Show Summary Details

Page of

Printed from Oxford Research Encyclopedias, Business and Management. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 06 July 2022

Cluster Evolutionfree

Cluster Evolutionfree

  • Nydia MacGregorNydia MacGregorManagement Department, Leavey School of Business, Santa Clara University
  •  and Tammy L. MadsenTammy L. MadsenManagament Department, Leavey School of Business, Santa Clara University


A substantial volume of research in economic geography, organization theory, and strategy examines the geographic concentration of interconnected firms, industries, and institutions. Theoretical and empirical work has named a host of agglomeration advantages (and disadvantages) with much agreement on the significance of clusters for firms, innovation, and regional growth. The core assertion of this vein of research is that geographically concentrated factors of production create self-reinforcing benefits, yielding increasing returns over time. The types of externalities (or agglomeration economies) generally fall into four categories: specialized labor or inputs, knowledge spillovers, diversity of actors and activity, and localized competition. Arising from multiple sources, each of these externalities attracts new and established firms and skilled workers.

Along with recent advancements in evolution economics, newer research embraces the idea that the agglomeration mechanisms that benefit clusters may evolve over time. While some have considered industry and cluster life-cycle approaches, the complex adaptive systems (CAS) theory provides a well-founded framework for developing a theory of cluster evolution for several reasons. In particular, the content and stages of complex adaptive systems directly connect with those of a cluster, comprising its multiple, evolving dimensions and their interplay over time. Importantly, this view emphasizes that the externalities associated with agglomeration may not have stable effects, and thus, what fosters advantage in a cluster will change as the cluster evolves. Furthermore, by including a cluster’s degree of resilience and ability for renewal, the CAS lens addresses two significant attributes absent from cyclical approaches.

Related research in various disciplines may further contribute to our understanding of cluster evolution. Studies of regional resilience (usually focused on a specific spatial unit rather than its industrial sectors) may correspond to the reorganization phase associated with clusters viewed as complex adaptive systems. In a similar vein, examining the shifting temporal dynamics and development trajectories resulting from discontinuous shocks may explain a cluster’s emergence and ultimate long-term renewal. Finally, the strain of research examining the relationship between policy initiatives and cluster development remains sparse. To offer the greatest theoretical and empirical traction, future research should examine policy outcomes aligned with specific stages of cluster evolution and include the relevant levels and scope of analysis. In sum, there is ample opportunity to further explore the complexities and interactions among firms, industries, networks, and institutions evident across the whole of a cluster’s evolution.


  • Technology and Innovation Management

Clusters: Introduction and Definitions

“Formally, clusters are defined as groups of geographically proximate, interconnected firms and associated institutions, operating in a field” and “linked via commonalities and complementarities” (Porter, 2000, p. 16). Work demonstrates that the presence of geographic concentrations of related economic activity plays an important role in the innovativeness and growth of regions (Audretsch & Feldman, 1996a, 1996b; Delgado, Porter, & Stern, 2010; Feldman & Audretsch, 1999; McCann & Folta, 2008). The literature identifies two general forms of geographic clusters, industry-focused and technology-focused; these two forms yield different resource profiles over time (St. John & Pouder, 2006). Industry-focused clusters revolve around a single industry such as semiconductors, hotels, wine-making, or automobiles (e.g., Baum & Haveman, 1997; Kalnins & Chung, 2004; Klepper, 2010; Martin, Salomon, & Wu, 2010; Wang, Madhok, & Li, 2013). In this form, the life cycle of a single regionally concentrated industry proxies for the evolution of a cluster. In contrast, technology-focused clusters, such as Silicon Valley, include multiple sectors and industries. As a result, differences in the activities and evolution of heterogeneous actors in a broad field give rise to a cluster’s development over time (Bergman, 2006; Delgado et al., 2010; Feldman & Audretsch, 1999; Porter, 1998; Rosenthal & Strange, 2003). In other words, the cluster’s evolution is not defined by any single industry’s life cycle or stage of development.

In addition to different characterizations of a cluster’s evolution, scholars also frequently vary in their operational definitions of clusters. Some differences stem from challenges in compiling data that includes a cluster’s multidimensional and multilevel attributes (Martin & Sunley, 2003). Nonetheless, these differences make it difficult to articulate a cohesive pattern of results from the empirical literature. In response, several scholars suggest quantitative approaches to specifying clusters and their unique spatial and thematic boundaries, inclusive of multiregion and cross-country considerations (see Alcácer, 2006; Alcácer & Zhao, 2016; Delgado, Porter, & Stern, 2016). Alcácer and Zhao (2012, p. 4) recommend that definitions consider at least three factors: “the measure of economic activity in a location, the geographic unit over which economic activity should be measured, and the concentration threshold required to classify a location as a cluster” and propose a methodology for identifying clusters organically based on economic activities within a geographic unit. Delgado et al. (2016) develop a methodology for comparing the efficacy of different sets of cluster definitions and in turn, provide a set of benchmark definitions for clusters in the United States.

