АНО «Центр междисциплинарных исследований» (ЦМИ)
Russian
| English
"Куда идет мир? Каково будущее науки? Как "объять необъятное", получая образование - высшее, среднее, начальное? Как преодолеть "пропасть двух культур" - естественнонаучной и гуманитарной? Как создать и вырастить научную школу? Какова структура нашего познания? Как управлять риском? Можно ли с единой точки зрения взглянуть на проблемы математики и экономики, физики и психологии, компьютерных наук и географии, техники и философии?"

«Leveraging complexity for ecosystemic innovation» 
Martha G.Russell, Nataliya V.Smorodinskaya

In this regard C. Wessner posits that the term ‘innovation ecosystem’ captures the complex synergies among a variety of collaborative efforts of large and small businesses, universities, research institutes, laboratories and venture capital firms, all involved in bringing innovation to market (National Research Council, 2007). In other relevant studies, this term is associated with an innovation business community, in which participants aim at a continual upgrading of their products, business processes, technologies and business models to flexibly meet rapid technological changes and gain market advantage (BSR Stars, 2013; Nallari and Griffith, 2013). Plainly speaking, the literature dealing with policies to facilitate competitiveness and innovation-led growth describes innovation ecosystems through the lens of their key function — to provide smooth and continual exchange of knowledge flows in bringing innovations to market.

3.4. Interpretations of localized and economy-wide innovation ecosystems

Though networks are not limited by geographical borders and can emerge as virtual structures, economists and sociologists have emphasized the importance of localization in the innovation process. The economic development strategies of various countries are all based on the underlying premise that co-location of partners matters for co-producing innovations, since geographical proximity of networked agents (particularly of firms and universities) is crucial for facilitating knowledge transfer, especially in knowledge-intensive industries (Owen-Smith and Powell, 2009; Powell and Giannella, 2010; Sölvell, 2009). As experience of Silicon Valley and similar innovative places shows, localization leads to very important agglomeration effects that enable cost reduction and support innovation (Carlino and Kerr, 2014). As stems from literature on complexity, collaborative networks of geographically co-located actors (such as regional clusters or macro-regional innovation hubs) have a greater innovation potential as ecosystems than organizationally dispersed networks in the form of production circuits, not tailored to the format of sustainable social integrity (f.e., global value chains) (Dicken, 2015)3. Taken together, this implies that national strategies for developing collaboration for the co-creation of new goods and values can be optimized at the level of localities and must also consider the contribution of proximity effects in the innovation dynamics of geographical areas.

In a strict economic sense, reflecting the complex nature of innovation in the twenty first century, a localized innovation ecosystem is a sustainable node of network communications among entrepreneurs, researchers and other institutional actors, which enables them to collectively generate knowledgemutually exchange it and transform it into commercial innovative assets through collaboration. As literature and practice tell, these entities often take the form of innovation clusters or university-industry partnerships emerging around specific industries and technologies, or even at the intersection of two or more distinctly different, yet overlapping sectors (Bramwell et al., 2012; Porter, 1998).

Economic and business studies on innovation clusters clarify that cluster actors collaborate under a common cluster project and create innovations by means of co-production and co-specialization (Eriksson, 2010; Hamdouch, 2007). Cluster literature, originating from M. Porter’s competitiveness theory (Ketels and Memedovic, 2008; Lindqvist, 2009; Porter, 1990, etc.), views clusters as innovation ecosystems that generate a combination of proximity effects and unique collaborative effects of continual innovation. Upon summarizing findings of cluster literature, N. Smorodinskaya describes innovation clusters in three interrelated dimensions: firstly, as a special class of agglomerations enjoying a glocal (global plus local) flow of resources; secondly, as a special class of collaborative networks relying on triple-helix pattern of collaboration; and thirdly, as a special class of economic projects realized through relational contracts (Smorodinskaya, 2015; Smorodinskaya and Katukov, 2015).

