Abstract

In this paper, we review the different definitions of resilience and their potential application in explaining the long-term development of urban and regional economies. We reject equilibrist versions of resilience and argue instead that we should seek an understanding of the concept from an evolutionary perspective. After discussing a number of such perspectives, we focus on the adaptive cycle model from panarchy theory to generate testable hypotheses concerning urban and regional resilience. Two case study city-regional economies are used to explore this model. We conclude that the evolutionary adaptive cycle model, though not without problems, warrants further study as a framework for analysing regional economic resilience.

Introduction

The recent interest in resilience has emerged in part at least as a reaction to specific extraordinary events and shocks that have prompted very particular types of public policy response. Indeed, as Vale and Campanella (2005) and Stehr (2006) point out, the perceived source of a disaster can have a material impact on the forms and mechanisms of resilience, including, e.g., the scale and scope of state response; rapid catastrophic disasters pose very different policy and organisational challenges from slow-paced and cumulative stresses (Foster, 2007). Somewhat similarly, regional and local economic development is far from a smooth and incremental process but is subject to all sorts of interruptions and disruptions: periodic economic recession, the unpredictable rise of major competitors elsewhere, unexpected plant closures, the challenges arising from technological change and the like. How regional and local economies respond and adjust to such disturbances and disruptions may well exert a formative influence on how they develop and evolve. The notion of ‘resilience’ would thus seem to be highly relevant to understanding the process and patterns of uneven regional development.

Of course this begs the question of what precisely we mean by regional and local economic resilience. Partly as a result of its comparatively recent emergence as an analytical concept there is, as yet, no universally agreed definition of resilience in economics or social science, let alone in regional or urban studies. In fact, some even contest the value of the notion of resilience altogether (Hanley, 1998). Some initial studies have recently appeared that attempt to outline how the idea of resilience might be defined in economics and in regional studies (e.g. Rose and Liao, 2005; Briguglio et al., 2006; Foster, 2007; Pendall et al., 2008; Hill et al., 2008; Ormerod, 2008), but this task is still far from complete. Our focus in this paper, therefore, is to examine how the notion might be given meaning in a regional or local economic context. We have problems with what we might call ‘equilibrist’ definitions that restrict the idea of resilience, explicitly or implicitly, to the ability of a system—in our case a regional or local economy—either to return to a pre-existing stable or equilibrium state or to move quickly to a new one. While not denying the value and possible relevance of such an equilibrist interpretation, our interest is much more in how far and in what ways resilience can function as an evolutionary concept. More specifically, we are interested in the idea of resilience as ‘adaptive ability’ since it is the differential ability of a region's or locality's firms to adapt to changes and shocks in competitive, market, technological, policy and related conditions that shape the evolutionary dynamics and trajectories of that regional or local economy over time.

Thinking about regional economic resilience

According to its strict Latin root, resilire, to leap back or to rebound, the idea of resilience refers to the ability of an entity or system to ‘recover form and position elastically’ following a disturbance or disruption of some kind. Most uses of the term in regional or urban applications refer to this idea of the ability of a local socio-economic system to recover from a shock or disruption. Thus, Foster (2007, 14) defines “regional resilience as the ability of a region to anticipate, prepare for, respond to, and recover from a disturbance”. Or again, Hill et al. (2008, 4) see resilience as “the ability of a region … to recover successfully from shocks to its economy that either throw it off its growth path or have the potential to throw it off its growth path”.

But beyond these rather broad statements, there is much ambiguity. For one thing, should the notion refer not just to a regional economy's ability to recover from a shock but also to the degree of resistance to that shock in the first place? After all, a regional economy that is hardly affected by a shock is much more likely to recover, and more quickly, than a regional economy that is severely weakened or disrupted by the shock. That is to say, should resilience also refer to the sensitivity or vulnerability of a regional economy to shocks? For another thing, there is the issue of whether the concept refers to the ability of a regional or urban economy to retain its structure and function despite the shock or disturbance to it or to the ability of a region or urban system to change its structure and function rapidly and successfully in response to a shock. Often, the two senses are combined, as in Hill et al. (2008, 4) where it is suggested that the resilience of a regional socio-economy refers to “the extent that its social structure of accumulation was stable or … the extent that it was able to make a rapid transition from one social structure of accumulation to another”. Then again, another issue is that the resilience of a region's economy is unlikely to be invariant over time: it may depend on the nature of the shock and may change over time as the structure and nature of a region's economy evolves.

The ambiguity that surrounds the concept of regional economic resilience is compounded by the fact that two definitions of the notion can be found in the ecological literature, where the idea has perhaps been most debated. The first and more traditional definition, so-called ‘engineering resilience’, concentrates on the stability of a system near an equilibrium or steady state, where resistance to disturbance and the speed of return to the pre-existing equilibrium are used to define the idea of resilience (e.g. Holling, 1973; Pimm, 1984). This seems closest to the notion of ‘elasticity’ or the ability of a system to absorb and accommodate perturbation without experiencing major structural transformation or collapse (McGlade et al., 2006). Under this definition, therefore, regional economic resilience would imply the retention of the region's pre-shock structure and function. The problem is that view carries with it the baggage of equilibrist thinking.

Indeed, the notion of engineering resilience bears a close affinity with the standard use of equilibrium in mainstream economics. In this instance, a shock or disturbance moves an economy off its equilibrium growth path, but the assumption is that self-correcting forces and adjustments eventually bring it back onto that path (see Figure 1(a)). So the relative resilience of different economies would be measured in terms of their susceptibility to being moved off their equilibrium paths (their ‘sensitivity’ to shocks) and their response times of recovery to equilibrium. The obvious problem with this definition is that if regional economic resilience is defined in terms of the ability of a regional economy to retain (return to) its equilibrium form and function following a major shock, it becomes difficult to reconcile the notion of resilience with the idea of regional economic evolution. The implication is that the more resilient is a regional economy, the less it would change over time, even in the face of various shocks. So, at best, this view of resilience would yield an evolutionary model based on the maintenance of structure and stability.

Figure 1.

Stylised Responses of a Regional Economy to a Major Shock.

