Environmental transition in clusters: empirical insights from the benchmarking study in Poland

Environmental sustainability has become one of the most important dimensions of competitiveness. Firms are increasingly expected to reduce emissions, improve resource efficiency, adopt circular solutions and respond to tightening regulatory, technological and market pressures. Yet, such transformation depends not only on the efforts of individual firms, but it requires access to knowledge, complementary capabilities, external partners and supportive institutional arrangements. In this context, clusters have attracted renewed attention as organisational environments that may help firms respond to transition pressures while sustaining or strengthening their competitive position. Environmental upgrading requires not only innovation but also collaboration, strategic direction and intermediary support capable of aligning heterogeneous actors around shared goals.

Accordingly, an emerging body of literature has begun to examine clusters not only as engines of productivity and innovation but also as potential vehicles of sustainability. However, there is still limited explanatory evidence on why some clusters engage more intensively in green transformation than others. Existing studies have produced important conceptual insights and case-based findings, but comparative benchmarking evidence remains scarce. This is particularly true for Central Europe, where cluster policy has played an important developmental role, yet the environmental dimension of cluster activity remains underexplored. Moreover, prior research has usually focused either on cluster organisations or on firms, while rarely combining both perspectives in a matched empirical design. As a result, there is still insufficient knowledge not only about the determinants of green transformation at the cluster level, but also about whether cluster-level environmental activity is recognised by members and translated into perceived firm-level benefits.

This gap is especially relevant in the Polish context. Poland has developed an extensive cluster policy framework, including the National Key Cluster (Krajowy Klaster Kluczowy, KKK) programme, which formally recognises selected clusters of strategic importance. At the same time, Polish firms and support organisations face growing pressure to adapt to the green transition, including energy transformation, circular economy requirements and ESG-related expectations. This makes Poland a useful setting for examining whether and under what conditions clusters can operate as intermediaries of environmental transformation, and whether institutional recognition, strategic formalisation and organisational capacity matter for sustainability-oriented competitiveness.

The objective of this study is to identify the determinants of green transformation in Polish clusters and to assess whether cluster-level environmental activity translates into positive perceptions and reported benefits at the member-firm level. More specifically, the study examines whether institutional status, strategic commitment, innovation capacity, cooperation intensity, resources and cluster size are associated with the scope of environmental actions undertaken by clusters. It also investigates whether clusters that are more active in green transformation are more positively evaluated by their members, more intensively used as providers of green services and more likely to be associated with reported environmental improvements in member organisations.

To address this objective, the study draws on matched survey data from 42 cluster coordinators and 733 member organisations collected within the 2024 benchmarking study of Polish clusters. At the cluster level, the analysis explains variation in a Green Transformation Index (GTI) based on six environmental actions reported by coordinators. At the member level, it examines whether more environmentally active clusters are associated with greater use of green services, more favourable evaluations of the coordinator’s efforts and stronger reported environmental improvements in member organisations.

The paper is structured as follows. The next section reviews the literature on clusters, eco-innovation and sustainability transitions and develops the analytical framework. The following sections describe the data and methodology, present the empirical results and discuss their implications for theory, policy and cluster management.

Clusters have long been linked to advantages that arise when enterprises and supporting institutions are situated in close proximity. The classical cluster theory emphasises localised specialisation, information spillovers, talent agglomeration, specialised suppliers and frequent interactions as crucial factors contributing to competitive advantage (Porter, 1998; Porter, 2000). Nevertheless, the literature has clarified that clusters do not inherently produce benefits. They evolve based on their capacity to regenerate, draw from external information sources and adapt to changing technological and market conditions (Menzel and Fornahl, 2010). This is especially pertinent to environmental sustainability, since change is often abrupt and generally necessitates novel patterns of cooperation, investment and collective learning.

Recent research on clusters has increasingly focused not just on their role as drivers of competitiveness but also on their potential to create conditions that stimulate and accelerate environmental transition (Hervás-Oliver et al., 2024). This change in perspective redirects the discourse from individual companies to the inherent collective capabilities of cluster frameworks. Clusters should not be viewed simply as collections of individual firms’ efforts towards environmental sustainability. Instead, they must be understood as a process arising from interdependencies, organised intermediary entities and shared infrastructures, all supported by broader institutional frameworks. The sustainability transitions literature offers a valuable perspective by emphasising the role of intermediaries in mitigating uncertainty, connecting stakeholders and facilitating the dissemination of new practices (Kivimaa et al., 2019; Mignon and Kanda, 2018). The critical issue is not the presence of a cluster, but whether the internal conditions are conducive to coordinated and cumulative environmental change.