Literature on business networks (Breschi and Malerba, 2005; Huggins and Izushi, 2007), as well as evolutionary geography and regional studies (Asheim et al., 2011; Cooke, 2001; ter Wal and Boschma, 2011) explore a diversified variety of localized public-private partnerships, which can be referred to as a class of triple helix collaborative networks. These studies usually emphasize the key role of intermediating institutions in enhancing collaboration in such networks, and hence, in advancing their innovation ecosystems. Intermediaries, in cluster literature also called institutions for collaboration (Ketels and Memedovic, 2008), apply considerable intentionality in effectively bridging multiple partners and directing them toward delivering a common agenda. For instance, in USA, localized ecosystems may involve a diverse spectrum of actors (from individual researchers to banks and large companies), with their intermediaries providing both public and private funding, or offering a platform for collaboration (National Research Council, 2012).

As an expression of both collaborative and competitive milieu, the innovation ecosystem concept heralds the newly emerging economic milieu, composed of various multiform and overlapping network partnerships. The population of these partnerships across a country’s economy makes up a national innovation ecosystem. For example, in the US, the term ‘innovation ecosystem’ appeared on the national policy agenda by 2005, as proposed by C. Wessner (2005) to highlight the non-linear nature of the innovation process. Unlike the practice of other nations in Europe or Asia, America’s national innovation system has not been conceived as a system intentionally planned or designed by the government, but rather as an extremely complex ecosystem characterized by myriad varieties of interactions among government agencies, universities, private industry, financiers, and intermediary organizations (National Research Council, 2012). In some large countries similar complex ecosystems appear at the regional levels, thus constituting regional innovation ecosystems (Ranga, 2011).

At the moment, a wide number of jurisdictions (Finland, Denmark, Korea, China, Sweden, UK, Australia, etc.), while following the front-running experience of USA, have established policies and institutions for advancing the formation of economy-wide ecosystems (both national and regional, as well as multinational) as means to accelerate their transition to a knowledge-based economy (Bramwell et al., 2012).

4. Properties of innovation ecosystems in terms of complexity science

4.1. The concept of complex adaptive systems

Literature on networks treats non-linear systems as complex adaptive systems(CAS), or complex dynamic systems (Jucevičius and Grumadaitė, 2014), which have been studied since mid-1980′s by complexity science, also called complexity theory4. Such systems, irrespective of their scale and origin (be they biological or social), have a universal holistic nature predefining their complexity: individual properties of constituents can be revealed only through aggregate properties of the whole system. Today, literature on complexity refers to CAS many natural and, increasingly, many artificial systems; the variety includes economies, ecologies, societies, human brain, developing embryos, ant colonies, computing systems, artificial intelligence systems, etc. All of them display common features and are often described under the interchangeable terms of complex adaptive, complex dynamic, or complex non-linear systems.

CAS is a dynamic open network of many heterogeneous agents acting in parallel in a complex, unpredictable (emergent) and mutually self-reinforcing way (Holland, 2002). In such networks, overall properties result from the aggregate behavior of individual agents; complexity results from the inter-relationship, inter-action and inter-connectivity of elements within a system and between a system and its environment; and control tends to be highly dispersed (Mitleton-Kelly, 1997). Coherent behavior in a national economy or in a local cluster arises from interactions among the agents themselves, with these interactions essentially implying both competition and cooperation (Chan, 2001), as well as birth and death of entities over the business cycle (Holling, 2001).

CAS is a dissipative structure, able to continually adapt in and evolve with its environment, as generated by the dynamic and self-reinforcing interactions of its agents. This implies that a complex system is not separated from its ever-changing environment, thus making up an ecosystem, in which each agent operates in an environment produced by its interactions with other agents (Elsner, 2015). As a result, CAS displays a spontaneous self-organization: the order can result from various complex feedbacks and mutually self-reinforcing interactions among a large number of agents at different levels of organization (Martin and Sunley, 2007). There is constant action and reaction of agents to what other agents are doing, based on differences in values and goals, and the boundary between the system and its environment is neither fixed nor easily identified (Chan, 2001).

In recent decades, complexity theory has attracted increasing interest of economists and evolutionary economic geographers as a novel and powerful mode of thinking, capable to bring a radical and long-overdue revision of the mainstream economic thought (Krugman, 1996; Metcalfe and Foster, 2004). This has led to the appearance of ‘complexity economics’, a new research stream creating an umbrella for those theoretical and empirical studies that can be directly or indirectly linked with complexity science (Beinhocker, 2006).

