Assessing Regional Innovation Strategies (RIS3) Through Competitive Public Project Networks: The Case of Aragón 2014–2020

Abstract

This study evaluates the effectiveness of Aragón’s Regional Innovation Strategy for Smart Specialization (RIS3) 2014–2020 by applying Social Network Analysis (SNA) to a comprehensive dataset of regional, national, and European competitive public projects involving Aragonese entities between 2014 and 2023. The analysis explores how collaborative structures—weighted by funding amounts—impact knowledge transfer, stakeholder participation, and strategic alignment within Aragón’s innovation ecosystem. Findings reveal a significant concentration of funding in European projects (86% of total ecosystem investment), contrasted with varying degrees of local and national involvement. Cohesion metrics show the high connectivity and closure of the European network, whereas regional calls, though exhibiting tighter density, remain comparatively fragmented and reliant on public research centres and universities. Centrality measures identify key bridging institutions, which facilitate advanced knowledge diffusion but also expose potential over-dependence on a small subset of actors. The analysis results highlight the value of integrating SNA metrics into RIS3 evaluations to better capture how resources, governance mechanisms, and institutional pressures interact. By offering a data-driven methodology that can be monitored continuously and policy recommendations, it aims to guide Aragón and other regions in refining their regional innovation strategies towards more inclusive, resilient, and globally competitive ecosystems.

1. Introduction

Over the past decade, the European Union’s Smart Specialization Strategies (S3 or RIS31) have aimed to transform regional economies into smart, sustainable, and inclusive systems by leveraging local strengths, as exemplified by Aragón’s 2014–2020 strategy targeting resource efficiency, competitiveness and wellbeing. These strategies seek to align research and innovation (R&I) with regional priorities to drive economic growth and social cohesion. However, traditional evaluations, including Aragón’s, rely heavily on metrics like R&D spending, patent filings, and collaborative project counts. The 2014–2020 RIS3 Aragón evaluation critiqued these as “abstract” due to their aggregation and lack of connection to specific actors (e.g., firms, universities), actions (e.g., project types), or tangible impacts (e.g., economic spillovers, knowledge diffusion), noting an “absence of correlation” with intended outcomes. This abstraction leaves policymakers and scholars with an incomplete picture of how collaborative structures underpin innovation ecosystems.

Existing research mirrors these limitations. Previous studies (Balland et al., 2019) assess RIS3 through output indicators (e.g., patents, employment), offering snapshots of innovation but neglecting relational dynamics. Similarly, other relevant works (Calvo-Gallardo et al., 2022b) applied Social Network Analysis (SNA) to Horizon 2020 projects, revealing centralized knowledge flows, yet focused solely on European-level initiatives, overlooking the multi-scale (regional, national, European) networks central to RIS3’s design. Moreover, prior SNA applications rarely weight ties by funding, missing a critical dimension of resource influence, and fail to link network properties to RIS3’s specific goals, such as fostering SME inclusion or interregional cooperation. These gaps—limited relational insight, single-scale focus, and disconnect between funding inputs and collaborative outcomes—hinder a comprehensive evaluation of RIS3’s effectiveness, leaving unanswered how well strategies translate into robust innovation ecosystems.

This study fills these gaps by employing SNA to analyze Aragón’s RIS3 (2014–2020) through a comprehensive dataset of 2464 competitive public projects (2014–2023) across regional, national, and European scales, totaling €5 billion in funding. Unlike traditional metrics, we move beyond abstract indicators by mapping multi-scale collaborative networks, with ties weighted by funding to reflect resource commitment. This approach captures both quantitative forces (e.g., investment distribution) and qualitative dynamics (e.g., knowledge transfer, stakeholder roles), offering a granular lens on Aragón’s innovation ecosystem. Specifically, it addresses three scientific contributions: (1) integrating regional, national, and European project networks, revealing multi-level collaboration patterns overlooked in prior work; (2) using funding-weighted SNA to link financial inputs to network outcomes, addressing the RIS3 evaluation’s call for better correlation; and (3) quantifying RIS3 effectiveness—defined as alignment with its action lines—through metrics like network centrality and cohesion, thus providing a replicable, data-driven framework beyond output-focused assessments. By doing so, this research not only refines our understanding of regional innovation systems but also delivers actionable insights for policymakers to enhance future strategies.

The structure of the paper is organized as follows. This introduction reviews the literature on RIS3 evaluation and examines the role of SNA in innovation policy, thereby establishing the theoretical foundation for our approach and outlining the research model and key questions. Section 2 describes the methodology, including details on the data and network metrics employed. Section 3 presents the findings from our analysis of participation, cohesion, and centrality, while Section 4 discusses the theoretical and policy implications. Finally, Section 5 offers the paper’s conclusions and outlines its limitations.

1.1. Regional Innovation Systems and Smart Specialization Strategies

Throughout this study, ‘innovation ecosystem’ denotes the network of interdependent actors—public, private, academic—that coordinate resources, knowledge, and expertise to produce innovations. Rather than focusing on isolated institutions, this ecosystem view underscores how collaborative relations underpin a region’s capacity for sustained technological and economic growth. Regional innovation systems are essential for fostering economic growth and technological advancement by promoting collaboration among universities, research institutions, businesses, governments, and civil society (Foray et al., 2021; Joint Research Centre (European Commission) et al., 2021). These systems leverage localized clusters of expertise to facilitate knowledge exchange and create synergies that align with broader socio-economic objective. By emphasizing place-based approaches, regions can capitalize on unique assets and capabilities, enabling more effective innovation processes.

In parallel, institutional impulses exert a powerful influence on how regional actors coordinate their innovation efforts. From an institutional theory perspective, shared norms and regulatory frameworks push organizations toward adopting common practices and processes to meet evolving expectations (Berrone et al., 2013; Gallego-Alvarez et al., 2017; Gao et al., 2019). In the context of regional innovation, these impulses are particularly visible in policies that incentivize collaboration or steer funding toward specific strategic sectors. For instance, Aragón’s reliance on European frameworks can be viewed as an outcome of regulatory pressures at the supranational level, shaping local priorities and prompting organizations to align their R&I activities with broader EU objectives. At the same time, normative and cognitive pressures, including stakeholder engagement and the promotion of shared environmental values, complement these regulations by fostering a culture of openness and inter-organizational trust (Ahmad et al., 2023; Arranz & Arroyabe, 2023).

In conjunction with institutional forces, social capital emerges as a central determinant of regional innovation success by embedding organizations within trust-based networks, accelerating the transfer of resources and expertise essential for sustaining advanced research and development (Rutten & Boekema, 2007; Zhang & Huang, 2022). Regions that develop strong internal linkages before attempting global-scale collaborations often achieve better outcomes, as localized trust, norms, and shared visions enhance both the speed and the depth of technology diffusion (Cappiello et al., 2020). Thus, following Rutten and Boekema, we view social capital as the ‘glue’ that fosters deeper collaboration among diverse entities.

The interplay between institutional structures and social capital thus underlies the resilience of innovation networks: while formal policies and funding set the strategic direction, high levels of social capital ensure that actual collaboration and knowledge-sharing take place on the ground. As a result, examining both formal institutional frameworks and the relational dimension of social capital offers a more nuanced view of how regional innovation ecosystems, like Aragón’s, can effectively harness their capacities for transformative growth.

Alongside these institutional and relational dimensions, evidence from European programs such as Horizon 2020 demonstrates that regional innovation systems also benefit from identifying strategic synergies and addressing critical gaps within their ecosystems. For instance, coal-transition research funded under Horizon 2020 highlighted the importance of aligning local capacities with transnational dynamics to create a transformative impact on regional economies (Calvo-Gallardo et al., 2022b).

In this context, Smart Specialization Strategies (S3), introduced by the European Union in 2010, mark a significant evolution of regional innovation systems by integrating research and innovation into regional development policies. (Foray et al., 2021; Joint Research Centre (European Commission) et al., 2021). These strategies, which focus on localization, prioritization, and participation, have reshaped regional agendas and proven instrumental in orienting resources toward sustainable and inclusive growth, while simultaneously addressing emerging challenges such as the green and digital transitions (CSIL et al., 2021, 2022).

The European Commission has reinforced the importance of robust governance mechanisms for S3 in the 2021–2027 programming period. New criteria emphasize up-to-date analyses of regional challenges, effective monitoring tools, enhanced stakeholder collaboration, and the fostering of international partnerships (CSIL et al., 2021; Joint Research Centre (European Commission) et al., 2021). This approach aligns S3 more closely with broader EU goals and leverages programs like Horizon 2020 and Horizon Europe to support cross-regional and cross-sector collaboration, accelerating innovation and knowledge transfer (Calvo-Gallardo et al., 2022b).

1.2. Evaluating Smart Specialization Strategies

Assessing the effectiveness of smart specialization strategies requires diverse methodologies that capture both qualitative and quantitative aspects of regional innovation. Traditional evaluation methods often relied on metrics like co-location and patent analysis to understand relatedness and complexity within regions. These methods focused on how regions could build upon existing capabilities to develop competitive advantages in targeted sectors (Balland et al., 2019).

However, such approaches have limitations, particularly in accounting for intangible assets like skills and knowledge bases. Recent critiques highlight that solely relying on tangible metrics may overlook the nuanced and diverse capabilities of regions (Buyukyazici, 2023; Nomaler & Verspagen, 2024).

Advancements in evaluation methods highlight the use of social network analysis (SNA) and relatedness-complexity frameworks to explore regional connections between industries, technologies, and human capital. For example, a recent study applied SNA to evaluate three Spanish regions’ RIS3 strategies (Calvo-Gallardo et al., 2022b). Moreover, the relatedness-complexity diagram helps map diversification opportunities and find high-value-added activities in regional contexts (Balland et al., 2019). However, this approach may overlook broader regional capabilities beyond measurable outputs.

To address these gaps, incorporating skill-relatedness and skill-complexity measures has been proposed. By using granular data on workplace skills, evaluations can better assess the alignment between regional capabilities and the sectors prioritized by S3 strategies. This approach helps determine whether regions are targeting sectors they are equipped to develop or pursuing overly ambitious goals disconnected from their existing capacities. For example, a study of Italian regions during the 2014–2020 period showed that leading regions selected sectors with strong skill-relatedness to their industrial bases, while lagging regions often chose more aspirational targets, leading to misalignment with their actual capabilities (D’Adda et al., 2020).

