The dark side of regional industrial path development: towards a typology of trajectories of decline

ABSTRACT

Over the past few years, scholarly debates on new path development have attracted increasing attention within the economic geography literature. This work distinguishes various trajectories of regional and industrial evolution. So far, these evolutionary trajectories have been mainly conceptualized as ‘positive’ forms of path development. However, in reality, many regions are undergoing phases that can be characterized as ‘negative’ trajectories. Despite their potentially detrimental social and political effects, negative pathways have to date largely been ignored in the extant literature. This paper seeks to shed light on the ‘dark side’ of path development by outlining a typology of ‘pathways of decline’. Three forms of negative pathways are identified, namely path contraction, path downgrading and path delocalization. Empirical illustrations are provided for each of them.

KEYWORDS:

• New path development models
• regional asset base
• trajectories of decline
• downgrading
• delocalization
• contraction

1. Introduction

Evolutionary economic geography (EEG) offers a powerful theoretical framework for understanding the socioeconomic evolution of regions and industries (Storper, 2011). Within this stream of literature, much attention has been devoted to elucidating the sources of what has been termed ‘new path development’, that is, the rise and growth of new economic activities and the successful rejuvenation of mature regional industries. Protagonists of this approach have provided a differentiated view on how such processes unfold, distinguishing between path renewal, diversification, importation and creation (Grillitsch, Asheim, & Trippl, 2018; Isaksen, 2015; Tödtling & Trippl, 2013; Trippl, Grillitsch, & Isaksen, 2018). So far, evolutionary trajectories have been predominantly conceptualized as ‘positive’ forms of path development, that is, growth paths.

This obscures the reality that many regions undergo phases that display negative trajectories, causing not only economic but also social and political challenges (Rodríguez-Pose, 2018). Yet, the ‘dark side’ (Phelps, Atienza, & Arias, 2018) of regional industrial path development has not been systematically investigated, even though negative trajectories in one region or industry are often closely intertwined with positive trajectories in other regions and industries. Moreover, periods of growth are often followed by periods of disarray (Beynon, Hudson, Lewis, Sadler, & Townsend, 1989; Chisholm, 1990). As argued already decades ago, the key driver of uneven development has shifted from an inter-sectoral spatial division of labour to an intra-sectoral division of labour reflecting asymmetric power relationships within the economy (Massey, 1979, 1984).

Arguably, it has to be acknowledged that many studies have examined declining industries and negative regional lock-in, especially those investigating the fate of old industrial regions (Cooke, 1995; Hassink, 2005, 2010) or negative trajectories in industrial districts in Italy and beyond (Giuliani & Rabellotti, 2018). However, to the best of our knowledge, even those studies that examine regional lock-ins pay primary attention to periods of renewal, while the strategies of particular actors and the mechanisms of change during periods of decline are hardly analyzed in detail (Bair & Werner, 2011a, 2011b; Beer, 2018; Cooke, 1989). Moreover, it has to be emphasized that so far only limited attention has been devoted to research of disruption, decline or disarticulation in services even though in developed countries the service sector is dominating in terms of employment and GDP (for recent exceptions, see Coe, Lee, & Woods, 2017; Murphy, 2019; Wu & Gereffi, 2019).

All these streams of literature and conceptual frameworks provide useful insights, enabling one to contribute to the existing new path development debate and also explicitly comprehend trajectories of socioeconomic decline, thus gaining a more complete understanding of evolutionary trajectories of regional industries. Thus, this paper seeks to develop a typology of negative pathways. We use examples from the extant literature and from our own research to illustrate the characteristics of different types of negative path development.

The paper is organized as follows. First, we outline and discuss the theoretical framework. Second, we identify three forms of negative pathways and link them to existing case studies of socioeconomic decline in regions. Third, a concluding section summarizes the main arguments of the paper and proposes avenues for future research.

