Re-imagining evolutionary economic geography

Re-imagining evolutionary economic geography—why, and why now?

It is now more than two decades since the idea of adopting evolutionary ideas in economic geography was first mooted (Rigby and Essletzbichler, 1997; Boschma and Lambooy, 1999). Since then, the paradigm has developed significantly, both theoretically and empirically (for some useful surveys, see, for example, Boschma and Frenken, 2006, 2018; MacKinnon et al., 2009; Boschma and Martin, 2010; Kogler, 2015; Pike et al., 2016; Henning, 2022; Martin and Sunley, 2023). Whether or not the paradigm of evolutionary economic geography (EEG) can be said to have yet reached ‘middle age’, it is certainly the case that its first flush of youth is now past. It is perhaps germane, therefore, to reflect on what has been achieved, and what ­remains to be done. A key question is whether the theoretical and empirical scope of EEG needs adjusting so that it is (better) suited to help understand the upheavals, crises and transformations that confront contemporary capitalism and its spatial configurations, and what those adjustments should be.

For the simple fact is that the circumstances under which geographers began to explore the relevance of evolutionary ideas and concepts for explaining change in economic landscapes have themselves changed dramatically. Since the beginning of the 21st century (if not before) the pace, nature and direction of capitalist economic development have been undergoing immense change: we seem to have entered a new age in which more or less continuous and momentous disruption is the ‘new normal’, what some even refer to as a permanent state of ‘polycrisis’ (Tooze, 2022; World Economic Forum, 2023). The list of upheavals that have combined and mutually reinforced one another is now well rehearsed—climate change, trade conflict (especially between the USA and China), financial crises (in 2008, and now again in 2023), the energy price crisis, demographic ageing, biodiversity loss, the accelerating artificial intelligence revolution, widespread distrust in democratic institutions and the rise of populisms of different kinds, increasing geopolitical tensions (especially Russia’s invasion of Ukraine), the COVID pandemic and its after-effects, and the growth in social inequality. All of these are impacting differentially not only across nations, but also across the regions and cities within them.

Do these ‘new realities’ require revisions and recalibrations of our concepts and theories, or are our existing explanatory frameworks entirely capable of grasping those realities? This question applies to all of our theoretical frameworks and perspectives, including EEG. Thus far, the main line of enquiry in EEG has been on industrial and technological aspects of regional economic evolution. More recently, the focus has also shifted downwards to the micro-level of individual firms or technologies. There has been much less attention, however, to such issues as social inequalities, the public sector or to political institutions, especially the state. Equally, there has been less attention on the ‘big processes’, ‘climacteric crises’ and ‘mega-trends’ that are arguably the key drivers of uneven regional and city development in the contemporary era. These are all topics in the analysis and understanding of which EEG can potentially contribute, though arguably by expanding its theoretical foundations and its empirical remit. The various articles in this themed issue of the journal are, in their various ways, directed to this challenge. But what is distinctive about EEG that makes it potentially valuable as a paradigm for interrogating the contemporary upheavals in global capitalism?

What is EEG? Clarifying its paradigmatic features

Defining EEG is by no means a straightforward task (see Boschma and Martin, 2010; Kogler, 2015; Martin and Sunley, 2023). Even in evolutionary economics, there is a plethora of self-declared approaches—what Dopfer and Potts (2004, 195) termed a ‘massive hybridisation of theory’. To some extent, evolutionary economic geographers face a still developing corpus of ideas. However, the principal guiding focus is on understanding how the spatial configurations of economic activity and materiality, at all geographical scales, changes over real historical time. As Boschma and Martin (2010, 6–7) put it, the basic concern of EEG is with ‘the processes by which the economic landscape—the spatial organisation of economic production, circulation, exchange, distribution and consumption—is transformed from within over time’. We need to extend this idea by recognising that as the spatial organisation of economic activity itself evolves and transforms, this can in turn influence and transform the very processes that drive that evolution. Furthermore, the transformation of economic landscapes can be driven by processes that are slow and cumulative, and by those that are episodic and highly disruptive, and of course by the interaction between the two.

EEG, then, is concerned with theorising the evolution of the economy over time from a geographical point of view. It has emerged as an identifiable and distinct body of theory and empirics—a paradigm—within economic geography. But it can also be seen in a broader sense as a ‘project’ to embed evolutionary ideas, concepts and processes more strongly within economic geography more generally (Martin and Sunley, 2023, 13). It should certainly be acknowledged that within economic geography there is a long tradition of considering the dimension of time, history and long-run regional development, especially within strands influenced by Marxism, Political Economy and Institutionalism. Hence there is the potential for fruitful exchanges of ideas and insights between EEG and these other strands within economic geography (see MacKinnon et al., 2009; Martin and Sunley, 2015; Essletzbichler et al., 2023).

Another way of seeing EEG, however, is to consider it as a distinct strand of work emerging from, and still related, to evolutionary economics, namely that which focuses on a particular aspect of economic evolution, that is, the spatial aspect (see Metcalfe, 2023; Dopfer, 2023). Yet another viewpoint would be to see EEG as spatially inflected economic history: EEG provides a set of concepts for theorising the spatial aspects of economic history, and economic history provides empirics and procedures to develop, test and modify theories within EEG (see Martin and Sunley, 2022; Martin et al., 2020).