The eco-innovation literature stresses that the capacity for innovation and environmental change are fundamentally interconnected. Environmental innovation does not transpire independently of an enhanced ability to generate, assimilate and use new knowledge (del Río González, 2009; Horbach et al., 2012; Hojnik and Ruzzier, 2016). Innovation-oriented firms and networks are generally better able to cope with environmental pressures, recognise opportunities stemming from sustainability issues and experiment with new products, processes and organisational configurations (De Marchi, 2012) . Thus, environmental performance must be integrated with the dynamics of innovation. This association may be particularly robust in cluster environments. An advanced innovation base typically offers expanded learning opportunities, enhanced knowledge exchanges and better circumstances for synergy among complementary competencies (Giuliani 2005; Giuliani and Bell, 2005). The environmental transition is precisely the type of process that depends on such capacities.

Technological innovation alone is insufficient, as it also necessitates organisational transformation, the establishment of new routines and the ability to reinterpret external challenges as strategic possibilities. From this perspective, creativity is not merely a characteristic among others but a fundamental condition that facilitates environmental enhancement. The literature clearly indicates that innovation alone cannot facilitate transformation. Due to the knowledge-intensive, uncertain and costly nature of environmental sustainability, collaboration frequently arises in eco-innovation research as a key explanatory element (Araújo and Franco, 2021). Collaboration allows firms and organisations to access external expertise, distribute risks, mitigate uncertainties and amalgamate resources that would be insufficient if used independently (De Marchi, 2012). Recent study indicates that the kind of collaboration is as significant as its mere presence. Various configurations of cooperative relationships provide distinct outcomes, and certain collaborative arrangements are more effective than others in fostering environmental innovation (De Marchi et al., 2022). Research linking active collaboration to a heightened eco-innovation orientation, along with studies emphasising the significance of networks involving enterprises, universities and public institutions, arrive at analogous conclusions (Diez-Martinez et al., 2023; Janahi et al., 2022).

This is especially pertinent for cluster research, as clusters are frequently defined solely by proximity that facilitates interaction, despite the contradiction that proximity does not necessarily lead to collaboration (Breschi and Lissoni, 2001). Access to local knowledge is determined by the quality of relationships and the roles that individuals have within networks (Giuliani, 2007; García-Villaverde et al., 2018). Certain players exhibit a markedly superior ability to assimilate knowledge and convert it into invention. At the cluster level, this indicates that we should not regard collaboration as a mere byproduct of clustered activity. Significantly, it is merely one of the methods via which clusters convert proximity into profitability. In areas with robust collaboration, collective responses to environmental concerns are more likely to arise and disseminate.

This is why the function of cluster organisations warrants particular consideration. Recent studies increasingly depict them as intermediaries that function beyond merely managing membership structures or reflecting local interests, as they evolve from broad-based democracies into tools of profit within a global capitalist economy. They can mediate interactions, support collaborative projects, link enterprises with external partners and aid in the conversion of generic sustainability objectives into concrete actions (Lis et al., 2020). Intermediary functions are crucial in sustainability transition studies, as they facilitate the alignment of diverse players and diminish the coordination costs associated with change (Kivimaa et al., 2019; Mignon and Kanda, 2018). This line of reasoning has been notably evident in the Polish setting. Lis and Mackiewicz (2023) demonstrate that clusters act as catalysts for green transformation, whereas Mackiewicz and Kuberska (2024) provide evidence that cluster organisations can motivate low-carbon and circular-economy initiatives, aid firms in surmounting transformation-related obstacles and generate spillover effects extending beyond their formal membership. Their findings suggest that greening in clusters is often systematically organised and institutionalised, rather than being a spontaneous effect of co-location. This interpretation is supported by new findings from Poland. Hegerty and Kowalski (2026) demonstrate that green transformation is significantly influenced by organised coordination and collective activity within clusters. Their research on firms in Poland, which highlights the twin transition, underscores the significance of cluster-based support systems and coordinated governance for sustainability-oriented development.