Agglomeration Externalities

The notion that organizations, whether similar or different, tend to concentrate economic activities geographically has a long tradition in organization studies, economics, and regional studies (Krugman, 1991; Marshall, 1920; Romer, 1986; Saxenian, 1994). A central premise of this literature is that when geographically concentrated factors of production are well established, the benefits of spatial proximity become self-reinforcing, yielding increasing returns over time (Krugman, 1991; Romer, 1986; Saxenian, 1994). These externalities, often referenced as agglomeration economies, arise from multiple sources and make a geographic locale or region attractive to established firms, new ventures, and workers. One category of externalities encourages the co-location of industries that require specialized labor and specialized inputs (Marshall, 1920). When similar types of firms agglomerate, they foster a pooled market for employees with industry-specific or specialized skills (Krugman, 1991). This thicker labor market allows firms to more efficiently fill employment vacancies. Likewise, the spatial concentration of firms and industries provides skilled workers access to more employment opportunities, reducing the chances that laid-off employees will remain unemployed. The lower risk exposure for workers attracts more workers, thereby advancing the benefits that accrue to clustered firms (David & Rosenbloom, 1990). Relatedly, local access to other specialized inputs reduces the transportation costs associated with intermediate goods while also providing opportunities for learning and knowledge sharing among firms, such as buyers and sellers, operating in adjacent industries (Harrison, Kelley, & Gant, 1996). A third agglomeration benefit stems from exposure to knowledge spillovers that arise via formal and informal interactions among actors operating in a cluster. Work demonstrates that geographically proximate firms benefit from an efficient flow of ideas, knowledge, and information (Alcácer, 2006; Jacobs, 1969; Saxenian, 1994). These conditions are especially attractive to resource-constrained start-ups. Another category of externalities is associated with the diverse mix of actors, activities, firms, technologies, and industries operating in a region. This diversity promotes the cross fertilization of activities within the region and, in turn, increases the likelihood of knowledge creation, recombination, and innovation (Delgado, Porter, & Stern, 2010; Dissart, 2003; Feldman & Audretsch, 1999; Frenken, Van Oort, & Verburg, 2007; see McCann & Folta, 2008).1 Additionally, a diverse portfolio of activities in a region makes it less dependent on a single industry’s viability and, in turn, less vulnerable to technological, institutional, or economic shocks. Finally, the degree of localized competition in a region is found to both positively and negatively influence regional innovative activity. For instance, studies show that firm performance tends to increase as the number of firms operating in a cluster grows (Folta, Cooper, & Baik, 2006), whereas other studies suggest that as a cluster becomes more congested, competition for valuable inputs ensues and, in turn, diseconomies of agglomeration arise (Zucker, Darby, & Brewer, 1999). In sum, extant research suggests that the main features fostering the success of firms and industries within a cluster are knowledge spillovers, specialized labor, a diverse set of actors and functions, as well as localized competition.

Implications for Understanding Cluster Evolution

Scholars theorize that the externalities that enhance a cluster’s attractiveness differ in strength as a cluster evolves (Folta, Cooper, & Baik, 2006; Martin & Sunley, 2011; Maskell & Malmberg, 2007; Menzel & Fornahl, 2010; Pouder & St. John, 1996). However, most empirical and theoretical work in strategy and organization theory assumes agglomeration economies have similar effects regardless of a cluster’s stage of development (e.g., emergence, growth, maturity, decline, renewal or transition, etc.) (exceptions, Folta et al., 2006; MacGregor & Madsen, 2016; Pouder & St. John, 1996; see McCann & Folta, 2008 for a review). Furthermore, studies typically overlook temporally specific conditions such as major events or shocks (whether technological, institutional, or market based) that interrupt a cluster’s course of development and the industries operating within it. Perhaps more concerning is that the majority of the empirical work does not consider temporality when constructing a data set for analysis. Stated another way, data collection and measurement ignore whether a cluster is in an emergence, growth, mature, or decline stage. As a result, many tests of the influence of agglomeration externalities lead to erroneous conclusions about those effects and the significance of a cluster’s development. Complicating matters, researchers in different disciplines typically analyze externalities that are salient to their fields, overlooking the full set of factors driving cluster evolution. For instance, several studies examine specialization, diversity, and localized competition effects but omit other effects (see Rosenthal & Strange, 2003). Other studies focus on localization economies and the asymmetric benefits of agglomeration but overlook diversity (e.g., Shaver & Flyer, 2000). Differences in the research and operational designs obscure the conditions under which agglomeration advantages (or disadvantages) persist or are more temporary in nature, providing an incomplete theory of cluster evolution. In sum, if agglomeration externalities differ in influence as a cluster evolves, then controlling for a cluster’s stage of evolution is warranted. Doing so is a critical step in enabling scholars to extract a coherent pattern of results from the empirical literature.

Studies also often frame the outcomes of agglomeration economies and diseconomies, such as firm founding and mortality, as inputs rather than outcomes of agglomeration dynamics. While entry and exit events affect the density of a cluster, density dynamics do not generate a self-reinforcing cycle of spatial benefits that yield increasing returns over time. In fact, work suggests that externalities associated with specialization promote founding during a cluster’s early stages of development but, beyond some threshold give rise to contestation and, in turn, the shakeout of firms (Feldman & Audretsch, 1999). In this case, the externality is the primary mechanism underlying changes in the cluster over time rather than structural dynamics.

Renewed interest in the phenomenon of the geographic agglomeration of firms emerged at the end of the 20th century. Theoretical and empirical work has named a host of agglomeration advantages (or disadvantages) with much agreement on the significance of clusters for innovation, regional growth, and competitive advantage (Porter, 2000). Still newer is the idea that clusters may not only afford benefits (and drawbacks) for firms but that these same effects may change over time.

Cluster Evolution: Different Views

Many studies on cluster evolution concentrate on cluster formation or emergence. This work is motivated, in part, by an interest in understanding how to build or mimic a productive cluster (for instance, see Bresnahan, Gambardella, & Saxenian, 2001; Saxenian, 1994). Yet, analysis suggests that factors and processes crucial to a cluster’s emergence differ from those critical to its functionality over time (Audretsch & Feldman, 1996b; Bresnahan, Gambardella, & Saxenian, 2001; Menzel & Fornahl, 2010). For instance, some work shows that externalities do not persist, whereas other studies conclude that some benefits become diseconomies as a cluster evolves (Holm & Østergaard, 2015; MacGregor & Madsen, 2016; Martin & Sunley, 2006). These findings yield at least two observations: (1) work on externalities alone is insufficient to explain how a cluster evolves and (2) a cluster’s development involves discrete phases, each with distinctive attributes. These observations motivate several questions: What are the life-cycle stages of a cluster? When do clusters grow, decline, or renew themselves? What are the primary drivers of a cluster’s evolution, within and across stages? The subsequent section highlights four approaches used to study cluster evolution.