These different perspectives are not exclusionary. The perspective adopted does however shape the manner in which one theorises the dimensions of space and time in economic evolution (see also Martin and Sunley, 2022; Chu and Hassink, 2023), and will also shape the cross-fertilisations with other bodies of work. Because different participants in the collective endeavour of EEG have approached its core question (how to theorise the evolution and transformation of economic activity and materiality from a geographical point of view) from different theoretical angles and through a variety of empirical research strategies, the body of theory within EEG has become quite diverse. It would indeed seem fair to characterise EEG theory as a ‘patchwork’ of concepts and ideas (Martin and Sunley, 2023).¹

A key question, therefore, is whether there should and can be a unifying theoretical framework to EEG to provide more coherence and guidance to its conceptual ideas and empirical enquiries. Some leading evolutionary economists—such as Geoff Hodgson, Sidney Winter, Ulrich Witt and Jan-Willem Stoelhorst—have argued that Generalised or Universal Darwinism provides the basis for an evolutionary perspective on the economy. The suggestion is that the three key precepts of variety, selection and retention (VSR), can constitute a ‘general model of evolution’ (see for example, Hodgson and Knudsen, 2010). The application of this model to any specific field, such as economics, will depend on context and require ‘auxiliary’ propositions and hypotheses in order to derive the economic equivalents of ‘population dynamics’, ‘phenotypes’ and ‘genes’ in evolutionary biology. Many evolutionary economists nevertheless remain wary of abducting evolutionary biological metaphors into economics (as indeed was Schumpeter (1954, 789), whose work is often invoked by evolutionary economists, and evolutionary economic geographers): see the discussion in Lawson (2003, Ch. 5) and Feldman (2023). In any case, most discussions of the relevance of Generalised Darwinism as a source of metaphors and ideas in evolutionary economics seem not to recognise that even within evolutionary biology the VSR model has been extended and superseded (Martin and Sunley, 2015). However, some evolutionary economists have moved away from the VSR model of Generalised Darwinism towards conceptual frameworks based on ‘complex adaptive systems’ and ‘self-organisation’ (for example, Foster and Hölz, 2004; Foster and Pyka, 2014; Witt, 2017).

What is essential to conceptualising the economic landscape as ‘evolutionary’ is to see it as hysteretic. That is, economic landscapes—regional and urban economies—embody the legacies of their pasts, and it is those legacies that shape the possibilities for current and future change and transformation. As Martin and Sunley (2022) argue, EEG is quintessentially and necessarily an historical science, given that it is concerned with how a particular spatial configuration of economic activity and materiality at any given moment has come to be what it is; that is, how it was produced through time (Martin and Sunley, 2022). History matters. In consequence, EEG needs to devote serious attention to the temporalities of capitalist evolution, to the fact that different processes of (spatial) economic change operate over different time scales, with different rhythms and different periodicities. To date, EEG has focussed mainly on slow cumulative change, and far less on episodic disruptions, critical events, major shocks, crises and climacteric disjunctures. Historically, such ‘big events’ have had major transformative effects on the spatial configurations and geographies of economic development. And the fact that we are precisely in another of these historical ‘hinge points’ highlights and exposes this theoretical and empirical lacuna within EEG inquiry.

At least five ‘big concepts’ or ‘conceptual building blocks’ of an evolutionary approach to economic geography can be identified (Figure 1), which together can be seen as defining components of an ‘evolutionary’ approach. And underpinning and intersecting across these conceptual pillars are some key processes that are needed to construct evolutionary explanations (see Table 1). Not all of the conceptual frameworks in Figure 1, have received the same attention, and within each there remains considerable scope for further elaboration and application. For example, in terms of history, by far the primary focus has been on the notion of path dependence (see Martin and Sunley, 2006, for a comprehensive discussion of this concept), and much less on causal history, sequence analysis and other such ideas which highlight the value of case studies and context-rich approaches.

Novelty—the generation of new variety—is a key consideration in every evolutionary science. Likewise, it occupies a central place in EEG. Indeed, it is this aspect that has attracted by far the most attention in recent years. The notion of ‘related variety’ seems to have become a sort of sine qua non of EEG (Frenken et al., 2007), and is used as the explanans of all sorts of regional economic change, from innovation, to structural and technological diversification, to growth.

This particular focus on ‘related variety’ is not unproblematic, however: not only are there issues as to how ‘related variety’ should be defined and measured, its causal influence will depend on what particular sectors or technologies are ‘related’, and on whether and when ‘related variety’ is more important than ‘unrelated variety’ and sectoral and technological modularity. Furthermore, ‘related variety’ is not just an independent variable, but also a dependent one: what causes related variety to emerge, when, and where? In any case, there is much more to the generation of novelty and how new pathways of regional economic development are created and evolve than can be explained simply by ‘related variety’.