Some studies, however, advise against using scale as a standalone explanation. Meta-analytic research indicates that the benefits of cluster location exhibit significant heterogeneity (Grashof and Fornahl, 2021), with structural advantages being pertinent solely when arranged efficiently. Recent evidence from Poland corroborates this assertion. Gancarczyk et al. (2025), in their examination of energy clusters in Poland, illustrate that the differentiation between transformative and less transformative clusters is determined not by formal existence or scale, but by the governance arrangements that facilitate actor coordination. The findings imply that sustainable industrial transformation is closely associated with ecosystemic governance models. It suggests that cluster size may be a possible catalyst for growth, contingent upon its correlation with enhanced coordination capacity and superior intermediation and mobilisation.

Sustainability should be regarded as a communal and relational process, rather than solely an enterprise-focused endeavour (Porter, 2000; Kivimaa et al., 2019). Secondly, as they enhance the capacity of clusters to seek, assimilate and incorporate sustainability-related knowledge (De Marchi, 2012; Araújo and Franco, 2021; De Marchi et al., 2022), innovation and collaboration appear to be the two most influential explanatory variables. Thirdly, although size and structural circumstances are significant, the implications are more likely to differ based on organisational and governance capacity rather than scale alone (Grashof and Fornahl, 2021; Karlsen et al., 2023; Gancarczyk et al., 2025).

Despite this progress, the sustainability role of clusters and industrial districts remains conceptually fragmented and empirically underdeveloped (Hervás-Oliver et al., 2024). Comparative benchmarking data have only rarely been used to explain why some clusters adopt environmental sustainability practices more intensively than others. This creates a clear research gap. Although innovation, collaboration, intermediary support and governance are repeatedly identified as relevant, there is still limited explanatory use of benchmarking-style data for sustainability analysis at the cluster level. The present study addresses this gap by examining whether differences in innovation, collaboration and cluster size help explain the degree of environmental sustainability adoption among Polish clusters.

The data for this study were drawn from the cyclical benchmarking survey of Polish clusters, conducted in 2024 on behalf of the Polish Agency for Enterprise Development (2025) and published as Benchmarking of Clusters in Poland 2024 edition (PARP, 2025).

The target population comprised active Polish clusters meeting predefined activity and maturity criteria (e.g. minimum number of members, demonstrated ongoing operations). During the recruitment phase, approximately 70 potentially active clusters were identified. After screening for eligibility (exclusion of entities with insufficient activity, too few members), the final sample included 42 clusters. The 42 participating clusters were distributed across various regions and included both long-established entities (18 founded before 2010, 16 between 2010 and 2014) and more recent ones (eight founded between 2015 and 2020), reflecting different phases of cluster life cycles. Dominant sectors aligned with PKD classifications, emphasising high-growth areas such as information and communication technologies (ICT), manufacturing, renewable energy and biotechnology.

Data were collected through two structured questionnaires executed from January to June 2025:

1. A detailed questionnaire addressed to cluster managers, capturing structural characteristics, strategic documents, financing sources, cooperation with R&D units, internationalisation efforts, environmental actions, resilience mechanisms and good practices.

2. A complementary questionnaire directed to individual cluster members (N = 733 valid responses), focusing on perceived benefits, participation in joint activities (innovation, marketing, internationalisation, competence development), utilisation of pro-innovation and green transformation services, improvements in firm performance (e.g. innovation output, digitalisation, green practices, export) and evaluation of cluster activities.

Both combined closed-ended items (multiple-choice, Likert scales, numerical counts) with selected open-ended questions for qualitative depth. Normalised composite indicators were computed for key dimensions (resources, processes, results, cluster’s impact, internationalisation) to enable benchmarking across clusters.

Although the sample size of 42 clusters is relatively small, it constitutes more than half of the active clusters in Poland, enhancing the study’s relevance despite limitations in generalisability. The member-level data (N = 733) enrich the analysis by providing a multi-perspective view (strategic vs operational).

Dependent variable, the GTI, was constructed as the unweighted sum of six binary indicators (0/1) reported by cluster coordinators, capturing the following environmental actions:

1. circular economy;

2. environmental certification or eco-labelling;

3. energy efficiency measures;

4. environmental R&D;

5. renewable energy adoption (OZE); and

6. low-emission or carbon-neutral initiatives. The index ranges from 0 to 6.

Cluster-level predictors included:

• National Key Cluster (KKK) status (binary);

• cluster size (number of members);

• log-transformed annual budget;

• presence of a formal green transformation strategy (binary);

• presence of an ESG strategy (binary);

• innovation index (standardised sum of R&D projects, product innovations, process innovations and technology transfers);

• cooperation index (standardised sum of formal cooperation agreements and forms of R&D-sector collaboration); and

• cluster age (years since establishment).

