Introduction

Numerous authors argue that innovations can substantially foster humanity’s transformation toward sustainability (Brodhag 2013; Dwyer 2013; Strambach 2017; Veiga Ávila et al. 2019; Pyka 2017; Mulder 2007; Ramzan et al. 2023). This idea has also found its way into application in the context of policy strategies, such as the European Union’s Green Deal (European Commission 2019) and the European Union’s current industrial strategy (European Commission 2020), as well as in the context of manuals and reports of international organizations such as the World Bank (World Bank Group 2014; World Bank 2021), and the OECD (OECD 2011, 2017). However, innovations do not positively influence sustainability per se (Schlaile et al. 2017; Schot and Steinmueller 2018). Rather, a positive social or environmental impact, or both, must be evident for an innovation to be considered conducive to sustainability. We can thus differentiate between conventional innovations, which primarily are supposed to increase efficiency, and sustainable innovations (see e.g., Rennings 2000; Adams et al. 2016). These different types of innovations are based in part on differing knowledge bases, as the inclusion of social or environmental objectives (in the case of sustainable innovations) requires further expertise compared to purely market-driven innovations (Hojnik and Ruzzier 2016; Horbach et al. 2013; Strambach 2017). As a result, these different types of innovation contribute very differently to the goal of sustainability transformation, with their impact ranging from hindering to supporting (Pyka 2023).

Sustainability science provides insights into what kind of knowledge is required for a transformation toward higher degrees of sustainability and what the corresponding knowledge processes look like. Three types of knowledge are widely recognized as being important, namely normative knowledge, systems knowledge, and transformative knowledge (ProClim 1997; Wehrden et al. 2017; Wiek et al. 2012; Trencher et al. 2017; Hart et al. 2015; Abson et al. 2014). Furthermore, specific aspects regarding creating, co-creating, and transferring knowledge for sustainability are deemed relevant. First, several authors emphasize the importance of interdisciplinary or transdisciplinary approaches (e.g., Abson et al. 2014, 2017; Batie 2008; Kates et al. 2001; Hart et al. 2015; Wehrden et al. 2017; Binder et al. 2015). Second, a problem- or solution-oriented approach seems especially conducive to creating and transferring relevant and applicable knowledge (e.g., Batie 2008; Abson et al. 2017; Kates et al. 2001; Clark et al. 2016; Wiek et al. 2012; Stephens et al. 2008). Third, the participation of all (or as many as possible) of the relevant stakeholders in the knowledge processes appears to be a key issue regarding sustainability transformations and associated innovation processes (e.g., Kates et al. 2001; Clark et al. 2016; Hart et al. 2015; Trencher et al. 2017; Urmetzer et al. 2018; Zilahy and Huisingh 2009; Abson et al. 2017).

Following the innovation system approach (Cooke 1992; Lundvall 1992; Nelson 1993), innovations can be understood as outcomes of interactive processes involving different actors, with knowledge flows and mutual learning between these actors being particularly important, as knowledge is at the heart of any innovation process (Lundvall 2016). In the case of sustainable innovations, the actors in those innovation systems come from business, government, academia, and civil society, as the relevant knowledge for sustainable innovations is widely dispersed in society (see e.g., Carayannis and Campbell 2010, Clark et al. 2016, or Ghassim 2018). Besides, the majority of sustainability innovations only contribute to sustainability targets because, from the very beginning, different stakeholders, knowledge fields, and regulatory authorities are involved. Therefore, innovation networks connecting these actors are, in most cases, a conditio sine qua non, as go-it-alone-strategies most likely lack the complementary and required pieces of knowledge. The knowledge of sustainability innovation is the outcome of the mutual learning process between the stakeholders during the innovation process, and it can hardly be imagined that single authors are in the position to solve the complex problems of developing relevant technologies that are accepted and adopted to finally widely diffuse to support the transformation of production and consumption systems toward sustainability.

The literature is rich in examples of difficulties that arise when such diverse actors collaborate in so-called innovation networks (confer Polk 2015; McCauley and Stephens 2012; Klenk and Hickey 2012; Pyka and Nelson 2018). Hence, to better understand learning and knowledge flows in sustainability-oriented innovation networks and the associated obstacles and drivers, it is helpful to examine the relations between the actors in these networks in detail. For this purpose, we look at different dimensions of proximity between actors. The concept of proximity originates from economic geography, where it is used for analyzing innovation processes in multi-actor constellations (Knoben and Oerlemans 2006). The underlying idea is that the different dimensions of proximity (geographical, cognitive, institutional, social, and organizational) enable knowledge exchange between two (or more) agents by providing means for resolving the underlying coordination problem of inter-organizational collaboration (Boschma 2005). Building upon the large body of scholarly work on proximity, we develop a proximity framework combining findings from sustainability science and innovation system theory. We then apply the framework to analyze selected examples from the literature, aiming to understand better the role of proximity in knowledge processes leading to sustainable innovations.

Our framework differentiates between macro-institutional and micro-institutional proximity, allowing for a more detailed analysis. The macro-institutional dimension refers to the overarching institutional framework, while the micro-institutional sub-dimension concerns the involved actors' internal rules, processes, and norms. We furthermore differentiate between systems-cognitive, normative-cognitive, and transformative-cognitive proximity based on the knowledge types supposed to be relevant for sustainability transitions. Establishing these sub-dimensions allows a more fine-grained examination of the complex relationships between the cognitive distances of two actors and the potential for knowledge exchange, learning, and innovation.

