Regional entrepreneurial ecosystems: learning from forest ecosystems

In more recent literature, we can observe resonance between the EE and FE management perspectives, whereby an objective function is imposed on the definition of the ecosystem. As discussed earlier, a number of studies base their definition of an EE by focusing directly on entrepreneurial activity and/or the value outcomes it produces (Acs et al., 2017; Roundy et al., 2018; Spigel & Harrison, 2018; Stam, 2015; Stangler & Bell-Masterson, 2015; Szerb et al., 2019). These views approach the definition of an EE as supporting entrepreneurial activity dependent on the context or setting and are consistent with FE studies in terms of an objective ecosystem perspective. In the case of an EE, the entrepreneurial actor and activities are given priority for management or governance within the EE to produce some objective measure of regional performance, usually, but not always, economic.

However, the extant literature seems to pay little attention to the alternate view of the FE literature, where there is no prioritisation of service or claims on the ecosystem but instead a focus on the stability or otherwise of the ecosystem to sustain, recover or evolve. In this regard there is no apparent observable (within the limits of this study) correspondence between the FE and EE where the holistic views of stability, change and succession of places come into focus. From this perspective, the lack of correspondence between the EE and FE concepts suggests an opportunity for reflection on the role of the entrepreneurial activity in a socioeconomic ecosystem for its influence on a region’s stability. Hence, we propose the REE concept is theoretically analogous to that of FEs with respect to the change mechanisms of a holistic ecology but differs to the extant models of the EE dominant in the literature focused on producing entrepreneurial activity and primarily economic or other objective outcomes.

Given that studies in FEs and REEs share an interest in spatially bound contexts, we next check whether there is potential correspondence between how the zones and boundaries of a forest are defined with similar extensions to the REE. Two concepts are important here in FE terms. The first being the climatic zone the forest is situated in that provides the conditions for particular forest characteristics, and the second concept is the breadth and range of the species that define the forest area.

In REE terms, we proposed earlier that a climatic zone may share similarity with the industrial arrangements that support business and entrepreneurial activity. In particular, the dominance of an industry profile may present an industrial zone boundary condition for the REE. However, the analogy is problematic. Just as the studies of forests are about how trees support the diversity of organic and inorganic relations, the REE definition suggests that the socioeconomic activity of industry supports the diversity of social and economic relations in regions. In turn, industry is influenced by the natural and institutional contexts or settings like the trees of forests are influenced by climatic zones. In short, if industry is a supporting socioeconomic activity arrangement in a region, it cannot also be the determining zonal characteristic of itself. Therefore, this literal interpreted analogy must be rejected.

In exploring the climatic zone proposition, an REE equivalent can also be considered through natural and institutional contexts. It suggests that the climatic zone equivalent may be either the natural advantages of a place or its institutional formal and informal ‘rules’. The natural conditions for an REE boundary may include conditions such as local climate (advantageous for primary industry, for instance) and/or factors such as deep ports or harbours (relevant to manufacturing, transport and logistics) or other local geographic factors (lakes or fertile valleys) that support a diversity (or in some cases dominance) of human activity and business forms. In these terms, the climatic zone of a FE’s micro-climate and geographic positioning may not be that dissimilar to the considerations necessary for an REE. That is, the natural environment may be foundational to the establishment of a socioeconomic community in some regions or be critical to sustaining economic activity such as may be evidenced in tourism-dependent locations.

The second consideration includes the social, cultural, economic and political context that may influence the boundary conditions for a socioeconomic colony (Spigel, 2017). This implies that an REE may have an extra layer of zonal conditions, being the institutional setting (North, 1990; Scott, 1987) that comprises the political, social and legal, formal and informal ‘rules’ that contribute to the context of new venture creation, opportunity, support and legitimacy, shaping and influencing the mental models of actors (Denzau & North, 1994). Neumeyer et al. (2018), who examined the issue of social boundaries in EEs, concluded that boundaries are subject to groups of like ventures (e.g. high tech, high growth or lifestyle ventures), the common forms of institutional support services (e.g. universities or government agencies) or individual level characteristics (e.g. race, ethnicity, or gender). This resonates soundly with the dispersion of a common tree species in forest ecology as defining a boundary. Therefore, the boundary of an REE may propositionally be also influenced by groups of linked economic actors (firms and transacting organisations), who are in turn influenced by institutional attributes that may be dominant within a community, and these may be influenced furthermore by the natural conditions.