4.2. Innovation ecosystems viewed as complex adaptive systems

Complexity economics sees an innovation-led economy as a complex adaptive system, or an ecosystem, constituted by innumerable knowledge flows and multiple inter-connections among diverse and heterogeneous elements that communicate through networks (Martin and Sunley, 2007). From this perspective, collaborative networksas well as network-based economiesgenerate innovation ecosystems as their inseparable and co-evolving environments and may display, either entirely or partially, the following generic properties of CAS (Chan, 2001; Martin and Sunley, 2007; OECD, 2009):

  • Basic network effect. Ecosystems are open-ended networks, in which each agent benefits in a non-linear way from any simple increase in the number of network nodes and participants. As a result, network-based ecosystems enjoy competitive advantages in dynamics as compared to linear systems;
  • Emergencyor non-determinate behavior. Ecosystems are almost unpredictable; they may behave in ways which don’t follow from their earlier state or from individual properties of their components. This corresponds to a non-linear behavior that generates abrupt removals of order by randomness, or stability by volatility, and vice versa;
  • Presence of feedback linkages and reflexive cyclesboth positive and negative. The pattern and level of interactions among the ecosystem agents matter more than their own characteristics and behavior of each individually. Quality and quantity of feedback linkages within an ecosystem determine its overall effectiveness, since agents reciprocally react in their behavior to the behavior of other agents, displaying high mutual interdependences;
  • Adaptabilityor capacity for adjustments. In the course of interactions, ecosystem’s agents modify their behavior, upon reacting to and considering the behavior of other agents, which adjusts the behavior of the whole ecosystem, implying its high adaptability to any changes;
  • Self-organizationself-regulation and self-governance. Ecosystems start to move and are advancing spontaneously, similar to self-adaptive living entities. They rely on a dispersed pattern of coordination through network nodes and don’t need either external intervention (typical for mechanical or static systems) or any governing center (typical for hierarchic systems). Accordingly, they obtain new sources for growth and achieve dynamic sustainability through internal, self-correcting structural transformations, which may involve a persistent agile recombination of shared assets circulating within an ecosystem;
  • Self-similarityor fractal-type recursions. Ecosystems can generate larger or smaller similarities at any scale level: networks form other networks with similar properties within different geographical and institutional spaces;
  • Holistic nature and synergy. An ecosystem’s behavior, its dynamics and innovativeness are an aggregated result of interactions among its agents (and not a simple summation of agents’ behaviors and performances), which implies synergy effects that enlarge an ecosystem’s productivity always to a greater extent than a sum of individual results of its participants.

The combination of these properties enables complex systems to generate “structural order” in non-equilibrium environments through interactive relationships of participants, i.e., by means of continual reciprocal adjustments of their individual non-linear behaviors on the basis of multiple feedbacks. The aggregate result of interactive and non-linear relationships among a significant number of networked agents is expressed in internal structural transformations in the ecosystem, thus lending it the ability for a spontaneous self-regulation(OECD, 2009).

We admit that all types of ecosystems presented above in Fig. 2 can share these common properties of CAS to a smaller or larger extent. All of them are populations able to self-organize and self-develop in a similar, agile manner, associated with inter-relationship of networked elements (Chan, 2001). However, we believe that ecosystems with the highest complexity in terms of pattern of interactions (like innovation clusters), and hencethe ones with the highest innovative potentialmay reveal the highest self-adaptive capacity, or which is to say, the strongest ability for sustainable development under high uncertainty.

Some essays suggest (Holling, 2001) that CAS are also distinguished by generativity as a generic feature, implying that innovation ecosystems display dynamic balances such as coexistence of creation and conservation, learning and continuity, or success and failure among participants, all them supporting an ecosystem’s evolution and dynamic sustainability. Basically, these and similar balances found in ecosystems — including inter-connectivity (coexistence of weak and strong horizontal ties), heterogeneity (a critical diversity of actors in terms of their functions and behaviours), or a certain relationship between the level of control and the level of functional independency (Carbonara, 2017) — should be seen as resultants of a CAS’s holistic nature and factors fostering its capacity for self-adjustments through interactive relationships of agents.