We find that sustainability-oriented innovations depict varying levels regarding the different proximity dimensions, with frequently low levels of micro-institutional and systems-cognitive proximity. Furthermore, normative-cognitive proximity seems to play an important role, while the other proximity dimensions can also serve as strong enablers for knowledge processes within these networks. Our results shed light on the actor constellations in sustainability-oriented networks. They give firms, government agencies, and user groups a valuable tool at hand when innovation networks are designed. Obviously, creativity and, with it, the potential contribution to sustainability targets ask for heterogeneous knowledge to be combined, which automatically increases the uncertainty of innovation as well as the need for close interaction and coordination. Considering the different proximities is an instrument that, on the one hand, focuses on the necessary distances in order to be creative and to develop sustainable novelties. On the other hand, the distances must not be too large to provide a common basis and understanding for the research ventures. Distances in some dimensions can finally be overcome by guaranteeing proximity in other dimensions. Keeping this in mind in the strategies to set up sustainability innovation networks potentially contribute to their success.

Our study provides contributions to both innovation studies and sustainability science through the introduction of a refined proximity framework. The differentiation of two proximity dimensions is a novel approach that allows for a refined analysis of actor constellations. This framework offers valuable insights into the dynamics of relationships among diverse partners within sustainability-oriented innovation networks and thus helps to better understand the associated knowledge processes and the occurrence of sustainable innovations as well as hindering constellations. For practitioners, the framework could prove useful to improve the performance of sustainability-oriented innovation networks and thus indirectly contribute to tackling environmental, social, and economic sustainability challenges.

The remaining paper is divided into six sections. “Sustainable innovations” and “Knowledge” discuss the theoretical background regarding sustainable innovations and knowledge for sustainability transformations. “The five dimensions of proximity” develops the proximity framework, which is applied in “Analysis of case studies”. In “Discussion”, we discuss the findings, and in the final section we offer some conclusions.

Sustainable innovations

There are several different notions for innovations that incorporate social or environmental objectives, such as green innovation, eco-innovation, environmental innovation, sustainability-oriented innovation, sustainable innovation, socio-ecological innovation, or socially responsible innovation, to give a non-exhaustive list. There is no single, universally applied definition for any of these terms. None of them is used in a strictly uniform manner, and there certainly is conceptual overlap between the different notions. For example, Rennings (2000) addresses eco-innovations as new products, processes, approaches, or practices that reduce environmental burdens and thus primarily contribute to the ecological aspects of sustainability. Adams et al. (2016) argue for sustainability-oriented innovations that imply changes to the norms and culture of an organization, resulting in new products and processes to generate positive social and environmental effects additionally to economic profits. Such extensive changes can concern all processes of an organization, implying concepts such as sustainable procurement (see for example Meehan and Bryde 2011), or sustainable human resource management (see for example Piwowar-Sulej et al. 2023). Mazzanti (2018) stresses eco-innovations’ role in attaining a symbiosis between different aspects of sustainability to preserve high living standards and competitiveness while avoiding environmental degradation. This argument is supported by Rao et al. (2023), who show that innovative environmental, social, and governance practices can positively influence the financial performance of firms. Dabard and Mann (2023) emphasize the procedural character of sustainability innovations that involve multiple actors and produce positive outcomes regarding social and environmental aspects within specific social–ecological contexts.

Despite the conceptual similarities, Franceschini et al. (2016) suggest caution in using those terms synonymously, as they are associated with different topics and scientific communities. However, there seems to be a common denominator for all these terms that distinguishes them from conventional innovations: the goal orientation given by the normative concept of sustainability (Strambach 2017). This means that the economic viability or market outcome is no longer the sole determinant of whether a novelty is considered a successful innovation. For a sustainable innovation, a positive environmental and social impact is equally decisive.

Many authors argue that the issue of sustainability requires transformative innovations that induce wide-reaching, systemic changes, given the complex, multi-layered nature of sustainability (Boons et al. 2013; Schlaile et al. 2017). This claim corresponds to the understanding of sustainable innovations as bundles of novelties (Dabard and Mann 2023), and more generally, it is linked to the argument that purely technological solutions will not suffice for achieving sustainability (Grunwald 2007; Sterman 2008), as the depletion of natural resources is at best postponed by new technologies (Goncz et al. 2007), which is also reflected in time lags of market-driven diffusion and application of new technologies (Jacobsson and Bergek 2011). Furthermore, rebound effects can offset efficiency gains from new technologies (Berkhout et al. 2000). They can even lead to higher overall levels of resource consumption if they induce additional economic growth (Zink and Geyer 2017). More generally, the sustainability impact of a new technology depends on its usage within the prevalent socio-economic and political framework conditions (Goncz et al. 2007; Grunwald 2007). There are countless examples of poorly targeted technological and process innovations that lead to unintended consequences and result in negative impacts on sustainability (Hauser et al. 2011). Thus, technological innovations must be accompanied by and embedded in social, institutional, and policy innovations to support and facilitate the sustainability transformation (Mulder 2007). Comprehensive economic concepts that integrate sustainability, like a sustainable circular economy (Geissdoerfer et al. 2017), or a knowledge-based sustainable bioeconomy (e.g., Pyka and Prettner 2018), take this into consideration. However, such broad concepts consequently imply numerous challenges, as Irfan et al. (2022) exemplary show for the case of introducing biomass energy in India, identifying twenty-four barriers in five categories (“technological and infrastructural”, “economic and financial”, “political and institutional”, “cultural and behavioral”, and “metrological”). This implies that countries can follow various pathways in their transition toward sustainability, despite being located in regions with cultural and socio-economic similarities, as exemplified by Alvarado et al. (2022) describing biocapacity convergence clubs in Latin America.