Industry Life-Cycle Model

Recent scholarship has embraced evolutionary models and methods to explain the development of clusters (e.g., Boschma & Martin, 2007; Potter & Watts, 2010). As noted, one view suggests that a cluster’s evolution mimics a traditional industrial life cycle with stages of birth, growth, maturity, and decline (Audretsch & Feldman, 1996b; Klepper, 2010; Potter & Watts, 2010; Wang et al., 2013). Initial conditions and the industry’s technology life cycle inform these stages. During an industry’s birth stage, firms typically are not spatially concentrated and a cluster does not exist. However, with industry growth, geographic agglomerations of firms and institutions may begin to emerge (Klepper, 2007). As the benefits of co-location become more pervasive, individual firms grow, advancing spatial concentration. This growth also attracts entry, either de novo or de alio (via spin-offs and spin-outs), increasing the overall size of an industry (Klepper & Graddy, 1990) and in turn, the cluster. After early and accelerated growth, path dependence may emerge from the accumulated advantages, benefiting the cluster. Over time, the increasing density of firms operating in the cluster gives rise to contestation and the product innovation rate declines (Audretsch & Feldman, 1996b, p. 253), signaling the beginning of the cluster’s decline.

Applying an industry life-cycle model to understand cluster evolution overlooks several critical attributes of clusters. First, clusters typically encompass multiple industries each evolving at their own rate. How spatially proximate firms and industries exploit this diversity distinguishes their growth and development from non-clustered firms and industries (Menzel & Fornahl, 2010). Diversity is also crucial to avoiding lock-in (to a previously successful and now overly rigid trajectory) and in turn, to a cluster’s long-run growth. On a related note, the interplay of heterogeneous actors and institutions operating in a cluster give rise to focal points of activity and underlie a cluster’s thematic boundaries. As such, changes in actors and their networks matter for a cluster’s ongoing development but the industry life-cycle model overlooks these micro dynamics (Fornahl, Hassink, & Menzel, 2015). Additionally, when faced with similar market and technological conditions, work indicates that some clusters thrive whereas others fade (Menzel & Fornahl, 2010) and that clusters belonging to the same industry can have distinct growth paths (Saxenian, 1994). Studies also demonstrate that clustered firms tend to outperform non-clustered firms at the beginning of the life cycle and suffer weaker performance in later stages (Audretsch & Feldman, 1996b; Potter & Watts, 2010). Finally, Potter and Watts (2010) demonstrate that, during the late stages of the industry life cycle, agglomeration economies foster diminishing returns to an economic region. In combination, these observations and findings imply that the drivers of a cluster’s evolution may not emulate the mechanisms moving an industry through its life cycle (Martin & Sunley, 2011).

Cluster Life-Cycle Model

An alternative view is that agglomeration advantages (or disadvantages) specific to clusters inform the stages of a cluster’s evolution and distinguish it from that of an industry life cycle (Iammarino & McCann, 2006; Martin & Sunley, 2011; Pouder & St. John, 1996). Although this view identifies similar life-cycle stages for a cluster (emergence, growth, maturity, and decline), the mechanisms prescribing each stage and influencing the shift from one stage to another differ. During emergence, de novo or de alio (spin-offs and spin-outs) firms co-locate in geographic space. As the benefits of agglomeration develop, additional entry ensues while weak competitors exit. Thus, the pool of resources and capabilities characterizing the cluster is continuously changing. These macro and micro dynamics, influenced by the extent of agglomeration advantages (or disadvantages), serve as key mechanisms moving the cluster through different life-cycle stages (Martin & Sunley, 2011). However, if these dynamics persist, a cluster may not evolve through all life-cycle stages; as the consequential shifting of the composition of firms and industries evolves there is little likelihood that the system will follow a neat life-cycle trajectory. Thus, conditions call into question the efficacy of employing the life-cycle model as a theory of cluster evolution. This observation has prompted scholars to take a closer look at the origins of cluster dynamics and their interrelationships (see Martin & Sunley, 2011, for a review). Despite refinements in understanding, a general conclusion is that the life-cycle model is a useful starting point for describing a cluster’s development but ultimately limited in its ability to fully explain cluster evolution.

Complex Adaptive Systems Model

In response to the concerns noted previously, emerging work develops a theory of cluster evolution based on complex adaptive system theory (Carbonara, Giannoccaro, & McKelvey, 2010; He, Rayman-Bacchus, & Wu, 2011; Martin & Sunley, 2007, 2011, 2012). These open, nonlinear dynamic systems comprise autonomous, heterogeneous actors (agents) whose functions and self-reinforcing interactions give rise to the system’s overall identity and fuel its evolution (Martin & Sunley, 2011; McKelvey, 1999; Simon, 1996). Clusters display many of the generic features of complex adaptive systems (Martin & Sunley, 2007). To begin, clusters include a variety of geographically proximate, interconnected firms and organizations with different goals and attributes—producers, buyers, suppliers, partners, institutions, and advisers—that seek to optimize their fitness. The interactions between and among various actors (such as specialized suppliers, firms in related industries, universities, and financial and legal institutions) may be competitive or cooperative. Each of these entities maintains its own development path. However, the co-evolutionary relationships among these different actors contribute to reciprocal and nonlinear change as each individual actor strives to continuously improve its resources and activities over time (Anderson, 1999; McKelvey, 1999). These interconnections and feedback loops foster locally based interdependencies, a hallmark of both complex systems and clusters. It follows that the reciprocal interactions among the shifting micro- and macro-level aspects of a cluster influence its functionality and progress. For instance, a recent case study of two clusters finds that positive feedback loops bolster the micro actions of actors (He, Rayman-Bacchus, & Wu, 2011). In this way, the complex adaptive systems model emphasizes the roles of endogenous and exogenous mechanisms at the firm, industry, and regional levels in pulling a cluster through various stages of development. Guided by this self-organizing process of adaptation, the effects of different mechanisms that underlie the system may change over time depending on the stage of its development (Anderson, 1999). For example, at one point, interconnectedness among firms may promote innovation and growth, but subsequently, high levels of interconnectedness may reduce innovation and trigger stagnation. Thus, path dependence, lock-in, and other boundary conditions may limit self-organizing processes (He et al., 2011; Martin & Sunley, 2006). This view emphasizes that the externalities associated with agglomeration may not have stable effects. In other words, what fosters advantage in a cluster will change as the cluster evolves. Further, the self-organizing and regeneration character of clusters as complex adaptive systems distinguishes them from the cluster life-cycle model whose stages adhere to a set order and trajectory: emergence, growth, maturity, and eventual demise.

