations that prompted the relational turn a decade or more ago. In any event, the literature that draws on the broad legacy of relational economic geography continues to grow, suggesting that many in the discipline see it continuing to offer a very fertile approach for understanding the space economy.