Knowledge

Most aspects of present economic systems are knowledge based (confer Foray 2004). Knowledge is the central resource in any innovation system. The agents’ ability within an innovation system to find new combinations (re-combinations) of existing knowledge elements forms the basis of the innovative performance of the system. The creation, co-creation, and transfer of knowledge are thus central processes for the emergence of innovations. And those processes require time and resources, as knowledge is sticky (Hippel 1994), path dependent (Dosi 1982), dispersed and cumulative (Foray 2004), and requires a certain level of absorptive capacity on the side of the receiving agents, which enables them to acquire new knowledge (Cohen and Levinthal 1990; Canter and Pyka 1998).

Three types of knowledge are deemed relevant for sustainability transition processes: systems knowledge, normative knowledge, and transformative knowledge (Wehrden et al. 2017; Trencher et al. 2017; Grunwald 2007; Rauschmayer et al. 2015). Systems knowledge refers to the Earth as a system consisting of various subsystems, such as the techno-economic and socio-environmental systems, which are interlinked and mutually influence each other. It covers all aspects of humanity’s production and consumption patterns (institutional, technological, organizational) and their effects on the natural environment. It implies analytical and descriptive knowledge about the currently given sustainability problems. Normative knowledge refers to the desired state of the earth system and all of its subsystems and anchors sustainability in decision-making processes. Transformative knowledge refers to the strategic and operational questions that explore which transition pathways are viable and promising for overcoming critical thresholds and for reaching the desired target state of the earth system. It is thus about societal change toward sustainability, about identifying feasible strategies for such change, taking into account the inevitable uncertainty about the chances of success and the actual costs of change, and with it, the likelihood of necessary adaptations and corrections in the future.

For analytical purposes, it is helpful to distinguish between three knowledge processes: creation, transfer, and application. Although in the context of most innovation processes, we can barely see a linear or sequential approach to these steps, as knowledge creation or co-creation, diffusion, and application happen rather simultaneously, in an iterative manner, whereby the transfer and application of (incomplete) knowledge generate feedback which serves to create new knowledge which is again transferred or applied, generating further feedback and so forth (Grunwald 2007).

Specific peculiarities of these knowledge processes have been identified regarding the wicked nature of sustainability (Rauschmayer et al. 2015; Schlaile et al. 2017). One is the emphasis on problem- and solution-oriented methods (e.g., Abson et al. 2017; Batie 2008; Kates et al. 2001; Wiek et al. 2012; Agamuthu and Hansen 2007). This should not be misunderstood to mean we no longer need curiosity-driven basic research. But for addressing the sustainability challenge, knowledge usable and applicable in different local contexts and institutional settings is required. Promising approaches in this regard seem to be experiments, such as sustainability transition experiments, policy experiments, or living laboratories (Luederitz et al. 2017; Evans et al. 2015; Leal Filho et al. 2023).

Another peculiarity of knowledge processes for sustainability is the apparent need to follow interdisciplinary or transdisciplinary approaches (Lang et al. 2012; Kates et al. 2001; Wehrden et al. 2017; Binder et al. 2015). The relevant knowledge for comprehensive solutions to sustainability problems requires expertise from a wide range of different scientific disciplines. It furthermore requires knowledge from diverse societal stakeholders, including an understanding of norms, values, and visions of sustainability. Hence, approaches for (co-)creating new knowledge must deal with different methodologies, ontologies, and epistemologies and overcome the obstacles in the communication building on specific disciplinary languages.

In close connection to the idea of inter- and transdisciplinary approaches is the concept of engaging all relevant actors in knowledge processes, resulting in co-design, co-production, and co-creation, which ideally leads to mutual learning of science and society (confer Zilahy and Huisingh 2009; Urmetzer et al. 2018; Trencher et al. 2017; Kates et al. 2001; Wittmayer and Schäpke 2014). The rationale for this is that the relevant knowledge for sustainability is not exclusively produced by the academic system. Furthermore, the normative nature of sustainability implies that all concerned stakeholders should be included to cover all relevant values and beliefs adequately. Put differently, formulating solutions for sustainability challenges that induce wide-reaching change seems more promising if the ones affected by the change are included in the process of finding these solutions.

The five dimensions of proximity

The innovation networks that are likely to produce innovations for the transformation toward sustainability include a range of diverse actors from business, academia, government, and civil society, who repeatedly interact in the creation and transfer of knowledge. Such collaborations with organizations from different institutional spheres with different disciplinary backgrounds are very demanding. Numerous aspects are to be considered to make such cooperation work. It starts with a common vocabulary and shared understanding of basic concepts, covers diverging interests, incentive structures, and constraints, and concerns power structures and different ideas about cooperation outcomes.

To better understand knowledge flows in sustainability-oriented innovation networks, analyzing the relations between the actors and the corresponding mechanisms for exchanging knowledge is helpful. The concept of proximity, stemming from economic geography, provides a powerful tool in this regard. According to Boschma (2005), there are five proximity dimensions, namely geographical, cognitive, institutional, social, and organizational, which all serve to understand better the mechanisms that resolve the coordination problem in inter-organizational collaborations. In the following, we briefly describe the five dimensions and how they are encompassed in our framework, including the differentiation of two dimensions. The resulting framework is shown in Fig. 1.