2. Conceptualizing pathways of decline

Recent debates on ‘new regional industrial path development’ (see Hassink, Isaksen, & Trippl, 2019, for a review) have enhanced the understanding of the conditions and mechanisms of longer-term regional structural change. Advocates of EEG models (Boschma, 2017; Boschma & Frenken, 2011; Boschma & Martin, 2007, 2010; Martin, 2010; Martin & Sunley, 2006, 2011) claim that the historically grown industrial and institutional structures have a major influence on future economic and innovation activities. Protagonists of the ‘new path development approach’ have identified various forms and mechanisms by which new industries come into being and old ones renew themselves, and they have argued that different types of regions (and their innovation systems) vary in their capacity to nurture these ‘positive’ forms of path development. Isaksen and Trippl (2016) distinguish between organizationally thick and diversified systems, organizationally thick and specialized systems, and organizationally thin systems. Organizationally thick and diversified regional innovation systems (RIS) are said to exhibit the highest capacity to promote path creation, while the typical development pattern in thick and specialized innovation systems comprises either path renewal or path extension. The most vulnerable regions, that is, organizationally thin RIS, are found to lack a critical mass of capable and innovative regional actors. They are also poorly endowed with support organizations, including universities, intermediaries and organizations capable of designing adequate innovation policy (Trippl, Zukauskaite, & Healy, 2019). It is argued that these regions show a limited capacity to develop new economic activities (Isaksen & Trippl, 2016; Trippl et al., 2018). Other studies highlight that these places are dependent upon the national level, both in terms of provision of financial support and empowerment of local intermediaries (Jeon & Phelps, 2018).

While these typologies are useful, as they spell out the characteristics and most likely evolutionary trajectories in different types of regions, they primarily cover phases of socioeconomic growth. Arguably, Isaksen and Trippl (2016) and Trippl, Asheim, and Miörner (2016) briefly introduce ‘path exhaustion’ (defined as the erosion of competitiveness due to negative lock-in) as one potential form of path development. However, they do not venture beyond the brief observation that the lack of renewal and experimentation might lead to stagnation and decline. Negative forms of path development are thus not sufficiently addressed in this body of research. The aim of this article is to address this gap.

Negative path development is defined here as the decline of a given regional industry (that is, functionally related firms and supportive actors and institutions) in terms of employment and capital accumulation, resulting in a drop in economic output and adverse modification of the region’s asset base. We argue that several conceptual approaches can be mobilized to outline the spectrum of possible trajectories of decline.

2.1. Insights from evolutionary economic geography

Three interrelated concepts that are widely employed within contemporary EEG research serve as useful points of departure: lock-in, path dependency and resilience (for detailed reviews of these notions, see for instance Boschma, 2015; Hassink, 2010). Scholarly work invoking the notion of negative lock-ins usually builds on a typology originally outlined in Grabher’s (1993) study of the decline of the Ruhr area (i.e. functional, cognitive and political lock-in). Hassink (2010) argues that regional lock-in represents a set of these closely interrelated types of lock-ins and he claims that regions with strong specializations in capital-intensive industries with high entry and exit barriers are particularly prone to severe lock-ins.

It should be stressed that while the concept of lock-in is often considered as a form of constraint on future development (even though some scholars explicitly distinguish between negative and positive lock-ins – see, for instance, Essletzbichler & Winther, 1999; Martin & Sunley, 2006), path dependence entails both negative and positive connotations (i.e. resources and capabilities may either constrain or enable current and future development). Finally, regional resilience is generally considered a desirable feature, not only as a form of recovery to a state preceding a shock, but more appropriately as a process enabling regions to ‘bounce forward’ and develop new growth paths (Boschma, 2015; Martin & Sunley, 2015).

We argue that conceptualisations of negative forms of path development can benefit also from insights from the literature on the adaptive cycle model (Martin & Sunley, 2011), which builds on the above-mentioned EEG concepts (lock-in, path dependency, resilience). This model points to a significant variation and unpredictability of evolutionary trajectories. The adaptive-cycle model underlines the importance of the recombination and reuse of resources and emphasizes the role of both path dependence and place dependence. Each evolutionary phase is said to be characterized by three dimensions: connectedness (defined as traded and untraded interdependencies among companies), capital/resource accumulation (accumulation of productive, knowledge and institutional capital) and resilience (capacity of firms to respond flexibly to internal and external disturbances). We believe that these three dimensions can be employed to capture the key features of negative pathways.

Since negative forms of path development are often closely interlinked with evolutionary trajectories in other regions, we see considerable value in developing a more nuanced view on the dimension of ‘connectedness’ by distinguishing between internal (regional) and external (inter-regional) connectedness of a given industry. We argue that varying forms and degrees of connectedness among companies may be observed within the region and at inter-regional level. For example, the notion of ‘truncated development’ points to a disruption of existing intra-regional networks, often induced by the inflow of FDI (Hayter, 1982), which is said to go hand-in-hand with a growing connectedness to companies in other regions, based on the integration of foreign-owned branch plants as well as local SMEs into global production networks (Pavlínek, 2018).