Notes: (a) Return of region to its pre-existing steady growth path following the shock; (b) and (c) region fails to resume former steady growth path after the shock, but settles on inferior path (d) region recovers from shock and assumes an improved growth path.

Figure 1.

Stylised Responses of a Regional Economy to a Major Shock.

Notes: (a) Return of region to its pre-existing steady growth path following the shock; (b) and (c) region fails to resume former steady growth path after the shock, but settles on inferior path (d) region recovers from shock and assumes an improved growth path.

The second definition, so-called ‘ecological resilience’, focuses on whether disturbances and shocks cause a system to move into another regime of behaviour. In this case, resilience refers to the magnitude of the shock of disturbance that can be absorbed before the system changes its structure and function and becomes shaped by a different set of processes (Holling, 1973). According to some authors, this definition opens up space for linking resilience with the idea of adaptability and is thus much richer in evolutionary scope (McGlade et al., 2006). But the notion of ecological resilience is not unproblematic. If ecological resilience is measured by the magnitude of the disturbance that can be absorbed before the system changes structure, this seems to imply that the bigger the shock required to change a system's structure and function, the more resilient that system would be deemed to be. So, again, a resilient regional economy would be one capable of absorbing and accommodating extreme shocks without any significant change to its form or function: we seem to be back to engineering resilience. It is difficult to see how this view of resilience imbues the construct with much evolutionary content.

If, on the other hand, resilience is interpreted in terms of how well a system adapts its structure and function in response to shocks, then, potentially, this does indeed open up much more scope for evolutionary analysis. But even here caution is needed. McGlade et al. (2006) argue that resilience “is concerned with the role of instabilities in pushing a system beyond a threshold or bifurcation point to a new stability domain” (149). The implication seems to be that the evolutionary dynamic is periodic in nature, in which episodic shocks cause a system to adapt from one ‘regime of stability’ to another. It is but a small step from this conception to the notion of multiple equilibria used in economics, the idea that there is no single unique equilibrium state or path of an economy but several possible states or paths and that an economy can be shifted from one such equilibrium to another by a shock. A resilient regional economy would then presumably be one that adapts successfully and either resumes, or better still improves, its long-run equilibrium growth path (as in Figure 1(d)). A non-resilient regional economy would presumably be one that fails to transform itself successfully and instead becomes ‘locked’ into an outmoded or obsolete structure, with a consequential lowering of its long-run equilibrium growth path (Figure 1(c)).

Under this view of resilience, the argument would seem to be that regional economic evolution follows a process of ‘punctuated equilibrium’, a succession of stable forms or steady growth paths the hysteretic movement between which is triggered by periodic shocks or major perturbations. But although the notion of punctuated equilibrium is widely used in discussions of biological and ecological evolution, is it a valid or sufficient model of how economic systems—national, regional or local—evolve? Indeed, are urban and regional economies ever in any equilibrium? While ecological systems if left undisturbed may, perhaps, attain equilibrium or stable states, economic systems are arguably different. Economic evolution depends on the actions of individual economic agents, who can learn, innovate and adjust their behaviour. For this reason, some evolutionary economists would argue that economies can never be in equilibrium. As Ramlogan and Metcalfe (2006) point out, economies are based on and driven by knowledge, and knowledge never stands still but constantly changes; thus economies can never stand still—capitalism is inherently a restless process.

In fact, appeal to multiple equilibria would not seem to be essential for developing the notion of regional economic resilience. Certainly, regional and urban economies exhibit stability and self-organisation. But, in line with the argument of Ramlogan and Metcalfe (op cit), stability and self-organisation are not the same as equilibrium (see also Martin and Sunley, 2006, 2007; Martin, 2010). To our mind, from an evolutionary perspective, the important attribute of regional economic resilience is the adaptive capacity of a local economy. What matters for the long-run success of a regional economy is the ability of the region's industrial, technological, labour force and institutional structures to adapt to the changing competitive, technological and market pressures and opportunities that confront its firms and workforce. What we have in mind is akin to what Schumpeter referred to as the notion of industrial ‘mutation’ that takes place via a process of ‘creative destruction’. And as he emphasised, industrial mutation (creative destruction) can occur both more or less ‘incessantly’ and also in rushes (gales of creative destruction). How regional economies adapt to major shocks thus remains of central importance; but how regional economies respond to major shocks, such as deep recessions, may itself be the product of a slower, more cumulative process of adaptation or ‘resilience building’. Put another way, any convincing theory of regional economic resilience must explain how a regional economy's resilience evolves as well as how its resilience impacts back on that economy's evolution.

Recently, some resilience theorists have begun to move closer in this direction of an evolutionary perspective and to consider the nature of constantly changing non-equilibrium systems (see Carpenter et al., 2005). Here resilience is considered as an ongoing process rather than a recovery to a (pre-existing or new) stable equilibrium state. This shifts the theoretical analysis from questions about how a system such as an economy is resilient to how it adapts through time to various kinds of stress. Such an approach does not require any necessary assumptions about equilibria and instead seeks to understand constant change rather than stability (Pendall et al., 2008, 127).

Regional economic resilience as an evolutionary process: in search of a conceptual framework

Linking the notion of regional economic resilience to that of adaptation, and setting the analysis within an evolutionary perspective, opens up a wide range of possible approaches. At least four conceptual frameworks for constructing an evolutionary account of regional economic resilience and adaptation can be distinguished: Generalised Darwinism (which emphasises variety, novelty and selection); path dependence theory (which focuses on historical continuity, ‘lock-in’ and new path creation); complexity theory (which highlights self-organisation, bifurcations and adaptive growth) and panarchy (which explicitly links resilience and ‘adaptive cycles’). Though distinctive, there are certain points of overlap between these frameworks.

An approach that draws on Generalised Darwinism would put particular stress on the role of variety in shaping regional economic resilience, e.g. in terms of structural (sectoral) variety and variation in firm behaviour, including the differential adaptability of local firms. Adaptability is about the potential to adjust to changing circumstances in an appropriate way. As Toulmin (1981) points out, there are three basic mechanisms by which an entity, such as a local firm, can change to become better adapted. One is the intentional response to the perception of circumstances; a second is homeostatic, the automatic following of specific rules in relation to target behaviours and a third is developmental, the cumulative unfolding of new behaviour patterns (such as innovation) within a specific set of constraints.