Accordingly, this view characterizes a cluster’s evolution as an adaptive cycle involving emergence, “growth, conservation, decline and release, and reorganization” (labels adapted from Martin & Sunley, 2011, p. 1307; see also Martin & Sunley, 2012). The stages include three fundamental attributes (each with different rates of change): “(1) the potential accumulated resources available to the system; (2) the internal connectedness of system components; and (3) resilience, a measure of system vulnerability to and recovery from shocks, disturbances, and stress” (Martin & Sunley, 2011, p. 1306). The growth phase entails a process of gradual, and relatively predictable, development fueled by the entry of firms, the accumulation of resources, an increasing interconnectedness among actors, and high adaptability. In the subsequent conservation stage, the system enters a period of stasis where resource accumulation slows and adaptability declines. A period of contraction (decline and release) follows where firms and resources depart, interconnections among actors erode, and the system’s heterogeneity dampens. Eventually, the system enters a reorganization phase, initiating a process of renewal and a new growth cycle.

While viewing cluster evolution as a complex adaptive system does address some of the limitations of the other conceptual approaches, it raises other questions and remains underdeveloped. For instance, the boundaries of a dynamic system, with multilevel interactions and exchange between agents inside and outside the cluster, shift over time. As Kauffman (1993) recounts, the process of the coevolution of actors entails the adaptations of each that then alter the environment for others (Carbonara, Giannoccaro, & McKelvey, 2010). Temporally specific boundaries present a challenge to defining who or what is in or out of a cluster. Similarly, to fully understand the dynamism at various scales requires examination at local, regional, national, and potentially international levels and across industries and sectors.

Emerging Views: Micro Dynamics of Cluster Life Cycles

As noted, life-cycle or cyclical approaches to understanding cluster evolution have been criticized for their inability to explain why clusters develop differently under similar conditions (Saxenian, 1994), and why some clusters and industries do not follow a life-cycle pattern (Piore & Sabel, 1984). Critics also argue that the general structural dynamics that inform the traditional life-cycle model provide an incomplete picture of a cluster’s development (see Fornahl, Hassink, & Menzel, 2015 for a review). In response, an emerging stream of work in economic geography advocates shifting attention to the micro dynamics of a cluster and argues that a cluster’s evolution is largely influenced by the interdependencies of three interrelated factors: actors, networks, and institutions. In this view, the composition and structure of actors, networks, and institutions differ across life-cycle stages. Furthermore, the behaviors, that is, the agency, of heterogeneous actors within the cluster drive these differences (Boschma & Fornahl, 2011; Martin & Sunley, 2011). Similar to a complex adaptive systems model, scholars argue that the co-evolution (i.e., reciprocal adaptations) of firm capabilities, network structures, and the institutional environment explains a cluster’s transition through different life-cycle stages (Maskell & Malmberg, 2007; Menzel & Fornahl, 2010; Ter Wal & Boschma, 2011).

These ideas resonate with work in strategy and organization theory that identifies the interplay of heterogeneous actors, networks, and institutions as crucial to a cluster’s formation and subsequent development (for instance, see Delgado, Porter, & Stern, 2010; Feldman & Audretsch, 1999; Porter, 1998; Saxenian, 1994). A related line of work in economic geography calls attention to an additional layer of network ties that inform a cluster’s evolution. In this case, small, and spatially independent, clusters are connected to other small clusters or networks across a broad geographic space. Interconnections in this “supranetwork” provide an extra source of inputs that might influence the activities within a cluster and, in turn, its progress (Boix, Hervás-Oliver, & De Miguel-Molina, 2015; Vittoria & Lavadera, 2014).


Table 1 summarizes the various approaches to studying cluster evolution. In our construal, complex adaptive systems (CAS) theory provides the most robust lens for developing a theory of cluster evolution for several reasons.

Table 1. Models of Cluster Evolution

Industry Life-Cycle Model

Cluster Life-Cycle Model

Complex Adaptive Systems

Cluster Life Cycle: Micro-Dynamics View

Cluster Form


Technology-Focused or Sector/Multi-Industry Focused

All Forms

All Forms

Stages of Development

Birth, Growth, Shakeout, Maturity, Decline

Emergence, Growth, Maturity, Decline

Emergence, Growth, Conservation, Decline, Release, Reorganization

Emergence, Growth, Maturity, Decline, Transformation

Theoretical Roots

Industry Evolution

Evolutionary Ecology

Agglomeration Dynamics

Complex Adaptive Systems Theory

Evolutionary Economics, Economic Geography, Network Theory, Institutional Theory

Primary Endogenous Sources of Evolution

Industry-Specific Structural Dynamics (Entry & Exit)

Firm Heterogeneity

Agglomeration Externalities

Different Rates of Change of Firms and Industries

Firm Heterogeneity

Self-Organizing Process of Adaptation (Self- Reinforcing Interactions Among Heterogeneous, Autonomous Actors)

Co-Evolution of Actors, Networks, and Institutions

Changes in Heterogeneous Actors Over Time

Sample Articles

Audretsch and Feldman (1996b), Klepper (2007), Klepper (2010), Potter and Watts (2010), Wang et al. (2013)

Iammarino and McCann (2006), Pouder and St. John (1996)

Carbonara, Giannoccaro, and McKelvey (2010), He, Rayman-Bacchus, and Wu (2011), Martin and Sunley (2007, 2011, 2012)

Boix, Hervás-Oliver, and De Miguel-Molina (2015), Boschma and Fornahl (2011), Vittoria and Lavadera (2014)

Specifically, the attributes of complex adaptive systems align with those of a cluster. In this way, the theory encompasses a cluster’s multiple evolving dimensions and their interactions over time. The CAS lens also considers a cluster’s degree of resilience (adaptability) and its ability to renew itself, two attributes absent from cyclical approaches. Despite these advantages, additional work is warranted to fine-tune our understanding of clusters as complex adaptive systems (see Martin & Sunley, 2011).