2.2. Insights from the theory of global value chains/global production networks (GVC/GPN)

The global production network (GPN) approach deals with organizational platforms through which various actors compete and cooperate for value creation and capture through geographically dispersed economic activity (Yeung & Coe, 2015). To contest the ‘inclusionary bias’ in existing GVC research, the disarticulation approach has been developed to conceptualize the processes of disconnection or expulsion of companies and regions from global value chains (Bair & Werner, 2011a, 2011b; Hough, 2011). Within the GPN framework, there is also a growing body of work on strategic coupling, decoupling and recoupling (Horner, 2014; MacKinnon, 2012; Yeung, 2016a, 2016b), which highlights that due attention should be given to dynamic relational processes between local firms and powerful global companies (i.e. lead firms of GPNs or multinational companies in general) often shaped by national or international regulatory regimes. These processes result in the formation of profoundly different types of power relationships, as well as connectivity between local and non-local actors, ranging from relatively symmetric relationships based on mutual dependence within ‘international partnership’ type of strategic coupling to ‘structural coupling’ via production platforms dominated by powerful buyers (Yeung, 2016b). Coe and Hess (2011) offer inter alia a typology of negative consequences representing the ‘dark side’ of strategic coupling resulting from profound power asymmetries which are typical for a bargaining process that results in uneven value capture, labour exploitation, disinvestment, displacement and social and class conflict. Coe and Hess (2011) conclude that strategic coupling with GPNs ‘is of course no guarantee of positive developmental outcomes’ and can lead to intra-regional disarticulations via break-up of existing cultural and economic networks and systems and can affect ‘the opportunities and livelihood of people and households, and hence raises serious questions about the nature and distribution of the value generated, enhanced and captured within the region’ (Coe & Hess, 2011, p. 134).

2.3. Unpacking the transformation of industries in regions: firm strategies, industry changes and the modification of the regional asset base

In order to comprehensively conceptualize negative forms of path development, it seems to be crucial to disentangle the underlying multi-dimensionality between the single-firm level, the industry level and the regional level. Studies on industrial districts in various countries (Alberti, 2006; Giuliani & Rabellotti, 2018; Hadjimichalis, 2006; Jeon & Phelps, 2018; Sammarra & Belussi, 2006; Schamp, 2005) point to a variety of economic challenges and identify a broad spectrum of firms’ adjustment strategies (such as outsourcing and delocalization, functional downgrading, engagement of ‘ethnic’ firms, downsizing, moves into market niches, etc.). In times of economic downturn or sharp competition, it might be rational (and in a sense positive) for firms to, for instance, downsize their portfolio of activities or relocate.

From regional and industry perspectives, however, such firm-level activities are most probably seen as negative because of the potential devaluation of assets and unemployment (regional level) or possible adverse effects on other firms (industry level). From an EEG point of view, negative forms of regional industrial path development in a given region will also influence its future trajectory by affecting the ‘regional environment’ and thereby weakening the region’s future potential for successful path development processes (Martin, 2010).

Recent work on regional asset modification (MacKinnon, Dawley, Pike, & Cumbers, 2019; Trippl, Baumgartinger-Seiringer, Frangenheim, Isaksen, & Rypestøl, 2019) might help to explore in more detail the underlying processes leading to what has been termed a ‘constraining environment’ for new path development. A broad set of assets, ranging from (i) natural assets to (ii) infrastructural and material assets, (iii) industrial assets, (iv) human assets, and (v) institutional endowments (MacKinnon et al., 2019; Maskell & Malmberg, 1999), is seen to be important for nurturing regional industrial path development. This ‘regional asset base’ is considered to be the outcome of past regional economic activities as well as national-regional power modalities (cfr. Rodríguez-Pose, 2018) and serves as a platform for future rounds of development (MacKinnon et al., 2019; Trippl, Baumgartinger-Seiringer, et al., 2019).

Current accounts of intra-regional and inter-regional path dependency (Martin & Sunley, 2006) are in accord with earlier understandings of the nature of ‘a regional problem’ proposed by Massey (1979) who argued that the geographical distribution of economic activity would be a result of the current division of labour, which is ‘overlaid on, and combined with, the pattern produced in previous periods by different forms of division of labour’, thus giving rise to a new form of spatial distribution of inequality (Massey, 1979, p. 235). This inequality then operates as a basis for the next round of investment (Massey, 1979). In the case of a negative pathway on the firm or industry level, the regional asset base is modified in a specific way, most probably rendering the region less favourable for future economic and innovation activities.

Obviously, the decline of an industrial path will imply some de-locking of existing assets, for instance human assets in the form of laid-off labour skills or infrastructural and material assets such as facilities. This underlines the relevance of exploring trajectories of decline in more detail from a system-level perspective, as it is the role not only of firm-level actors but also of system-level actors to identify, harness and valorize the value rooted in (de-locked) regional assets (MacKinnon et al., 2019). This could entail the modification of assets to fit the needs of multinational firms (MacKinnon, 2012), leading to path transplantation (Isaksen & Trippl, 2017). Other potential strategies could strive to alter de-locked (redundant) assets in ways that boost the potential of path creation or diversification (Trippl, Baumgartinger-Seiringer, et al., 2019).