Fig. 1
figure 1

Multidimensional proximity framework (CP = cognitive proximity)

Full size image

Geographical proximity

The idea that location influences the outcomes of economic processes is rather old and dates back to Alfred Marshall (1920) and the subsequent works on external economies of scale leading to industrial districts (Capello 2014). Geographical proximity concerns the physical distance between two agents. It can be expressed by a measure of length, travel time, or another appropriate scale, depending on the specific circumstances of the analyzed setting and distinctive features such as national borders (Boschma 2005). The co-location of two or more actors fosters innovation by facilitating direct interaction and face-to-face exchange (Fritsch and Schwirten 1999), which also simplifies the transfer of tacit knowledge (Knoben and Oerlemans 2006). Hence, high geographical proximity is conducive to knowledge exchange and innovation. However, there are reasons for including specific agents in an innovation network, even if they are situated far away. This is the case when finding solutions to sustainability challenges (even for context-specific, local challenges) requires specialized knowledge elements that are dispersed among a few geographically spread agents (Janssen and Abbasiharofteh 2022).

Cognitive proximity

Cognitive proximity is associated with shared mental models and a shared understanding of the world (Boschma 2005) and is closely connected to absorptive capacity (Cohen and Levinthal 1990) and the cumulative nature of knowledge. Put simply, two agents cannot exchange knowledge if they do not understand each other. Hence there needs to be a certain overlap in their knowledge bases. Nevertheless, if two agents possess nearly the same knowledge, there is little to learn from each other. Thus, the relationship between cognitive proximity and the potential for knowledge exchange and innovation appears to have an inverted U-shape, with the peak representing the optimal cognitive distance between two agents (Wuyts et al. 2005; Nooteboom et al. 2007). This raises questions about finding this point of a promising distance (or proximity) and how to translate this into actor constellations in sustainability-oriented innovation networks.

Building upon the insight mentioned above on the three types of knowledge that are particularly important for the transition toward sustainability, we propose to further differentiate this dimension into three sub-dimensions: normative-cognitive, systems-cognitive, and transformative-cognitive proximity. This fine-tuning allows for a more accurate analysis of the cognitive proximity dimension and its influence on knowledge exchange and innovation.

Figure 2 illustrates the methodological and theoretical approach behind the refined proximity framework and the subsequent analysis of examples from the literature. The findings from sustainability science and innovation system theory, discussed above, fed into the differentiation of two proximity categories and thus helped to elaborate a refined framework.

Fig. 2
figure 2

Conceptual approach

Full size image

Institutional proximity

Institutional proximity considers the differences and similarities regarding the institutional framework under which two or more agents operate. It refers to rules and regulations, shared norms of conduct, routines, or established practices (Boschma 2005) and can also refer to the internal institutional logic of the involved actors covering aspects such as incentive structures, constraints, norms, or organizational culture (Ponds et al. 2007). On the one hand, institutional proximity can be conducive to knowledge exchange and innovation, as the collaboration of two agents is more straightforward if they operate under the same legal framework, have similar incentives, and follow the same codes of conduct. On the other hand, sustainability-oriented innovation networks, by their very nature, include organizations from different institutional spheres (business, government, academia, and civil society) and thus automatically depict low levels of proximity regarding their internal institutional logic.

Thus, we distinguish in our framework between micro-institutional proximity, referring to the internal institutional settings of the involved actors, and macro-institutional proximity referring to the overarching institutional conditions. This differentiation implies that macro-institutional proximity is linked to geographical proximity, as many institutions, such as rules, regulations, cultural norms, and habits, are place bound. However, they are not congruent, as we understand geographical proximity in a purely physical sense (confer Werker et al. 2019). That means that two actors can be geographically close but still have a low level of institutional proximity, for example, if a country border separates them or if they belong to different regions with different cultural and institutional settings.

Organizational proximity

The (co-)creation and transfer of knowledge come along with uncertainties and the risk of opportunistic behavior regarding the appropriability, use, and misuse of knowledge (confer Hennart 1993). This requires control mechanisms to ensure ownership rights and safeguard appropriate benefits for investing resources in knowledge creation and exchange. From transaction cost economics, we know different forms of governance in this regard, such as markets, hierarchies, or hybrid forms, which all have different incentive structures and different mechanisms for enforcement and penalization (Williamson 1989). Pyka (2002) shows the usefulness of Williamson’s approach for understanding the dynamics of innovation networks as a hybrid between hierarchies and markets). Organizational proximity refers to such inter-organizational constellations and pertains to the extent to which two (or more) actors are linked via a formal arrangement (Boschma 2005). It thus refers to the level of independence and the extent of authority that can be exercised in organizational arrangements, or in other words, it is established through the existing power structures and hierarchies in intra-organizational or inter-organizational constellations (Mattes 2012). The economic or financial dependence between two actors has a strong influence in this regard (Kirat and Lung 1999).

The level of organizational proximity of a specific arrangement thus appears to be located along a continuum (Boschma 2005). For example, a low level of organizational proximity is given when two fully independent actors with no formal ties at all collaborate, and a high level would correspond to the case of two firms belonging to the same parent company. In between lie different arrangements such as loosely coupled networks or formal cooperation agreements such as research joint ventures, etc.

Organizational proximity is assumed to benefit knowledge exchange and innovation by reducing the risk of opportunisms and uncertainty about the partner’s motives when two actors collaborate (Boschma 2005; Hansen 2015). Too much proximity on the other hand may result in excessive bureaucracy and a lack of flexibility, impeding the ability to learn and innovate. Overly formal rules and strict hierarchies presumably do not reward new ideas and novel approaches. Close, inward-looking organizations or networks of organizations might have limited access to new sources of knowledge. In contrast, informal, voluntary collaborations without concrete obligations provide flexibility in defining objectives, committing resources, and involving new partners. However, this flexibility can be a drawback, as partners may unilaterally end their engagement. The absence of formal commitments may also lead to purely self-interested actions, possibly misusing shared information and compromising cooperation.