Despite these methodological advancements, challenges persist. Evaluations often reveal a disconnect between the ambitious objectives of S3 policies and their practical implementation. Regions may imitate the priorities of others instead of crafting strategies tailored to their unique strengths. Furthermore, many assessments focus predominantly on economic outputs, such as innovation performance or patent rates, without adequately considering social impacts or the quality of stakeholder engagement in the entrepreneurial discovery process. Additionally, studies that have employed social network analysis based on ecosystems created by competitive innovation projects often fail to include a comprehensive list of regional and national projects, limiting their scope and effectiveness (Calvo-Gallardo et al., 2022b; Radosevic, 2017).

These limitations highlight the need for a more comprehensive evaluative framework that incorporates social network analysis and the dynamics of key actor collaborations.

1.3. The Case of Aragon

As of 2023, the Autonomous Community of Aragón in Spain has a population of approximately 1,348,206 inhabitants, accounting for about 2.8% of the national population. Following data from the Spanish National Statistics Institute, the region’s Gross Domestic Product (GDP) is estimated at €41.76 billion, representing 3.1% of Spain’s total GDP. This results in a GDP per capita of €31,051, positioning Aragón among the higher-ranking regions in Spain in terms of economic output per person.

According to the same official source, Spain’s total R&D expenditure amounted to €19.33 billion in 2022, representing 1.44% of the national GDP. Although Aragón historically invested around 0.9% of its GDP in R&D, this figure rose to 1.07% in 2022, and—while not yet detailed in official records—it appears to be continuing upward in 2023 and 2024, thus narrowing the gap with the national average.

The Regional Innovation and Smart Specialization Strategy (RIS3) for Aragón (2014–2020) was developed as a regional framework to leverage Aragón’s unique strengths and position it as a smart, sustainable, and inclusive economy. Developed in alignment with EU and national policies, RIS3 Aragón focused on three main strategic priorities: resource efficiency and sustainability (energy and water), competitiveness (logistics and transport materials), and quality of life (agro-industry, tourism, and health). The strategy emphasized the integration of Key Enabling Technologies (KETs) such as nanotechnology and advanced materials to enhance regional innovation capabilities. Its implementation included specific actions across talent development, entrepreneurship, SME support, knowledge transfer, interregional cooperation, and internationalization, all designed to maximize the impact of public and private investment in research and innovation (Aragón Government, 2015).

The evaluation of RIS3 Aragón revealed a mixed picture of achievements and shortcomings. It underscored the importance of regional priorities but highlighted issues in implementation, monitoring, and coordination. The strategy defined six lines of action—talent and training, entrepreneurship, SME support, knowledge transfer, interregional and intersectoral cooperation, and internationalization—each supported by various tools, including grants, subsidies, and public procurement of innovation. Despite these structured efforts, the evaluation process faced challenges, such as insufficient leadership, limited integration of priorities, and a disconnection between strategic goals and tangible outcomes (Aragón Government, 2020).

Quantitative data showed moderate success in achieving specific goals. However, a decline in the number of innovative companies and stagnation in R&D investment as a percentage of GDP—falling short of the 1.30% target—highlighted ongoing structural issues in the regional innovation ecosystem. The lack of clear correlations between actions and outcomes further complicated the evaluation process, necessitating a reliance on qualitative assessments to derive actionable insights.

One of the most significant difficulties identified during the evaluation of RIS3 Aragón was the absence of robust governance and monitoring mechanisms. The strategy suffered from a lack of clear ownership and accountability among stakeholders, which undermined the ability to ensure continuous and systematic implementation. Additionally, the abstraction of indicators and their limited relevance to actual progress hindered effective tracking of the strategy’s success. This was compounded by insufficient alignment between the strategy’s lines of action, tools, and overarching goals.

The fragmented nature of stakeholder coordination posed another challenge. While entrepreneurial discovery processes aimed to involve a wide array of participants, their implementation was uneven, with limited engagement from key sectors. This lack of inclusivity restricted the ability to generate cohesive and impactful innovation policies (Laranja et al., 2022).

1.4. Research Model

The European Union’s emphasis on smart specialization has led regions like Aragón to develop and implement Regional Innovation Strategies for Smart Specialization (RIS3) to leverage their unique strengths and capabilities (Foray et al., 2021; Joint Research Centre (European Commission) et al., 2021). However, assessing the effectiveness of these strategies requires comprehensive methodologies that capture the complex dynamics of innovation ecosystems.

In this context, competitive public project networks emerge as vital components in understanding the structure and dynamics of regional innovation systems. By facilitating partnerships among universities, research institutions, businesses, and government agencies, these networks promote knowledge exchange, innovation, and economic development (Balland et al., 2019). Previous research has highlighted the importance of network analysis in revealing the interconnectedness of actors within innovation systems at regional and national levels, focusing on the role of particular stakeholders like high education institutions or remarking the importance of social capital and embeddedness in the regional innovation success (Rutten & Boekema, 2007; Tomasi et al., 2024).

This way, this study demonstrates that the EU Framework Programmes create supranational research networks that foster relational spillovers by encouraging collaboration among regions with varying levels of R&D intensity. This approach aligns with EU Cohesion Policy goals and provides insights into how collaborative networks can reduce regional disparities in innovation capacity, an objective relevant to Aragón’s RIS3 goals (Di Cagno et al., 2016).

Moreover, the difficulties encountered in evaluating RIS3 Aragón highlight the need for alternative evaluation frameworks that go beyond traditional metrics. A social network-based approach, as proposed in this paper, offers a more dynamic and interconnected perspective. By analysing the cohesion and centrality of networks formed by competitive public projects, such a method can provide deeper insights into the structure and functioning of regional innovation ecosystems. This approach facilitates the identification of key actors, including connectors, influencers, and gatekeepers, and their roles in driving knowledge transfer and innovation. Moreover, it enables a more nuanced understanding of how regional policies influence network dynamics and strategic alignment.

However, existing studies that apply social network analysis to evaluate the effectiveness of RIS3 strategies remain somewhat limited. For instance, while Calvo Gallardo et al. employed SNA to investigate RIS3 outcomes, they focused exclusively on H2020 initiatives. Expanding the analysis to include additional European programs—alongside national and regional calls—would yield a more complete view of the innovation ecosystem that emerged under RIS3. Moreover, the current literature has yet to incorporate entity-specific funding as a weighting mechanism for the ties between projects and participants, thus constraining the scope and depth of network-based evaluations.

Given these gaps, the present study applies these analysis techniques to assess the collaborative public project networks involving Aragón’s entities at the regional, national, and European levels from 2014 to 2023. By examining participation data, network cohesion properties and centrality measures, the research aims to provide a nuanced understanding of Aragón’s innovation ecosystem and evaluate the effectiveness of the RIS3 2014–2020 strategy.

The analysis of these networks offers insights into how entities interact, the strength of their collaborations, and the roles they play within the innovation system. The cohesion properties reveal the overall connectedness and resilience of the networks, indicating how well entities collaborate to achieve common goals. Centrality measures highlight key actors within the networks, such as connectors, influencers, or gatekeepers, whose positions can significantly impact knowledge flow and innovation diffusion. Finally, weighted ties between projects and entities provide a more transparent view of how financial resources are distributed and exploited (Chen et al., 2024; Kolleck, 2013; Li et al., 2024; Muller & Peres, 2018).

In this paper, the research model refers to our integrated framework for evaluating Aragón’s RIS3 strategy, which comprises three elements: (1) a multi-level dataset of competitive public projects (regional, national, European), capturing who collaborates with whom; (2) Social Network Analysis (SNA) metrics—such as cohesion, centrality, and weighted funding ties—that reveal the structure and intensity of collaborations; and (3) policy benchmarks derived from the RIS3 objectives. By systematically comparing SNA findings to the strategy’s stated goals, the model offers a granular, evidence-based approach to identifying strengths, gaps, and potential improvements in the regional innovation ecosystem. By exploring these aspects, the research addresses the following questions:

• RQ1: How do the features of competitive public project networks at European, national, and regional levels from 2014 to 2023 reveal the structure and dynamics of Aragón’s innovation ecosystem?

We define ‘competitive public project networks’ as consortia created through open, merit-based funding calls at the regional, national, and EU levels. This differs from nominative or direct funding, as project participants must formally apply and be evaluated on innovation merit, strategic alignment, and feasibility. By focusing on these vetted collaborations, we obtain a reliable snapshot of structured, policy-driven innovation activities in Aragón.

Understanding the features of these networks is foundational to comprehending the nature of Aragón’s innovation ecosystem. This question aims to uncover patterns of collaboration, the extent of connectivity among entities, and the interplay between different levels of projects. By analysing network structures, the study seeks to identify strengths and weaknesses within the ecosystem, such as highly connected clusters or isolated nodes, which can influence the region’s capacity for innovation and economic development.

• RQ2: In what ways do the cohesion and centrality properties of the networks shaped by these competitive public projects reflect the effectiveness of Aragón’s RIS3 2014–2020?

Our study examines specific network attributes—such as cohesion properties (network density and clustering coefficients) and centrality measures (including betweenness centrality)—to evaluate whether collaborative networks align with the RIS3 strategy’s objectives of fostering inclusive stakeholder engagement, cross-level knowledge flows, and sector-specific specialization. By mapping these social network analysis (SNA) metrics against the policy aims of the RIS3 2014–2020 agenda, we assess the resilience and robustness of the innovation ecosystem and identify key entities that facilitate collaboration and knowledge transfer, ultimately determining whether the observed network structures support long-term regional competitiveness and innovation.

• RQ3: What lessons can be learned from the evaluation of RIS3 2014–2020 that can inform future regional innovation strategies in Aragón?

Building on the findings from the previous questions, the third research question focuses on deriving actionable insights for future policy development. By identifying gaps, inefficiencies, and areas of strength within the current innovation ecosystem, the study aims to provide evidence-based recommendations for refining Aragón’s future strategies. This could include suggestions for enhancing governance structures, fostering collaboration among specific entities and targeting support to areas with high potential for innovation but low current connectivity.

Through this research model, the study contributes to a more granular and data-driven evaluation of regional innovation strategies. By applying social network analysis to a comprehensive dataset of competitive public projects, it offers a novel approach to assessing the effectiveness of RIS3 strategies. This methodology not only complements existing qualitative evaluations but also provides quantifiable metrics that can guide policymakers in enhancing the innovation ecosystem’s robustness, connectivity, and sustainability.

2. Materials and Methods

2.1. Data

To comprehensively evaluate Aragón’s Regional Innovation Strategy for Smart Specialization (RIS3) during the period 2014–2020, this study undertakes an extensive analysis of the region’s innovation ecosystem. Incorporating both the original RIS3 Aragón strategy document and its evaluation report from 2020 (available online in the following url: https://www.aragon.es/-/ris-3-y-s3-aragon-estrategia-de-investigacion-e-innovacion-para-una-especializacion-inteligente, accessed on 4 March 2025), this study examines all competitive public projects at the regional, national, and European levels that included at least one partner from the Aragón region over the last ten complete years (2014–2023).