Related Research and Future Directions

Resilience of Regions and Clusters

Economic geographers have become increasingly interested in an evolutionary view of the resilience of geographic regions, roughly defined as the extent to which a region or cluster is able to absorb and/or recover from a shock over the long run (Boschma & Martin, 2007, 2010; Bristow & Healy, 2014; Hill et al., 2012; Martin, 2012; Simmie & Martin, 2010; for a recent review, see Boschma, 2015). The nascent work in this area varies in focus, with some studies attending to critical events such as national disasters and others concentrating on specific economic jolts to a region. Although the evolutionary view of regional resilience remains “a work in progress” (Boschma, 2015, p. 734), and regions and technology-focused clusters are not necessarily equivalent, conclusions from this work are relevant to cluster evolution. The emerging literature provides at least three different views of regional resilience. The first view draws from engineering where resilience is the ability of a system to return to a preexisting stable equilibrium state after a shock (Boschma, 2015, p. 735). Building on ecology, a second view also emphasizes a region’s ability to change in response to a shock but invokes multiple equilibria where a region or cluster shifts from one steady state to another over time. Economic geographers challenge these perspectives, arguing that they tend to overlook how changes in elements that make up a region or cluster (actors, organizations, structures, institutions, etc.) affect its long-term evolution (Martin, 2012). The criticisms of the first two approaches inspired an evolutionary view of regional resilience where resilience is an ongoing process rather than recovery to a previous, or new, equilibrium state. In this view, resilience includes a trade-off between adaptation, a region’s ability to change within existing growth paths, and adaptability, a region’s ability to develop new growth paths. Thus, the emphasis is on a region’s ability to sustain long-term development in response to a shock (Boschma, 2015; Martin & Sunley, 2015; Simmie & Martin, 2010). In sum, this stream of work suggests that a region’s industrial diversity, rather than related diversity (e.g., adjacent linkages among its industries), dissipates risk and lowers the probability that a sector-specific shock will negatively influence a region’s economy as a whole (Boschma, 2015; Diodato & Weterings, 2015; Dissart, 2003; Frenken, Van Oort, & Verburg, 2007). While intriguing, these ideas merit further conceptual and empirical examination to understand when and where they are applicable and how they integrate with other views of cluster evolution.

Role of Shocks

Shocks, institutional, technological, market-based, or natural, are sudden events that dramatically alter the operations, structure, and evolution of an industry, cluster, or region. Such disturbances also may emerge gradually, precipitated by a sequence of events (Dosi, 1982). For instance, a natural disaster may occur abruptly, whereas a significant technological discontinuity might stem from either a gradual process or a single punctuated event. Surprisingly, empirical work in organization theory and strategy overlooks the influence of temporally specific conditions such as shocks, whether exogenous or endogenous, on a cluster’s development (exceptions, Buenstorf & Fornahl, 2009; Holm & Østergaard, 2015; MacGregor & Madsen, 2013, 2016).2

As noted, when geographically concentrated factors of production are well established, the benefits of spatial proximity become self-reinforcing, yielding increasing returns over time (Krugman, 1991; Romer, 1986; Saxenian, 1994). However, a major shock can disrupt the established industrial activities in a cluster and, in turn, the dynamic process of increasing returns that supports the cluster’s growth. As a result, the positive economies associated with the geographic concentration of different types of activities may differ in influence before and after a cluster experiences a shock. Scholars have recently begun to explore how shocks affect a cluster’s evolution. For instance, in an empirical study of a technology-focused cluster (Silicon Valley) before and after a major shock (the Internet bust of 2000), MacGregor and Madsen (2016) show that the shock disrupts the self-reinforcing benefits associated with agglomeration. In particular, the type of diversity-based economies that positively influence growth before the shock matter less in the cluster’s post-shock recovery period and vice versa. Exploring regional resilience, Holm and Østergaard’s (2015) study of the Danish Information and Communications Technology (ICT) sector before and after the Internet bust demonstrates that geographic areas populated with small and young ICT service firms grew faster than other areas pre and post shock. Their findings also reveal that healthy regional growth before a shock does not necessarily translate to strong regional resilience post-shock. Despite progress, this area of inquiry remains ripe for further analysis. For one, the implications of different types of shocks on agglomeration dynamics and a cluster’s progress are underexplored. Additionally, a deeper understanding of how the timing and scope of a shock affects new firm formation, industry churn, and a cluster’s recovery is warranted. Advancing our understanding of the type of agglomeration economies that matter most before and after a cluster experiences a fundamental shock also is crucial to policymakers pursuing regional development initiatives (Essletzbichler, 2007).

Policy Implications

Since the 1990s, the success of well-known regional industrial clusters has garnered much attention (Krugman, 1991; Porter, 1998; Saxenian, 1994). In particular, policymakers across nations, states, and regions have implemented a variety of measures in hopes of bolstering existing regional industrial clusters or shepherding emerging clusters in order to improve economic outcomes (Brenner & Schlump, 2011). As an example, the European Commission encourages regional stakeholders to implement policies that leverage clusters in “promoting regional industrial modernisation, supporting the growth of SMEs and encouraging smart specialization” (European Commission, 2016).

The intent is that cluster policies stimulate the theorized mechanisms that produce advantages for firms and industries within the cluster. They typically complement traditional sector or science, technology, or innovation policies by adding a spatial focus. Besides general aid for cluster activities and infrastructure, the set of conceivable policy measures includes direct and indirect financial subsidies (for R&D or start-up), operational assistance, education, and networks and cooperation (Brenner & Schlump, 2011). Nishimura and Okamuro (2011) note, however, that more recently the emphasis has narrowed, concentrating on networking and coordination support. This shift reflects the recognition that innovation, in particular, is embedded in a nexus of interaction and collaboration among a variety of actors (such as buyers and suppliers, employees, research institutions, and regulatory agencies) (Aranguren, De La Maza, & Parrilli, 2014; Lundvall, 1992). Is it clear, though, that policy measures such as these bring about the desired outcomes?