In terms of a recent classification of the existing analytical ecosystem perspectives, presented in Tsujimoto et al. (2017), our treatment of innovation ecosystems, which relies on the CAS concept, is closer to the perspective associated with multi-actor networks as a phenomenon of non-linear world. It should be also noted that innovation ecosystems and natural ones, although both displaying common CAS’s properties, are by no means the same, since their complexity rests on different principles of self-organization and self-adaptation. When describing an innovation-led economy, complexity economics focuses on the key formative role of knowledge flows, rather than drawing any biological analogies (Martin and Sunley, 2007). This literature connects economic growth in the twenty first century with persistent emergence of innovation and with continual self-correcting structural changes through which an economy obtains new sources for growth and adapts itself to the global upgrading of technologies.

4.3. Innovation clusters viewed as complex adaptive systems

According to a classical definition, clusters are groups of geographically co-located companies and associated institutions, engaged in a particular field of related industries, and linked through various types of externalities (Porter, 1998, 2003). Modern cluster literature views innovation clusters as collaborative networks initiated through a common project of triple-helix actors (Sölvell, 2009). It directly interprets clusters as complex dynamic systems, highlighting their unique synergy effects (European Commission, 2013).

A mathematical formalization of the triple helix model (Ivanova and Leydesdorff, 2014) confirms that triple-helix networks may form a very sophisticated ecosystem of social communications and functional interdependences, which has a dynamic nature of non-linear fractal structures and can provide continuous upgrading required for innovation-driven growth. It follows, therefore, that this pattern of collaboration may extend the three institutional pillars to a Quadruple Helix, a Quintuple Helix, and even N-tuple helices (Ahonen and Hämäläinen, 2012; Leydesdorff, 2012). Such complex systems are now increasingly recognized as a typical way to create knowledge, disseminate it across economies and transform it into new values. As argued above (part 2.2), triple helix networks are characterized by more advanced and complex interaction patterns in terms of agility, dynamism, and co-production of innovation on a continual basis. Successful innovation clusters demonstrate these advantages (Porter, 1998; Rullani, 2002).

Successful clusters involve collaborative partners of various profiles, who are engaged in co-production of innovative goods and values (Fig. 3), while staying free to join and leave the open-end cluster network (Ketels, 2012). Each competitive cluster relies on a certain critical mass of participants, provided by the presence of three key categories of actors: firstly, representatives of all the three triple helix actors, co-located in the given territory; secondly, venture capital investors and financial sponsors from private, government or international sectors; and thirdly, the cluster organization as a specialized cluster coordinator (Lindqvist et al., 2013). At critical moments over time, the fluidity of a complex adaptive system and the coordinating work of a cluster organization enable a co-located group of companies to evolve as a self-governed and self-sustainable innovation ecosystem.

Fig. 3

 

Fig. 3. The complexity of an ecosystem in a regional innovation cluster.

Source: authors’ design, based on: (Napier and Kethelz, 2014).

As complex adaptive systems, clusters are displaying a holistic model of relationships between geographic proximity and industrial competitiveness (European Commission, 2013). Their high innovativeness is fueled by localized externalities generated within a cluster’s ecosystem through several dynamic balances, such as inter-firm co-opetition, coexistence of specialization and diversification of activities, an agile and mobile combination of local and global resource flows, etc. (Smorodinskaya, 2015).

In terms of generated innovation synergy, the complexity of a cluster ecosystem arises from the following features of its organizational and institutional design (Smorodinskaya and Katukov, 2016).

First, evolution of a highly cohesive milieu enthused through collaboration of co-located triple helix actors, and embracing a certain critical mass of networked partners, both in terms of quantity and quality. Successful clusters develop an ecosystem of dense functional linkages among the micro- and macro-level partners of various profiles, from incumbent firms and innovative SMEs to service organizations and financial institutions (Napier and Kethelz, 2014). Collaborations of triple helix actors can reproduce self-similarities on a variety of scales, involving other agents within and outside the cluster to form more complex helices.

Second, development of a cluster under the discipline of a common project(cluster initiative), launched jointly by two or more triple helix actors (Sölvell et al., 2003). Cluster initiatives are network projects realized through collaboration of cluster participants. Collaboration is ultimately oriented toward implementation of joint business projects integrated into global or other transborder value chains (Smorodinskaya et al., 2017), but the success of these projects and the very competitiveness of a cluster directly depend on the intentional development of collaborative interactions as such (Lindqvist et al., 2013).