This extensive dataset enables a thorough examination of participation and collaboration patterns within Aragón’s innovation networks. By analysing these projects and their associated entities, this study aims to reveal the structure and dynamics of the region’s innovation ecosystem, assess the effectiveness of the RIS3 strategy, and provide insights to inform future regional innovation policies.

2.1.1. Data Sources, Collection and Preparation

All the funding programs selected for this study comprise competitive public calls, funding opportunities where multiple independent entities submit proposals to obtain financing based on the merit of their projects and alignment with specified objectives. These calls differ from nominative funding, where projects are directly financed without a competitive selection process, often designated to specific entities or initiatives without the requirement for proposal submission and evaluation. With this established, the primary data sources for this study were:

• Regional Project Databases: Information on competitive public projects funded at the regional level was obtained from the Aragón government’s official repositories and databases, including those from the Aragón Development Institute (IAF for its initials in Spanish).
• National Project Databases: Data on national-level projects involving Aragón entities were collected from Spain’s Ministry of Science and Innovation and related national funding agencies, like the centre for industrial technology development (CDTI for its initials in Spanish).
• European Project Databases: European-level project data were sourced from the European Commission’s databases, including the Horizon 2020 and Horizon Europe frameworks. The CORDIS database (Community Research and Development Information Service) has been particularly instrumental, providing detailed information on projects, participants, funding, and outcomes.

Following the consultation of these databases, in Table A1, Table A2 and Table A3, that can be found in Appendix A, are shown the calls considered at the three mentioned geographical levels.

All data were accessed and collected between June and November 2024, ensuring the inclusion of the most recent project information up to the end of 2023.

For the purposes of this study, only projects that involved at least one participant from the Aragón region were selected. While the vast majority of the included projects were collaborative public initiatives—defined as those involving multiple independent entities collaborating toward common research and innovation objectives—a small proportion of competitive projects executed by a single entity were also considered. Examples of such projects include those funded under the SME Instrument of H2020, which were included as they reflect the performance of these entities in specific domains.

Regarding the entities’ selection criteria, those belonging to the Aragón region were identified by their tax identification number (NIF, by its initials in Spanish) and their associated addresses. Following this methodology, any entities from outside Aragón that participated in any of the selected projects were also included in the study, as they are directly related to the Aragonese R&I ecosystem through collaborative efforts.

This dataset provides a structured and replicable framework for assessing regional innovation dynamics. While it does not capture purely informal partnerships (e.g., local meetups or unstructured alliances), it ensures a high degree of reliability in mapping formalized, policy-driven networks. Future research could complement this approach with qualitative data or survey-based methods to explore additional informal collaborations, further enriching our understanding of the broader innovation ecosystem.

To process and prepare the data for analysis, several meticulous steps were undertaken to ensure accuracy and relevance to Aragón’s innovation ecosystem:

• Data Cleaning: After the initial extraction of data from regional, national, and European project databases, a rigorous cleaning process was implemented to eliminate redundancies, inconsistencies, and inaccuracies. This involved verifying the validity of entity identifiers and ensuring consistent formatting across different datasets. Entities from the Aragón region were confirmed based on their NIF and associated addresses. To enhance accuracy, the dataset was cross-referenced with official regional directories and national databases, such as records from the Spanish Ministry of Science and Innovation. For European projects, cross-referencing with the European Commission’s Participant Identification Codes (PICs) and the Unique Registration Facility (URF) database helped identify and rectify any discrepancies. This thorough validation ensured that both Aragón-based entities and external collaborators were accurately represented in the dataset.
• Network Construction: Based on the cleaned and validated dataset, networks were constructed to represent the collaborative relationships within Aragón’s innovation ecosystem. Entities were represented as nodes, and their collaborative ties—established through joint participation in projects—were depicted as edges. The networks encompassed projects at the regional, national, and European levels from 2014 to 2023, forming the foundation for subsequent social network analysis aimed at revealing the structure and dynamics of the ecosystem.
• Attribute Assignment: Relevant attributes were assigned to each node (entity) to facilitate a nuanced analysis of their roles and significance within the networks. Attributes included the type of entity (e.g., higher education establishments (HES), large private companies (PRC)), their role in projects (coordinator or participant), the level of project participation (regional, national, or European), and the funding received.
• Data Validation: To ensure the integrity and accuracy of the constructed networks, random samples were cross-checked against the original databases. Any discrepancies identified during this process were promptly addressed. This validation step was essential to confirm that the networks accurately represented the collaborative activities and relationships within Aragón’s innovation ecosystem over the specified period.

Through these steps, the data were prepared for further analysis, ensuring that the insights derived are precise and relevant in the context of assessing Aragón’s innovation ecosystem and the effectiveness of its RIS3 2014–2020 strategy.

Using data from regional, national, and European competitive public project databases is aligned with the first research question. With the focus on projects involving at least one participant from the Aragón region—spanning multiple funding levels and including mainly collaborative projects—the study aims to comprehensively understand how the features of these networks reveal the structure and dynamics of Aragón’s innovation ecosystem. Examining entities’ type participation across different levels and projects scale provides a holistic view of the ecosystem’s interconnectedness, collaboration patterns, and the roles of various actors, which are critical for evaluating the regional innovation strategy’s impact and guiding future policy development.

2.1.2. Entities Participations Attributes

In categorizing the participating entities based on their intrinsic characteristics and primary activities, it has been identified six distinct groups:

• Public Sector (PUB): This category includes national, regional, and local governmental authorities, as well as energy agencies.
• Higher Education Establishments (HES): Primarily composed of universities engaged in teaching and research activities.
• Research Organizations (REC): Encompasses two primary types of entities: publicly funded national research centres and research and technology organizations, which are predominantly private, non-profit institutions.
• Private Companies:

  ○

  Large Private Companies (PRC): Companies exceeding the thresholds for small and medium-sized enterprises.

  ○

  Micro, Small and Medium-sized Private Companies (PRC-SME): Enterprises classified as micro, small, or medium-sized according to the criteria specified in the EU Recommendation 2003/361 (European Commission, 2003).

• Others (OTH): Includes sector-level associations and may also comprise certain research institutes legally constituted as associations.

Regarding the roles of entities within projects, it is pertinent to note that each consortium is led by one entity acting as a coordinator, while the remaining consortium partners are considered participants, regardless of whether they are full partners signing the grant agreement or third parties with a legal and/or capital link to another beneficiary. This classification approach has been adopted because the presence of third parties is negligible and thus will not be separately considered in this study.

To identify entities participating in project networks at regional, national, or European levels, each participation was tagged with the corresponding scale level. Furthermore, recognizing that a partner’s contribution to a project is not uniform—varying significantly if, for example, an entity receives substantial funding compared to serving as a member of the project’s advisory board without monetary compensation—each participation was tagged with the amount of funding received from the funding authority, the budget allocated if the partner participates as a subcontractor, or an amount of one euro if no monetary compensation was provided. This methodology allows the links between entities in the different networks to be weighted based on the expected contribution from each entity to the project.

Awarded funding provides a reliable and policy-relevant basis for weighting ties, allowing for a clear assessment of how financial resources shape collaboration dynamics. While alternative metrics—such as patent co-authorship, scientific publications, or technology-readiness assessments—could offer additional perspectives, our funding-weighted method directly addresses a key policy concern: the relationship between financial investment and network connectivity. Nevertheless, future research could explore these complementary indicators to provide an even more nuanced understanding of innovation impact.

2.2. Methodology

Several studies have explored the use of SNA to assess innovation system performance (Abreu & Nunes, 2020; Calvo-Gallardo et al., 2021; Calvo-Gallardo et al., 2022b; Fernandez de Arroyabe et al., 2021; Morisson et al., 2020; Rodríguez Ochoa et al., 2023). While these studies have predominantly analysed projects funded under the European Framework Programme for Research and Innovation, they have largely overlooked the integration of competitive research and innovation projects spanning regional, national, and European levels. Moreover, existing research has not incorporated the funding received by each entity as a weighting mechanism to quantify the ties between projects and entities, thereby limiting the scope of network analysis in this context.

Empirical evidence from various fields demonstrates a positive correlation between the performance of innovation systems and the connectivity of related networks. This relationship emphasizes the efficiency of networks as mechanisms for achieving improved sustainable development outcomes and higher productivity rates as well as fostering breakthrough innovation (Di et al., 2024; Hu et al., 2024; Q. Liu et al., 2025). These metrics provide valuable insights into how collaborative structures align with policy objectives, offering a robust, data-driven understanding of the regional innovation ecosystem. While establishing a direct causal link between network features and innovation success is beyond the current study’s scope, our approach delivers critical evidence of how strategic partnerships and funding flows contribute to building a resilient and dynamic innovation network.

The structure and impact of collaboration networks within a particular field are pivotal to the success of research programs (Plex Sulá et al., 2024). Such research networks facilitate the exchange of information and sharing of experiences, which helps to prevent overlapping actions and fragmentation of activities—critical challenges in enhancing the European Union’s R&D performance (European Commission, 2010). Furthermore, evidence indicates a positive correlation between well-structured project teams and the impact of their innovation outcomes (Z. Liu et al., 2025).

In this context, social network analysis is essential for addressing the second research question, as it offers a framework to quantify network cohesion and centrality. By examining these metrics, a data-driven methodology can be developed to evaluate the effectiveness of Aragón’s RIS3 2014–2020 in a more rigorous, quantitative manner. The insights gained from applying SNA can then shed light on how the initiatives outlined in this RIS3 have influenced the regional innovation ecosystem.

For this analysis, it has been employed UCINET software (S. Borgatti et al., 2002) to assess the relationship between the project networks that have shaped the Aragonese research and innovation ecosystem over the past decade. UCINET is a highly regarded software platform for the analysis of social network data. Created by Borgatti, Everett, and Freeman, it is specifically designed to facilitate the examination of sociometric data, providing researchers with robust tools to investigate complex relational structures and the patterns that underpin them.

Furthermore, the insights derived from the SNA can inform the development of more effective innovation policies and programs, in line with this study’s third research question. Regional public authorities may incorporate this methodology into the evaluation of their regional innovation strategies, while the European Commission could adopt SNA-based indicators to assess the performance of regional smart specialization strategies. By focusing on network cohesion and the relationships among key actors at European, national, and regional levels, the analysis offers a foundation for guiding future efforts to strengthen the innovation ecosystem and enhance its overall performance.