Social proximity

Social proximity refers to the interpersonal connections among individuals within cooperating organizations, built on trust stemming from kinship, friendship, mutual sympathy, or past work experiences (Boschma 2005). Rooted in Granovetter's (1985) idea that economic interactions are embedded in social networks, social proximity is assumed to foster knowledge exchange and innovation in more informal but strongly reciprocal network relationships. Similar to organizational proximity, it brings clarity to the motivations of cooperating partners, reducing the risk of opportunistic behavior. Informal mechanisms based on social relations reduce transaction costs and establish reciprocity (Balland et al. 2022). These trust-based contacts can replace inefficient formal arrangements. Insufficient social proximity may breed mistrust and complex formal transactions, hindering knowledge (co-)creation and transfer. Conversely, excessive social proximity might lead to overlooking opportunism and excluding actors with whom no prior working experiences exist, resulting in a lock-in that hampers access to new ideas and complementary knowledge (Boschma 2005).

The degree of social proximity between two actors depends on factors like their personal acquaintance, emotional closeness, feelings of obligation, and the level of trust between them (Huber 2012; Hansen 2015). Trust in general plays a crucial role in collaborative innovation and interactive learning processes, as emphasized by various authors highlighting its significance in facilitating exchange and cooperation (see for example Velenturf and Jensen 2016; Dovey 2009; Hardwick et al. 2013). Trust is essential for both the sender and receiver in a transmission process. The sender must be trusted to provide valuable and accurate information, while the receiver must be trusted not to misuse it. Given that learning and knowledge creation involve costs and commitment of resources, trust, in the sense of anticipating a positive outcome, becomes a prerequisite. The involvement of various agents, such as companies, universities, state agencies, and NGOs, with quite different organizational cultures, elucidates the importance of trust for sustainable innovations.

Interdependencies of the different proximity dimensions

The different dimensions of proximity are interlinked (Mattes 2012; Heringa et al. 2014). Two effects of interdependencies between the various dimensions can be distinguished: one is reinforcement, and the other effect is overlap or substitution (Hansen 2015). Reinforcement means that a high level in one proximity dimension can favor a high level in another. For example, geographical proximity can be conducive to building trust and thereby increase social proximity (Delgadillo et al. 2021). Another example is a high level of organizational proximity that can reinforce micro-institutional proximity in the long run if the organizational arrangements lead to shared codes of conduct, norms, and values (Addy and Dubé 2018). The effect of substitution or overlap refers to the fact that one proximity dimension can replace another or show a certain degree of correlation based on shared, underlying characteristics. An example of substitution are informal, trust-based rules for cooperation (social proximity), which might replace formal contractual arrangements (organizational proximity). Another example is that a high level of mutual understanding diminishes the need for frequent face-to-face contact, meaning that a high level of cognitive proximity can facilitate long-distance collaborations and thus substitute for geographical proximity. Overlap can, for instance, be seen between geographical and macro-institutional proximity, as many rules and norms are place specific.

Analysis of case studies

In the next step, we use our proximity framework to analyze selected examples reported in the literature. Table 1 provides an overview of the selected articles, their topics, and the actor constellations described. The papers were chosen according to the following criteria: All articles discuss the emergence of sustainable innovations, addressing social, environmental, and economic aspects. The selected examples furthermore all provide a substantial presentation of details regarding two overarching parameters. Firstly, each case portrays the involved actors, elucidating their relationships and the dynamics of their interaction. This facilitates an analytical exploration of the different proximity dimensions. Secondly, the papers explicate the knowledge processes underpinning sustainable innovations, encompassing the dynamics, challenges, and obstacles that occurred during knowledge creation, co-creation, and transfer. Furthermore, the amalgamation of the diverse articles spans a wide array of thematic fields, industries, and socio-economic settings, providing a broad representation of sustainable innovations in varied contexts.

Table 1 Sample selection from the literature
Full size table

Transnational development of a soil sensor

Baek et al. (2019) analyze a collaborative design project between a social enterprise from Myanmar and a university from South Korea. The project’s objective was to create a soil sensor to improve sustainability in agriculture. The sensor would enable farmers to measure different soil qualities, allowing them to use inputs more efficiently and increase productivity, leading to poverty alleviation as a positive economic and social impact of this innovation. The main partners of this international cooperation, the Myanmar social enterprise and the Korean university, brought complementary knowledge bases to the project. Furthermore, farmers in both countries were questioned to capture their needs, meaning that their knowledge was incorporated as well, although to an inadequately low extent. Ultimately, the project did not provide the intended results, which Baek et al. (2019) mainly ascribe to socio-economic, technological, and cultural differences. Through the lens of our proximity framework, we can see low levels of normative-cognitive and transformative-cognitive proximity between the actors. On the Myanmar side, the soil sensor and its application were considered a social innovation, while the South Korean partners considered it mainly a technological innovation. The Myanmar partners (especially the Myanmar farmers, who were only marginally included) had different ideas about applying the sensor due to other farming practices and mindsets. The large distance between the partners (low geographical proximity) made frequent direct contact impossible. The cognitive distance could thus not be overcome, despite the extensive use of information and communication technology (ICT). The two countries’ different socio-economic and institutional conditions (low macro-institutional proximity) also played a role. Different understandings of the expected social impact of the desired innovations and their application existed initially and could not be resolved but were rather magnified. Organizational and social proximity were also low and thus did not facilitate the exchange of (implicit) normative and transformative knowledge, which would have been required.