The combination of the natural and institutional settings proposes limits to various socioeconomic activity and therefore suggests there is no direct one to one correspondence between the FE and REE at this level despite the apparent metaphoric similarities. Unlike the FE model developed in this article, the REE has an additional institutional layer on top of the natural contexts that influences the form, type and spread of socioeconomic activity in a regional boundary. The boundary conditions of an REE are in effect set by the range of natural and institutional settings that support the particular community in focus for the study.

With respect to the change mechanisms imposed on the REE, we propose that the FE model equivalence falls short in its explanations of change for a socioeconomic region.

At a global level, like in an FE, an REE could be affected by factors generated externally to the REE boundary that are beyond the control of the actors within a region. For example, global political economies may alter economic conditions (erection of trade barriers or sanctions) or opportunities (free trade agreements) or introduce advances in technology. However, unlike our literal interpretation of the analogous conceptual model above that suggests global power (in terms of political economy) and technology replace the millennial-paced climate change as a factor, a realist perspective will accept that climate change is a highly relevant influence in the REE model. We note that both millennial-paced terrestrial and aquatic change due to shifts in climate will and can dramatically affect the socioeconomic activity in a particular region or area. Therefore, the longer-term changes in an REE include climate and the geopolitical and technological influences. Although we note that geo-political along with technological influences are also potentially more frequent and fall within shorter temporal cycles than centuries or decades, we therefore propose two further origins of more frequently experienced shocks for an REE that may cause or induce change. The first we might call political/economic (produced by external political and economic shifts of which the global financial crisis would be an example) and, the second, technological (produced externally by advances and changes in technology like information technology or artificial intelligence) (see Fig. 3).

In FE terms, we confirm the presence of more frequent shocks (century or decade intervals) as termed either allogenic or biogenic, and each has a distinct form either physical (wild fires and floods) or biological (Covid-19 or insect plagues) that will influence the regional socioeconomic community. Therefore, we retain these aspects in the REE model.

The FE literature also introduces an autogenic change process, and as noted previously, this offers perhaps the most significant analogous transference to the REE. While acknowledging the diversity of views on entrepreneurship, two recent broad reviews suggest something very specific in terms of what entrepreneurship offers an ecosystem. The first, outlined by Audretsch et al., (2015, p. 709) suggests that entrepreneurship is inherently multi-level and involves the ‘static element [that] creates an optimal environment for entrepreneurial action, and the dynamic elements [that] are embodied in the spillover impacts’. The second, by Mayer et al. (2018), suggests ‘the creation of new enterprises or the creation of entrepreneurial activities in existing enterprises … generate indeterminate uncertainty’. Embedded in both these views is the notion that the ‘entrepreneurial’ is not only about the static view to optimise entrepreneurship, but it also includes its influence on change within the bounded study of the EE. In other words, we suggest it is an autogenic change in a similar way to the terms of the FE literature.

When entrepreneurship is viewed as a mechanism of change, the study of an REE should then examine the influence of entrepreneurial activity on the changes in the structure of elements contributing to the socioeconomic community. The role of new ventures and entrepreneurial activity within a socioeconomic ecosystem is to seize new and emerging opportunities either through knowledge spill-over or by responding to experienced or perceived changes and disruptions external to the region. Hence, entrepreneurship creates and/or responds to conditions of uncertainty that alter the dominant socioeconomic activity within the ecosystem boundary. This framing is an alternative to the EE thinking about entrepreneurship comprising an entrepreneurial actor responsible for economic and other ‘value creating’ outcomes. By contrast, an REE framing suggests that ‘the entrepreneurial’ within a regional socioeconomic system is an endogenous change agent responsible for shifts in socioeconomic composition, structure and function that may be observed in many dimensions, in various ways and by many disciplines.