In this study, the innovation ecosystem is shaped as a two-mode structure, in which entities are connected to projects and the link strength between an entity and a project corresponds to the amount of funding that an entity receives within that project. When projecting this two-mode network into a one-mode network of entities, the tie weight between two entities reflects the sum of their respective funding in the shared project(s), thereby emphasizing the collaborative strength of their participation. Although it is possible to also generate a one-mode network of projects linked by common entities, this paper focuses exclusively on the entity-based network, given its primary interest in examining how participant interactions—and their associated funding—shape the cohesion and dynamics of the regional innovation ecosystem. Figure 1 illustrates an example of these networks.

Figure 1. Illustrative example of a 2-mode network of entities and projects and its associated 1-mode network of entities and its corresponding weights.

In order to address the first and second research questions, it is first necessary to conceptualize the ecosystem as a network where each node corresponds to a distinct entity, and each edge denotes a collaborative relationship. This approach not only illustrates the underlying topology of the network but also highlights the factors that may influence the evaluation of regional smart specialization strategies.

Additionally, each node is assigned specific attributes—entity type (HES, REC, PRC, PRC-SME, OTH), geographical origin, and project role (coordinator or participant). This categorization, which directly relates to RQ2, enables an examination of how diverse entity characteristics shape network behavior, influence cohesion and centrality measures, and affect the broader implementation of RIS3 policies.

Two separate analyses are carried out for the resulting one-mode networks: a network-level assessment of cohesion metrics and a node-level computation of centrality metrics (S. P. Borgatti et al., 2023; Wasserman & Faust, 1994). Cohesion metrics shows the network’s degree of interconnectedness, which is critical to understanding the potential for knowledge exchange and leveraging specialized research capacities within the region’s innovation ecosystem. In this sense, the following cohesion metrics have been analysed:

• Average Degree: Calculated by determining the mean degree across all nodes. The degree of a node corresponds to the number of direct connections it possesses. This metric serves as an indicator of the overall activity within the network.
• Average Distance: Represents the mean distance among all pairs of reachable nodes. The distance between two connected nodes is defined as the length of the shortest path, measured by the number of edges. This value reflects the network’s compactness or dispersion.
• H-Index: Captures the highest number of nodes that each have at least an equivalent or greater number of connections to other nodes. It provides a measure of network cohesion, particularly by mitigating the influence of outliers.
• Density: Computed by dividing the sum of all connections by the maximum possible number of connections. In a weighted network such as this one, it is the sum of all tie values divided by the total possible ties, thereby offering a sense of how densely the network is interconnected.
• Connectedness: Evaluates the degree to which nodes are directly or indirectly interconnected, typically assessed as the average number of steps between any two nodes. Higher connectedness is often associated with more effective knowledge sharing and collaboration.
• Transitivity/Closure: Reflects the likelihood that nodes in a network form complete triads. A triad is considered complete if node A is connected to node B, node B is connected to node C, and node A is also connected to node C.
• Diameter: Identifies the greatest geodesic distance (i.e., the longest shortest path) among all connected nodes in the network, signifying the network’s overall spatial extent.
• Average Tie Strength Between Groups: Involves calculating the mean of weighted connections between nodes with different attributes, offering insight into how strongly distinct categories of nodes are linked.

Centrality measures provide deeper insight into the relative influence and connectivity of individual entities within the innovation network. They are vital for pinpointing principal actors and potential bottlenecks, thereby informing decisions related to resource distribution and stakeholder engagement. In this sense, the following centrality metrics were considered:

• Degree: In a weighted network, the degree of a node is determined by summing all tie values linked to that node. This metric indicates the immediate probability that a node will receive information or resources—particularly relevant for gauging how knowledge flows through the network.
• Closeness: Calculated as the mean of the shortest path lengths between a particular node and all other nodes. It measures how centrally situated a node is with respect to the entire network, reflecting its potential to quickly interact with multiple entities.
• Eigenvector: Serves as a measure of a node’s prominence within the network, assigning scores based on the prestige of the nodes to which it is connected. Nodes with high-scoring neighbours confer greater influence on the node itself.
• Betweenness: Quantifies how frequently a node lies on the shortest paths between pairs of other nodes. This value indicates the extent to which a node exercises control or brokerage over information flows across the network.

These metrics collectively offer a structured approach to evaluating both the global and individual dimensions of the network, enabling a richer understanding of how collaborative projects and key entities contribute to the regional innovation ecosystem.

3. Analysis and Results

Following the methodology presented, this section provides the results of the following three levels of investigation conducted:

• Participation Analysis: An initial assessment focused on the number of projects in which each entity participated, the extent to which they acted as coordinators, and the total funding they received. This step provides a foundational understanding of how various stakeholders engage with the regional, national, and European calls.
• Cohesion Properties: Next, the networks formed by regional, national, and European calls, as well as the aggregate network comprising all projects, were scrutinized for their cohesion metrics. This examination addresses the interconnectedness of the innovation ecosystem, highlighting how entities cluster, collaborate, and potentially influence the diffusion of knowledge and resources.
• Centrality Measures: Finally, the centrality properties of entities in each of the aforementioned networks were analysed to identify key actors, potential bottlenecks, and collaboration pathways. By discerning the roles of individual entities and the strength of their ties, the study elucidates how leadership, brokerage, and influence patterns emerge within the innovation networks.

Through these three levels of analysis, the subsequent sections offer a detailed perspective on the structure and performance of Aragón’s competitive public projects, facilitating a more nuanced evaluation of the RIS3 2014–2020 strategy and providing insights for future policy refinement.

3.1. Participation Analysis

In this section are shown the data gathered from the projects considered for the study during the period 2014–2023. Thus, Table 1 shows the main figures from the networks addressing different geographic scope levels:

Table 1. Data from the projects analysed during the period 2014–2023.

Calls Scope	Number of Projects	Number of Unique Participants	Number of Participations	Total Funding Received	% Over the Network’s Total Funding
Europe	781	6712	13,134	4,325,344,254€	85.95%
Spain	1126	1118	3163	669,228,934€	13.30%
Aragon	557	932	1828	38,047,347€	0.76%
Totals	2464	8233	18,125	5,032,620,536€	100%

In total, the analysed calls include 2464 projects performed by 8233 distinct entities, with a total number of participations (established as the participation of one entity in one project) of 18,125. These projects suppose a total funding of 5032 million euros. It is also relevant to mention that almost 86% of the ecosystem funding comes from European calls, while less than 1% comes from Aragón region calls.

In Table 2 is shown the funding achieved by Aragonese entities along the analysed period:

Table 2. Funding allocation in Aragonese entities per network.

Calls Scope	Total Funding Received	Aragonese Participants Funding	% of Aragonese Participants Funding over Total
Europe	4,325,344,254€	331,558,113€	7.67%
Spain	669,228,934€	398,675,398€	59.57%
Aragon	38,047,347€	35,409,934€	93.07%
Totals	5,032,620,536€	765,643,445€	15%

Here can be seen that more than 93% of the regional funds goes to Aragonese entities, and almost 60% of the funds from Spanish national projects with at least one partner from Aragón region, but less than 8% of the European projects with Aragonese presence. This may be happening as consortia in Spanish calls are normally reduced, thus not including several partners from different Spanish regions, but in Europe it is very unusual to find consortia with more than two or three entities from the same region, considering that the European Commission looks for the implementation and replication of the solutions elaborated in various European regions which are complementary in terms of location or socioeconomic indicators.

Then, Table 3 shows the funding allocation among the different Aragonese entities depending on its NUTS3 province location.

Table 3. Funding distribution among the three NUTS3 provinces in Aragón region.

Calls Scope	Aragonese Participants Funding	Zaragoza Province Entities % Funding	Huesca Province Entities % Funding	Teruel Province Entities % Funding
Europe	331,558,113€	91.01%	7.51%	1.48%
Spain	398,675,398€	80.45%	12.00%	7.56%
Aragon	35,409,934€	86.00%	8.73%	5.36%
Totals	765,643,445€	85.28%	9.90%	4.83%

When observing the total distribution, it is relevant to note that the funding distribution over ten years of projects is different to the GDP share of the three provinces, as both Teruel and Huesca count with 9% and 18% respectively, while Zaragoza accounts for 73%. This suggests that the concentration of innovation agents in the capital province of Aragón is creating a synergic effect in the attraction of funding to Zaragoza.

Finally, Table 4 shows the number of Aragonese entities that are coordinating initiatives at different scope level. There can be seen that when the ratio of coordination by Aragonese entities is similar to the funding achieved in projects coming from regional calls, the Aragonese partners needs to coordinate initiatives in a major degree to receive the same funding in projects coming from national and European calls.

Table 4. Aragonese entities coordination figures.

Calls Scope	Total Nnumber of Project Coordinators	Number of Aragonese Coordination Actions	Aragonese Coordination Actions Ratio
Europe	781	232	29.71%
Spain	1174	896	76.32%
Aragon	365	352	96.44%
Totals	2320	1480	63.79%

Regarding the entities type, Table 5, Table 6 and Table 7 shows the funding achieved, the number of participations in projects and the number of coordination activities carried out by each entity type.

Table 5. Funding distribution by entity type.

Calls Scope	HES	OTH	PRC	PRC-SME	PUB	REC
Global	20.13%	6.03%	21.29%	17.49%	6.62%	27.54%
Europe	22.50%	6.73%	17.74%	14.54%	7.71%	29.97%
Spain	5.95%	1.72%	44.76%	34.57%	0.01%	12.99%
Aragon	10.21%	5.16%	11.98%	52.99%	3.56%	16.10%

Table 6. Participations distribution by entity type.

Calls Scope	HES	OTH	PRC	PRC-SME	PUB	REC
Global	18.71%	12.87%	17.03%	23.59%	6.90%	20.55%
Europe	21.99%	11.28%	19.35%	17.56%	7.76%	21.80%
Spain	12.17%	13.88%	14.89%	38.54%	0.19%	20.30%
Aragon	7.66%	22.54%	3.99%	41.08%	12.69%	12.04%

Table 7. Coordination activities distribution by entity type.

Calls Scope	HES	OTH	PRC	PRC-SME	PUB	REC
Global	10.08%	14.95%	21.71%	33.43%	4.61%	15.21%
Europe	25.58%	7.29%	11.00%	13.17%	5.75%	37.21%
Spain	0.00%	21.21%	34.16%	41.99%	0.17%	2.47%
Aragon	9.32%	11.23%	4.66%	49.32%	16.44%	9.04%

In Table 5 is shown that research centres are receiving the main funding share both globally and when considering the European projects networks. Nevertheless, when observing the national projects network, large companies are receiving the main part of the funds while SMEs are getting more than half of the total funding in regional projects network. This makes sense as European calls count with a great focus on technology readiness levels (TRLs) from 4 to 8, matching the scope of work of a regular research and innovation agent, while Spanish and Aragonese calls are mainly focused on funding companies, and the regional Aragón Development Institute (IAF for its initials in Spanish) is devoted to fund micro to medium companies.