International cooperation in organic farming

Haas et al. (2016) analyze a case of co-innovation involving a community of Nepalese smallholder farmers, a Nepalese ethical business run by a German entrepreneur, an Austrian university, and an Austrian firm. Further stakeholders, such as government agencies, were involved as well, however, to a smaller extent and only during parts of the collaboration. The partners brought complementary knowledge, expertise, and skills into the cooperation, partly due to the different institutional spheres to which they belong. Those complementary knowledge bases provided promising ground for the development of sustainable innovations. Using our framework, this setting can be described by low levels of systems-cognitive and micro-institutional proximity. At the same time, there were high levels of normative-cognitive and transformative-cognitive proximity as all partners had similar ideas about the environmental and social impacts of the intended innovations, their measurements, and how to achieve change. Established certificates for organic agricultural products also played a role in this regard. The establishment of the innovation network was facilitated by social proximity, as several partners had (bilateral) prior working experiences with each other. Despite the large physical distance between the Austrian and the Nepalese partners, the knowledge exchange did function well, not least because one of the Nepali partners (the German founder of the Nepalese ethical business) had the same cultural background as the European partners, what can be described as a situation in which one actor brought the macro-institutional experience to other partners, thus creating a setting with medium macro-institutional proximity and low geographical proximity. The German entrepreneur in Nepal thus played a decisive role, as “this kind of communication was easier than an intercultural dialog without him acting as a bridge might have been” (Haas et al. 2016). A certain degree of organizational proximity was established when the Nepalese smallholder farmers applied for funding from a German development aid organization, making the ethical business a formal marketing partner. This was done after the innovation network was established for some time and provided additional routine for the network’s activities.

Innovations for sustainable tourism

Higuchi and Yamanaka (2017) describe the collaborative development of sustainable innovations in the tourism industry by a tour operator and a university, systematically incorporating user feedback (tourists). Both partners brought specific knowledge into the cooperation, depicting a low level of systems-cognitive proximity. A difficulty, especially in the early phase of collaboration, was the equally low level of transformative-cognitive proximity. The university needed time to fully comprehend the firm’s business interests, what constitutes usable knowledge for a sustainable tourism operation, and how to reach users in this context to achieve a sustainable impact. The tourist operator needed time to adequately understand the motives of the university and the value and character (rigorous nature) of scientific knowledge. Nevertheless, the cooperation developed very well and led to satisfying results for both sides, culminating in innovations that received positive feedback from tourists. As there was no formal organizational arrangement (low organizational proximity) and no former experience of cooperating (low initial levels of social proximity), the decisive factor was the high geographical proximity and a basic shared idea of the intended sustainability effects (medium level of normative-cognitive proximity). The partners met frequently, with university members even participating in the daily business of the tour operator. This built up trust (increasing levels of social proximity) and allowed the partners to reach a level of cognitive proximity conducive to developing sustainable innovations.

Sino-German cooperation in the green building sector

Strambach (2017) analyzes transnational Sino-German innovation projects in the green building sector, highlighting differences between the two countries in this branch. The German green construction sector is described as an example of a well-functioning, almost industry-wide knowledge network (a sectoral innovation system, Malerba 2002) in which numerous actors from the whole country and different institutional spheres (low micro-institutional and low geographical proximity) combine their diverse knowledge bases for collaborating and developing innovations (low systems-cognitive proximity). The collaborative efforts of the network are supported by an intermediary organization (high organizational proximity) which promotes regular exchange (high social proximity). The collaboration has created high normative-cognitive and transformative-cognitive proximity, also manifested in a widely recognized voluntary quality certificate for sustainable buildings. The network is quite successful and provides comprehensive input to legislation in this area, thereby influencing the macro-institutional framework of this branch.

In contrast, the Sino-German cooperation projects examined by Strambach (2017), which were all situated in China with frequent face-to-face contact with the partners on-site, encountered several challenges. The main challenges were different sustainability concepts and the corresponding technologies and artifacts (low normative-cognitive proximity), as well as differing evaluations of cost and time issues and the subsequent planning approaches (low transformative-cognitive proximity). Differences that Strambach (2017) associates with fundamental differences in macro-institutional settings in both countries, shaped by different institutional trajectories (low macro-institutional proximity). The German firms involved in these projects applied different strategies to cope with these challenges. One strategy was knowledge transfer via heads by directly employing Chinese personnel and thus bridging the macro-institutional distance. Some firms following this approach relied on Chinese staff with an educational background in Germany (meaning a lower internal macro-institutional distance within the German firm). Others focused on Chinese personnel with experience in the construction sector in China (lower macro-institutional distance to partners in China). The second strategy that firms applied was to increase social and organizational proximity by intensifying and formalizing relations with selected Chinese partners, primarily of a similar size and type, meaning partners with high micro-institutional proximity.

Sustainable urban development in a cross-border region

Valkering et al. (2013) examine collective innovation processes for sustainable urban development in a European cross-border region. The area, called Euregio Meuse-Rhine, covers parts of Belgium, Germany, and the Netherlands and is relatively small, implying a high geographical proximity. In this region, several neighborhoods with different official languages, cultures, and institutional settings cooperated within the frame of a project funded by the EU, implying a low level of macro-institutional proximity and a medium level of organizational proximity. Various actors from business, academia, state, and civil society were involved in the funded project, bringing their specific expertise to the project (low systems-cognitive proximity). In the preparation phase of the project, an adequate level of proximity regarding the normative-cognitive dimension was established by formulating an overarching objective, which was detailed via the formulation of specific practical challenges and targets. The main obstacles to collaboration were associated with low macro-institutional proximity and manifested in mundane issues such as language barriers. The geographical proximity and the corresponding opportunity for regular face-to-face meetings helped to overcome this challenge. The project furthermore showed that cross-border cooperation was easier when actors of similar type or with similar roles cooperated, thus showing that high micro-institutional proximity can mitigate low macro-institutional proximity.