Variety might be expected to influence regional economic resilience and adaptability in several ways. In regional economics, the degree of local sectoral variety (diversity) is often claimed to influence the vulnerability of a regional economy to exogenous shocks, with regions having a more diversified economic structure being less prone to shocks, or at least more able to recover from them, than more economically specialised regions, which are not only prone to idiosyncratic sector-specific shocks but lack the breadth of economic activity to offset such adverse disturbances. Sectoral variety is also alleged to influence innovative activity among local firms, though opinion is divided on whether a diversified industrial structure is more conducive to innovation than a specialised structure (the ongoing debate over the Jacobsian versus Marshall–Arrow–Romer theories). In its turn, innovation is key to the generation of local economic variety. The importance of variety for understanding regional economic resilience is that the existence of variation of types of firm and of firm behaviour in a region operates in conjunction with the mechanisms of firm selection (the competitive survival or failure of firms) to impart adaptation of the regional ensemble of firms as a whole over time.

How the notion of path dependence—itself problematic (see Martin and Sunley, 2006; Martin, 2010)—bears on the question of regional resilience and adaptation is open to different interpretations. One of the key features of standard path dependence theory is the notion of lock-in, the process whereby an economy—say a regional economy—becomes ‘locked into’ in a particular trajectory of economic development through the operation of self-reinforcing localised increasing returns effects. In the canonical account of this model (e.g. David, 2005, 2007), the assumption is that an external shock is required to ‘de-lock’ the economy from this path (in David's work, such a path is assumed to be one of a number of possible multiple equilibria). So one interpretation might be that a regional economy is resilient if it is able to maintain its ‘locked-in’ development path even when disturbed by an external shock of some kind: lock-in is thus seen as a positive attribute of a regional economy. This is akin to the notion of engineering resilience discussed above. A different interpretation, however, would be to regard lock-in as a negative attribute, as holding back the adaptation of the regional economy to a shock. The implication in this instance is that path-dependent lock-in undermines a regional economy's resilience. Whether a regional economy suffers from ‘negative lock-in’ thus becomes an ex post matter, evident once a shock has occurred (see also Setterfield, 1997).

There is a third way in which path dependence might be useful in conceptualising regional economic resilience and one that has a closer bearing on the issue of adaptation. Standard path dependence theory has relatively little to say about where paths of regional economic development come from, how they arise. Indeed, the typical assumption is that such paths originate in random, happenstance events—hence, the frequent argument that path dependence is concerned with how random or adventitious events can shape the course of history. But there is both good empirical evidence and strong conceptual grounds for arguing that new paths are often shaped by old paths (see Martin and Sunley, 2006; Simmie et al., 2008; Martin, 2010). The emergence of a new local industry may not be due to ‘chance’ or ‘historical accident’ but stimulated or enabled—at least in part—by the pre-existing resources, competences, skills and experiences inherited from previous local paths and patterns of economic development. These inherited conditions shape the environment in which purposive or intentional experimentation and competition occur among local agents (or shapes its attractiveness to agents from elsewhere). Contrariwise, in other places—precisely for reasons arising from the specifics of their past economic development—the local environment may be less conducive to, perhaps even a ‘constraining’ force on, the emergence of new technologies and industries. This might arise, e.g., because the specific inherited knowledges and resources are not easily recombined or converted into new competences (Maskell and Malmberg, 2007) or because of their previous success, existing industries have bid up local land rents, prices and wages to levels that deter new entrepreneurial activity (Brezis and Krugman, 1997). Path dependence may thus act to enable or constrain regional economic adaptation in response to a shock.

While these ideas from Generalised Darwinism and path dependence theory certainly provide some scope for thinking about regional economic resilience as adaptation, they by no means exhaust the range of possible approaches. In fact, the logical step is to turn to those theoretical frameworks that attempt to deal explicitly with the evolutionary dynamics of complex adaptive systems. There are strong grounds for arguing that regional economies represent complex adaptive systems with emergent patterns of behaviour and organisation. Complex adaptive systems are characterised by several key features (Martin and Sunley, 2007). Typically such systems have functions and relationships that are distributed across system components at a whole variety of scales, giving the system a degree of connectivity. The boundary between a complex adaptive system and its environment is neither fixed nor easy to identify, making operational closure difficult, and the system subject to constant exchange with its environment. Complex adaptive systems are characterised by non-linear dynamics because of complex feedbacks and self-reinforcing interactions among components, with the result that they are often characterised by path dependence. They are also characterised by emergence and self-organisation: i.e. there is a tendency for macroscale structures and dynamics to emerge spontaneously out of microscale behaviours and interactions of system components. And this same process of self-organisation imbues complex systems with the potential to adapt their structures and dynamics, whether in response to changes in the external environment (e.g. external shocks), or from within through co-evolutionary mechanisms or is response to ‘self-organised criticality’.

Certain implications are claimed to follow from these features. Of particular relevance to the idea of regional economic resilience is the argument that complex adaptive systems are characterised by two conflicting tendencies: on the one hand, there is tendency in such systems towards increasing connectedness and order (or interrelatedness) among system components; but, on the other hand, increasing connectedness and order tend to reduce the adaptability of the system to changes in environmental conditions. This implies that there is a trade-off or conflict between connectedness and resilience: the more internally connected is a system, the more structurally and functionally rigid and less adaptive it is.