The data presented in Table 6 highlights that the participation of private companies (including both PRC and PRC-SME categories) in the European calls network accounts for approximately 37%. This is significantly lower than the average participation share of companies in European innovation programs, such as Horizon 2020, where the average stands at 50% (Calvo-Gallardo et al., 2022a). This discrepancy indicates that Aragonese companies are participating in European innovation programs at a rate approximately 35% lower than the average for European companies.

Table 7 further reveals notable differences in the entities taking on leadership roles within networks shaped by different project levels. In the European project networks, research centres emerge as the primary coordinators of initiatives. In contrast, SMEs predominantly lead both national and regional project networks, although the presence of large companies in the national projects network is almost as relevant as the one of SMEs. These distinctions underline the varying dynamics and capacities of entities across distinct levels of innovation networks.

Following the data by type of entity, Table 8, Table 9 and Table 10 shows the top 20 entities with higher numbers in funding achieved, participation in projects and coordinated initiatives respectively.

Table 8. Top 20 entities by funding achievement in the Aragonese innovation ecosystem 2014–2023.

#	Entity Name	Type	Country	Funding Achieved
1	Universidad de Zaragoza	HES	ES	84,209,468€
2	Fundacion CIRCE Centro de Investigacion de Recursos y Consumos Energeticos	REC	ES	68,347,537€
3	Fraunhofer Gesellschaft zur Forderung der Angewandten Forschung EV	REC	DE	49,206,276€
4	Instituto Tecnologico de Aragon	REC	ES	45,027,528€
5	Agencia Estatal Consejo Superior de Investigaciones Cientificas	REC	ES	39,749,567€
6	BSH Electrodomesticos Espana SA	PRC	ES	39,158,626€
7	Centre National de la Recherche Scientifique CNRS	REC	FR	38,493,145€
8	Commissariat a l Energie Atomique et aux Energies Alternatives	REC	FR	34,846,116€
9	Interuniversitair Micro-Electronica Centrum	REC	BE	34,109,976€
10	Ethniko Kentro Erevnas Kai Technologikis Anaptyxis	REC	EL	30,502,159€
11	Fundacion AITIIP	REC	ES	29,412,491€
12	Chalmers Tekniska Hogskola AB	HES	SE	29,079,143€
13	Fundacion Tecnalia Research & Innovation	REC	ES	29,029,615€
14	Forschungszentrum Julich Gmbh	REC	DE	27,753,082€
15	Airbus Operations SL	PRC	ES	23,242,762€
16	Consiglio Nazionale delle Ricerche	REC	IT	22,841,540€
17	Deutsches Zentrum fur Luft—Und Raumfahrt EV	REC	DE	22,270,031€
18	Stichting Wageningen Research	REC	NL	22,195,454€
19	Fondation Innovative Medicines for Tuberculosis (IM4TB)	OTH	CH	22,172,625€
20	Ministero della Salute	PUB	IT	21,797,028€
Top 20 Weight				14%

Table 9. Top 20 entities by number of projects in the Aragonese innovation ecosystem 2014–2023.

#	Entity Name	Type	Country	Number of Projects
1	Universidad de Zaragoza	HES	ES	484
2	Instituto Tecnologico de Aragon	REC	ES	304
3	Fundacion CIRCE Centro de Investigacion de Recursos y Consumos Energeticos	REC	ES	165
4	Agencia Estatal Consejo Superior de Investigaciones Cientificas	REC	ES	156
5	Centro de Investigacion y Tecnologia Agroalimentaria de Aragon	REC	ES	101
6	Fundacion AITIIP	REC	ES	98
7	Fraunhofer Gesellschaft zur Forderung der Angewandten Forschung EV	REC	DE	79
8	Fundacion para el Desarrollo de las Nuevas Tecnologias del Hidrogeno en Aragon	REC	ES	61
9	Ethniko Kentro Erevnas Kai Technologikis Anaptyxis	REC	EL	59
10	Fundacion Tecnalia Research & Innovation	REC	ES	57
11	Centre National De La Recherche Scientifique Cnrs	REC	FR	52
12	Consiglio Nazionale Delle Ricerche	REC	IT	47
13	Fundacion De Innovacion Y Transferencia Agroalimentaria De Aragon	REC	ES	44
14	Universitat Politecnica De Valencia	HES	ES	43
15	Commissariat a l Energie Atomique et Aux Energies Alternatives	REC	FR	42
16	Asociacion Cluster para el Uso Eficiente del Agua-ZINNAE	OTH	ES	41
17	Fertinagro Biotech SL	PRC	ES	40
18	RINA Consulting Spa	PRC	IT	40
19	Universidad Politecnica de Madrid	HES	ES	39
20	Asociacion Cluster de Automocion de Aragon	OTH	ES	36
Top 20 Weight				11%

Table 10. Top 20 entities by number of coordination actions in the Aragonese innovation ecosystem 2014–2023.

#	Entity Name	Type	Country	Coordination Actions
1	Universidad de Zaragoza	HES	ES	67
2	Fundacion CIRCE Centro de Investigacion de Recursos y Consumos Energeticos	REC	ES	37
3	Agencia Estatal Consejo Superior de Investigaciones Cientificas	REC	ES	34
4	Fundacion para el Desarrollo de las Nuevas Tecnologias del Hidrogeno en Aragon	REC	ES	33
5	Asociacion Cluster para el Uso Eficiente del Agua-ZINNAE	OTH	ES	32
6	Asociacion Cluster de Automocion de Aragon	OTH	ES	30
7	Cluster Español de Productores de Ganado Porcino	OTH	ES	30
8	Asociacion Espanola de Fabricantes Exportadores de Maquinaria para Construccion, Obras Publicas y Mineria	OTH	ES	28
9	BSH Electrodomesticos Espana SA	PRC	ES	28
10	Fertinagro Biotech SL	PRC	ES	24
11	Instituto Tecnologico de Aragon	REC	ES	23
12	Cluster de la Salud de Aragón (ARAHEALTH)	OTH	ES	22
13	Asociacion Logistica Innovadora de Aragon	OTH	ES	20
14	Fundacion AITIIP	REC	ES	19
15	Asociación Cluster de la Energía de Aragón	OTH	ES	18
16	Cluster de Empresas de Tecnologías de la Información, Electrónica y Telecomunicaciones de Aragón	OTH	ES	17
17	Cluster Maquinaria Agrícola de Aragón	OTH	ES	14
18	Aragon INNOVALIMEN	OTH	ES	12
19	Bit & Brain Technologies SL	PRC-SME	ES	12
20	Sociedad Anonima Industrias Celulosa Aragonesa	PRC	ES	12
Top 20 Weight				22%

Table 8 shows that the University of Zaragoza (UNIZAR) leads with €84.2 million in funding, showcasing its pivotal role as a major actor in securing resources for regional innovation. Research centres like Fundación CIRCE (CIRCE) and the Instituto Tecnológico de Aragón (ITA) also demonstrate substantial funding success, reflecting their capacity to attract significant investment for cutting-edge projects. However, the presence of entities from outside Spain, such as Fraunhofer Gesellschaft in Germany and Centre National de la Recherche Scientifique (CNRS) in France, highlights the integration of international institutions into the Aragonese ecosystem. Notably, the top 20 entities account for 14% of the total funding, indicating that a significant proportion of resources is concentrated among key actors.

When analysing project participation in Table 9, UNIZAR emerges again as the most involved entity, participating in 484 projects. This reinforces its position as a cornerstone of the regional innovation system. ITA and CIRCE also show strong engagement, participating in 304 and 165 projects, respectively. Interestingly, this table highlights the involvement of smaller clusters and specialized companies, such as the Cluster for Efficient Water Use (ZINNAE), which participated in 41 projects, and Fertinagro Biotech, with 40 projects. This diverse participation suggests a balanced contribution from academic, industrial, and research-oriented entities, fostering a multidimensional innovation ecosystem. The top 20 entities collectively represent 11% of total project participation, indicating broader engagement beyond these leading players.

Finally, the coordination actions shown in Table 10 provide critical insights into leadership dynamics within the network. UNIZAR leads again with 67 coordination actions, reflecting its dominant role in orchestrating initiatives. Similarly, CIRCE plays a significant leadership role with 37, although it is relevant that ITA occupies a modest 11st place event if as a public research entity should be promoting innovation initiatives at least at regional level. An interesting observation is the prominent role of other types of organizations (OTH) like regional clusters, such as ZINNAE, the Automotive cluster of Aragón (CAAR), and the Cluster for Health in Aragón (ARAHEALTH), which collectively coordinate numerous initiatives. These clusters’ involvement highlights the importance of sector-specific organizations in driving collaborative efforts and aligning projects with regional specialization areas.

The overlap of entities across the three tables demonstrates the importance of key players in multiple dimensions of the innovation ecosystem. For example, UNIZAR ranks first in all three categories, highlighting its multifaceted contributions to funding acquisition, project participation, and leadership. Similarly, CIRCE and ITA maintain high rankings, underscoring their strategic roles in fostering innovation.

However, the data also points to certain imbalances. For instance, private companies such as BSH Electrodomésticos España and Fertinagro Biotech show strong funding and project participation respectively, but relatively fewer coordination actions. This suggests that while private entities contribute significantly to resources and implementation, leadership roles in project coordination may still be predominantly held by public and research institutions.

Moreover, the analysis of Table A4, Table A5 and Table A6 from Appendix A, allows a more granular understanding of the varying dynamics and performance of entities within the European, Spanish national, and Aragonese regional networks. This additional analysis highlights not only the diversity of actors but also their differential engagement and leadership across calls scopes.

In terms of funding, European calls are led by CIRCE, which secures 61.8 million euro, reflecting its strong position in attracting international resources. UNIZAR and international entities such as Fraunhofer Gesellschaft and CNRS also emerge as key players, showcasing the robust integration of Aragón’s research centres into the European innovation landscape. On the national level, large industrial players like BSH Electrodomésticos España and Airbus Operations SL take the lead, emphasizing the centrality of private companies in securing Spanish funding. Meanwhile, in regional calls, Universidad de Zaragoza dominates with 3.4 million euro, accompanied by more localized actors such as ITA and Certest Biotec.

Project participation data further highlights the critical role of the UNIZAR across all levels, leading in European (153 projects), national (218 projects), and regional (113 projects) networks. CIRCE and ITA follow closely in European and national networks, while regional project participation sees increased representation from agricultural and cooperative organizations such as Centro de Investigación y Tecnología Agroalimentaria De Aragón (CITA) and Unión de Agricultores y Ganaderos De Aragón (UAGA).