Sustainable innovations in the forest products industry

Van Horne and Dutot (2017) examine sustainable innovations in the forest products industry in the Canadian Province of Quebec, involving actors from government, industry, academia, and intermediary organizations (high macro-institutional proximity but low geographical proximity as actors are spread throughout the large region). Following a quadruple helix approach (Carayannis and Campbell 2010), the authors emphasize that all these actors play significant roles in creating sustainable innovations by bringing complementary skills, expertise, and knowledge into the innovation system. At the same time, the cognitive distance between the actors is one of the main challenges, illustrating the inverse U-shaped relation of cognitive proximity and innovation. The high complementarity of knowledge bases provides ample room for creating new and innovative knowledge combinations. But at the same time, the differences in the knowledge bases exacerbate knowledge exchange due to insufficient mutual understanding. Looking into detail, much of the problem in the case of van Horne and Dutot (2017) seems to be caused by low transformative-cognitive proximity, as the academic researchers struggled to understand the issues in the industry and the kinds of solutions that work for firms. Firms, as well as intermediary organizations, had difficulties absorbing the methods and approaches of the universities. The low micro-institutional proximity, resulting from the diverse actors, is seen as a prerequisite for the successful creation of sustainability innovations but can also impose serious challenges. Different time horizons, goals, values, motivations, and modes of working on problems, made collaboration a demanding endeavor. Several innovation projects struggled with tensions between the pursuit of scientific excellence and the applicability of solutions. Van Horne and Dutot (2017) identify the direct engagement of university personnel in firms as a promising strategy to cope with these difficulties. Researchers who spent time embedded in the industry could be the key to successful knowledge exchange and mutual learning. This strategy builds upon increasing geographical proximity (and partly social and organizational proximity) to mitigate the lack of cognitive proximity and facilitate the transfer of implicit knowledge.

Discussion

Figure 3 gives an overview of the initial levels of the five proximity dimensions for the examples from the literature. The collaborative project analyzed by Baek et al. (2019), which produced results below expectations, is the sole one having only low to medium levels in all proximity dimensions. The other examples have high levels of proximity in at least two dimensions and a medium level in at least one more dimension. This suggests that predominantly low levels of proximity can seriously impede knowledge exchange and innovations, and at the same time, high levels in all dimensions neither seem necessary nor favorable. A combination of high and low levels in different proximity dimensions seems more conducive to innovations. It thus appears to be about having a balanced level across the different proximity dimensions (Ievoli et al. 2019).

Fig. 3
figure 3

Initial levels of the proximity dimensions for the examples from the literature

Full size image

As mentioned before, a minimum level of cognitive proximity seems to be a prerequisite for knowledge exchange (confer Boschma 2005; Huber 2012). At the same time, insights from sustainability science point out that sustainable innovations require inter- and transdisciplinary knowledge from various actors and a comprehensive inclusion of relevant stakeholders in the corresponding knowledge processes. Both are reflected in the examples from the literature, as they all depict low levels of systems-cognitive and micro-institutional proximity. In other words, deriving sustainable innovations that involve complex objectives, such as positive social and environmental impacts in addition to economic benefits, requires broad expertise in systems knowledge, which is reflected in the corresponding innovation networks that include diverse members from business, academia, government, and civil society. Still, they all need some basic common ground for exchange.

Scrutinizing further the issue of cognitive proximity, we can see that normative-cognitive proximity seems to play an important role: similar ideas about the sustainability impacts of the intended innovations seem to be of high importance. This sub-dimension almost appears to be a prerequisite for facilitating knowledge exchange and innovations and seems related to the significance of problem- and solution-oriented approaches in knowledge generation. The importance of a shared understanding regarding the (normative) sustainability goals in innovation networks was already observed by others such as Coenen et al. (2010), Akpo et al. (2015), Bünger and Schiller (2022), and Bisseleua et al. (2018), and also fits similar findings of other authors who address the importance of explicitly including norms and values in the processes of (co)-creating and transferring sustainability knowledge (Abson et al. 2017; Wiek et al. 2012). A low initial level of normative-cognitive proximity seems to require intensive exchange about the desired outcomes already at very early stages.

Transformative-cognitive proximity also seems highly relevant and is linked to focusing on specific problems and possible viable solutions. This sub-dimension concerns strategies for addressing sustainability problems and is connected to the relevance and applicability of (new) knowledge. It is about the mechanisms that make a novelty economically viable while simultaneously producing social and environmental benefits, thus making it a sustainable innovation. This appears to be a realm of sector-specific knowledge, as firms, academic institutions, civil society actors, and state agencies all have their peculiar knowledge about sustainability transformations and their ideas about what kind of knowledge is needed to arrive at such strategies. Rigorous scientific understanding created by a university might be of little value for a firm and its market-oriented practices. The transformative-cognitive sub-dimension also seems to be linked to the normative-cognitive one in the sense that the latter is a prerequisite for the former. If there is no normative-cognitive proximity, there is also no transformative-cognitive proximity. This seems plausible as actors need a shared understanding of the common goal before agreeing on how to get there together.

Organizational and social proximity seem to work as important auxiliary dimensions, which is a similar finding to that of Mattes (2012), who, however, also sees geographic proximity in this regard, which we do not support based on our results as we see different mechanisms at work. Both organizational and social proximity can be powerful enablers for knowledge exchange and innovation. Trust especially seems to be an important factor, which is in line with the findings of other authors regarding the role of trust in knowledge processes (Easterby-Smith et al. 2008; Roux et al. 2006).