While cluster policy initiatives are associated with some benefits for firms and industries within a focal region, rigorously demonstrated causal linkages are rare. On the one hand, rich case-based and correlational studies have collected evidence that some advantages can be gained via targeted cluster policies. On the other hand, the same empirical challenges for generally understanding the effects of agglomeration advantages (or disadvantages) in regional industrial clusters apply here: little coherence to outcomes studied, varied units of analysis due to the complex multileveled phenomena, and little attention to temporal dynamics. For instance, individual studies have examined and found a positive relationship between specific cluster policies and innovation activity (Falck, Heblich, & Kipar, 2010), firm productivity (Aranguren et al., 2014), and participation in key knowledge networks (Fornahl & Morrison, 2015) for firms within the cluster. One study of the outcomes of cluster policies in Japan serves as an exemplar. Nishimura and Okamuro (2011) showed that national networking and coordination support programs fostered new collaborative networks among firms within Japanese clusters. Additionally, networking and coordination support and direct R&D subsidies increased sales transactions, new products, and process innovation. The results, however, are complicated. R&D subsidies, while strongly tied to improving the technological capabilities of participating firms, nonetheless had a smaller marginal effect on commercial success and innovative activity than the indirect/coordination programs. In other work, Fornahl and Morrison (2015) demonstrate that German biotechnology firms in clusters disproportionately benefit from R&D subsidies not due solely to agglomeration economies (such as knowledge spillovers), but also because they are favored to receive national R&D subsidies (compared to those outside the cluster) and are centrally positioned in global R&D collaborative networks. These intriguing findings begin to unpack difference between actors inside and outside clusters, effects of policy programs at different administrative levels (EU vs. national), and distinctions between desired outcomes (such as R&D intensity and network affiliation).

Thus, several studies on cluster policies are promising, yet the findings underscore how much remains to be explored. Invoking the view that clusters embody the qualities of complex adaptive systems (CAS) suggests that future research on cluster policy falls along the lines of CAS characteristics (see Martin & Sunley, 2007, 2011). Specifically, studies on the outcomes of a variety of cluster policies must address the effects for the heterogeneous set of actors, the multilevel interactions, the stage of development and temporal dynamics, and the various important outcomes for different units of analysis (firms, region, or cluster). Furthermore, the evolutionary nature of clusters points to the likely evolution of the boundaries of these systems.3 Each of these aspects could expand or shrink over time (distinct from the evolutionary stage of development of the cluster). For example, the growth of the urbanized area in which a cluster resides could coincide with the geographic spread of firms, institutions, and labor. In another possibility, clusters may maintain linkages to other similar clusters or larger networks, stretching the relevant span of interaction. Examples include connected creative clusters (Boix, Hervás-Oliver, & De Miguel-Molina, 2015), international R&D networks (Fornahl & Morrison, 2015), and technological similarity (Caragliu & Nijkamp, 2016). Regardless how defined, future analysis on the relationship between clusters and policy needs “sufficient scope for interaction and coherence to enhance collective external economies,” the critical advantages of clusters (Asheim, Smith, & Oughton, 2011, p. 885; Martin & Sunley, 2003). The dearth of work evaluating the intersection of cluster policies and cluster life-cycle is impetus for further examination.


A variety of emerging research across various disciplines may inform our understanding of cluster evolution. The mechanisms that promote regional resilience (typically characterized as part of a geographic area rather than its associated industries) may have similarities to the reorganization mechanisms associated with clusters viewed as complex adaptive systems. Parallel studies on discontinuous shocks (sudden events that alter the previous developmental trajectory and structure of clusters) may also help explain the manner in which clusters emerge and ultimately renew their long-term progression. Lastly, to offer the greatest theoretical and empirical traction, research on the relationship between policy initiatives and cluster development needs to correspond to the stages of cluster evolution and include the relevant levels and scope of analysis. Certainly the interest and significance of clusters for global, regional, and local growth, as well as the potential opportunities for firms in clusters, motivate substantial, multidisciplinary future research.