Studies that illustrate this thinking are found in the allied fields of industry agglomerations, clusters and complex systems. When considering local economic structures and agglomeration economies, the creation of innovative start-ups varies in response to the different local development conditions (Capozza et al., 2018). Innovative start-ups have also been instrumental in both creating diversification (through capitalising on opportunities influenced externally to the community boundary) and specialisation (seizing opportunities to build and reinforce strengths found within a community). Similarly, studies of cluster formation (the dynamic and emergent state pre-agglomeration and observable clusters) is seen to be influenced by strong entrepreneurial influences such as start-ups, network building, forming actor linkages and new knowledge that enables branching of local industry and propels exogenously introduced uncertainty (Li, 2018). Transition studies in complex systems have also directed attention toward the internal mechanisms or power and politics among the networks of both human and non-human agents (Kok et al., 2021). These issues raised across disciplines are more likely to come into focus when considering entrepreneurship as a change mechanism in a regional community through the REE framing of analysis as regions move from one socioeconomic order to another.

Due to limitations of space, we wish to avoid the deep analysis of each of the highlighted configurational elements to instead examine whether each of the type of elements can be sufficiently described or grouped into composition, structure, and function. To explore this further, we can observe the various standpoints on regional-industrial configurations to consider the consistency or otherwise of the types of elements.

The research relating geographies and industry has a long heritage, and two competing primary narratives emerged from the foundations. Some authors argued that concentration of firms in the same business domain (Arrow, 1962; Marshall, 1890; Romer, 1986) is preferable in a region to drive growth. Others argued the opposite to suggest diversity in urban business and industry has greater merit (Jacobs 1969). The former authors represent the MAR theory, which argues firms in close proximity experience spill-over effects that fuel growth in the regional economy (Glaeser et al., 1992). The latter argument, often accredited to Jacobs (1969), presents the case for variety and diversity in industry that inspires innovation and growth (Glaeser et al., 1992). These two arguments have been more recently re-framed into positions of regional related variety within sectors and unrelated variety between sectors (Frenken et al., 2007). Further advances have supported smart specialisation that focuses on the region’s capacity to ‘leverage existing strengths, to identify hidden opportunities and to generate novel platforms upon which regions can build competitive advantage in high value-added activities’ (Balland et al., 2019, p. 1252). However, despite these advances, there remains a paradox in regional innovation policy and industrial policy (Hassink, 2020; Muscio et al., 2015; Oughton et al., 2002) and a puzzle about the extent to which entrepreneurial processes discover or exploit discovered opportunities (Balland et al., 2019) that remains unsolved. The use of ecological studies to re-frame the problem into an REE explicitly brings into focus the study of the entrepreneurial process and how this process is configured to influence the change dynamics in a region. The REE potentially fills the void left by economic geographers in understanding the regional-industrial dynamics.

Many authors have compared and contrasted different systemic-based views of geographies over the last decade or so. Notably, Cao and Shi (2021) conducted a thorough systematic review to conclude—in agreement with Tallman et al. (2004)—that the thread of argument uniting these studies is that there are geographic boundaries and location specific advantages. The variations between the studies lie in the attention given by each of the author(s) to a number of factors, such as focus and locus of action (O’Connor et al., 2018); the environment, actors, key determinants of system performance and the impacts at micro and macro levels (Pilinkiene & Maciulis, 2014); the transactional costs, resource and knowledge capabilities and distinctions in power and control (Pitelis, 2012); or governance, knowledge, industry, actors, resources and benefits (Spigel & Harrison, 2018). Each of the studies though is concerned with the various parts that differentiate a particular framing of a system. None recognise that the ecological framing is concerned with how the parts work in concert to affect the whole. Adopting an REE approach highlights that there are differences between the wholes (the regional community) and leverages the depth of knowledge accumulated in ecological studies to examine the way communities regionally defined differently configure the parts rather than reducing the differences between the systems to the parts themselves.