In this paper, we seek to answer the following research questions: (i) What types of negative forms of path development can be identified? (ii) What are the differences between these pathways in terms of capital accumulation, connectedness and resilience? (iii) In what ways is the regional asset base modified as a result of these different regional industrial pathways, and how does asset modification affect future rounds of regional industrial path development?

3. Towards a typology of trajectories of decline

A large number of factors and processes can lead to negative forms of path development and the socioeconomic decline of regions. Within the EEG literature, these factors and processes have often been subsumed under the broad notion of lock-in (see above). More often than not, the forms of lock-in are interrelated and tend to reinforce one another (Hassink, 2010). Consequently, we emphasize that the pathways outlined below should be considered as ideal types (in a Weberian sense) of adverse trajectories, which in reality are often mutually interrelated and overlapping. In order to understand how various forms of lock-in are linked to negative pathways, we argue that due attention needs to be paid to the severity and longevity of lock-in processes¹ and the type of response to development challenges by key stakeholders. In addition, in line with Boschma (2015), it has to be acknowledged that regions are made up of individuals, organizations, industries, networks and institutions, each of which could have their own evolutionary trajectories. Therefore, any path development model deals primarily with prevailing or dominant pathways only.

Following MacKinnon et al. (2019) and Trippl, Baumgartinger-Seiringer, et al. (2019), we also recognize the need to pay due attention to the role played (or not played) by different types of actors, including firms and non-firm actors (such as universities, regional development agencies, labour unions, policy actors, and so on). Due to space restrictions, in this article our focus is on economic actors (companies), but we see considerable value in including other actor groups in future analyses (see the concluding section). Firm actors can be foreign or home-grown, and they may differ in terms of the strategies they use to respond to the decline of their industry.

Building on the conceptual discussion outlined above, we develop a typology of negative forms of path development that draws a distinction between three main types, namely path contraction, path downgrading and path delocalization, even though we acknowledge that in reality industries in regions can follow multiple pathways. Thus, we admit that the cases we provide as examples of particular negative pathways may also exhibit features of other ideal-types. Finally, we argue that each of these pathways can in the worst case result in the most radical form of decline, i.e. in the disappearance of a given industry in a particular region.

3.1. Path downgrading

This pathway is underpinned by two major strategies of key companies in a given regional industry. First, path downgrading may be the result of what we call ‘adjustment strategy’, which is the removal of higher value-added functions such as R&D (i.e. downgrading in terms of functions performed). Second, path downgrading can be the outcome of what is termed ‘re-specialisation strategy’, more precisely re-specialization in low-cost production (see, for instance, the case of South African wine producers, Ponte & Ewert, 2009) and – related to that – reorientation towards serving less demanding markets (i.e. downgrading in terms of market segments).

Functional downgrading can occur through the inflow of FDI driven by low-cost motives, while strategic functions are being performed abroad. This often takes the form of turning the acquired local companies into branch-plants of multinational companies without higher-level functions, and it frequently results in the destruction of previous local linkages and networks (Hayter, 1982). Emergence of branch-plant economies in less developed or peripheral regions can represent a powerful mechanism, replicating or even deepening the downgrading of economies of these regions, especially under the condition of a shifting scale of the division of labour (from national to international) and increasing dependence from foreign direct investment in peripheral regions (Phelps, 1993). Particular processes resulting in deepening the dependency of branch-plant economies upon the decisions taken in core regions (such as functions to be performed, labour recruitment schemes or centralized purchasing, see, for instance, Thwaites, 1978). This often results in fragile regional development processes, not least because regional industrial growth is orchestrated by powerful external actors, seeking higher returns from their investments (Pavlínek & Smith, 1998).

Another source of downgrading could be an intense competitive pressure, which local companies in a given industry are unable to withstand, leading them to reorient their strategy towards low-cost production of standard goods or components. This strategy was often adopted in Central Eastern European countries when the former state-owned companies that had been primarily locked into the Eastern European markets for several decades found themselves uncompetitive after the sudden trade liberalization (Blažek & Csank, 2016). This was often accompanied by a disruption or dissolution of pre-existing regional networks and an increase of external connectedness (Květoň & Blažek, 2018; Pavlínek, 2018).

A distinctive case of social downgrading (i.e. devaluation of the quality of jobs in terms of work conditions, remuneration and job stability) as a result of the introduction of crowdsourcing and crowdemployment as documented for the ICT sector by Flecker and Schönauer (2016). Most vulnerable to such practices are workers in regions with limited employment opportunities.