  • Alcácer, J. (2006). Location choices across the value chain: How activity and capability influence collocation. Management Science, 52, 1457–1471.
  • Alcácer, J., & Delgado, M. (2016). Spatial organization of firms and locations through the value chain. Management Science, 62(11), 3213–3234.
  • Alcácer, J., & Zhao, M. (2012). Local R&D strategies and multilocation firms: The role of internal linkages. Management Science, 58, 734–753.
  • Alcácer, J., & Zhao, M. (2016). Zooming in: A practical manual for identifying geographic clusters. Strategic Management Journal, 37, 10–21.
  • Anderson, P. (1999). Complexity theory and organization science. Organization Science, 10, 216–232.
  • Aranguren, M. J., De La Maza, X., & Parrilli, M. D. (2014). Nested methodological approaches for cluster policy evaluation: An application to the Basque Country. Regional Studies, 48, 1547–1562.
  • Asheim, B. T., Smith, H. L., & Oughton, C. (2011). Regional innovation systems: Theory, empirics and policy. Regional Studies, 45, 875–891.
  • Attaran, M., & Zwick, M. (1987). The effect of industrial diversification on employment and income: A case study. Quarterly Review of Economics and Business, 27, 38–54.
  • Audretsch, D., & Feldman, M. (1996a). R&D spillovers and the geography of innovation and production. American Economic Review, 86, 253–273.
  • Audretsch, D., & Feldman, M. (1996b). Innovative clusters and the industry life cycle. Review of Industrial Organization, 11, 253–273.
  • Baum, J. A. C., & Haveman, H. A. (1997). Love thy neighbor? Differentiation and agglomeration in the Manhattan hotel industry, 1898–1990. Administrative Science Quarterly, 42, 304–338.
  • Bergman, E. M. (2006). The sustainability of clusters and regions at Austria’s accession edge. In Z. Bochniarz & G. B. Cohen (Eds.), The environment and sustainable development in the new Central Europe. New York: Berghahn.
  • Blau, P. (1977). Inequality and heterogeneity. New York: Free Press.
  • Boix, R., Hervás-Oliver, J. L., & De Miguel-Molina, B. (2015). Micro-geographies of creative industries clusters in Europe: From hot spots to assemblages. Papers in Regional Science, 94, 753–772.
  • Boschma, R. (2015). Towards an evolutionary perspective on regional resilience. Regional Studies, 49, 733–751.
  • Boschma, R., & Fornahl, D. (2011). Cluster evolution and a roadmap for future research. Regional Studies, 45, 1295–1298.
  • Boschma, R., & Martin, R. (2007). Editorial: Constructing an evolutionary economic geography. Journal of Economic Geography, 7, 537–548.
  • Boschma, R. A., & Martin, R. (2010). The handbook of evolutionary economic geography. Cheltenham, UK: Edward Elgar.
  • Brenner, T., & Schlump, C. (2011). Policy measures and their effects in the different phases of the cluster life cycle. Regional Studies, 45, 1363–1386.
  • Bresnahan, T., Gambardella, A., & Saxenian, A. (2001). “Old economy” inputs for “new economy” outcomes: Cluster formation in the new Silicon Valleys. Industrial and Corporate Change, 10(4), 835–860.
  • Bristow, G., & Healy, A. (2014). Regional resilience: An agency perspective. Regional Studies, 48, 923–935.
  • Buenstorf, G., & Fornahl, D. (2009). B2C-bubble to cluster: The dot-com boom, spinoff entrepreneurship, and regional agglomeration. Journal of Evolutionary Economics, 3, 349–378.
  • Caragliu, A., & Nijkamp, P. (2016). Space and knowledge spillovers in European regions: The impact of different forms of proximity on spatial knowledge diffusion. Journal of Economic Geography, 16, 749–774.
  • Carbonara, N., Giannoccaro, I., & McKelvey, B. (2010). Making geographical clusters more successful: Complexity-based policies. Emergence: Complexity and Organization, 12, 21.
  • David, P. A., & Rosenbloom, J. L. (1990). Marshallian factor market externalities and the dynamics of industrial localization. Journal of Urban Economics, 28, 349–370.
  • Delgado, M., Porter, M. E., & Stern, S. (2010). Clusters and entrepreneurship. Journal of Economic Geography, 10, 1–24.
  • Delgado, M., Porter, M. E., & Stern, S. (2016). Defining clusters of related industries. Journal of Economic Geography, 16(1), 1–38.
  • Diodato, D., & Weterings, A. B. R. (2015). The resilience of regional labour markets to economic shocks: Exploring the role of interactions among firms and workers. Journal of Economic Geography, 15(4), 723–742.
  • Dissart, J. C. (2003). Regional economic diversity and regional economic stability: Research results and agenda. International Regional Science Review, 26, 423–446.
  • Dosi, G. (1982). Technological paradigms and technological trajectories: A suggested interpretation of the determinants and directions of technical change. Research Policy, 1, 147–162.
  • Edquist, C. (2005). Systems of innovation: Perspectives and challenges. In J. Fagerberg, D. Mowery, & R. Nelson (Eds.), The Oxford handbook of innovation (pp. 1–26). Oxford: Oxford University Press.
  • Essletzbichler, J. (2007). Diversity, stability and regional growth in the United States, 1975–2002. In K. Frenken (Ed.), Applied evolutionary economics and economic geography (pp. 203–239). Cheltenham, UK: Edward Elgar.
  • European Commission. (2016). Cluster policy: Smart guide to cluster policy.
  • Falck, O., Heblich, S., & Kipar, S. (2010). Industrial innovation: Direct evidence from a cluster-oriented policy. Regional Science and Urban Economics, 40, 574–582.
  • Feldman, M. P., & Audretsch, D. (1999). Innovation in cities: Science-based diversity, specialization and localized competition. European Economic Review, 43, 409–429.
  • Folta, T. B., Cooper, A. C., & Baik, Y. S. (2006). Geographic cluster size and firm performance. Journal of Business Venturing, 21, 217–242.
  • Fornahl, D., Hassink, R., & Menzel, M. P. (2015). Broadening our knowledge on cluster evolution. European Planning Studies, 23, 1921–1931.
  • Fornahl, D., & Morrison, A. (2015). Another cluster premium: Innovation subsidies and R&D collaboration networks. Research Policy, 44, 1431–1444.
  • Frenken, K., Van Oort, F., & Verburg, T. N. (2007). Related variety, unrelated variety and regional economic growth. Regional Studies, 41, 685–697.
  • Harrison, B., Kelley, M., & Gant, J. (1996). Innovative firm behavior and local milieu: Exploring the intersection of agglomeration, firm effects, industrial organization, and technological change. Economic Geography, 72, 233–258.
  • He, Z., Rayman-Bacchus, L., & Wu, Y. (2011). Self-organization of industrial clustering in a transition economy: A proposed framework and case study evidence from China. Research Policy, 40, 1280–1294.
  • Hill, E., Clair, T. S., Wial, H., Wolman, H., Atkins, P., Blumenthal, P., . . . Friedhoff, A. (2012). Economic shocks and regional economic resilience. In N. Pindus, M. Weir, H. Wial, & H. Wolman (Eds.), Building resilient regions: Urban and regional policy and its effects (Vol. 4, pp. 193–274). Washington, DC: Brookings Institution.
  • Holm, J. R., & Østergaard, C. R. (2015). Regional employment growth, shocks and regional industrial resilience: A quantitative analysis of the Danish ICT sector. Regional Studies, 49, 95–112.
  • Iammarino, S., & McCann, P. (2006). The structure and evolution of industrial clusters: Transactions, technology and knowledge spillovers. Research Policy, 35, 1018–1036.
  • Jacobs, J. (1969). The economy of cities. New York: Random House.