It is particularly instructive that borrowing from the biological form of ecological studies (FE) draws attention to different understandings of what changes are important (Chazdon et al., 2016) in an ecology. This is equally reflected in change for the industrial economy and socioeconomic systems that sustain and/or evolve the socio-industrial economy within a particular community. While studies of forests can observe the same resources, which resources selected, how they are arranged and accessed and for what purpose they are used is where the variation in emphasis of the studies is found. We argue that these same configurational differences are also defining factors in distinguishing REEs.

Whether our interests are with analysing industrial districts, clusters, innovation systems, the triple helix, innovation ecosystems, business ecosystem, digital business ecosystem or the EE, the principal reason for doing so is to better understand and design interventions that sustain or change the community and business and/or industry circumstances. Changes affect social standards, living conditions and wealth associated with the community of analysis. Changes in socioeconomic conditions within a geographic region are defined by changes in various elements of analysis of the socio-industrial-economic relationships. Here, we suggest that at a regional level, industry is to be interpreted as the industriousness of the socioeconomic community and not necessarily as a particular industry sector. Regions will invariably comprise multiple industry sectors and at times they may be dominated by an industry sector. However, we intend by industry to represent the socioeconomic activity of the community and the various elements that comprise the socioeconomic infrastructure that supports a community.

Similar variations abound in in the studies of FEs with respect to terminology, measurement, constructs, and methods. While this correspondence of diversity in views and opinions suggests a metaphoric similarity, it does not drill down to the specific correspondence of configurational analysis relating to composition, structures, and functions.

In the EE literature we can find a vast variation in what may be considered elements. Some authors describe the entrepreneurial ecosystem as consisting of individuals with entrepreneurial attitudes, abilities and aspirations (Acs et al., 2014), but others consider a broader pool of elements such as people, roles, infrastructure, organisations and events (Regele & Neck, 2012). Stam (2015) describes a broad set of interdependent actors and Spigel (2017) notes more narrowly nascent entrepreneurs but then broadly adds ‘and others’. Mason and Brown (2014) stress the entrepreneurial nature of both actors and organisations and yet others simply note a group of actors (Cohen, 2006) or a community of complex agents (Roundy et al., 2018) as systemic elements. Although some definitions neglect any specification of what may be defined as elements, it would seem actors are implied as those who are influenced as in Audretsch and Bilitski’s (2016) definition and, as being one of the elements that combine in complex ways in Isenberg’s (2010) definition. Presumably, elements include resources such as finance, labour, skills, and advice that are potentially re-allocated (Acs et al., 2014) or recycled (Mason & Harrison, 2006; Spigel, 2020; Spigel & Vinodrai, 2020) through exchanges within the EE.

These definitional stances represent a variety of elements that do not address more specifically the issues of correspondence with composition, function and structure. It is apparent that the extant EE literature has not moved sufficiently to disentangle the relationships or categorisations of elements, as is highlighted by one notable study:

most studies of the context of entrepreneurship have been qualitative case studies that provide rich descriptions of entrepreneurial ecosystem constructs and elements, but do not provide information about how these constructs and elements are related to each other and to entrepreneurial outcomes. (Stam & Van de Ven, 2021, p. 828)

However, Stam and Van de Ven (2021) attempt to address this by studying the relationship between the combination of ten elements and a productive entrepreneurship output measure using statistical analysis, drawing data from twelve regions in the Netherlands. While they do not find or conclude a distinction among the elements that may specifically correspond with composition, structure and function, they do claim the following:

…that the entrepreneurial ecosystem elements are mutually interdependent and co-evolve in a territory. There is strong interdependence in general, and in three clusters of elements in particular. Talent, entrepreneurial culture and support services are strongly correlated, both simultaneously and over time. The same counts for knowledge and leadership (in innovation projects), also reflecting interdependencies in the knowledge economy. We also find strong interdependencies, both simultaneous and over time, between physical infrastructure and demand. (Stam & Van de Ven, 2021, p. 828)