In coordination actions, European projects are predominantly led by UNIZAR and CIRCE, with 36 and 34 initiatives, respectively. This leadership emphasizes their strategic importance in fostering cross-national collaborations. In Spanish networks, cluster organizations such as ZINNAE and CAAR emerge as central actors, reflecting the critical role of sectoral hubs in driving innovation. By contrast, the coordination pattern in regional projects is more decentralized, with UNIZAR and the Agencia Estatal Consejo Superior de Investigaciones Científicas (CSIC) holding the foremost positions, while smaller private entities, despite coordinating fewer initiatives, also contribute significantly to the ecosystem’s collaborative landscape.

The comparative analysis of the funding achievements, project participation, and coordination actions across European, national, and regional networks reveals notable patterns regarding the consistency of key entities within Aragón’s innovation ecosystem. Only five maintain a position in the top 20 across all levels of funding and within the overall network. These entities collectively underline the multifaceted contributions of Aragón’s research and innovation institutions, while also highlighting disparities in their performance across different funding scopes.

• UNIZAR stands out as the most consistent performer, ranking at the top in all three funding levels and receiving the highest overall funding globally and regionally. This highlights its central role as a cornerstone of Aragón’s innovation ecosystem, effectively leveraging its academic and research capacity to secure resources and lead initiatives across diverse networks.
• CIRCE exemplifies the specialization of certain entities, achieving the top position in funding from European calls while ranking significantly lower—13th and 19th in the national and regional networks, respectively. This disparity reflects its strong alignment with European priorities but highlights a need for greater integration into national and regional innovation efforts.
• ITA demonstrates significant versatility, securing second place in the regional network and fourth nationally. However, it ranks lower, at 12th, within the European network. Despite this, it remains a key player globally, reflecting its capacity to adapt and contribute across multiple innovation scales.
• CSIC holds a unique position as a national institution appearing in the top 20 across all networks. However, when considering only its regional branches, its performance is more limited, ranking 4th in the regional network. This indicates that its national-level dominance does not fully translate into local impact, suggesting potential areas for enhancing regional alignment.
• Similarly, Fundación AITIIP (AITIIP) achieves its best rankings in the regional (9th) and national (10th) networks yet derives more than two-thirds of its funding from European projects, where it ranks 17th. This reflects the duality of its role as both a regional innovator and an internationally engaged actor.

Beyond these five key entities, organizations such as CITA and Fundación para el Desarrollo de las Nuevas Tecnologías del Hidrógeno en Aragón (FHA) display notable contributions within specific networks. However, CITA lacks a presence at the national level, and FHA does not appear in the top 20 for either national or regional funding, suggesting a more specialized or localized focus.

A broader observation is that only UNIZAR and ITA maintain a significant presence across European, national, and regional networks. The remaining major participants exhibit limited roles in national and regional networks, drawing attention to an uneven distribution of influence and involvement. This fragmentation highlights the need for targeted strategies to foster greater integration and balanced participation across all levels of Aragón’s innovation ecosystem.

3.2. Cohesion Properties

The cohesion properties of the regional, national, European, and global networks reflect the differing scales and dynamics of these innovation ecosystems. The data in Table 11 reveals that the European network is significantly larger than the regional and national networks, with seven times the number of nodes compared to the regional network and more than nine times the number of ties. Although the ratio between node count and funding (see Table 1) remains consistent for both the European and national project networks, the total funding in the European network is disproportionately high—approximately one hundred times greater—than that of the regional network. This pronounced disparity reflects a higher concentration of resources and larger-scale initiatives characterizing European-level collaborations.

Table 11. Networks of entities cohesion properties.

Properties	Regional Network	National Network	European Network	Global Network
Number of nodes	932	1118	6712	8233
Number of ties	5232	7170	62,538	379,172
Average Degree	6	6	9	46
H-Index	22	28	184	246
Density	0.006030	0.005741	0.001388	0.005595
Connectedness	0.4786	0.7746	0.0103	0.9289
Closure	0.2017	0.1366	0.8763	0.4038
Diameter	9	7	9	10

From a density perspective, the regional network displays a slightly higher density (0.00603) compared to the national network (0.00574), indicating marginally tighter interconnectedness among entities. The European network, while larger in size, exhibits a lower density (0.00139), reflective of its expansive scale and the challenges of maintaining strong interconnections across such a broad scope. The global network, which integrates all three geographic levels, presents an intermediate density (0.0056), influenced by the sparse connections at the regional and national levels.

In terms of connectedness, the European network demonstrates exceptional integration (0.8763), highlighting its ability to facilitate high levels of interaction and collaboration among entities. Conversely, the national and regional networks are less connected (0.7746 and 0.4786, respectively), suggesting greater fragmentation at smaller scales. The global network achieves the highest connectedness (0.9289), underscoring the integrative role of European-level projects in bridging regional and national actors.

The regional network also shows a slightly higher closure (0.2017) compared to the national network (0.1366), indicative of better local clustering and the formation of cohesive subgroups. However, the European network stands out with a remarkably high closure (0.8763), reflecting its robust clustering and collaborative efficiency. The global network, with a closure of 0.4038, demonstrates an intermediate capacity for clustering, shaped by the interplay of diverse networks.

The analysis of tie strengths across the different networks further illustrates the dynamics of collaboration among several types of entities (PUB, HES, REC, PRC, PRC-SME, and OTH). In the global network (Table 12), the strongest average tie strengths are observed between public institutions (PUB) and universities (HES), as well as between universities and research centres. This pattern is consistent with the findings at the European level, as European projects constitute a sizeable portion of the global network.

Table 12. Average tie strengths between the different entity types in the global network.

	PUB	HES	REC	PRC	PRC-SME	OTH
PUB	1.283 × 1010	1.458 × 109	6.758 × 109	2.684 × 108	1.187 × 108	7.101 × 108
HES	1.458 × 109	2.577 × 1010	1.891 × 1010	2.642 × 109	1.200 × 109	1.229 × 109
REC	6.758 × 109	1.891 × 1010	1.621 × 1010	2.343 × 109	1.161 × 109	1.615 × 109
PRC	2.684 × 108	2.642 × 109	2.343 × 109	1.201 × 109	3.939 × 108	2.570 × 108
PRC-SME	1.187 × 108	1.200 × 109	1.161 × 109	3.939 × 108	2.027 × 108	1.629 × 108
OTH	7.101 × 108	1.229 × 109	1.615 × 109	2.570 × 108	1.629 × 108	2.438 × 108

At the European level (Table 13), universities (HES) and research centres (REC) dominate the collaborative landscape, with the strongest ties observed between these two categories. This dynamic is aligned the European Union’s focus on fostering high-impact research through robust academic-research partnerships. The substantial collaboration between universities and research centres reflects the EU’s emphasis on advancing fundamental and applied research as a foundation for innovation.

Table 13. Average tie strengths between the different entity types in the European projects network.

	PUB	HES	REC	PRC	PRC-SME	OTH
PUB	1.834 × 1010	1.787 × 109	8.430 × 109	3.545 × 108	2.212 × 108	1.033 × 109
HES	1.787 × 109	2.707 × 1010	2.021 × 1010	2.968 × 109	1.909 × 109	1.533 × 109
REC	8.430 × 109	2.021 × 1010	1.755 × 1010	2.668 × 109	1.869 × 109	2.050 × 109
PRC	3.545 × 108	2.968 × 109	2.668 × 109	1.430 × 109	6.591 × 108	3.455 × 108
PRC-SME	2.212 × 108	1.909 × 109	1.869 × 109	6.591 × 108	4.821 × 108	3.085 × 108
OTH	1.033 × 109	1.533 × 109	2.050 × 109	3.455 × 108	3.085 × 108	3.608 × 108

In the national network (Table 14), research centres (REC) emerge as the central actors, demonstrating the strongest connections both internally and with universities. This prominence highlights the role of research-driven initiatives in shaping the national innovation ecosystem. The ties between research centres and large companies (PRC) are also notable, pointing to a strong emphasis on scaling innovation through industrial applications.

Table 14. Average tie strengths between the different entity types in the Spanish national projects network.

	PUB	HES	REC	PRC	PRC-SME	OTH
PUB	0	4.902 × 103	6.734 × 104	2.174 × 103	4.831 × 105	6.089 × 104
HES	4.902 × 103	1.019 × 109	1.538 × 109	2.648 × 109	3.230 × 108	1.926 × 107
REC	6.734 × 104	1.538 × 109	8.725 × 109	2.359 × 109	4.579 × 108	3.159 × 107
PRC	2.174 × 103	2.648 × 109	2.359 × 109	2.034 × 109	3.630 × 108	4.283 × 106
PRC-SME	4.831 × 105	3.230 × 108	4.579 × 108	3.630 × 108	6.475 × 107	1.772 × 107
OTH	6.089 × 104	1.926 × 107	3.159 × 107	4.283 × 106	1.772 × 107	4.916 × 107

Finally, at the regional level (Table 15), the strongest ties are observed between universities (HES) and large companies (PRC), reflecting a focus on applied research and innovation with direct local impact. These ties emphasize the importance of academia-industry partnerships in driving regional development.

Table 15. Average tie strengths between the different entity types in the Aragonese regional projects network.

	PUB	HES	REC	PRC	PRC-SME	OTH
PUB	6.908 × 105	6.157 × 104	9.068 × 104	4.688 × 104	3.011 × 104	1.517 × 105
HES	6.157 × 104	2.130 × 107	1.930 × 107	1.208 × 108	5.268 × 107	1.065 × 105
REC	9.068 × 104	1.930 × 107	2.770 × 106	1.028 × 108	3.533 × 107	2.381 × 105
PRC	4.688 × 104	1.208 × 108	1.028 × 108	6.745 × 106	2.858 × 107	1.898 × 105
PRC-SME	3.011 × 104	5.268 × 107	3.533 × 107	2.858 × 107	3.864 × 106	5.957 × 105
OTH	1.517 × 105	1.065 × 105	2.381 × 105	1.898 × 105	5.957 × 105	9.303 × 105

The cohesion properties across the networks reveal important distinctions in the dynamics of collaboration at different geographic scales. While the regional and national networks show localized strengths, such as tighter clustering and impactful academia-industry partnerships, the European network demonstrates superior integration, connectivity, and closure. These characteristics of the European network not only enhance its capacity for large-scale innovation but also influence the structure and performance of the global network.