The differentiation between micro- and macro-institutional proximity seems fruitful: macro-institutional proximity appears as an important framework aspect that shapes knowledge processes by providing a set of established rules, norms, and values. A lack of macro-institutional proximity can pose serious challenges to knowledge exchange and appears to be linked to different aspects of cognitive proximity, such as inherently different understandings of concepts or simply different languages. Micro-institutional proximity can apparently substitute macro-institutional proximity to a limited extent, as two similar organizations (for example, two firms of similar kind and size or two universities) cooperate more easily across macro-institutional borders. Figure 4 illustrates graphically the differing levels of proximity across the six examples from the literature.

Fig. 4
figure 4

Graphical comparison of the different levels of the five proximity dimensions in each example

Full size image

The literature on proximity describes numerous substitution effects, namely between social and geographical proximity, organizational and geographical proximity, social and organizational proximity, and between geographical and cognitive proximity (Hansen 2015; Dubois 2019; Velenturf 2016; Ghassim 2018). Based on the few examples from the literature we used to test our proximity framework, we can neither systematically confirm nor contest these findings but see supporting indications. At the same time, all examples that report successful knowledge exchange and the occurrence of sustainable innovations also describe a high level in more than one proximity dimension, which raises questions about the extent of substitutability between the different dimensions.

Hansen (2015) describes another interdependency between different proximity dimensions: an overlap between geographical and institutional proximity. Based on our differentiation in the sub-dimensions macro-institutional and micro-institutional proximity, we believe this overlap to be relevant only for the macro-institutional sub-dimension and only to a limited extent, as it does not apply to vast regions, such as Quebec in the example of van Horne and Dutot (2017). This overlap can also be canceled out by actors that “bring along” parts of a macro-institutional background from one geographic spot to another, such as the entrepreneur in the example of Haas et al. (2016), who was very well familiar with European-oriented norms, values and codes of conduct, but had his business set up in Nepal, on-site with the smallholder farmers.

Conclusions

The relationship between proximity and knowledge for sustainable innovations is complex and multifaceted. The different proximity dimensions all influence knowledge exchange between partners in an innovation network. Various combinations of high and low levels of the different proximity dimensions can be observed. They exhibit interdependencies, and the levels of certain dimensions can increase over time, showing that proximity must be understood as a dynamic concept (Balland et al. 2022).

Based upon the five dimensions of proximity (Boschma 2005) and incorporating findings from sustainability science and innovation systems theory, we developed a refined proximity framework that differentiates the institutional proximity dimension into two sub-dimensions, the macro-institutional referring to the overall institutional framework and the micro-institutional dimension, referring to the institutional setting of the involved actors themselves. The macro-institutional dimension helps to explain why geographically close partners might face obstacles when cooperating, while institutions of a similar kind more easily collaborate across large distances and different regulatory and cultural settings. We furthermore divide cognitive proximity into three sub-dimensions: systems-cognitive, normative-cognitive, and transformative-cognitive proximity. This helps to scrutinize further the complex, inverted U-shaped relation between differences in knowledge bases and innovative performance.

We found that sustainability-oriented innovation networks frequently depict low levels in systems-cognitive and micro-institutional proximity. These low levels of proximity need to be combined with high levels in other proximity dimensions to enable knowledge exchange and the development of innovations. Especially normative-cognitive proximity (and following on from this transformative-cognitive proximity) seems to be important in this regard. A shared vision of the positive impacts of intended innovations significantly facilitates collaboration. However, our study does not allow us to conclude whether this is enough for innovation networks to produce the expected results. All cases that reported successful developments of innovations are characterized by high levels of proximity in at least two dimensions. Hence it might be necessary also to have a high level in another proximity dimension. This is presumably more relevant for networks with low normative-cognitive (and low transformative-cognitive) proximity. These questions are on our future research agenda.

Our study contributes to the field of innovation studies and sustainability science by providing a refined proximity framework that offers insights into the relations of the different partners in sustainability-oriented innovation networks. By differentiating cognitive and institutional proximity dimensions into further sub-dimensions, we believe to add to the existing theory. This expansion, based on the combination of findings from different fields of analysis, appears to be a prolific approach for further investigating sustainability-oriented innovation networks and their dynamics. This contributes to a better understanding of the knowledge processes leading to sustainable innovations and serves as a means for tackling various sustainability challenges. The societal benefit of this research has to be seen in promoting actor constellations in innovation networks that are conducive to effective knowledge co-creation and transfer.

We suggest that our framework and the associated insights can be used as a tool for identifying partners that could fit into existing innovation networks. Based upon the insights of our study, an existing network could assess whether the inclusion of a specific partner might be conducive to knowledge exchange or whether this will likely evoke obstacles caused by the relative proximity to the other partners in the network. Our results emphasize the necessary distances required for fostering creativity and the development of sustainable innovations. On the other hand, these distances should not be excessively large, as a certain proximity is crucial for establishing a shared ground and mutual understanding. While distance in some dimensions may present challenges, proximity in others helps overcome such barriers. Bearing this in mind when formulating strategies for establishing sustainability innovation networks has the potential to enhance their overall success.

The shortcomings of our study surely lie in the limited number of examples from the literature used to derive conclusions about the applicability of our framework and the prevalent combinations of the different proximity dimensions in sustainability-oriented innovation networks. However, a larger number of examples would have exceeded the scope of this paper, and the cases we scrutinize already allowed for broad conclusions. Nonetheless, a quantitative approach could be the next step to test and validate this study’s findings and examine the above questions about minimal requirements regarding combinations of proximity dimensions that allow networks to exchange knowledge and develop sustainable innovations.