  • Kalnins, A., & Chung, W. (2004). Resource seeking agglomeration: A study of market entry in the lodging industry. Strategic Management Journal, 25(7), 689–699.
  • Kauffman, S. A. (1993). The origins of order: Self-organization and selection in evolution. New York: Oxford University Press.Klepper, S. (2007). Disagreements, spinoffs, and the evolution of Detroit as the capital of the US automobile industry. Management Science, 53, 616–631.
  • Klepper, S. (2007). Disagreements, spinoffs, and the evolution of Detroit as the capital of the US automobile industry. Management Science, 53, 616–631.
  • Klepper, S. (2010). The origin and growth of industry clusters: The making of Silicon Valley and Detroit. Journal of Urban Economics, 67, 15–32.
  • Klepper, S., & Graddy, E. (1990). The evolution of new industries and the determinants of market structure. RAND Journal of Economics, 21(1), 27–44.
  • Krugman, P. R. (1991). Geography and trade. Cambridge, MA: MIT Press.
  • Lundvall, B. A. (1992). National systems of innovation. Towards a theory of innovation and interactive learning. London: Pinter.
  • MacGregor, N., & Madsen, T. L. (2013). Recovery following disruption to an ecosystem: The effects of the Internet bust on community evolution. Journal of Leadership and Organization Studies, 20, 465–478.
  • MacGregor, N., & Madsen, T. L. (2016). Aftershocks: Exploring cluster heterogeneity and agglomeration dynamics after the Internet bust. Working paper. Santa Clara University.
  • Marshall, A. (1920). Principles of economics. London: Macmillan.
  • Martin, R. (2012). Regional economic resilience, hysteresis and recessionary shocks. Journal of Economic Geography, 12, 1–32.
  • Martin, R. L., & Sunley, P. J. (2003). Deconstructing clusters: Chaotic concept or policy panacea? Journal of Economic Geography, 1, 5–35.
  • Martin, R. L., & Sunley, P. J. (2006). Path dependence and regional economic evolution. Journal of Economic Geography, 6, 395–437.
  • Martin, R. L., & Sunley, P. J. (2007). Complexity thinking and evolutionary economic geography. Journal of Economic Geography, 4, 16–45.
  • Martin, R. L., & Sunley, P. J. (2011). Conceptualizing cluster evolution: Beyond the life cycle model. Regional Studies, 45, 1299–1318.
  • Martin, R. L., & Sunley, P. J. (2012). Forms of emergence and the evolution of economic landscapes. Journal of Economic Behaviour and Organisation, 82, 338–351.
  • Martin, R. L., & Sunley, P. J. (2015). On the notion of regional economic resilience: Conceptualization and explanation. Journal of Economic Geography, 15, 1–42.
  • Martin, X., Salomon, R. M., & Wu, Z. (2010). The institutional determinants of agglomeration: A study in the global semiconductor industry. Industrial and Corporate Change, 19, 1769–1800.
  • Maskell, P., & Malmberg, A. (2007). Myopia, knowledge development and cluster evolution. Journal of Economic Geography, 7, 603–618.
  • McCann, B. T., & Folta, T. B. (2008). Location matters: Where we have been and where we might go in agglomeration research. Journal of Management, 34, 532–565.
  • McKelvey, B. (1999). Avoiding complexity catastrophe in coevolutionary pockets: Strategies for rugged landscapes. Organization Science, 10, 294–321.
  • Menzel, M. P., & Fornahl, D. (2010). Cluster life cycles—Dimensions and rationales of cluster evolution. Industrial and Corporate Change, 19, 205–238.
  • Nishimura, J., & Okamuro, H. (2011). Subsidy and networking: The effects of direct and indirect support programs of the cluster policy. Research Policy, 40, 714–727.
  • Ozer, M., & Zhang, W. (2014). The effects of geographic and network ties on exploitative and exploratory product innovation. Strategic Management Journal, 36, 1105–1114.
  • Paniccia, I. (1998). One, a hundred, thousands of industrial districts: Organizational variety in local networks of small and medium-sized enterprises. Organization Studies, 19, 667–699.
  • Paruchuri, S., & Ingram, P. (2012). Appetite for destruction: The impact of the September 11 attacks on business founding. Industrial and Corporate Change, 21, 127–149.
  • Piore, M. J., & Sabel, C. F. (1984). The second industrial divide: Possibilities for prosperity. Basic Books.
  • Potter, A., & Watts, H. D. (2010). Evolutionary agglomeration theory: increasing returns, diminishing returns, and the industry life cycle. Journal of Economic Geography, 11, 417–455.
  • Porter, M. E. (1998). Clusters and the new economics of competition. Harvard Business Review, 76(6), 77–90.
  • Porter, M. E. (2000). Location, competition, and economic development: Local clusters in a global economy. Economic Development Quarterly, 14, 15–34.
  • Pouder, R., & St. John, C. H. (1996). Hot spots and blind spots: Geographical clusters of firms and innovation. Academy of Management Review, 21(4), 1192–1225.
  • Romer, P. M. (1986). Increasing returns and long-run growth. Journal of Political Economy, 94, 1002–1037.
  • Rosenthal, S. S., & Strange, W. C. (2003). Geography, industrial organization, and agglomeration. Review of Economics and Statistics, 85, 377–393.
  • Saxenian, A. (1994). Regional advantage: Culture and competition in Silicon Valley and Route 128. Cambridge, MA: Harvard University Press.
  • Shaver, J., & Flyer, F. (2000). Agglomeration economies, firm heterogeneity, and foreign direct investment in the United States. Strategic Management Journal, 21(12), 1175–1193.
  • Simmie, J., & Martin, R. (2010). The economic resilience of regions: Towards an evolutionary approach. Cambridge Journal of Regions, Economy and Society, 3, 27–43.
  • Simon, H. A. (1996). The sciences of the artificial. Cambridge, MA: MIT Press.
  • St. John, C. H., & Pouder, R. W. (2006). Technology clusters versus industry clusters: Resources, networks, and regional advantages. Growth and Change, 37, 141–171.
  • Ter Wal, A. L., & Boschma, R. (2011). Co-evolution of firms, industries and networks in space. Regional Studies, 45, 919–933.
  • Vittoria, M. P., & Lavadera, G. L. (2014). Knowledge networks and dynamic capabilities as the new regional policy milieu: A social network analysis of the Campania biotechnology community in southern Italy. Entrepreneurship & Regional Development, 26, 594–618.
  • Wang, L., Madhok, A., & Li, S. X. (2013). Agglomeration and clustering over the industry lifecycle: toward a dynamic model of geographic concentration. Strategic Management Journal, 35, 995–1012.
  • Zucker, L. G., Darby, M. R, & Brewer, M. B. (1999). Intellectual capital and the birth of U.S. biotechnology enterprises. NBER Working Paper No. 4653, Cambridge, MA: National Bureau of Economic Research.


  • 1. This externality is informed by Jacobs’s (1969) work on urbanization economies.

  • 2. In a related line of inquiry, Paruchuri and Ingram (2012) examined the influence of an exogenous shock, 9/11, on the recovery of a regional economic system, New York City. They found that the formation of new businesses in close proximity to NYC grew at a faster rate after the attack as compared to before the attack and that the growth in business formation was faster in closer proximity to Manhattan as compared to geographic areas more distant from the epicenter of the shock.

  • 3. Edquist (2005) states that boundaries could be based on geography (even as administrative boundaries may be economically meaningless), on sector, or on the system’s activities or functions.