The ecological model of ‘adaptive cycles’ seeks to reconcile this contradiction or conflict through the idea of ‘panarchy’, a model that posits a four-phase process of continual adjustment in ecological, social and environmental systems. Each phase is characterised by varying levels of three dimensions of change: (i) the potential of accumulated resources available to the system; (ii) the internal connectedness of system actors or components and (iii) resilience, a measure of system vulnerability to shocks disturbances and stresses, with high resilience associated with phases of creative and flexible response (Petersen, 2000; Holling and Gunderson, 2002; Pendall et al., 2008). Translated into a regional economic context, the accumulated resources would include the competences of individual firms, the skills of local workers, institutional forms and arrangements, physical and soft infrastructures (such as business and work cultures) and the like: these of course would depend on previous forms and structures of economic and social development in the region. Internal connectedness would relate to patterns of traded and untraded interdependencies among local firms, including supply inputs, horizontal interfirm divisions of labour in production, local networks of trust, knowledge spillover, formal and informal business associations, interfirm patterns of labour mobility and so on. These likewise would be shaped by previous economic developments in the region. Creative and flexible response would depend on the innovative capacity of local firms, on entrepreneurial capabilities and new firm formation, institutional innovation, access to investment and venture capital, willingness of workers to reskill and similar factors.

The adaptive cycle model applied to a regional economy might take the form shown in Figure 2. The cycle has two loops: one relating to the emergence, development and stabilisation of a particular economic structure and growth path (exploitation to conservation) and another relating to the eventual rigidification and decline of that structure and growth path and the opening up of new potential types of activity and growth sources for exploitation (release and reorganisation). Pendall et al. (2008) describe the movement between these phases in a regional economy in the following way. In the exploitation phase, regional growth develops and productive, human and knowledge capital are accumulated as new local industries exploit comparative advantages and various external economies of localisation. But as this growth continues, the connectedness between the various components of the regional economy increases, and the pattern of development becomes increasingly rigid: its resilience to potential shocks declines (Figure 3). Thus if such a shock occurs, structural decline and loss of growth momentum are likely to follow. Firms close or move out of the region, the degree of connectedness declines and agglomeration of localisation economies lose their impact. Old patterns of production and institutional form unravel and resources are released. This opens up the possibility for a second release–reorganisation loop, characterised by innovation, experimentation and restructuring, as new types of activity begin to emerge. Connectedness is low, the potential for the creation of new paths high, the trajectories of development open, and thus resilience high. As the particular forms of new activity and new technologies are exploited, new comparative advantages develop and a new round of regional growth and accumulation is set in motion.

Figure 2.

A four-phase adaptive cycle model of regional economic resilience.

Source: Adapted from Holling and Gunderson (2002) and Pendall et al. (2008).

Figure 2.

A four-phase adaptive cycle model of regional economic resilience.

Source: Adapted from Holling and Gunderson (2002) and Pendall et al. (2008).

Figure 3.

Resilience as a process: variations in resilience across the adaptive cycle.

Figure 3.

Resilience as a process: variations in resilience across the adaptive cycle.

A further, interesting feature of complex systems under this panarchy model is linkages across scales—across both time and space—through two principal mechanisms. Larger scales affect smaller ones (akin to downward causation or ‘supervenience’ in some theories of emergence in complex systems), with longer term, region-wide processes and features (such as local government, the region's educational system, physical infrastructures, etc.) shaping if not determining interactions and outcomes at smaller scales (i.e. at the firm level). But smaller scales also act back on larger scales, particularly during the release phase. The resilience of a regional economy thus depends both on the longer term, region-wide processes and on shorter term more microscale processes and on how these interact. For example, a failure of a region's workforce to upgrade its educational attainment and skill levels may hinder the local emergence of new firms orientated to new technologies and new types of worker. In addition, conditions in the national economy can affect those in regional economies. A national recession can adversely affect conditions in different regional economies. Or again, national policies to promote entrepreneurship and new business formation can impact on a locality's economic structure. And of course competitive pressures or knowledge transfers originating at the international and global levels can stimulate local economic change. So local variety and novelty can be generated by processes operating at a whole range of spatial scales.

The panarchy model of adaptive cycles is obviously highly suggestive. It has the advantage of linking key attributes and processes of regional development, such as innovation, the dynamics of capital accumulation and the mechanisms that generate connectedness between and among local firms and institutions, with the notion of resilience and of emphasising the need to examine how the processes and mechanisms at different scales. But the model is not without limitations and unresolved problems. It is important to remember that the model was developed to analyse ecological systems, the evolution of which may well be adequately conceptualised as long periods of stability, even stasis, interrupted by major external shocks. Is this a reasonable conceptualisation of how regional economies evolve? Regional economies consist of collections of agents and institutions that learn and can change their behaviour even in the absence of major shocks or disturbances. Further, the adaptive cycle approach to understanding resilience implies that regional economic development has an ineluctable, inner logic: that regional development necessarily follows the phases that are argued to characterise adaptive cycles and that the evolution of a regional economy can be adequately represented in terms of such cycles. The fact is that the adaptive cycle model functions as a conceptual framework rather than as a theory or a set of testable hypotheses. As Swanstrom (2008) argues, ultimately, the value of the panarchy conceptual framework for studying regional economies for will rest on its ability to suggest new hypotheses about regional economic resilience that can be tested by empirical case studies. It is to some initial steps in this direction that we now turn by briefly examining two contrasting examples of city-regional development and adaptation.

Empirical exploration: two city region case studies

To explore how the adaptive cycle model might help to think about regional economic resilience, two UK city regions were selected that have experienced quite different economic histories and outcome over the past 40–50 years. The city region of Cambridge is widely regarded as one of the UK's most successful innovative, high-tech and knowledge-based local economies. It is the type of economy that might be expected to be adaptable and resilient. The city region of Swansea, on the other hand, has had a quite different industrial history to that of Cambridge and has struggled to recover from the loss of its former economic role and to adapt to the changes taking place in the national and global economy.

Adaptation in city-regional economies might be expected to take years if not decades. We therefore sketch the economic changes that have taken place in these two city regions (defined in terms of travel-to-work areas) from the 1960s until the late-2000s (see Table 1), and attempt to categorise these changes and developments into the four phases postulated by the panarchy model summarised in Figure 2. Over this time, the high-tech economy of Cambridge developed through what can arguably be identified as three phases of the adaptive cycle model: from the reorganisation of the local economy around a new high-tech development path through a phase of exploitation and growth to a phase of conservation. In contrast, the Swansea economy appears to have experienced six phases, as it first lost its traditional mining industries, turned to foreign direct investment (FDI) in electronics and subsequently lost that as well. Further, over this 40-year period both economies were subject to external ‘slow-burn’ stresses such as changes in markets, technology and policy regimes. In addition, they were also confronted by the shock of two national economic recessions (early-1980s and early-1990s). Of particular interest is how far the resilience of the two city regions to these major downturns can be explained in terms of the adaptive phases described by the panarchy model.