The same scope exists for expanding work on a complex systems approach to EEG. Following the lead set by Hildago and his associates (Hidalgo and Hausmann, 2009; Hausmann et al., 2014), thus far the idea of complexity used in EEG has been that of what might be called ‘compositional complexity’, that is, on the pattern, density and complementarities of the network configurations of related products, industries or technologies within and between individual regions and places. This notion of ‘economic complexity’ is very similar to that of ‘related variety’. The claim is that the greater is compositional and configurational ‘complexity’ of a regional economy, the more innovative, productive, prosperous and faster growing the region will be. Hidalgo (2021, 95) himself has even gone as far as to claim that: ‘Measures of economic complexity explain and predict international and regional variations in income, economic growth, income inequality, gender inequality and greenhouse emissions’. Such a claim is grossly exaggerated (and reductionist!). And some see economic complexity as a ‘new paradigm’ in EEG (Balland et al., 2022). But viewing regional economies in terms of their ‘economic complexity’ is not the same as viewing them as evolving, ‘complex adaptive systems’. The latter perspective would conceptualise regional economies as highly open, interactive dynamic systems, characterised by non-linear processes, differential adaptability and emergence effects, and typically out of equilibrium (see Martin and Sunley, 2007; also Feldman, 2023). This is an area of EEG where much remains to be explored, yet which again would seem of particular relevance at a time of repeated profound systemic shocks to national and regional economies alike.

Likewise, the notion of emergence, itself closely associated with complex adaptive systems, has not received the attention it deserves. By emergence is meant the processes by which the macro-level features and behaviour of a system are derived from, but are not simply reducible to, the actions and interactions of micro-level agents (see Sawyer, 2005). At the same time, such macro-level outcomes exert downward causation on the system’s micro-level components. Thus an explicit role is given not just to the behaviour of economic agents, but to the ‘higher-level’ context (including institutions) within which such ­behaviour takes place. An emergentist view on the evolution of economic landscapes thus eschews any crude division between, or simple upward aggregation from, the micro and the macro. Economic landscapes are characterised by numerous processes and forms of emergence: for example, from markets to clusters to local entrepreneurial ecosystems to cities (see Martin and Sunley, 2012). How these emerge from the individual actions of micro agents and feed back to influence those actions is a central dimension of an evolutionary approach.

Arguably, the over-riding or ultimate goal of EEG is to explain how city and regional economies adapt over time. The notion of ‘adaptability’ in this context is far from straightforward, but it would surely have to do, first, with the capability of a region’s or city’s firms, workers and ­institutions to remain well adapted to the changing competitive, technological, market and regulatory/policy environment in which they operate; and second, with the processes by which such adaptation is achieved. All of the conceptual building blocks in Figure 1 bear on these two interrelated notions. The differential adaptability of different regions and cities underpins their differential development over time. Why are some regions and cities more successful than others in adapting to changing economic conditions and opportunities? Does the more successful adaptability of some regions and cities actually hinder the adaptability of others (akin to the idea of combined and uneven regional development)? These sorts of questions go to the heart of an evolutionary approach to studying economic landscapes. They also have critical implications for policy. How cities and regions adapt to the new phase of economic development based on green technologies and artificial intelligence technologies will almost certainly be uneven, unless policies are explicitly designed to ensure a spatially (and socially) inclusive transition.

Re-imagining EEG in light of the polycrisis: bringing in the state and policy

The multiple interconnected upheavals referred to earlier, including climate change, financial crises, demographic ageing, geopolitical tensions, the COVID-19 pandemic and social inequality, have sparked a renewed interest in the role of governance, institutions and the state. However, the so-called polycrisis has to some extent subsequently led to a ‘polycleavages’ predicament, that is, when highly relevant issues dominate the general discourse simultaneously and as such divide rather than unify potential stakeholders in their search for solutions (Zeitlin et al., 2019). This is, of course, less than ideal given the significant transformations that are currently unfolding at an unprecedented pace that might only compare to very few previous highly disruptive times, for example, the industrial revolution. Significant technological, and associated institutional and organisational transformation, emanating from novel methods of production as articulated in the notion ‘Industry 4.0’ (Lasi et al., 2014), along with the rapid diffusion of artificial intelligence-based tools across industrial sectors and the public in general, further exaggerate present uncertainties, but equally offer ample opportunities to positively transform economic and social life more generally. In order to tackle challenges and to ensure that benefits are equally distributed, that vulnerable populations are protected, and that these transformations do not result in a collapse of the entire socio-economic fabric which is already on unstable grounds, requires a more active and diverse role of the state than has been previously acknowledged, for example, as regulator, animateur, purchaser and lead-user.