Gibbon and Ponte (2005) reveal a distinctive trend of agricultural farms in several African countries towards downgrading through greater specialization in more price competitive offerings of products of basic quality. Overall, the level of autonomy of regional companies in a given industry may decline significantly as a result of downgrading and, if the conditions change, the way back can be very challenging or even impossible. Thus, the level of resilience tends to decrease significantly. However, in some cases, downgrading can enable significant employment growth as well as an improvement in economic performance, at least in the short-term (Ponte & Ewert, 2009).

Generally, it can be claimed that path downgrading is most likely to occur in regional industries that fail to accumulate sufficient know-how and capital. If companies that operate in these industries are unable to safeguard a favourable cost-capability ratio (Yeung & Coe, 2015) or, more broadly, if they fail to develop and sustain firm-specific routines that are equally efficient as those of competing firms (Schamp, 2005), an intense market pressure might push them further downwards to serve yet lower market segments and/or to perform mere low-value-adding production activities.

As a result of withdrawing from upper market segments or from the final market for a given industry, the regional asset base is also likely to diminish in quality. While, in some cases, single firms might profit from following a downgrading strategy, the regional innovation system and its asset base will most likely be weakened. In particular, industrial assets in the form of technology or firm competencies or human assets (labour skills) might be lost, as observed in many CEE regions, where R&D functions and other activities with higher added-value disappeared during the 1990s, because they were dispensable for routine manufacturing operations. Losing such ‘creative assets’ might be particularly detrimental to a region’s future potential, as they are seen as especially important for economic prosperity (Florida, 2014). Additionally, local and extra-regional innovation networks and linkages might be destroyed due to changing trade structures. This could lead to an overreliance on localized routines, i.e. ‘spatial myopia’, and thereby future lock-ins (Maskell & Malmberg, 2007; Zhu, He, & Zhou, 2017). Path downgrading processes in a regional industry might thus set in motion a downward spiral, rendering the regional asset base less favourable for future higher-value economic activities and thereby severely damaging long-term regional competitiveness.

3.2. Path contraction

The second trajectory is referred to as path contraction. Path contraction implies a shrinkage in the size of the regional industry brought about by withdrawal from some market segments or market territories and a gradual re-specialization of existing companies in a limited number of products, niches or activities. The process of re-specialization of companies reflects not only accumulated portfolios of their capabilities (rather than resources) and the level of their interaction with the regional institutional environment, but also socialized and emergent bargaining processes between diverse actors within a given company (Fuller & Phelps, 2018). Thus, the portfolio of economic activities performed in a regional industry becomes increasingly specialized. However, in contrast to path downgrading, where the companies usually lose a significant part of their autonomy, path contraction implies that the key companies retain their know-how and high-value-added functions, including organization of production in other regions or countries, but limit their own production activities to specialized and demanding and hence often low-volume niches – see Treado’s (2010) study of the steel industry in Pittsburgh, South Birmingham’s automotive industry during the 1970s and 1980s (Smith, 1989), and the Teeside steel and chemical industry in the same period (Beynon et al., 1989, see Table 1). Path contraction has also been identified in the financial sector in Luxembourg, whose elite joined forces with national regulators to transform the city into a global centre specialized in the management of investment funds while low value-adding activities have been standardized to facilitate their offshoring (Dörry, 2015). Similar processes have been observed in the field of business services and software development, where less demanding activities have been transferred from metropolitan regions in advanced countries to capital regions in Central and Eastern Europe (Hardy & Hollinshead, 2016).

Table 1. The key dimensions of particular evolutionary pathways of a decline.

Path	Firm-level		Key dimensions			Region-level: Modification of the regional asset base		Empirical illustrations
	Strategies	Main drivers	Connectedness	Capital accumulation	Resilience	Main alterations	Future potential	Example	Reference
Down-grading	Removal of higher value-added functions	FDIs driven by low-cost strategies / intensive competitive pressure	Internal: decrease; external: increase	declining	declining	Loss of key assets	Risk of being trapped in lower-quality segments	South East of Bulgaria (apparel, 1990s) Non-metropolitan regions in Philippines (voice-based services, 2000–2013) Daegu (textile, 1980–2005)	Pickles, Smith, Bucěk, Roukova, and Begg (2006) Kleibert (2016) Hassink (2010)
Contraction	Reduction of product diversity / re-specialization	Weak competitiveness in the global economy	Internal: decrease; external: increase	declining	varying	Risk of over-specialized assets	Diminishing opportunities for diversification	Pittsburg (steel, 1980–2005) South Birmingham (automotive, 1970–1980s) Teeside (steel & chemicals, 1960–1980s) Luxembourg (financial services, 1990–2014)	Treado (2010) Smith (1989) Beynon et al. (1989) Dörry, (2015)
De-localisation	Relocation to more favourable locations	Better cost-capability ratio, more suitable frameworks or availability of resources at other locations	Ultimately declining (internally & externally)	Dis-investment	sharply declining	Broad destruction and de-locking of assets	Severe effects on long-term development potentials	La Laguna (textile, 2000–2010) S. Korea (retail, 1999–2015) S. Australia & Victoria (automotive, 2000–2017)	Bair & Werner (2011b) Coe et al. (2017) Beer (2018)