Table 1.

Adaptive cycles in the Cambridge and Swansea city region economies, 1960–2005.

 Cambridge Swansea 
1960 1 Reorganisation phase ‘Cambridge Consultants’ formed, by group of newly graduated scientists and engineers from the university 1 Release phase Decline of local extractive industries 
1969 Mott report published, recommending the founding of a Science Park  
1970 Decision by Trinity College to develop the country’s first Science Park  
1971 Outline planning permission granted for Cambridge Science Park  
  2 Reorganisation phase 
1973 First company, Laser-Scan, moved onto the Science Park FDI by Sony electronics factory at Bridgend 
1976  WDA established 
  FDI by Japanese electronics firms including Hitachi and Panasonic 
1979 By end of 1970s, around 25 companies located on the Science Park  
1980 National recession trough Last coal mine closed in Swansea County 
1981  UK’s first Enterprise Zone designated in Swansea 
 2 Exploitation phase 3 Exploitation phase 
1985 ‘Cambridge Phenomenon’ report published by SQW FDI in consumer electronics, e.g. Hitachi and Panasonic 
1987 St John’s Innovation Centre establishedUniversity publishes its first IP policy for Research Council funded research  
  4 Conservation phase 
1991 1993 National recession trough During 1990s, cluster of hi-tech companies in the Cambridge area grew to around 1200 companies employing around 35,000 people 5 Release phase Loss of competitiveness based on low cost labour 
1996  Planned LG investment does not arrive 
 3 Conservation phase 6 Reorganisation phase 
1997 Eastern Region Biotechnology Initiative established  
1999 Cambridge Network formed, as a voice for the high-technology business communityGreater Cambridge Partnership formed WAG formed 
2001 University revises its IP policy for externally funded research Cambridge Angels group formed Technium Programme launched 
2004 Cambridge Science Park now home to more than 70 R&D companies on the 152-acre site, including new clusters such as photonics, nanotechnology and materials science  
2005 New IP policy adopted by the university Publication of ‘Swansea 2020’, economic regeneration strategy for the city 
Post-2005 Slow down of innovation and growing concern over growth momentum of high-tech cluster  
 Cambridge Swansea 
1960 1 Reorganisation phase ‘Cambridge Consultants’ formed, by group of newly graduated scientists and engineers from the university 1 Release phase Decline of local extractive industries 
1969 Mott report published, recommending the founding of a Science Park  
1970 Decision by Trinity College to develop the country’s first Science Park  
1971 Outline planning permission granted for Cambridge Science Park  
  2 Reorganisation phase 
1973 First company, Laser-Scan, moved onto the Science Park FDI by Sony electronics factory at Bridgend 
1976  WDA established 
  FDI by Japanese electronics firms including Hitachi and Panasonic 
1979 By end of 1970s, around 25 companies located on the Science Park  
1980 National recession trough Last coal mine closed in Swansea County 
1981  UK’s first Enterprise Zone designated in Swansea 
 2 Exploitation phase 3 Exploitation phase 
1985 ‘Cambridge Phenomenon’ report published by SQW FDI in consumer electronics, e.g. Hitachi and Panasonic 
1987 St John’s Innovation Centre establishedUniversity publishes its first IP policy for Research Council funded research  
  4 Conservation phase 
1991 1993 National recession trough During 1990s, cluster of hi-tech companies in the Cambridge area grew to around 1200 companies employing around 35,000 people 5 Release phase Loss of competitiveness based on low cost labour 
1996  Planned LG investment does not arrive 
 3 Conservation phase 6 Reorganisation phase 
1997 Eastern Region Biotechnology Initiative established  
1999 Cambridge Network formed, as a voice for the high-technology business communityGreater Cambridge Partnership formed WAG formed 
2001 University revises its IP policy for externally funded research Cambridge Angels group formed Technium Programme launched 
2004 Cambridge Science Park now home to more than 70 R&D companies on the 152-acre site, including new clusters such as photonics, nanotechnology and materials science  
2005 New IP policy adopted by the university Publication of ‘Swansea 2020’, economic regeneration strategy for the city 
Post-2005 Slow down of innovation and growing concern over growth momentum of high-tech cluster  

As Table 1 shows, in the 1960s and 1970s, the Cambridge economy was going through an historical reorganisation in the composition and orientation of its economy. The transition from being a service-based market town and home to one of the UK's foremost universities to a high-tech, high-growth and innovative economy began around 1960 with the formation of Cambridge Consultants by a group of newly graduated scientists and engineers from the university. It was followed nearly a decade later by the Mott report that recommended the establishment of a science park, for which planning permission was granted in 1971. In contrast, during the 1960s, the Swansea economy was experiencing a major phase of release and decline. By the beginning of the 1970s, many of the area's traditional mining industries (coal, iron, copper, tin and zinc) that had formed the basis of its former economic development were in steep decline or had already closed. These closures were driven by international competition in the availability of raw materials and more efficient production processes elsewhere. The closures released a pool of low-skilled, low-wage labour onto the local labour market.

In Cambridge, the 1970s saw the continuation of reorganisation with the establishment of several new high-tech enterprises and the emergence of an identifiable high-tech cluster. The first company, Laser-Scan, moved onto the new Trinity College science park in 1973, and by the end of the 1970s, the reorientation of the Cambridge economy around high-tech activity was well underway, with some 30 companies located on the park. They were mainly based on IT activities, specifically electronic equipment and instruments. The 1970s also saw the Swansea economy beginning to embark on a phase of reorganisation. With the closure of its traditional mining industries, it was forced by international competition to seek new forms of economic activity. It did this by turning to FDI and offering relatively cheap land and labour. One of the first significant inward investments in the electronics industry was the Sony facility in Bridgend in 1973, east of Swansea. This was the first major Japanese investment in Wales and acted as a catalyst for other inward investors, particularly in the electronics sector.