Third, formation of a membership-based cluster organization that guides the cluster initiative, aiming to strengthen the cluster’s overall innovativeness. Such organization takes the form of a specialized internal network within the cluster ecosystem, which incorporates the majority of cluster actors via formal membership. The membership implies certain commitment plus fees plus relational contracts, i.e. specified-term agreements on shared rules, shared objectives, and directions of mutual activity (Bathelt and Glückler, 2011). The cluster organization lends an institutional structure and a communication platform to the cluster, as well as creates special intermediaries (institutions for collaboration) that care for the continual face-to-face coordination of plans, interests and complementary activities, aiming to sustain the intensity of triple-helix collaboration, the above-mentioned dynamic balances, and a critical level of mutual understanding and trust within the ecosystem.

Finally, a regime of collaborative governance built by the cluster organization for orchestrating and developing the cluster. This regime makes an alternative to traditional patterns of governance, introducing a collective decision-making, in which investment priorities, lines of business activity and conventions are defined through interactive consensus-building among functionally interdependent (and hence, commercially interested) network actors (Ansell and Gash, 2007). As follows from CAS’ features and argued by Porter (Porter, 1990), the stronger are interactive linkages and feedbacks within a networkthe greater are benefits to participants from the commonly produced value. Therefore, coordinating efforts of the cluster organization are focused both on enhancing the social integrity of cluster actors (building trust and facilitating collaboration) and on successful realization of their common business projects (Smorodinskaya, 2015). This cultivates a new, leadership style of heterarchical coordination, based on relational contracts and peer-to-peer collaborative interactions, in which common development strategies (for advancing the overarching cluster initiative) and current economic decisions are shaped jointly and interactively, by means of negotiations and reconciliations among all stakeholders (OECD, 2001). Typically, two institutions for collaboration (a cluster governance team and a cluster management group) concentrate their activities on continually removing inner and outer gaps in communication, aiming to bridge cluster actors for a smooth co-creation of innovations and bringing them to market (PwC, 2011; Sölvell, 2015).

Properties of CAS help to highlight the most general collaborative mechanism of self-supportive growth. The basic capacity of clusters for continual co-production of new goods and values is fostered through synergies derived from creative and complementary reshuffling (agile combining and recombining) of shared assets, both tangible and intangible, in numerous novel configurations (Sölvell, 2009). This capacity is simultaneously supported by the complexity of a cluster ecosystem, the presence of a highly cooperative, and a highly competitive business milieu that nurtures the generation, survival and deployment of novel assets through those re-combinations (Padgett and Powell, 2013). As a result, well-developed clusters enjoying the complexity of triple helix partnerships can critically enhance productivity, decrease uncertainty, and flexibly start new venture business projects to meet the rapidly changing market demands (Delgado et al., 2010; Jackson, 2008; Ketels, 2012).

4.4. The complex ecosystem-based industrial landscape

Complexity economics helps to coin a new ecosystemic vision of organizational design of economies, implying that new goods and values are increasingly created in collaborative ways, through formation of network-based ecosystems that display properties of CAS. As mentioned above, this vision is gaining support in various streams of studies, and particularly, in economic sociology (f.e., Smith-Doerr and Powell, 2005; Padgett and Powell, 2013).

An ecosystem perspective suggests a holistic view of regional or national innovation systems, and hence, of economy-wide production systems, highlighting not just the functional roles of their constituent entities but also the pattern and dynamics of interactions among them (Bramwell et al., 2012). In the term “ecosystem”, the prefix ‘eco’ (in relation to ‘system’) emphasizes the non-linear nature of innovation and the key role of collaboration in generating it (Townsend et al., 2009). In a non-linear environment, development is based on fractal-type recursions, in which new collaborative networks reproduce basic properties of CAS at any scale, thus progressively increasing the organizational complexity of modern economies. As a result, economies of all levels, from local to global, are gradually assuming the design of open, highly interconnected, self-organizing, emergent and adaptive systems (Martin and Sunley, 2007; OECD, 2009).

In other words, the industrial landscape of knowledge-based economies is emerging as a manifold variety of innovation-inducing ecosystems (network-based organizations, agglomerations, communities, areas, etc.). The milieus of these ecosystems, initially emerging at the level of localities, can further grow in size and complexity through non-linear integration with other ecosystems. Successful innovation ecosystems can generate new networks around themselves and develop collaboration with each other, thus leading to the appearance of more complex and robust innovation ecosystems, embracing clusters of clusters and networks of networks. In some essays, this complex ecosystemic landscape is compared with “a multilevel, multi-modal, multi-nodal, and multi-agent system of systems” (MacGregor and Carleton, 2012).