Source: Own compilation.

Path contraction implies that the level of connectedness of a given industry in the region tends to decrease, while the external connectedness can remain high or may even increase. By contrast, overall capital accumulation is likely to decrease due to a reduction in production capacities. Nevertheless, the key companies tend to make new investments in order to enhance their position in selected and demanding niches. Likewise, resilience of the remaining companies, which succeed in re-specialization in particular phases of production and/or in particular market niches, might increase.

This evolution concerns primarily large endogenous firms or even lead firms in global production networks, which struggle to reposition themselves within the global economy. However, given the prominent role of large companies in some regional economies, their suppliers might also be forced to narrow the portfolio of performed activities to reflect the changing demand from their larger customers. If the contraction of an industry into a narrower spectrum of activities does not lead to a recovery in economic performance, and if further negative changes occur, overall destruction of the regional industry may follow. On the other hand, this type of path development can also come with a significant functional upgrading of companies in a regional industry that successfully re-specialise in selected market niches and high-value-added functions, despite the overall shrinkage in the volume of production.

Path contraction activities are also likely to downsize the regional asset base. As outlined above, the external connectedness of a regional industry might shift, but it could also stay rather high, making the risk for ‘spatial myopia’ less severe. Similarly, re-specialization of key companies might in some cases lead to a stronger alignment of complementary assets and thereby to shrinking communication costs, stronger synergetic effects and possibly increasing returns. However, even though there might be some cases in which path contraction activities lead to (short-term) regional benefits, we assume that path contraction activities will harm a region’s long-term competitiveness, especially in cases when assets become ‘over-specialised’.

A declining range of competencies at the firm level will diminish the scope for knowledge re-combination and is therefore likely to impact negatively on regional development opportunities, as the regional variety of (related) industries shrinks (Boschma & Frenken, 2011). In cases of less successful re-specialization or over-specialization at the firm level, the subsequent effects might be severe (e.g. closure of connected firms), which might ultimately dissolve the whole regional industry, releasing or destroying an even broader range of (human, infrastructural or/and industrial) assets.

3.3. Path delocalization

The third type of negative path development encompasses relocation of key economic activities in a given regional industry, often followed by further disinvestment and brain drain processes. Delocalization is frequently epitomized as a move performed by footloose multinational companies in their quest for yet cheaper locations (Labrianidis, Kalantaridis, & Dunford, 2011). Arguably, this is contingent upon the level of their institutional embeddedness (Fuller & Phelps, 2018). Generally, in addition to technological and market developments, delocalization of foreign-owned and sometimes even of home-grown companies can also occur due to plant- and area-related factors (Bertoncin, Pase, Quatrida, & Scroccaro, 2018; Fothergill & Guy, 1990; Watts, 1991). Flecker and Schönauer (2016) showed how delocalization has been actively produced in the case of ICT services via modularization of digital work as well as through organizational restructuring processes. Thus, there is a close link between the design of labour processes, the quality of work and the spatial division of labour (Flecker & Schönauer, 2016).

Arguably, large-scale relocation of labour-intensive phases of production or whole industries to lower-cost locations has already been documented in the literature on deindustrialization of developed countries/regions (cfr. Fröbel, Heinrichs, & Kreye, 1978) and in the literature on global Fordism emphasizing the highly variegated nature and impact of industrialization of Third World countries (Lipietz, 1982). Delocalization may well lead to an overall dissolution of the regional industrial fabric and can result in the complete destruction of the regional industry (Bair & Werner, 2011b). Delocalization of leading firms may force their suppliers to follow them (Bertoncin et al., 2018), leading – in the worst case – to a disruption of the economic fabric of the region under consideration. Alberti’s (2006) study of the apparel industry in the industrial district of Como suggests that delocalization of economic activities may also accelerate the district’s decline through ‘self-created’ competitors, which benefitted from the transfer of tacit knowledge, skills and technology from Italy to China.