Not only does the Stam and Van de Ven (2021) study reveals the close correspondence with the inter-relatedness and interdependency of elements as we find in the FE literature, but the three sets of clustered elements suggest a correspondence between compositional, functional, and structural elements. For instance, talent (compositional) clusters with entrepreneurial culture (institutional) and support services (suggestive of a functional element). In other words, when talent is organised by an informal institutional cultural setting and supported by services, one may expect an entrepreneurial output. Similarly, when knowledge is part of the composition of the EE and is facilitated by functional leadership, one may expect a productive entrepreneurial output. Lastly, when there are sufficient physical infrastructure (compositional) elements combined with functional demand, it will likely give rise to an entrepreneurial output.

While we suggest a concordance between the REE and FE configurations, at the conceptual level, further research is needed to be more explicit about what makes up each of the configurational elements in an REE context before any degree of confidence can be expressed in this proposed correspondence. Lastly, the definitional stance and realist ontology also exposes the need for case studies that examine the specifics of regions, their configurations and inter-relatedness as well as empirical studies of data collected across regions to draw attention to the parts within any region that deserve closer scrutiny with respect to their interdependency.

The fifth and final check seeks to resolve whether there is a rational correspondence between the FE concept of structural complexity and biodiversity with the REE concepts of structural complexity and social diversity. We have suggested that socioeconomic activity within REE communities may be characterised by the particular configuration of socioeconomic compositional, structural, and functional elements. An EE has been described as a complex adaptive system that brings about new order from the bottom up through interactions among multilevel structures, processes and attributes of the system (Roundy et al., 2018). An EE approach to an REE implies it is inescapably multilevel by its very nature. It follows that there is potentially correspondence with the structural complexity found in FE.

In common with the social sciences of organisational and regional studies, entrepreneurship has long since recognised the existence and interaction of multiple levels (Davidsson and Wiklund 2001). Reflecting the extant literature, the three levels often considered in the social sciences of organisational, regional and entrepreneurship studies (see Kozlowski & Klein, 2000; Lichtenstein 2011; Roundy et al., 2018; Sardeshmukh, O’Connor, & Smith, 2020) are:

The macro level aligns well with the institutional characteristics of social, cultural, political and economic reference points. However, it is readily acknowledged that the term ‘macro’ has instability in its meaning when considered across disciplines, whereby macro could mean anything from industries, interfirm networks, regional clusters, strategic groups, economies, alliances, populations, organisations, fields, societies, labour markets, nations and suprasocietal structures depending upon the disciplinary field referenced (Molloy et al., 2011). From this perspective it is quite reasonable to claim that the ‘macro’-level is not simply a unidimensional level but describes a supra-set of interacting contextual ‘rules’.

Roundy et al. (2017) argue that macro level coherence is built upon shared values and goals, common business models and venture similarities, institutional thickness, community logics and trust. It is apparent in this view that what is referred to as macro represents aggregated formal and informal ‘rules’ or institutions. This has resonance with the extracted FE analogy to the institutional setting. The macro perspective seems to support the idea of a set of conditions within which socioeconomic activities take place. We contend that the macro level accounts for the formal and informal setting of the ecosystem.

At the opposite extreme is the micro level of analysis. In complexity sciences, the bottom-up evolution of new order is termed ‘emergence’ (see Kozlowski & Klein, 2000 for a thorough coverage of the topic at the organisational level and Lichtenstein, 2011, for an equally relevant discussion to entrepreneurship). Of importance for this work is the definitional and compositional view of the micro and whether it may reflect the social diversity indicator relevant for an EE.

A clear link between the micro level and diversity is apparent in Roundy et al., (2017, p. 101) who wrote ‘[d]iversity can manifest in the types of ventures in an EE; however, its genesis is at the individual – “micro” – level of the EE’. Sardeshmukh, O’Connor and Smith (2020) argue that at a micro level, both divergence and non-linear complex dynamics are critical attributes that may be represented by diversity in populations. Perhaps more generally, Bhawe and Zahra (2019, p. 443) make it plain that ‘[e]cosystems that support greater heterogeneity of entrepreneurial activities are vibrant, dynamic, and responsive to changes in global business environments. A lack of heterogeneity cannot buffer firms against changes in technology or customer tastes’. Therefore, the social diversity is potentially a critical indicator of the extent of diversity in socioeconomic activity.