A recent OECD report argues (OECD, 2015a) that during the nearest decades the world economy will progressively increase its internal inter-connectedness and complexity. Already today it should be seen as a complex non-linear system, in which micro-level interactions are generating macro-level transformations not equivalent to simple aggregate results of those interactions. Meanwhile, studies describing the new complex world of the twenty first century posit that the majority of global-level transformations will emerge endogenouslywhile national economies will have to operate in a continually changingmulti-equilibrium environment (Silim, 2012). These findings give grounds to believe that over time, rather in the long-term prospect, the global economy, as well as its national, sub-regional or macro-regional components will become an agile and self-structuralizing mixture of inter-connected and overlapping collaborative partnerships - some local, some global — in which networks are generating new clusters and inter-cluster linkages of various spontaneous configurations, and vice versa on a continual basis — toward mutual economic, and hopefully societal, advantage.

This heterarchical, cohesive yet super-volatile and super-competitive organizational complexity makes a crucial departure from the ‘hierarchy-market’ dichotomy of industrial era.

5. Ecosystem versus system approach to economic development

5.1. Comparing traditional and complexity economic thinking

The transformation of economic systems into network-based ecosystems provides the organizational basis for their transition to innovation-driven model of growth. This transformation is expected to leverage the total productivity in economies through innovation synergy effects, thus enabling companies and territories to generate a higher additional income (value added) than they could obtain under a traditional, less complex design. Meanwhile, the self-organization capacities of ecosystems do not render intentional interventions redundant (Elsner, 2015). Market forces themselves cannot automatically provide the needed pace of emanation of the new industrial landscape, even in the most liberalized economies. Rather, economies of all types, and especially emerging market economies, need active intentional efforts and program-based support directed toward enabling and facilitating the ecosystemic transformation.

A comparison of ecosystems to systems approaches focuses on the complexity economics dealing with non-linear realities of the twenty first century versus traditional conceptual models of the pastwhich dealt with linear development. The ecosystem approach alters a more simplistic, mostly mechanistic perception of economic systems, putting forward a set of new principles in economic thinking and economic policy making, relevant for non-linear development (Table 1).

Table 1
System approach (traditional thinking) Ecosystem approach (complexity thinking)
Economic dynamics Linear systems — closed, static, in equilibrium Non-linear systems — open, dynamic, dissipative
Emergence and synergy Macro-level growth patterns are formed by linear summation of individual decisions of homogenous agents, with few synergies occurring spontaneously Macro-level growth patterns emerge nonlinearly, out of synergies generated by dynamic network interactions of various heterogeneous agents at micro-level
Network interactions Network relationships are inessential, agents interact indirectly through market price mechanisms Network relationships are essential, economic systems of all levels (from local to global) are seen as network-based ecosystems meant for innovation
Predominant model of economic governance and adaptation Hierarchic model: a rigid, centralized organization governed by administrator through top-down decisions. The economy lacks feedback linkages for self-adjustment to changing environment and, hence, has low capacity for adaptation Heterarchical model: a dispersed agile network with spontaneous self-organization, self-regulated through horizontal coordination of network nodes and collaborative consensus-building. The economy gets self-adaptable through interactive communication of agents, their feedbacks, their learning and proactive reciprocity
Interpretation of innovation Limited endogenous capacity of economic system, dependent on a complex of its available resources. Requires external incentives or exogenous sources, not connected with a system’s social and structural transformations. Implies linear process of knowledge flow, from science to industry (‘mode 1’ in knowledge creation) Sustainable endogenous capacity of economic system, based on internal incentives and new sources, arising from a system’s ability for continual self-correcting structural changes. Implies non-linear process of knowledge flow (‘mode 2’), relying on interactive communication of various agents, as well as continual, systemic process (‘mode 3’), arising from proliferation of collaborative networks and their ecosystems
Model of producing innovations (goods, values, technologies) Linear models of innovation (‘technology push’ and ‘demand pull’), driven by technological developments of individual firms Interactive model: co-creation of innovations by networked agents through their collaboration within a generated ecosystem of linkages and assets
Interpretation of innovation systems (regional, national, macro-regional) Non-cohesive organizational structures that depend on involving a certain critical mass of agents, talents and new infrastructure Holistic social communities, or ecosystems, with properties of complex adaptive systems, depending on a certain critical mass of interactive inter-linkages among networked agents
Institutional and business environment for innovation Creation of new institutions, technologies and industries is higher priority than enhancement of cohesive context for a smooth dissemination of innovations across sectors and regions Priority is given to continual improvements in environment, with the purpose to eliminate barriers and provide incentives for more business networks, more collaboration, more cohesion, and continual knowledge spillovers across and around the economy
Focus of strategies for innovation and growth To develop R&D and national innovation system by supporting its agents and infrastructure elements, with no focus on collaboration and its innovation synergy effects To promote localized ecosystems across the economy and enhance their innovation synergy effects by facilitating the dynamics of interactions and collaboration within-between networks