Generally, delocalization has been enhanced by a growing ‘footlessness’, especially of some manufacturing industries and services (such as voice-based and back-office services) in the contemporary globalized economy, and by the increase in vertical fragmentation of the production process, which has led to a widespread formation of global production networks in manufacturing as well as in services (e.g. Coe, 2014; Dörry, 2015; Flecker & Schönauer, 2016; Kleibert, 2016). Offshore services such as various voice-based, back-office services and software development are particularly footloose due to low sunk costs as well as due to limited local linkages (Hardy & Hollinshead, 2016; Kleibert, 2016).

However, delocalization has been documented also in other segments of the tertiary sector such as in retail. For example, Coe et al. (2017) unpack the multiplicity of causes leading to Tesco’s departure from South Korea in 2015. The authors point to severe processes of corporate restructuring and disruption in retail caused inter alia by new internet giants such as Amazon and Alibaba (cfr. also Wu & Gereffi, 2019).

In addition to the above-mentioned exit of FDI due to the emergence of new competitors with better cost-capability ratios or due to technological change (Yeung & Coe, 2015), another set of factors deserves attention. Delocalization may also be triggered by a broad spectrum of changes in the external environment as well as inside the region. For example, there is evidence for a profound influence of national economic regulation and policies, e.g. implementation of ‘Go up policy, go west policy, go out policy’ in China (Zhu & Pickles, 2014), changes in regulatory frameworks for international trade, environmental protection, etc.

Path delocalization may affect the level of connectedness in distinct ways. Connectedness within the region tends to decrease, while external connectedness can even increase in the early phase of delocalization before its drop in subsequent phases if the process of delocalization gains momentum. By contrast, capital accumulation as well as the resilience of the regional industry in question is likely to decrease due to the reduction in production capacities and capabilities.

Consequently, it can be argued that, given the high intensity of evolutionary dynamics when new industrial spaces frequently challenge the established industries in advanced regions, as well as the sharp international competition in the current, highly globalized economy, delocalization is bound to be a widespread phenomenon across a large variety of industries and services. According to Labrianidis et al. (2011) and Kleibert (2016), regions specialized in manufacturing and service industries with a high labour intensity, easy entry and relatively low value-added are particularly endangered by delocalization. On the other hand, as has been shown by Alberti (2006), delocalization can, at least under certain conditions, represent a mechanism allowing for catching-up processes by regions previously lacking any tradition in a particular industry. Subsequently, these emerging economic spaces can become strong competitors for regions with established traditions in a given industry.

This negative pathway might likely be the most severe form for the region’s future development potential in terms of its asset base, given the high probability of an overall dissolution of the regional industrial path. Especially in regions that show a strong specialization in industries facing delocalization, the danger of triggering a subsequent chain reaction is high. This might include a brain drain process (human assets), leaving supplier-firms (industrial assets) and further investments towards other regions that offer more promising prospects (MacKinnon, 2012). Additionally, strong alignments between the asset base and the industry that leaves the region might delay restructuring efforts (Grabher, 1993; MacKinnon, 2012). Another problem stems from the fact that regional asset bases oriented towards labour-intensive tasks and low value-added are already equipped with relatively weak potential for new path creation (Isaksen & Trippl, 2016). This makes these regions even more prone to long-term development struggles. Furthermore, natural assets are of special significance, as their depletion can be an important cause for path delocalization. Evidence from regions in developing countries (Auty, 1993) suggests that such a situation poses particular restructuring challenges (‘resource curse’).

As mentioned above, the modification of institutional assets (regulatory frameworks) might also be a reason for path delocalization. However, this relationship is ambiguous. In the GPN and FDI literature, the issue of power asymmetries between multinational corporations and regional actors have gained attention (Coe & Hess, 2011; Dawley, 2011). Due to their ‘footlessness’, multinational companies can essentially bargain with different regions (see MacKinnon, 2012; Phelps & Wood, 2006, for an analysis of how various intermediaries or coordinating agents acting on behalf of inward investors deal with local interests). This might lead to asset modifications complementary to the needs of powerful multinational companies, especially in regions with a rather weak asset base and therefore low bargaining power. However, such endeavours to adapt the asset base to the needs of powerful firms in order to prevent delocalization might come at the expense of other actors (Phelps, 2000, 2009), essentially leading to dependencies and a constraining environment for new path development.

To summarize, negative regional industrial path development may come in different forms, ranging from path downgrading to path contraction and delocalization. Arguably, this is an analytical distinction. In reality, one might observe a coalescence of the ‘ideal types’ discussed above (see, for instance, Alberti, 2006, for an analysis of the adverse trajectory of the apparel industry in Como, which points to a combination of path contraction and path delocalization). Table 1 provides an overview of the key dimensions and features for each trajectory and it points to empirical findings from case studies. Each of the three trajectories is illustrated by several cases, indicating that none of the pathways is confined to a specific sector but can rather be observed in various industrial and regional contexts as well as for different time periods.