Although EEG scholarship has been interested in what policy should look like, it still lacks an evolutionary conception of the state. Public policy under an evolutionary lens needs to be tackled, not only from a normative angle, but also a positive one that understands policy-making and politics themselves as evolutionary processes. Failure to do this is likely to result in unrealistic and ineffective policy prescriptions. In much of the evolutionary ­economics literature, and also in EEG, there is an implicit assumption that policy is exogenous and detached from evolutionary socio-economic systems, as if performed by a rational policymaker able to control the system by choosing from a toolbox of policy instruments, overlooking the evolutionary dynamics of diversity, selection and path dependency in policy processes and political systems. Similar to innovations, policies are influenced by pressures and constraints that determine the selection between competing ideas and solutions to problems, leading to the legitimisation of some and the rejection of others (John, 1999; Kerr, 2002). Policy is also subject to path dependencies (Pierson, 2000) resulting from historical legacies, institutional inertia, learning effects and adaptive mutual expectations. Past policy choices shape and condition future policy decisions, potentially constraining the range of options available for policy action (Uyarra, 2010).

The exploration of policy as evolving in path-dependent trajectories has largely been overlooked in EEG (although see, for example, Valdaliso et al., 2014). Analysis of policy impacts over time are often at best exercises in comparative statics, neglecting how ‘policy histories’ shape, co-evolve or adapt to processes of economic and regional structural change. This limits our ability to understand the multiple roles the state plays in new path creation (Dawley et al., 2014; Flanagan et al., 2023). A better understanding of actual policy-making and policy evolution would allow a more incisive understanding of the opportunities for policy action available to regions within multiscalar institutional environments. As Essletzbichler et al. (2023) suggest, the dominant focus in EEG on ­micro-meso ­interactions at the expense of the macro scale has led to incomplete accounts of regional evolution and a tendency to neglect the links between subnational and national policy contexts and actors (see also MacKinnon et al., 2009; Dawley et al., 2014).

Adopting an evolutionary perspective involves a fundamental appraisal of policy action (Metcalfe and Georghiou, 1997), including an acknowledgement of the degrees of freedom that policy actors have to influence and direct the evolution of economies (Lambooy and Boschma, 2001). Attempts at steering or control by policy actors are made still more difficult by the fact that they are themselves part of the system that they are trying to influence (Flanagan and Uyarra, 2016, Weller and Beer, 2022). A key implication is that, when trying to derive policy implications, EEG scholars should not ignore the complexity of governance and politics, but instead grapple with the implications of this complexity and support cautious experimentation and policy learning, recognising the intricacies and uncertainties involved, and foster open discussions about the trade-offs and inherent conflicts among policy objectives, such as environmental sustainability, equity and economic growth.

EEG as a driver for challenge-oriented innovation policy

Evolutionary Economics has played a prominent role in articulating the rationales for government intervention in science, technology and innovation policy (Metcalfe, 1994; Nelson et al., 2018). Recent approaches to industrial policy (Rodrik, 2007; Mazzucato, 2013), mission-oriented policies (Mazzucato, 2018), transformative policy (Schot and Steinmuller, 2018) and policies to tackle societal challenges (Edler and Fagerberg, 2017; Kuhlman and Rip, 2018) have called for a more transformative agency of the state, exercised through mixes of roles and tools, including market shaping interventions. In these policy accounts, however, geographical aspects of economic evolution are often not given full consideration.

This is disappointing considering that the EEG paradigm offers significant insights into how the present geographies of innovation have come about and continue to unfold, and as such provides much-needed explanations regarding the factors that enable or prohibit place-bound configurations of economic activity and resulting benefits. Despite rather implicit contributions to fundamental policy debates, one particular policy arena where EEG actually does excel and has been quite influential is in regional innovation policy-making.

The regional innovation system concept has been adopted in many regions as a framework for the design and implementation of regional innovation policies (Asheim et al., 2011; Coenen et al., 2017). More recently, this policy interest in EEG has been given a boost by Smart Specialisation policy in the European Union (Foray, 2015). This represents a form of place-based innovation policy in which sectoral-technological connectedness along with an emphasis on structural evolution are key elements for regional economic development. Essentially, the core idea is to prioritise new domains of specialisation in regions that complement and leverage their local capabilities and related activities.

While these may be considered steps forward in policy thinking, EEG is still struggling with crucial questions, especially where it concerns old industrial and peripheral regions with weak capabilities and poor institutions (Ferreira et al., 2021) which often find themselves trapped in stagnant or declining industry sectors (Evenhuis, 2016). How to evolve out of such ‘lock-ins’ remains a largely unresolved issue. Economic diversification has been considered as a potential solution for less developed regions to escape constraints, but how the interplay of related and unrelated varieties guided by evolutionary principles actually drives economic growth in regions characterised by weak institutional settings, and thus a lack of policy interventions that potentially initiate structural change, is again something the relevant literature has only began to disentangle (Pylak and Kogler, 2021). One of the main questions in this regard is to what extent can policy actions shape and reshape evolutionary trajectories and initiate significant structural transformations towards more advanced development stages? Place-based capabilities are usually firmly grounded in the history, culture and economic activities of a particular place and thus are difficult to change in the short run without major interventions. Nevertheless, paradigm shifts introduced by significant scientific discoveries or massive investments rooted in collective actions, for example, the recent push by the European Union towards the twin green and digital transition (Muench et al., 2022), might give rise to opportunities to escape path dependency and diversify into novel economic activities based on capabilities that are linked to declining sectors. For instance, green diversification in regions is often embedded in local capabilities, even those based on so-called ‘dirty’ activities, which suggests that some combination of related and unrelated economic diversification is important for green transitions. (Tanner, 2016; Van den Berge et al., 2020; Froy et al., 2022).