The observed variations in tie strengths across entity types reflect the differing roles and priorities at each level. Regional collaborations emphasize practical, localized innovation, as evidenced by the strong ties between universities and large companies. National networks, driven by research centres, focus on scaling and expanding innovation capabilities. At the European and global levels, the dominance of academic and research collaborations highlights the pursuit of cutting-edge research and global competitiveness.

These insights emphasize the importance of tailoring innovation strategies to the unique characteristics of each network. For Aragón, leveraging the strengths of regional and national networks while fostering deeper integration into European collaborations could enhance the region’s overall innovation ecosystem and ensure its competitiveness across all levels.

3.3. Centrality Properties

The centrality measures provide a detailed understanding of the roles and influence of entities within the networks at global, European, national, and regional levels. The data in Table 16 reveals significant variations in the average values of degree, eigenvector, betweenness, and closeness centrality across these scopes, reflecting the structural characteristics of each network and the distinct dynamics governing their interactions.

Table 16. Average centrality properties of the entities’ networks by call scope.

Calls Scope	Average Degree	Average Eigenvector	Average Betweenness	Average Closeness
Global	1.463 × 1013	1.020 × 10-3	6.913 × 103	2.791 × 104
Europe	1.786 × 1013	1.250 × 103	5.592 × 103	1.846 × 104
Spain	4.942 × 1011	1.037 × 103	8.768 × 102	4.633 × 103
Aragon	5.743 × 109	1.270 × 103	4.708 × 102	6.242 × 103

In the European network, the average eigenvector centrality is the highest, indicating that entities in European calls are more embedded in cohesive and influential clusters. This aligns with the European Union’s emphasis on fostering high-impact, collaborative research through tightly integrated partnerships. The average closeness centrality in the European network is higher than the one in national and regional networks, reflecting the larger size and complexity of the European innovation landscape.

National and regional networks, by comparison, exhibit significantly lower values for all centrality measures. The average degree and eigenvector centralities are particularly diminished, reflecting the smaller and more localized nature of these networks. However, the average closeness centrality in the regional network exceeds that of the national network, suggesting better connectivity and accessibility among entities at the regional level. This may reflect a higher degree of collaboration and integration among local actors, consistent with the regional network’s focus on addressing localized challenges.

At the global level, the inertia of the European projects network, too big in comparison with the national and regional ones, is highlighted, with all their centrality values closer to the European projects network ones.

The centrality measures presented in Table 17, Table 18, Table 19 and Table 20 reflect the roles and influence of top 20 entities specifically within the global network, encompassing interactions across regional, national, and European levels. These measures are distinct in that they do not segment performance by geographic scope, focusing instead on the overarching network structure. This distinction provides a comprehensive view of the interconnectedness and strategic importance of key actors.

Table 17. Top 20 entities with higher values of degree centrality.

#	Entity Name	Type	Country	Degree
1	Chalmers Tekniska Hogskola AB	HES	SE	2.86976 × 1015
2	Forschungszentrum Julich Gmbh	REC	DE	2.67705 × 1015
3	Centre National de la Recherche Scientifique CNRS	REC	FR	2.27249 × 1015
4	The University of Manchester	HES	UK	2.10091 × 1015
5	The Chancellor Masters and Scholars of the University of Cambridge	HES	UK	2.002 × 1015
6	Ministero della Salute	PUB	IT	1.75529 × 1015
7	EBRAINS	REC	BE	1.72445 × 1015
8	Commissariat a l Energie Atomique et aux Energies Alternatives	REC	FR	1.65635 × 1015
9	Consiglio Nazionale delle Ricerche	REC	IT	1.5603 × 1015
10	Fondation Innovative Medicines for Tuberculosis (IM4TB)	OTH	CH	1.53304 × 1015
11	Ecole Polytechnique Federale De Lausanne	HES	CH	1.50884 × 1015
12	Deutsches Zentrum fur Luft—und Raumfahrt EV	REC	DE	1.41643 × 1015
13	Interuniversitair Micro-Electronica Centrum	REC	BE	1.36352 × 1015
14	Agence Nationale de la Recherche	PUB	FR	1.23701 × 1015
15	Fondazione Istituto Italiano di Tecnologia	REC	IT	1.23516 × 1015
16	Technische Universitaet Dresden	HES	DE	1.01165 × 1015
17	Instituto de Salud Carlos III	REC	ES	9.98088 × 1014
18	Institut Pasteur	REC	FR	9.6498 × 1014
19	Ministero dell’universitá e della Ricerca	PUB	IT	9.36666 × 1014
20	Ministere de la Sante et de la Prevention	PUB	FR	9.07356 × 1014

Table 18. Top 20 entities with higher values of eigenvector centrality.

#	Entity Name	Type	Country	Eigenvector
1	Forschungszentrum Julich Gmbh	REC	DE	0.47871
2	EBRAINS	REC	BE	0.29343
3	Fondation Innovative Medicines for Tuberculosis (IM4TB)	OTH	CH	0.26458
4	Chalmers Tekniska Hogskola AB	HES	SE	0.26238
5	Centre National de la Recherche Scientifique CNRS	REC	FR	0.25930
6	The University of Manchester	HES	UK	0.21185
7	Ecole Polytechnique Federale de Lausanne	HES	CH	0.20693
8	Commissariat a l Energie Atomique et aux Energies Alternatives	REC	FR	0.20059
9	The Chancellor Masters and Scholars of the University of Cambridge	HES	UK	0.17946
10	Consiglio Nazionale delle Ricerche	REC	IT	0.17560
11	Institut Pasteur	REC	FR	0.14216
12	Universitetet I Oslo	HES	NO	0.13528
13	Servicio Madrileno de Salud	PUB	ES	0.13251
14	Eidgenoessische Technische Hochschule Zuerich	HES	CH	0.13068
15	Fondazione Istituto Italiano di Tecnologia	REC	IT	0.11256
16	Universite d’aix Marseille	HES	FR	0.11102
17	Interuniversitair Micro-Electronica Centrum	REC	BE	0.10657
18	Athina-Erevnitiko Kentro Kainotomias Stis Technologies Tis Pliroforias, Ton Epikoinonion Kai Tis Gnosis	REC	GR	0.09761
19	Ruprecht-Karls-Universitaet Heidelberg	HES	DE	0.09718
20	Technische Universitaet Muenchen	HES	DE	0.08980

Table 19. Top 20 entities with higher values of betweenness centrality.

#	Entity Name	Type	Country	Betweenness
1	Universidad de Zaragoza	HES	ES	5,640,607
2	Fundacion CIRCE Centro de Investigacion de Recursos y Consumos Energeticos	REC	ES	3,238,695
3	Instituto Tecnologico de Aragon	REC	ES	3,116,968
4	Fraunhofer Gesellschaft zur Forderung der Angewandten Forschung EV	REC	DE	1,823,370
5	Agencia Estatal Consejo Superior de Investigaciones Cientificas	REC	ES	1,638,812
6	Ethniko Kentro Erevnas Kai Technologikis Anaptyxis	REC	GR	1,264,860
7	Centro de Investigacion y Tecnologia Agroalimentaria de Aragon	REC	ES	1,061,048
8	Consiglio Nazionale delle Ricerche	REC	IT	1,006,249
9	Commissariat a l Energie Atomique et aux Energies Alternatives	REC	FR	743,577
10	Fundacion AITIIP	REC	ES	720,315
11	Fundacion Tecnalia Research & Innovation	REC	ES	717,912
12	Centre National de la Recherche Scientifique CNRS	REC	FR	661,888
13	Teknologian Tutkimuskeskus VTT OY	REC	FI	625,391
14	Alma Mater Studiorum—Universita di Bologna	HES	IT	602,574
15	Universidad Politecnica de Madrid	HES	ES	590,396
16	Danmarks Tekniske Universitet	HES	DK	449,857
17	Gobierno de Aragon	PUB	ES	449,018
18	Fundacion para el Desarrollo de las Nuevas Tecnologias del Hidrogeno en Aragon	REC	ES	443,341
19	Nederlandse Organisatie voor Toegepast Natuurwetenschappelijk Onderzoek TNO	REC	NL	412,515
20	Technische Universiteit Delft	HES	NL	321,293

Table 20. Top 20 entities with lower values of closeness centrality.

#	Entity Name	Type	Country	Closeness
1	Universidad de Zaragoza	HES	ES	17,548
2	Fundacion CIRCE Centro de Investigacion de Recursos y Consumos Energeticos	REC	ES	18,277
3	Fraunhofer Gesellschaft zur Forderung der Angewandten Forschung EV	REC	DE	18,338
4	Instituto Tecnologico de Aragon	REC	ES	18,405
5	Agencia Estatal Consejo Superior de Investigaciones Cientificas	REC	ES	18,538
6	Consiglio Nazionale delle Ricerche	REC	IT	18,796
7	Ethniko Kentro Erevnas Kai Technologikis Anaptyxis	REC	GR	18,939
8	Centre National de la Recherche Scientifique CNRS	REC	FR	18,992
9	Fundacion Tecnalia Research & Innovation	REC	ES	19,107
10	Commissariat a l Energie Atomique et aux Energies Alternatives	REC	FR	19,111
11	Teknologian Tutkimuskeskus VTT OY	REC	FI	19,133
12	Alma Mater Studiorum—Universita di Bologna	HES	IT	19,173
13	Danmarks Tekniske Universitet	HES	DK	19,187
14	Technische Universiteit Delft	HES	NL	19,453
15	Karlsruher Institut fuer Technologie	HES	DE	19,474
16	Universidad Politecnica de Madrid	HES	ES	19,491
17	RISE Research Institutes of Sweden AB	REC	SE	19,641
18	Eidgenoessische Technische Hochschule Zuerich	HES	CH	19,643
19	Technische Universitaet Dresden	HES	DE	19,674
20	Nederlandse Organisatie voor Toegepast Natuurwetenschappelijk Onderzoek TNO	REC	NL	19,680

In Table 17 is shown the top 20 entities with higher degree centrality. The top positions in degree are held by major European research centres and universities (e.g., Chalmers Tekniska Hogskola, Forschungszentrum Jülich or CNRS), which often participate in broad consortia. Their prominence in the global network largely stems from their engagement in multiple high-profile European projects, granting them widespread direct connections. Among Spanish entities, only the Instituto de Salud Carlos III appears in the top 20. In contrast, Aragonese entities do not appear in this ranking, as their partnerships tend to cluster within Spanish or region-specific calls rather than spanning the broader European landscape.

According to Table 18, Forschungszentrum Jülich, EBRAINS, and Fondation Innovative Medicines for Tuberculosis (IM4TB) occupy top spots in eigenvector centrality. High eigenvector values mean that these nodes not only have many direct connections, but those connections are also with other well-connected entities, amplifying their “influence” in the broader network. Aragonese entities with a high number of connections may still see lower eigenvector centrality if those connections are concentrated in smaller or less influential partners. In this sense, linking up with big transnational research centres can substantially raise an entity’s eigenvector score.