Sector (based on dominant PKD section) was used as a control variable.

Member-level variables were aggregated to cluster level. Three variables were derived from member responses:

1. green service utilisation rate (proportion reporting “yes” to using cluster green services);

2. coordinator green assessment (mean score on a 1–3 scale, 1 = low, 3 = high; “don’t know” excluded); and

3. reported environmental improvement rate (proportion reporting actual improvement in their organisation’s environmental performance; “don’t know” excluded).

Given the small sample size and non-normal distribution of most variables, non-parametric methods (Spearman rank-order correlations and Mann–Whitney U tests) were used for bivariate analyses. For the multivariate analysis, negative binomial regression was applied because the GTI is a non-negative count variable with a bimodal distribution. Bootstrap standard errors (1,000 replications) were computed to ensure robustness.

Member-level data were analysed using Spearman correlations and linear mixed-effects models (members nested within clusters) to examine alignment between coordinator-reported GTI and members’ perceptions. Missing data were handled via listwise deletion for regression models and pairwise deletion for bivariate tests. Variance inflation factors were examined to assess multicollinearity.

The GTI was constructed as the unweighted sum of six binary indicators of environmental actions implemented by each cluster (range 0–6). The six actions included: adoption of circular economy principles, environmental certification or eco-labelling, energy efficiency measures, environmental R&D activities, renewable energy adoption (OZE) and low-emission or carbon-neutral initiatives.

Across the 42 surveyed clusters, the distribution of the GTI was bimodal. Fourteen clusters (33.3%) scored in the lower range (0–2), while 28 clusters (66.7%) scored 3 or higher. The mean GTI was 3.21 (SD = 2.31) and the median was 3.5.

Clusters holding National Key Cluster (KKK) status (n = 17) exhibited substantially higher levels of green transformation activity (M = 5.06, SD = 0.97) compared with non-KKK clusters (n = 25; M = 1.96, SD = 2.11). The difference was highly significant (Mann–Whitney U test, p < 0.001) with a large effect size (rank-biserial r ≈ 0.75). Similarly, clusters possessing a formal green transformation strategy or an ESG strategy showed markedly higher GTI scores.

Spearman rank-order correlations revealed consistent positive relationships between the GTI and several cluster characteristics (Table 1). The strongest associations were found for KKK status (ρ = 0.647, p < 0.001), log(budget) (ρ = 0.574, p < 0.001), the innovation index (ρ = 0.514, p < 0.001) and the cooperation index (ρ = 0.539, p < 0.001). Cluster size showed a moderate positive correlation, while cluster age exhibited a weaker positive trend.

Because the GTI is a non-negative integer count variable with a bimodal distribution, negative binomial regression models were estimated. To ensure robustness given the small sample size (N = 42, effective N ≈ 34–37 after listwise deletion of missing values on strategy and budget variables), bootstrap standard errors based on 1,000 replications were used. Predictors were z-standardised where appropriate. Variance inflation factors remained below 2.0 in the final specification, indicating no serious multicollinearity problems.

The results of the negative binomial regression (Table 2) indicate that National Key Cluster status is the strongest predictor: KKK clusters are expected to have 314% higher GTI scores than non-KKK clusters, holding other variables constant (IRR = 4.14, bootstrap p < 0.001). Possession of a formal green transformation strategy is associated with a 97% increase in the expected GTI (IRR = 1.97, p = 0.012), while an ESG strategy is linked to a 103% increase (IRR = 2.03, p = 0.008). Both the innovation index and cooperation index exert statistically significant positive effects, as does the log-transformed budget. After controlling for these factors, cluster size shows no independent association with green transformation activity.

Model diagnostics indicated acceptable fit (pseudo-R² ≈ 0.53). Sector dummies (based on dominant PKD section) were tested as additional controls but did not materially change the main findings and were therefore omitted from the final parsimonious model for clarity.

As robustness checks, Poisson GLM and OLS regressions with HC3 standard errors were also estimated. These models produced qualitatively similar patterns, although standard errors were larger and some coefficients fell to marginal significance, consistent with the known challenges of multicollinearity and small sample size in the original specifications.

To complement the coordinator perspective, responses from 733 individual cluster members were aggregated to the cluster level (n = 42). This allowed us to examine the degree of alignment between the coordinator-reported GTI and members’ perceptions and experiences of green transformation activities.