Questions have also been raised in relation to the ability of regions to effectively implement place-based policies. According to Foray (2019), the way the entrepreneurial discovery process is organised and implemented will influence the ability of regions to diversify successfully. However, there is still little understanding of how that might work. Despite considerable efforts towards the identification of potential diversification options, there is increasing evidence of implementation challenges, particularly the ability of regional actors to prioritise key sectors or technological domains (Di Cataldo et al., 2022; Gianelle et al., 2020; Marrocu et al., 2023). Institutional and political leaders (Battilana et al., 2009) that trigger new initiatives, mobilise resources, promote collective action, build legitimacy and implement institutional change (Sotarauta and Pulkkinen, 2011; Uyarra and Flanagan, 2022), have been found to play a crucial role. Such agents of change operate in institutional contexts that vary widely across regions, differentiated by local governance cultures (Kroll, 2015) and quality of government (Rodríguez-Pose and Di Cataldo, 2015). A better understanding is needed of who these agents of change are, specifically the role of policy entrepreneurs and implementers, and their degrees of freedom given the context they operate in at various spatial scales. All of that of course places enormous demands on the ability of public institutions at multiple levels to design and implement effective innovation policies (Morgan and Marques, 2019).

There is also a need in EEG to consider broader policy mixes that go beyond traditional supply-side interventions. Societal challenges require policy intervention on many fronts that need coordinated policy action by multiple policy actors, across multiple levels and policy domains. New-generation innovation policy approaches, such as mission oriented (Mazzucato, 2018) and transformative innovation policies, place a greater focus on the demand side, where legitimacy, demand articulation, market formation and institutional change play key roles (Weber and Rohracher, 2012; Boon and Edler, 2018). This could supplement the supply-driven policy approach (with a prime focus on local capabilities) that EEG scholars often promote in the context of Smart Specialisation policy (Flanagan et al., 2023). A main focus of recent approaches is on interactions and interdependencies between different policy domains, as they affect the feasibility and effectiveness of policy objectives. Evolutionary scholars have been active in the policy mix literature, especially in policies on innovation and sustainable transition (for example, Flanagan et al., 2011; Kivimaa and Kern, 2016; Rogge and Reichardt, 2016, Matti et al., 2017) as well as the ‘Geography of Transitions’ literature (for example, Hansen and Coenen, 2015), but further steps are needed for it to provide (more) useful holistic insights to address current predicaments (Magro and Wilson, 2019).

EEG and the problem of unsustainable development

Interesting questions arise when we consider EEG in relation to the issue of sustainability. By many accounts, sustainability problems are endogenous to a model of capitalism that has relied, at the micro-level, on the governance of firms based on a shareholder value maximisation approach (Friedman, 1970; Stiglitz, 2019), which has attributed primacy of shareholder profits over the value generated for society at large. This approach has had negative consequences for the extent to which companies are able to contribute to equitable forms of economic development.

On the one hand, it has led companies, especially large multinational enterprises (MNEs), to use business models that have proved successful because they could free ride on environmental or social costs (Henderson, 2020), especially through operations in poorly regulated developing countries (Giuliani and Macchi, 2014). At a macro level, this laissez faire model was expected to create societal well-being based on the idea that companies would promote economic growth and would also contribute to fix their negative externalities through tax payments. However, in reality, this expectation has not been met. Economists have recently documented that some 36% of MNE’s profits are shifted to tax havens globally (Tørsløv et al., 2023), suggesting that instead of seeking to minimise their harmful impacts on society and the natural environment, for example, through proper payment of local taxes, corporate governance decisions are more concerned with protecting profits and shareholder value, and companies engage in a raft of strategies so as to geographically distribute their profits in ways that minimise their tax liabilities (see, for example, Pistor, 2019). As a result, valuable resources are lost which would otherwise contribute to expenditure on local public services.

Against this background, questions arise as to whether EEG has so far developed a sufficient theoretical apparatus to tackle or understand this corporate behaviour. How is the ‘evolutionary’ approach, using notions such as path dependency, relatedness, complexity or adaptability, fit to explain the dark side of corporate decision making? Do we need other theories or other theoretical constructs in combination with the conventional ones to explain these decisions?