Source: authors’ elaboration based on: Beinhocker (2012); Bramwell et al. (2012); Holland (2002); Martin and Sunley (2007); Townsend et al. (2009).

Modern economies and production systems of all levels are seen as network-based ecosystems intending innovation. They are open non-linear spaces(milieus) that are undergoing persistent transformations, and hence, far from static equilibrium. Instead, they search for a dynamic equilibrium, relying on their ability to proactively respond to changing environment, and to modify their behaviours as experience accumulates (Holland, 2002). If in linear economies agents interact indirectly through market price mechanisms, in non-linear economies agents continually learn from each other and adapt, communicating interactively (on the basis of reciprocal actions and feedbacks) within constantly changing networks. Networked agents, their interactions, and feedback linkages within-between networks are the source of the constant novelty that imbues the non-linear economy with its evolutionary momentum (Beinhocker, 2012; Bramwell et al., 2012).

In traditional systems, macro-level patterns are formed by a linear summation of individual decisions of homogenous agents. Unlike that, ecosystems are holistic: their micro- and macroeconomic spheres are closely interrelated, composing non-separable parts of economy made up of networked and interactively communicating agents. This implies self-development of economies, in which the emerging pattern of growth at the macro-level is induced and self-supported endogenously, through dynamic network interactions of heterogeneous agents at the micro-level.

Traditional systems are based on rigid hierarchic governance, or top-down decision-making by central bodies. Unlike that, agile collaborative networks and their ecosystems enjoy the advantage of self-organization and self-regulation, referred to as resiliency (Crespo et al., 2014), an adaptive nature typical for CAS. This implies that the more the modern economies will advance in their transition to the ecosystem organizational design, the more they will move from hierarchic to collaborative governance, and hence, increase their self-adaptability to rapid changes. One can imagine that heterarchical models of social coordination, based on horizontal consensus-building, will gradually evolve across the world, making up a creative functional alternative both to market coordination and to the system of administrative orders. The powerful global competition for the speed in innovation, which has replaced the traditional model of competition for resources in various local markets, will serve to spur on agents to unite themselves in networks and develop collaboration.

Traditional economies rely on linear processes of knowledge flow from science to industry, often identified in literature as ‘mode 1’ in knowledge creation (Gibbons et al., 1994), and usually driven by ‘technology push’ or ‘demand pull’ on the part of individual firms (Godin, 2006). Linear models perceive innovation as an exogenous force, independent from a system’s social and structural transformations. On the contrary, in modern complex economies, innovation is perceived as endogenous capacity of a systemresulting from the agility of networks - their facility for persistent structural reconstructions. As noted earlier in this paper, ecosystems can obtain new sources for growth and achieve dynamic sustainability through internal, self-correcting structural changes — rather than through top-down intervention of any centralized bodies, or from an external intervention, as typical for traditional systems. No surprise that developed economies and a growing number of developing economies are promoting a non-linear model of knowledge creation (‘mode 2’), driven by interactive communication of various networked agents across institutional and geographical borders (Gibbons et al., 1994). And since the turn of the century, the most advanced countries are cultivating an even more complex innovation, seen as a continual, or systemic process (‘mode 3’), which results from and simultaneously predefines further proliferation of ecosystems, or an increasing organizational complexity of the economy (Carayannis and Campbell, 2009; OECD, 2015b).