Three aggregated variables were analysed:

1. Green service utilisation rate – the proportion of members who reported using the cluster’s green transformation services.

2. Coordinator green assessment – the mean evaluation of the coordinator’s actions in the field of green transformation on a three-point scale with responses “don’t know” treated as missing and excluded from the cluster mean.

3. Reported improvement rate – the proportion of members who declared that their organisation’s environmental performance had improved as a result of cluster membership, excluding “don’t know” responses.

Table 3 presents Spearman rank-order correlations between the GTI and the three aggregated variables.

The results show that clusters with higher GTI scores are characterised by significantly greater utilisation of green services and receive more favourable evaluations from their members. These positive correlations remain significant in linear mixed-effects models that account for the nested structure of the data (members within clusters).

However, no statistically significant relationship was found between the GTI and the reported improvement rate. This discrepancy points to a notable implementation gap: although more environmentally active clusters successfully deliver and are recognised for providing green services, these efforts have not yet translated into measurable environmental improvements at the level of individual member organisations. This finding highlights the importance of moving beyond service provision towards more effective mechanisms of knowledge transfer, implementation support and outcome monitoring within clusters.

This study contributes to the growing debate on whether clusters and industrial districts act as effective vehicles for sustainability-oriented competitiveness. The findings suggest that green transformation in Polish clusters is determined less by structural characteristics alone and more closely related to institutional quality, strategic intentionality and the ability to orchestrate collective action. In this regard, the results support arguments that clusters should not be treated as automatically beneficial environments for sustainability innovation, but rather as organisationally mediated systems whose outcomes depend on governance, intermediation and absorptive capacity.

The first key finding is that a major determinant is National Key Cluster (KKK) status. Its impact persists even when controlling for variables related to budget, innovation, cooperation and strategy, suggesting that KKK status represents more than just formal recognition. It mirrors a wider institutional benefits, related to enhanced legitimacy, better coordination mechanisms, increased access to policy support and a more developed ability to mobilise members around common priorities. Thus, from a competitiveness perspective, it indicates that the institutional embeddedness and renowned organisational excellence can enhance a cluster’s ability to respond to sustainability pressures and convert them into strategic opportunities. It thus seems to indicate that green transformation is not only related to firm-level adjustment, but it is also about the quality of meso-level governance.

Both formal green transformation and ESG strategies have positive effects with consistent direction. These findings suggest that the green transition is more advanced where environmental priorities have been explicitly enshrined in cluster governance. This is, theoretically, significant because it marks out strategy not as some kind of symbolic statement but rather a means for unambiguously reducing uncertainty, coordination among heterogeneous actors and legitimising investment in new capabilities. In such cluster settings, where member organisations differ in size, interests and readiness for change, formal strategic orientation may be particularly valuable in ensuring that collective efforts are aligned. The findings, therefore, underpin the argument that sustainability-related competitiveness is born when cluster organisations interpret diffuse environmental pressures into structured agendas and concrete priorities.

The findings further indicate that innovation and cooperation are the main enabling conditions of green transformation. This is in line with the eco-innovation literature that emphasises that pollution prevention typically requires access to knowledge, absorptive capacity for complementary competences and experimentation under uncertainty. In the milieu of a cluster, these mechanisms are reinforced through networked interaction and intermediary support. More innovative and cooperative clusters are more capable to develop, transfer and aggregate practices towards sustainability. This strengthens the argument that competitiveness in contemporary clusters increasingly depends on dynamic capabilities rather than static locational advantages. Green transformation is thus not in contradiction with competitiveness but rather it is increasingly one of its expressions.

By contrast, cluster size loses significance once governance, strategy, innovation, cooperation and resources are taken into account. This is an important finding allowing to question simplistic assumptions suggesting that larger clusters are better placed in terms of achieving sustainability related goals. The bivariate relationship between size and GTI suggests that scale might matter descriptively, but the multivariate results indicate that its effect is indirect. It is true that larger clusters may have more resources or wider networks, but size alone does not result in transformation. This confirms the claim that proximity and density must lead to an activation of organisation. For research on competitiveness, it also means that the quality of coordination matters more than the number of actors.