To start answering these questions, we should first conceptualise as ‘deviant’ (Earle et al., 2010) organisational behaviour that deviates from established social norms or legal prescriptions, oftentimes reflected in decisions that privilege corporate or individual gains over societal well-being and that result in significant collective or individual harm—for example, environmental depletion, child labour, slavery, exploitation of workers and discrimination, and other forms of human rights harm. Second, we should acknowledge that deviant behaviours are not events simply resulting from aberrant corporate CEOs or intentionally dishonest decision makers: we know that good firms do bad things (Mishina et al., 2010) and that ‘deviant’ organisational conducts may be more the norm than the exception (Palmer, 2012), depending on several conditions, including the institutional contexts where firms operate (Giuliani et al., 2023). Third, bad decisions at the micro-level influence the development path of industrial clusters, regions and other ecosystems (Giuliani, 2016), which can thrive in certain domains (for example, innovation, product complexity, growth and job opportunities) while failing in others. Silicon Valley, economic geographers’ most acclaimed innovation hot spot (Saxenian, 1994), is now blamed for being a powerhouse of inequality and social injustice; its high-tech products are questioned for the unsustainable sourcing of raw materials (typically critical minerals) in developing countries, while also criticised for their ­negative societal impacts (Feldman et al., 2021). Lastly, micro-level ‘deviant’ decisions are dynamically related to macro-level structures such as the form (‘variety’) of capitalism prevailing in a country (Hall and Soskice, 2001), as well as to the type of institutions that are set up to prevent or punish undesirable behaviours by powerful companies or elite groups (Acemoglu and Robinson, 2012).

By considering these elements as key pillars of a new understanding of economic spaces, we call for a reconsideration of the EEG theoretical apparatus to ensure that it is fit to predict these negative effects and integrate them with the conventional—and generally positive—view that companies play to foster development processes within and across geographical space. Behavioural theory, if not some of its constructs, are particularly helpful to predict bad decision making. While EEG is akin to behavioural theory in considering organisations and their managers as boundedly rational (Simon, 1955), the uncertainty of the external environment leads decision makers to proceed via heuristics and trial-and-error, and that companies operate in their business context through the search of a satisficing rather than an optimal goal (Cyert and March, 1963). It has generally used these conceptual pillars to predict dominant patterns (such as path dependency, development through related vs. unrelated variety, increasing compositional complexity) on an assumption of legitimate behaviours, including innovation trajectories, location choices, investment choices, etc. Yet, as mentioned above, illegitimate or deviant choices are non-negligible and largely overlooked in EEG literature. Their explanation requires considering insights from other scientific domains, focussing on micro-level behaviours, such as behavioural psychology, as well as on disciplines taking a more macro-level stance focussing on varieties of capitalism and institutions.

As concerns the former, recognising ethical judgements in decision making is crucial. Scholars investigating the behavioural psychology of decision making suggest that managers are susceptible to moral or ethical dissonance (Kelman and Baron, 1974), which means that a clash may arise between the aspiration to benefit from unethical behaviour and the need to maintain a positive moral self-image (Nieri et al., 2023). Dissonance theory (Festinger, 1962) suggests that individuals have a natural psychological need for consonance and consistency, which implies that ethical dissonance is an uncomfortable state which individuals try to lessen. This is generally done by either reducing the behaviour—hence moderating action in order to reduce the sensation of discomfort—or by continuing the behaviour while activating a psychological process of self-justification or self-denial of the harm (Lowell, 2012) which leads to the perpetuation of the activity and of its normalisation through time (Sykes and Matza, 1957). The normalisation of deviant behaviours is one of the reasons why certain practices are persistent and not sporadic, and the explanation of normalisation processes is also not just based on individual decision making but it rather has a group-thinking and collective dimension which reinforces normalisation processes by making ‘deviant’ or highly risky behaviour to appear as socially acceptable and, therefore, legitimate (Janis, 1972; Vaughan, 1996). This is highly relevant for regional or other sub-regional economies (for example, clusters) as some local normalisation dynamics can negatively development processes (Giuliani, 2016). A clear example of ‘deviant’ behaviour is surely the activities of banks in the years leading to the global financial crisis, when they dramatically expanded the issuance of high-risk mortgage debt to ‘subprime’ households, bundled that debt into complex securities traded on global markets, then increased mortgage interest rates, resulting in repayment defaults, and the potential collapse of the banking system. Sheer avarice and greed on the part of banks and bankers became ‘normalised’, until the boom burst. The recession that followed impacted highly uneven geographically, as did the decade of fiscal austerity imposed by states to recoup the cost of bailing out the banks. More than three decades of financial deregulation facilitated this reckless ‘market’ behaviour by banks.

Augmenting EEG with elements of behavioural psychology seems one of the ways through which one can better understand such events, and why some places are more likely than others to face unsustainable development paths. Recent EEG research has already started to point out how personality traits can influence urban growth, as some traits are more likely associated with entrepreneurial initiative (Garretsen et al., 2019). However, there is still room to expand this perspective to explain unsustainable development processes rather than growth. Moreover, at the organisational level, scholars have used performance feedback as a behavioural mechanism to explain why some firms are more willing to take risky decisions leading to harmful business conducts (Harris and Bromiley, 2007; Giuliani et al., 2023), while neo-institutional theory has been widely used to explain MNEs’ acceptance of certain deviant practices as a survival strategy, especially when they are the norm in countries with lower regulatory standards (Surroca et al., 2013). While a full overview of the theoretical approaches to predict companies’ socially and environmentally unsustainable business practices is beyond the scope of this editorial, we suggest that EEG could benefit from refining its theoretical toolbox in order to be able to improve our understanding of the dynamics of unsustainable decision making.