4. Conclusions

In this paper, we seek to contribute to a better understanding of the ‘dark side’ (Phelps et al., 2018) of regional industrial path development. Scholarly work in this field has thus far mainly focused on explicating how (different types of) positive path development (that is, the rise and growth of new industries and the successful renewal of traditional ones) unfolds across space and over time. Systematic research on the negative pathways is missing.

We complement existing typologies of positive evolutionary pathways by identifying and conceptualizing three distinctive trajectories of decline, namely path downgrading, path contraction and path delocalization, The research emanates from the well-established argument that while in the past the main driver of uneven development has been the spatial division of labour based on inter-sectoral specialization of regions this has been latter superseded by within-sector specialization (Massey, 1979). Our arguments build on early work by Massey (1979) on spatial divisions of labour and insights provided by the global value chain/global production network frameworks, which conceptualize the drivers of uneven development under the paradigm of vigorous vertical disintegration of production (Yeung, 2016a, 2016b; Yeung & Coe, 2015). We also employ key concepts of evolutionary economic geography (lock-in, path dependence and resilience) as well as the core dimensions of the adaptive cycle model suggested by Martin and Sunley (2011), that is, connectedness, capital accumulation and resilience, and the role of economic actors in ‘producing’ these outcomes. Further, we draw on recent work on asset modification processes (MacKinnon et al., 2019; Trippl, Baumgartinger-Seiringer, et al., 2019) to better understand how negative forms of path development affect the wider regional asset base and thus the region’s future potential for path development.

Overall, the above-outlined typology of evolutionary trajectories of decline reflects different firm strategies in a given regional industry shaped by their power positions in national and/or global economic systems. It should be underlined that in practice the evolutionary trajectory of a particular industry in a particular region might consist of multiple and swinging shifts. Moreover, it is important to stress that – in the most severe cases – all three types of negative pathways might lead to an overall destruction of a given regional industry and the related institutional set-up, resulting in job losses and a profound alteration of the regional economic structure. However, it is vital to note that the pathways discussed above may also entail positive features. For example, path downgrading can lead to an increase in sales and to recovery of the regional industry, and path contraction can in some cases be associated with functional upgrading despite the shrinkage in total production.

Overall, there are strong reasons to argue that regions differ markedly in their capacity not only to cushion the adverse effects of declining paths but also to redeploy regional assets in novel ways to open up new growth trajectories reflecting inter alia the spatial hierarchy of ownership and control under the current paradigm of vertical disintegration of production. Regions with thick and diversified innovation systems and with capacities to embed powerful actors can be assumed to be best placed in this regard, while places with highly specialized or thin system structures may face severe challenges.

Arguably, more conceptual and empirical research is required to gain a deeper understanding of the sources, mechanisms and processes of negative path development in different types of regions. Future work in this field might seek to identify other distinctive trajectories of decline to complement those discussed in the present paper. We also see considerable value in developing a more dynamic approach to grasp various sequences of pathways of growth and decline in particular types of regional contexts.

Another issue for future research is to adopt a multi-actor approach to study the (changing) role of non-firm stakeholders. In addition, more conceptual and empirical work is needed to better understand the mechanisms that underpin various forms of negative path development, especially the conditions under which negative dynamics could be broken. Moreover, detailed studies might reveal variations across different types of regions and industries and comprehend influences of multi-scalar institutional environments.

Finally, a key challenge for future research is to elaborate on how policy at various spatial scales could target the dark side of path development in different spatial contexts. Declining industrial paths can easily turn the fate of entire regions, transforming them into disadvantaged and left-behind places, suffering from economic, social and political problems (Rodríguez-Pose, 2018). Developing adequate strategies and responses to negative forms of path development is thus a core task for policy.

Acknowledgement

The authors wish to thank to both referees and the editor for their insightful feedback. Usual disclaimer applies.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

Financial support of Grant Agency of The Czech Republic (no: 17-06621S) is greatly acknowledged. This work has also been supported by the Charles University Research Centre Program UNCE/HUM/018.

Notes

1 Arguably, other factors such as depletion of natural resources (Campling, 2012), environmental catastrophes like earthquakes, tsunamis, floods, hurricanes, and man-made pollution (Martin, 2012), or sudden disruption of economies and international trade, for example by civil unrest (Hough, 2011), can play a key role, as shown by studies employing a disarticulation perspective (Bair & Werner, 2011a, 2011b).