This micro perspective deals with the interests of the individual entrepreneurial actor that both drives and responds to change and uncertainty induced in the REE. Socially diverse entrepreneurial actors are those who reactively and proactively drive autogenic change while at the same time keeping socioeconomic activity running. Therefore, we proffer that populations of individuals form communities aggregate to represent a new industry or participate in existing industries and in turn contribute to other industries and communities that co-exist and are co-located in a region. Meanwhile, the entrepreneurial among them create ventures (whether social, commercial or other) that sustain or change the ecosystem’s industrial dynamics. The greater the social diversity of that community, the greater the potential diversity of ventures will be; and in turn the more evident divergence and non-linear complex dynamics will be, and the more vibrant and responsive the change resulting from shifts in technology, consumer tastes, etc. Hence the micro level for an REE seems to represent the social diversity that corresponds with the biodiversity of an FE.

Considering the concepts of macro and micro levels, we suggest a correspondence with the institutional setting at the macro level and the social diversity at the micro level. However, that leaves a question about whether and how the meso level contributes further to our understanding of the REE. It is apparent that the meso level facilitates interactions (Sardeshmukh, O’Connor and Smith, 2020) and intermediary processes (Roundy et al., 2018) that provide a ‘socially situated cognitive bridge’ (Goswami et al., 2018) between the macro and micro levels. Therefore, the meso level would appear to be an essential component of the conceptual model of an REE, facilitating the transactional alignment between the macro institutional setting and the micro diversity of social actors. This suggests that, in an REE context, a critical configuration of compositional, structural and functional elements is needed that accounts for the connectedness between the macro and micro levels and we might call this structural-organisational complexity.

This nested meso level is particularly distinctive when considering the analogous extraction from the FE and increases the complexity of the ecosystem analysis. In essence, the meso level comprises the organisations that form and/or exist to employ and/or engage the micro level social actors in activities that are in some way regulated by the macro level institutional context. These organisations are a species of their own contending for the resources of the ecosystem. Unlike FEs, the individual communities of species in an REE (humans) form organisations that industrialise and impose conscious and willing actions that compete for and consume resources from within and beyond the ecosystem boundary, at scales sometimes well beyond the needs of the REE communities. FE species (non-human) do not create an artificial resource (money) to facilitate exchange. They do not hold aspirations or desires for goods and services nor create organisations to produce goods and services for monetary exchange. The meso level in this way is peculiar to an REE and is not directly found in a FE. Where FEs do account for human habitation, humans are considered as one species within the ecosystem to be treated as part of the balance of the system (e.g. a home for indigenous people). Alternatively, the FE itself is considered as a resource for the organisations that use the forest resources for an industrial purpose (e.g. a timber resource). However, in an REE there exists any number of organisations that co-exist with the natural environment and the communities of human inhabitants. Organisations are part of the ecosystem analysis, providing an economic engine that sits between the social level of individuals and their communities and the setting of institutional contexts that support and in some way regulate them. The meso level includes organisations, whether they be private or public, for profit or not for profit, service providers, intermediaries, manufacturers or miners. Organisations are a species of their own that reside within the REE. They provide structural ‘habitats’ for socioeconomic activity enacted by humans and also behave as an organism, providing complex arrangements that employ and/or otherwise engage the micro level actors, in competing for and/or providing resources for the REE and beyond.

This leads to the following revised conceptual model of the REE (Fig. 3), derived through closer and realist considerations of the definitional level, the boundary and zonal setting, the dynamics of change, the configuration of activities and the dynamic indicators. It advances beyond the simply constructed interpreted analogous view to extend the conceptual arrangement of ideas that may equate to an understanding of what an REE is and how it may be analysed.