Likewise, there is scope for EEG to extend this area of research by considering how varieties of capitalism or, even more specifically, varieties of neoliberalism (Birch and Mykhnenko, 2009) shape (and are shaped by) the proclivity of companies to deviate from norms. Earlier research in business ethics has suggested that national business contexts characterised by shareholder pre-eminence, ­typically the most advanced forms of liberal market economies, create a favourable environment for prioritising profits over societal well-being (Matten and Moon, 2008). However, coordinated market economies or different varieties of state capitalism (Musacchio et al., 2015) may also experience pervasive misconduct owing to the opacity of their institutions and to the political ties between economic and political spheres favouring favouritism, lobbying and. Hence, more research examining how contemporary varieties of capitalism, and their related dominant ideologies, influence the evolution of (un)sustainable development patterns at regional and other geographical spaces appears to be both timely and relevant.

An expanded ontology for re-imagining EEG

The expansion of the theoretical remit of EEG along the lines discussed in the previous sections, to make it more suitable to address the key transformations of today, requires a reconsideration of the paradigm’s ontological foundations. The use of ontology here is given to mean a particular view and interpretation of the world, or of the system of interest, its main components, processes and features, how these are structured and organised, how they function, and what, therefore, requires explanation. Every research paradigm or research programme is based on an ontology of some sort, even if that ontology is rarely acknowledged or explicitly specified. There are, no doubt, some evolutionary economic geographers who will argue that discussions of ontology are bound to be inconclusive and hence it is better to just focus on doing empirical research—a sort of position that EEG is simply what evolutionary economic geographers say they are doing. This is not an altogether satisfactory response. The topics for—and approach to—empirical research will be guided by an implicit view of how the world works. Moreover, findings from empirical research only take on significance in light of a broader interpretation of what is going on in the world.

The prevailing ontology in EEG emphasises activities and processes at the micro-level (innovation, entrepreneurship and firm growth), which are conditioned by available technologies, skills and knowledges. These then give rise to—and are in turn impacted by—meso-level patterns, such as the emergence or stagnation of industries or clusters in a certain region, or the long-term performance or resilience of places. We need to expand this ontology in a number of ways, so it provides a more appropriate basis for theorising current trends and challenges shaping the economic landscape. We do not offer a complete specification of this expanded ontology, but we do want to outline three aspects of such an expanded ontology.

First, the ontology should encompass a broad view of endogenous development within regions. The evolution and transformation of a particular region is determined by changes happening in many interrelated domains, not only in the technological and industrial base, but also in other aspects of the regional economy, such as the labour market, institutional arrangements, and the domain of policy-making. Table 2 offers various examples of processes and mechanisms of evolution in different domains of a local economy. Over the past decade or so, this agenda has indeed been taken up within EEG, through a stream of articles that have extended the concepts of path development and path creation in regional economic evolution (see for example, Hassink et al., 2019; MacKinnon et al., 2019; Grillitsch and Sotarauta, 2020). These articles foreground the different mechanisms of path dependency (and lock-in) at play within and across various domains and connect these to how the economy of a region evolves (see Martin and Sunley, 2006; Martin, 2010).

Several articles within this themed issue take this agenda forward. Frenken et al. (2023) synthesise various insights about the development of technologies, industries and institutions into an encompassing framework for understanding regional economic development. The articles by Elekes et al. (2023), Henning and Kekezi (2023) and Benner (2023), look at the interactions between labour market dynamics and the characteristics and evolution of other aspects of the regional economy (industrial structure, institutions and policies), Moreover, these articles explicitly assess the outcomes in terms of addressing inequalities (that is, inclusion and upward mobility within the labour market). Menzel (2023) sheds additional light on the process of market construction, and how this impacts on the evolution of industries. The article by Vela Almeida and Karlsen (2023) shows that in some contexts the informal economy and unpaid labour will also be relevant for the evolution of the industries in a place. The contribution by Froy (2023) makes the case to also include the role of the spatial configuration of the (changing) built environment into the conceptualisation of the economic evolution of regions and cities, a hitherto neglected topic. The commentary by Feldman (2023) suggests that EEG can draw on the work on Entrepreneurial Ecosystems to enrich its conception of—and approach to—endogenous economic development within places.

Second, the evolution of a particular region should be considered in relation to a much broader space that encompasses actors, structures, flows and processes at ­various levels of scale beyond the place in question (Kedron et al., 2020). That is, the prevailing focus in conceptualising the evolution and transformation of the spatial economy, as mainly an interplay between activities, processes and factors at the micro-level, with patterns at the meso-level (that is, the evolution of a regional economy and its various components), should be complemented by a consideration of the relations with other places and the dynamics at the macro-level. EEG has indeed made strides in considering how the evolution of the economy within places is influenced by the structures and relations in which a region is embedded (for example, Kogler et al., 2023; Boschma and Capone, 2015; Boschma, 2017; Trippl et al, 2018; Boschma, 2022, Rocchetta et al., 2022; Frenken et al., this issue). The article by Shin et al. (2023) adds valuable new insights into the knowledge flows between locations and the impacts these have on the evolution of technological capabilities within a region via the ‘innovative footprint’ of multi-location firms, and then how this contributes to the evolutionary diversification trajectories in their respective home regions. The article by Morrison (2023) directs attention to the importance of flows of high-skilled migrants between places, for the knowledge, competences and skills available across regions, which will then have repercussions for innovation and the evolution of local industries.