Both economies were then subject to the external shock of the deep national recession of the early-1980s. By the onset of the recession the structure of both economies had evolved significantly from what they had been in the 1960s. According to the adaptive cycle model outlined in Figure 2 above, the structural reorganisation occurring in both economies should have led to increasing resilience to external shocks. But the two economies reacted rather differently to the shock of the recession. Figure 4 shows how the employment growth paths of the two city regions were affected. Just prior to the downturn, the two city regions had almost identical employment levels (around 215,000). The recession impacted much more severely on the Swansea economy than it did on Cambridge, and the latter recovered much more strongly and quickly than the former. Indeed, in Swansea employment had not fully returned to its 1980s level even a decade later.

Figure 4.

Employment growth paths in the two city regions, 1980–2008.

Source: Cambridge Econometrics.

Figure 4.

Employment growth paths in the two city regions, 1980–2008.

Source: Cambridge Econometrics.

One of the main differences in the resilience of the two economies following the shock of the recession can be seen in the changes taking place in manufacturing employment (Figure 5). In Cambridge, manufacturing employment declined slightly during and after the recession but it soon recovered and by the mid-1980s had exceeded its 1980 level. By way of contrast, following the recession, manufacturing employment in Swansea showed no signs of regaining its pre-1980 level. The shock of the early-1980s recession exposed the different nature of adaptation and resilience in the two case study economies. The steady market-led development of innovation, high-tech small and medium sized enterprises and early cluster development in Cambridge over the 20 years from 1960 did seem to provide it with the resilience to weather the industrial downturn of the recession. In contrast, by the late-1970s, the economy of Swansea had not been able to reorganise its economy sufficiently to be able to withstand the impact of the recession. The extent of restructuring based on the introduction of medium-sized FDI branch plants was not sufficient to provide it with the sort of economic base to survive the shock of the downturn, and its manufacturing base not only declined faster but also failed to recover. The evidence suggests that the Swansea economy was noticeably less resilient than its Cambridge counterpart.

Figure 5.

Manufacturing employment growth paths in the two city regions, 1980–2008.

Source: Cambridge Econometrics.

Figure 5.

Manufacturing employment growth paths in the two city regions, 1980–2008.

Source: Cambridge Econometrics.

Following the early-1980s recession, the economies of both Cambridge and Swansea entered a growth phase based on the exploitation of their existing industries. In the case of Cambridge, the foundations proved to be far stronger than those in Swansea. In Cambridge, the rate of high-tech new firm formation increased significantly and was marked by the creation of new pathways by branching out into life sciences and in particular biotechnology. This was driven by a mixture of endogenous cutting-edge university research and organic growth within the sector. The Swansea economy for its part became increasingly reliant on the external knowledge brought into the area by foreign-owned electronics companies. Encouraged by the Welsh Development Agency, in addition to Sony, multinational companies such as Panasonic and Hitachi opened branch plants in the area. At the height of this phase, some 50% of televisions and 75% of video cassette recorders (VCRs) produced in Europe were made in south Wales, located mainly in and around Swansea.

This success led to a certain degree of conservation in the local Swansea economy, in the sense this notion is used in the panarchy model. The branch plants became locked into the technologies distributed to them by their Japanese-owned parent companies. Although this marked a temporary period of stability, it was also characterised by increasing rigidity in an industry driven globally by rapid technological change, which in the case of Swansea's production plants was in the companies’ research laboratories located back in Japan. This local technological lock-in and external dependence reduced the resilience of the local economy and soon proved fatal to the survival of the consumer electronics industry in the Swansea area. The run up to the 1991 recession saw the international demise of both cathode ray tube (CRT) technology-based television and VCR-based recording devices. Sony closed one of its Swansea plants, moving production to Barcelona where local expertise in the new digital and plasma technology helped them respond to the changing market. This was followed after the recession by the eventual closure of the Panasonic and Hitachi manufacturing plants. All that remains in Swansea today are retail outlets for the products of these companies that are manufactured elsewhere.

The early-1990s recession impacted on both economies, as shown in Figure 3. Both areas experienced a sharp contraction in manufacturing employment (Figure 5), but Cambridge once again proved more resilient, with a recovery in jobs in those sectors, at least up to 2000. In Swansea, manufacturing employment failed to recover at all: the contraction proved permanent. And the slowdown in service employment growth during the recession continued until the late-1990s. Only then did steady growth in this sector take place (Figure 6). In Cambridge, although the recession halted the rapid expansion in service jobs that occurred over the 1980s, the interruption was short lived, and from 1992 onwards, the economy resumed strong employment growth in this sector.

Figure 6.

Service sector employment growth paths in the two city regions, 1980–2008.

Source: Cambridge Econometrics.

Figure 6.

Service sector employment growth paths in the two city regions, 1980–2008.

Source: Cambridge Econometrics.

One of the key contributions to the relatively higher adaptability and resilience in the Cambridge economy was the ability to continually branch out of the existing specialised industrial sectors. New development pathways were created out of a strong endogenous knowledge platform grounded in advanced mathematics and computing. The 1990s saw a continued growth in the high-tech economy of the city region, with a strong upward rise in the number of new firms formed each year (Figure 7), and the formation of several new subcluster specialisms, including ink-jet technologies, telecoms, biosciences, informatics and clean technologies. By the end of the decade, the high-tech cluster comprised more than 300 such enterprises, employing perhaps as many as 20,000 people. There is some sign, however, that this phase of exploitation, rapid expansion and accumulation of high-tech assets, may possibly have come to an end and that local high-tech development may have entered a conservation phase of distinctly slower growth. Both the rate of new firm formation and the rate of innovation have slowed appreciably since 2000, and local business leaders and associations have expressed repeated concern that the cluster seems to have lost its momentum.

Figure 7.

Number of new firms established in the Cambridge high-tech cluster, 1970–2004.

Source: Library House.

Figure 7.