The member-level evidence provides an important second layer of analysis. Higher GTI scores are associated with greater utilisation of green services and more favourable member assessments of cluster coordinators, which indicates that environmentally active clusters are recognised by their members and succeed in making relevant services visible and accessible. At the same time, the absence of a significant relationship between GTI and reported improvements in firms’ own environmental performance reveals a clear implementation gap. This suggests that service provision and positive evaluation do not automatically translate into operational change at the enterprise level. In other words, cluster organisations may be effective in creating an enabling environment for green transformation without yet securing measurable firm-level outcomes.

This gap is especially noticeable because it draws attention to a key tension in the sustainability–competitiveness nexus. Collective support structures can lead to better awareness and coordination of support services, and increase service uptake through wells cameras, yet these intermediate outputs may not translate into deeper-level shifts in firm routines, investment patterns or even performance. The explanation is that environmental evolution takes longer time frames than those covered in cross-sectional benchmarking data. Another is that member firms differ substantially in absorptive capacity, financial constraints and strategic readiness, limiting the conversion of cluster support into tangible results. The findings therefore suggest that competitiveness gains from green cluster initiatives may be contingent, uneven and mediated by firm-level implementation capabilities.

The study also produces some practical implications. The findings suggest that policymakers should position cluster support programmes towards governance quality and strategic formalisation, focusing on elements that facilitate innovation and cooperation before scale or numbers. The strong role of KKK status indicates that such selective, high-quality institutional support can make a difference. For cluster managers, the findings suggest that competitiveness-enhancing green transformation entails more than simply providing services, as it also demands follow-through mechanisms that assist firms with implementing solutions in practice. More task-specific modes of advice, piloting, peer-learning formats and the monitoring of outcomes on member-level may be required to bridge cluster initiative with firm level effect.

Lastly, the study adds to the literature by combining coordinators’ and members’ evidence in a matched sample, which allows for more fine-grained analysis of how environmental action on the part of clusters relates to firm perceptions and outcomes. However, the limitations should be recognised as well. The cross-sectional design precludes causal claims, the 42 clusters’ sample is modest and GTI captures breadth not depth of environmental activity. Nevertheless, the findings offer convergent evidence that green transformation in clusters depends mostly on institutional capacity, strategic orientation, innovation and collaboration. In this sense, the paper is in line with an emerging view of clusters as, potentially, intermediaries of sustainability-oriented competitiveness, but it also demonstrates that how collective action translates into firm-level transformation remains uneven.

This research indicates that the determinant of green transformation in Polish clusters is institutional quality, strategic commitment, innovation capacity, as well as cooperation intensity and resource availability. National key cluster (KKK) status was discovered to be the most significant predictor of environmental engagement, although methods for formalised green transformation as well as ESG strategies were also very important factors. The results are informative of the conclusion that sustainable-oriented transition in clusters does not constitutes a self-evident, automatic outcome from agglomeration but rather a co-ordinate mechanism formed by governance rules, intermediation and practical ability to activate collective potential.

The study further shows that environmentally more active clusters are treated as such by cluster members. Those with higher GTI scores reported greater frequency of green services used in the past year and had more positive evaluations of the coordinator’s work. Simultaneously, these cluster-level activities did not on average correlate with reported progress of member organisations on environmental best practice implementation. This indicates a longstanding implementation gap between sustainability-oriented support is offered and firm-level outcomes that are achieved.

The findings advance knowledge on cluster competitiveness by demonstrating that green transformation must be seen as part of the wider lattice of capabilities around which contemporary clusters are built. These results underlined the perception that competitiveness is even more a function of cluster organisations in their role as mediators, aligning disparate actors and translating sustainability demands into joint strategic responses. But they also show that the competitiveness impacts from such actions are partial at best, absent stronger mechanisms for firm-level absorption and implementation to accompany them at the cluster level.

From a practical point of view, the findings suggest that cluster and innovation policy need to focus more on quality of governance, strategic formalisation and support for co-operation- and innovation-related activities. For cluster managers, the findings highlight the need to move beyond service provision toward more implementation-oriented instruments, including tailored advisory support, pilot initiatives, peer learning and systematic monitoring of member-level outcomes.

The study provides rare matched evidence from both coordinators and member organisations in the context of a Central European economy. Later research would do well to extend these findings using longitudinal designs or cross-national comparative samples and more direct measures of firm-level environmental performance. This would not only clarify what drives green transformation in clusters but also under which conditions it translates into measurable gains towards organisational sustainability and long-term competitiveness.

This paper was enriched thanks to the Polish Agency for Enterprise Development – the owner of the database resulting from a survey with cluster managers and cluster members which was used for the analysis