Nevertheless, more is required in order to adequately conceptualise the macro-level patterns, trends, processes, structures and upheavals in the global economy, as the broader space in which regional economies operate, and which in turn are also shaped by what happens within regions. Moreover, as noted earlier, also the state, arrangements of governance and the variegated ways in which capitalism is organised, should be accorded a place in the extended ontology of EEG, as these cannot be seen as separate from and outside of the economic sphere, and are themselves subject to evolutionary processes and mechanisms. Several calls and programmatic statements have already suggested this (see MacKinnon et al., 2009; Essletzbichler, 2012; Martin and Sunley, 2015).

Essletzbichler et al. (2023) restate and augment these earlier statements. Moreover, they show how patterns, flows and processes within the global economy can be explicitly integrated into the EEG paradigm, which then leads to different insights into the drivers of economic development in regions and cities. Also the contribution by Bieri (2023) takes the conceptualisation of the nature of the capitalist spatial economy and its evolution, in new directions. He builds on the work of Joseph Schumpeter to incorporate money and finance into an evolutionary and geographically sensitive framework, which points to promising new avenues for research within EEG given the financialized nature of contemporary capitalism.

Third, the ontology of re-imagined EEG needs a broad view of the processes of evolution and transformation within the economic landscape. That is, how the various actors, assets, structures, flows, patterns, etc. shape the evolution of different aspects of the spatial economy, at different scales, and with different temporalities; and how these processes are also interconnected. As outlined above, a defining characteristic is that EEG is about understanding the dynamics of the spatial economy. History, novelty, adaptability, complexity and emergence are key building blocks in this respect. Furthermore, as also discussed, there is considerable promise in enriching the micro-foundations of EEG by drawing on insights from behavioural psychology, to account for the ways in which actors act as they do, giving rise to the persistence of certain patterns or to unintended outcomes.

The commentaries by Metcalfe (2023) and Dopfer (2023) outline how evolutionary economists view the processes of evolution and transformation in the economy, and how this view can be extended by evolutionary economic geographers. In their article Chu and Hassink (2023) on the other hand point to the importance of rooting the theorisation of such dynamics within EEG, much more strongly and more explicitly within geography. They argue that the spatial aspects of these dynamics should be brought to the fore, and not be treated as an afterthought.

In the past decades there have been important advances in EEG in understanding and researching the ‘interdependence’ and ‘co-evolution’ between different processes in various domains and/or at different scales of the spatial economy (Schamp, 2010; Pike et al., 2016; Evenhuis, 2017; Gong and Hassink, 2019; Frangenheim et al., 2020; Benner, 2022; Sotarauta and Grillitsch, 2023). All of the articles in this themed issue contain, to some degree or another, new insights on exactly this issue.

Concluding thoughts

To make EEG better suited to theorise and interrogate how the economic landscape is being reshaped from within, as well as from without, in light of a number of challenges, crises and transformations, a number of adjustments are needed in the paradigm. These adjustments concern the conception of the state, and the role of policy to effectuate and direct the innovations and changes that are needed to cope with the challenges and transformations we are facing. Furthermore, also our theorisation of the behavioural foundations needs to be expanded, to help understand why actors act in the ways they do, including when such actions are harmful for individual and collective well-being, and thus why it may be difficult to achieve the outcomes we desire. Moreover, these possible advancements in the body of theory of EEG will require that we expand and refine the ontological framework on which EEG rests as a paradigm.

Underlying this call for a re-imagining of EEG is an unequivocally normative premise, rooted in meliorism: a belief that, like any other approach in the social sciences, issues of social justice, social inclusion and sustainability should underpin the questions we ask and the explanations we construct (Martin, 2021). That is to say, there should be an explicit axiological dimension to EEG, a ­normative position that it should not only be a body of knowledge for understanding economic growth and development within and across regions, but founded on a concern to explain why and when the evolutionary processes it studies generate socially and spatially divisive outcomes, and thence what policies are needed to avoid or correct such inequalities. If ever such a normative disposition is required it is surely now, a time of historic turmoil and transformation.

Acknowledgements

The authors are grateful to three referees for their comments on an earlier version of this Editorial. This themed issue on Evolutionary Economic Geography greatly benefited from the presentations and vivid discussions that took place at the dedicated workshop at University College Dublin, June 2023. In this regard, Dieter F. Kogler would like to acknowledge financial support from the European Research Council under the European Union's Horizon 2020 Research and Innovation Programme [grant agreement number 715631, ERC TechEvo]. The usual disclaimer applies.

References

Footnotes