Number of new firms established in the Cambridge high-tech cluster, 1970–2004.

Source: Library House.

The Swansea economy fared much less well following the early-1990s recession. The flow of FDI that had begun to provide the possible basis for a new phase of exploitation and growth ceased and promised additional arrivals of overseas investments, such as that by the electronics giant LG, did not take place. It is arguable whether the economy ever embarked on a secure or sustained path of exploitation and growth based around a new consumer electronics industry. Rather, it would seem that this potential development path failed before it really took off, and as a result the economy moved quickly to a further phase of release and decline. The FDI withdrew and the ability of the local economy to adapt to these changes was hampered by lock-in to a culture of industrial subsidies dependency and an expectation of lifetime employment in traditional industries. This combination proved too difficult to be overcome by free market adaptations. Instead it has required intervention by the Welsh Assembly Government (WAG) to develop a public strategy to move Wales (and Swansea) up the value chain, by focusing development on the knowledge economy.

Since the end of the 1990s, in line with this wider regional strategy, a new attempt to reorganise the Swansea economy has been underway, initially based on the provision of start-up incubator and support units, an initiative called Technia, in partnership with the Welsh Development Agency (now incorporated into the WAG) and Swansea University. The implementation of this strategy has been slow, however, and the first Technium was not opened in Swansea until 2001. There are currently a total of nine across Wales. Their aim is to provide infrastructure and business support to new innovative companies and to provide space for them to develop their ideas and grow into successful knowledge-related businesses. The scale of each Technium is quite small. Each of them contains around only 10–12 firms. How resilient the two city region economies will prove to be to the deep recession of 2008–2009 remains to be seen.

Summary and conclusions

In this paper, we have sought to review and analyse the concept of regional economic resilience. Our starting position is a rejection of equilibrist approaches. This is because we think that the firms, organisations and institutions that comprise regional economies are continually changing and adapting to their economic environments. These changes are increasingly driven by the creation, acquisition and commercial exploitation of new knowledge. These processes are never in equilibrium.

Following these arguments, we turned to an evolutionary theoretical perspective. This emphasises adaptation and change as key processes in the development of regional economies. We argue that these processes are the bases of regional economic resilience. For the purposes of this paper, we explored the applicability of the panarchy model, developed in ecological science, for further analysis and developed a four-phase adaptive cycle model of regional economic resilience. This postulated that adaptation in regional economies follows a sequential cycle of innovation and restructuring, growth and the seizing of opportunities, stability and increasing rigidity, followed by a release phase and eventually the repetition of the cycle over various periods of time. Each phase of the cycle is associated with different degrees of resilience, connectedness and capital accumulation or release.

It should be stressed that this is primarily a descriptive model and does not contain any theoretical explanations of the causes of each phase of adaptation nor what drives an economy from one phase to the next. To illustrate possible explanations for such adaptation and change, we turned to two contrasting case studies. These were selected according to their presumed illustration of contrasting degrees of adaptation and resilience.

What we found was that over the 45-year period of study, the Cambridge high-tech economy would appear to have gone through only three phases suggested by adaptive cycle model: reorganisation, exploitation and possibly the onset of conservation. It arrived at each of the two major recessions covered by this study having developed its resilient capacity over the preceding one or two decades. The long-term development of adaptive and resilience capacities in Cambridge were driven by the conscious decisions of local entrepreneurs making use of endogenously created new knowledge. This was facilitated by the co-evolution of the attitude of Cambridge University to commercial exploitation of intellectual property rights and the facilitation of the development of science parks and incubation units on land owned by the different colleges.

The experience of Swansea has been quite different and far less successful. The Swansea economy seems to have gone through one and a half cycles of the model. The first cycle, starting around the 1960s, saw the decline of the traditional extractive industries and the loss of their endogenous knowledge base. This was followed in the 1970s by the beginnings of reorganisation driven by the conscious decisions of public policy and based on the attraction of FDI in the form of Japanese consumer electronics. The FDI brought with it sufficient codified external knowledge for the establishment of manufacturing branch plants based on VCR and CRT technologies. According to the adaptive cycle model, this reorganisation phase should have been accompanied by increasing resilience. This did not prove to be the case. The shock of the early-1980s recession led to a decline in manufacturing employment from which the Swansea economy never fully recovered. The exploitation of foreign-owned electronics was resumed after the recession, again driven as much by public policy as by market forces. This sector then entered a short conservation phase becoming locked into increasingly outdated technologies. As a result, it arrived at the recession of the early-1990s with declining resilience. The shock of the recession exposed the weakness of relying on exogenous knowledge generated by multinational companies. Instead of transferring new technological knowledge to Swansea, they transferred it to new plants in Spain and Eastern Europe. As a result, manufacturing employment fell again in Swansea as the economy entered a second phase of capital disinvestment and release.

These case studies—albeit only briefly discussed here—suggest that endogenous sources of new knowledge combined with market driven and conscious entrepreneurial decisions could be among the key factors for understanding regional economic resilience. In addition, the co-evolution of facilitating institutional environments is also significant. Conversely, where existing sources of knowledge are locked into particular stable technologies, long-term adaptive capacity and resilience tend to decline. In such circumstances, reliance on external sources of new knowledge that come with the mass production branch plants of FDIs is likely to be only a short-term fix. It can lead to declining or the lack of endogenous sources of new knowledge for local potential entrepreneurs and to a consequential decline of the capacity of the local economy to adapt to external market forces. Institutional hysteresis and unchanging cultures can also contribute to a lack of economic resilience.

Our analysis suggests that further investigation of the adaptive cycle model may be a useful way of conceptualising and analysing regional economic resilience. Clearly, the model is not unproblematic or without limitations. It is of course only one way of thinking about how a local or regional economy evolves, and there are issues about abducting a model from one disciplinary field, in this instance the study of ecosystems, to another, in our case regional economic development. Regional economies may be analogous to ecosystems in certain respects, but they are quite different in others. Nevertheless, the model has the virtue that it sets the issue of resilience of a system—such as a local economy—in the context of the evolution of that system: resilience as a process rather than an unchanging characteristic. For this reason alone, it seems worthwhile pursuing the panarchy adaptive cycle model in more depth, with the aim of giving it a more explicit regional economic form, as one way of trying to cast light on the important but somewhat elusive notion of regional economic resilience.

The authors gratefully acknowledge the financial support of The National Endowment for Science, Technology and the Arts (NESTA) for the empirical work reported in the paper. A full exposition of this research is available from NESTA published as: Simmie, J., Carpenter, J., Chadwick, A. and Martin, R. (2008) History Matters: Path Dependence and Innovation in British Cities, London: NESTA. The authors also acknowledge the helpful comments from two anonymous referees.

References

Brezis
ES
Krugman
PR
Technology and the Life Cycle of Cities
Journal of Economic Growth
 , 
1997
, vol. 
2
 (pg. 
369
-
383
)
Briguglio
L
Cordina
G
Bugeja
S
, et al.  . 
Conceptualizing and Measuring Economic Resilience
 , 
2006
 
Mimeo, Department of Economics, University of Malta
Carpenter
SR
Westley
F
Turner
M
Surrogates for resilience of social-ecological systems
Ecosystems
 , 
2005
, vol. 
8
 (pg. 
941
-
944
)
David
PA
Dopfer
K
Path dependence in economic processes: implications for policy analysis in dynamical systems contexts
The Evolutionary Foundations of Economics
 , 
2005
Cambridge, UK
Cambridge University Press
(pg. 
151
-
194
)
David
PA
Path dependence and historical social science: an introductory lecture
 , 
2007
 
Paper presented at the Symposium on Twenty Years of Path Dependence and Qwerty-Effects, Russian University–Higher School of Economics, Moscow, 13 May 2005. SIEPR Policy Paper No. 04–022
Foster
KA
A Case Study Approach to Understanding Regional Resilience
 , 
2007
 
Working Paper 2007–08, Institute of Urban and Regional Development, University of California, Berkeley
Hanley
N
Resilience in social and economic systems: a concept that fails the cost-benefit test?
Environment and Development Economics
 , 
1998
, vol. 
3
 (pg. 
244
-
249
)
Hill
EW
Wial
H
Wolman
H
Exploring Regional Economic Resilience
 , 
2008
 
Working Paper 2008–04, Institute of Urban and Regional Development, University of California, Berkeley
Holling
CS
Resilience and stability of ecological systems
Annual Review of Ecological Systems
 , 
1973
, vol. 
4
 (pg. 
1
-
23
)
Holling
CS
Gunderson
LH
Gunderson
L
Holling
CS
Resilience and adaptive cycles
Panarchy: Understanding Transformations in Human and Natural Systems
 , 
2002
Washington, DC
Island Press
(pg. 
25
-
62
)
McGlade
J
Murray
R
Baldwin
J
, et al.  . 
Garnsey
E
McGlade
J
Industrial resilience and decline: a co-evolutionary approach
Complexity and Co-Evolution: Continuity and Change in Socio-Economic Systems
 , 
2006
Cheltenham
Edward Elgar
(pg. 
147
-
176
)
Maskell
P
Malmberg
A
Myopia, knowledge development and cluster evolution
Journal of Economic Geography
 , 
2007
, vol. 
7
 (pg. 
603
-
618
)
Martin
RL
Roepke lecture in economic geography—rethinking regional path dependence: beyond lock-in to evolution. Lecture delivered at the annual meeting of the Association of American Geographers, Las Vegas, March 2009
Economic Geography
 , 
2010
(pg. 
1
-
27
)
Martin
RL
Sunley
PJ
Path dependence and regional economic evolution
Journal of Economic Geography
 , 
2006
, vol. 
6
 (pg. 
395
-
438
)
Martin
RL
Sunley
PJ
Complexity thinking and evolutionary economic geography
Journal of Economic Geography
 , 
2007
, vol. 
7
 (pg. 
573
-
602
)
Ormerod
P
Resilience after local economic shocks
Applied Economics Letters
 , 
2008
(pg. 
1
-
5
)
Pendall
R
Foster
KA
Cowell
M
, et al.  . 
Resilience and Regions: Building Understanding of the Metaphor
 , 
2008
Mimeo Institute of Urban and Regional Development, Ithaca, NY: Cornell University
Petersen
G
Political ecology and ecological resilience: an integration of human and ecological dynamics
Ecological Economics
 , 
2000
, vol. 
35
 (pg. 
323
-
336
)
Pimm
SL
The complexity and stability of ecosystems
Nature
 , 
1984
, vol. 
307
 (pg. 
321
-
326
)
Ramlogan
R
Metcalfe
JS
Garnsey
E
McGlade
J
Restless capitalism: a complexity perspective on modern capitalist economies
Complexity and Co-Evolution: Continuity and Change in Socio-Economic Systems
 , 
2006
Cheltenham
Edward Elgar
(pg. 
115
-
146
)
Rose
A
Liao
S-Y
Modelling regional economic resilience to disasters: a computable general equilibrium models of water service disruptions
Journal of Regional Science
 , 
2005
, vol. 
45
 (pg. 
75
-
112
)
Setterfield
M
Should economists dispense with the notion of equilibrium?
Journal of Keynesian Economics
 , 
1997
, vol. 
20
 (pg. 
47
-
76
)
Stehr
SD
The political economy of urban disaster assistance
Urban Affairs Review
 , 
2006
, vol. 
41
 (pg. 
492
-
500
)
Simmie
J
Martin
RL
Carpenter
J
, et al.  . 
History Matters: Path Dependence and Innovation in British City-Regions
 , 
2008
 
Research Report, London: National Endowment for Science, Technology and the Arts
Swanstrom
T
Regional resilience: a critical examination of the ecological framework
 , 
2008
 
Working Paper 2008–07, Institute of Urban and Regional Development, University of California, Berkeley
Toulmin
S
Jenson
UF
Harré
R
Human adaptation
The Philosophy of Evolution
 , 
1981
London
Harvester Press
(pg. 
176
-
195
)
Vale
L J
Campanella
T J
The Resilient City
 , 
2005
New York
Oxford University Press