Table 19 reveals that UNIZAR leads with the highest betweenness centrality, highlighting its critical role as a bridge between otherwise disconnected parts of the network. Other Aragonese entities, such as CIRCE and ITA, also rank highly, demonstrating their intermediary roles in facilitating knowledge and resource flows. At the European level, entities like Fraunhofer Gesellschaft and CNRS hold significant intermediary positions, reinforcing their importance in linking diverse actors across the innovation ecosystem.

Finally, Table 20 shows that UNIZAR again ranks highly, reflecting its accessibility and integration within the regional and global networks. Other Aragonese institutions, such as CIRCE and ITA, maintain again strong positions, highlighting their ability to reach other entities efficiently. At the European level, entities like Fraunhofer Gesellschaft and Consiglio Nazionale delle Ricerche (CNR) demonstrate high closeness centrality, pointing up their strategic positions within densely connected clusters.

The centrality measures reveal the contrasting dynamics governing influence in the global network. Degree and eigenvector centrality metrics highlight the dominance of well-funded European institutions, particularly in sectors like health and well-being, which benefit from substantial European investments. In contrast, betweenness and closeness centrality measures emphasize the strategic importance of Spanish and, particularly, Aragonese entities as facilitators and hubs within the network.

The absence of Spanish entities in the top 20 for degree centrality stresses the need for stronger integration of national actors into high-budget, highly connected European projects. However, the prominence of Aragonese entities in betweenness and closeness centrality metrics demonstrates the region’s potential to function as a bridge between networks and to improve its global visibility.

These insights reinforce the importance of strategically enhancing Aragón’s participation in high-budget European initiatives, while continuing to leverage the bridging roles of its key actors. By fostering deeper collaborations and aligning with influential sub-networks, Aragonese entities can solidify its position within the global innovation ecosystem and maximize the impact of its Regional Innovation Strategies for Smart Specialisation.

4. Discussion

This paper set out to examine the structure and properties of Aragón’s innovation ecosystem, using SNA to evaluate the effectiveness of the region’s RIS3 2014–2020. The objectives were threefold: (i) to identify how the collaborative public project networks at different levels (regional, national, European) shape the structure and dynamics of the innovation ecosystem, (ii) to assess whether cohesion and centrality properties reflect the effectiveness of the RIS3 2014–2020, and (iii) to propose lessons that can inform future regional innovation strategies. This section discusses the results considering these objectives, connecting them with the known limitations and gaps of RIS3 Aragón 2014–2020, and offering insights into how SNA contributes to a more nuanced evaluation framework.

From a theoretical implications perspective, this study underscores the interplay between institutional impulses, social capital, and the efficacy of regional innovation systems. By drawing on institutional theory, the analysis shows how formal policies, regulations, and normative pressures guide collaborative behaviours and steer funding toward strategic priorities—embodied in Aragón’s RIS3 framework. At the same time, high levels of social capital ensure that organizations move beyond merely fulfilling policy mandates, actively embedding knowledge exchange and mutual trust into their everyday practices. This synergy—where institutional structures set the direction for R&I investment, and social capital facilitates deep, trust-based collaboration—forms the backbone of robust regional innovation systems.

In line with similar works (Calvo-Gallardo et al., 2022b; Fernandez de Arroyabe et al., 2021; Kang & Hwang, 2016; Rodríguez Ochoa et al., 2023), applying SNA metrics to Aragon’s networks confirms that centrality and cohesion measures can reveal whether an innovation ecosystem meets policy-based goals, such as those in the RIS3 Aragón 2014–2020. Indeed, our findings show how governance mechanisms and relational assets—particularly social capital—underpin resilience, inclusiveness, and transformative capacity in regional innovation networks. Taken together, this synthesis offers both a theoretical foundation and practical guidelines for refining policy and fostering more integrated approaches to regional development.

In this sense, when facing the first research question, the analysis of the different project networks revealed a multilayered innovation ecosystem in Aragón, where diverse actors operate at regional, national, and European scales, involving a total of 2464 projects, 8233 entities, and 18,125 participations. While the European network accounts for nearly 86% of overall funding, the regional and national networks also play distinct roles in promoting localized and mid-level collaborations, respectively.

The study’s findings indicate that a select group of Aragonese organizations such as UNIZAR and ITA plays a pivotal role in bridging the local, national, and European scales, thereby facilitating knowledge transfer and resource flows across diverse funding programs. This bridging capacity is critical for boosting Aragón’s adaptability and competitiveness, particularly when aligning regional objectives with larger-scale European initiatives. Notably, although the European network offers greater funding volumes and larger consortia, institutions such as UNIZAR, CIRCE, and ITA repeatedly engage in multiple projects, demonstrating their ability to channel international expertise back into the regional ecosystem. However, it is mostly publicly funded entities (e.g., UNIZAR, ITA) that maintain strong involvement in both European and regional calls; in contrast, private R&D agents like CIRCE and AITIIP encounter barriers that limit their participation at the regional level. Consequently, moving beyond this relatively narrow group of bridging institutions is essential if Aragón is to reinforce its industrial competitiveness and broaden the benefits stemming from large-scale collaborative efforts.

These observations suggest that, despite the presence of several robust and well-connected entities, the region’s innovation ecosystem remains fragmented. Many clusters and SMEs with specialized areas of expertise tend to focus on local and national calls, leading to underrepresentation in larger European initiatives. While this orientation provides certain benefits—such as leveraging localized knowledge and fostering niche collaborations—it also creates challenges by limiting integration with broader innovation flows and potentially hindering Aragón’s long-term economic and technological development.

One contributing factor may be that the RIS3 Aragón 2014–2020 and its evaluation placed relatively more emphasis on specific European funding mechanisms (e.g., EAFRD, ESF, ERDF, POCTEFA) while devoting less attention to the European Framework Programme (Horizon 2020 and Horizon Europe). This latter set of programs accounts for more than three-quarters of the total ecosystem funding over the past decade (78%), underscoring a mismatch between strategic focus and the actual distribution of resources.

Overlooking the roles of key actors and the significance of major European programs can impede the effective diffusion of knowledge acquired at the European level into the regional innovation network. In future RIS3 evaluations, social network analysis focused on cohesion and centrality metrics could mitigate this shortcoming by highlighting structural imbalances and illuminating the interplay among weaker and stronger participants. This perspective aligns with previous studies emphasizing that S3 evaluations must identify and address “structural holes” in networks, thereby supporting more balanced and inclusive regional ecosystems (Pinto et al., 2024).

Adopting SNA, also allows a more data-driven perspective on the robustness of Aragón’s innovation ecosystem, examining aspects like network density, connectedness, closure, and the presence of potential “weak spots.” Thus, the following paragraphs are dedicated to answer the second research question.

Network cohesion metrics (Table 11) underline that while the European network achieves superior connectivity and closure, the regional network exhibits a smaller scale but shows moderately higher density. This suggests that local collaborations can be more tightly knit, potentially reinforcing specialized activities. At the same time, the disproportionate gap between European and local funding place emphasis on one drawback of RIS3’s reliance on external resources: if local players do not strategically leverage those European consortia, they risk underperformance in global competitiveness. Encouraging more thorough integration of local actors with large-scale European opportunities would improve system resilience (Ma et al., 2024), which is a significant element missing from the RIS3 evaluation.

Additionally, the analysis reveals as well certain structural shortcomings, such as the limited representation of SMEs in European networks and the comparatively smaller roles of large private companies in local calls. These findings mirror the RIS3 Aragón evaluation, which lamented the fragmented participation of different sectors. By identifying sub-networks or isolated nodes (e.g., specialized agricultural cooperatives that rarely participate in calls beyond the regional level), policymakers can develop targeted support or incentives to enhance connectivity, thereby addressing the “lack of correlation” that has historically impeded strategic progress.

Another key advantage of SNA lies in its ability to identify central actors—whether they are connectors, influencers, or gatekeepers (Díez-Vial & Montoro-Sánchez, 2020; García Muñiz & Cuervo, 2018; Reis et al., 2023)—and examine their alignment with the strategic goals set forth in RIS3.

As the betweenness and closeness results demonstrate (Table 19 and Table 20), certain entities function as pivotal “brokers,” (Ehnert, 2023; Haas, 2014; Terhorst et al., 2024) notably UNIZAR, CIRCE and ITA. Their high betweenness points to an ability to link diverse project consortia, bridging distinct thematic or geographical areas. These entities appear well-positioned to fulfil strategic tasks in knowledge transfer, entrepreneurial discovery processes, and cross-sector collaboration—objectives that the 2014–2020 evaluation found to be partly unmet.

While the presence of these brokers is undoubtedly beneficial, the data also highlight a risk: over-dependence on a handful of institutions. Many coordination actions and bridging positions are concentrated among a select group of universities and research centres, echoing the “dependence on individual actors” critique from the RIS3 evaluation. Overreliance on these key players can perpetuate imbalances, leaving other segments of the innovation ecosystem behind. Addressing this gap through capacity-building programs and encouraging more private actors (especially SMEs) to lead projects could help distribute leadership more evenly and deepen the region’s collective innovation capacity (Almuzel et al., 2024; Roux et al., 2023). This has already been remarked in previous studies, where is stated that for SMEs, access to gatekeepers provides indirect exposure to international markets and expertise, helping them overcome resource limitations for direct internationalization. However, the benefits for SMEs depend on their ability to engage with central actors in the domestic network, as gatekeepers prioritize collaboration with firms that can complement their own objectives (Díez-Vial & Montoro-Sánchez, 2020).

Another major objective of RIS3 is to facilitate knowledge transfer to strengthen innovation performance. Drawing on the SNA findings, it becomes clear that participation data, which tracks each entity’s involvement in regional, national, or European calls, provides a direct indication of the channels through which knowledge can flow. This analysis makes it possible to gauge the extent to which Aragonese organizations—especially those identified as critical “brokers” through betweenness or closeness centrality—are effectively diffusing advanced expertise. Furthermore, the cohesion metrics (e.g., density, transitivity) reveal localized sub-networks that could either expedite or impede this diffusion. If a bridging institution, for instance, interacts closely with an otherwise insular cluster of SMEs, there is a higher likelihood that cutting-edge methods acquired from large-scale European consortia will filter down to smaller firms, elevating the region’s R&D intensity. Conversely, a lack of strong ties or minimal overlap between European-level and local project partners can lead to “knowledge leakage”, where innovations gained abroad do not take root in Aragón. This is aligned with previous works suggesting that the structure and cohesion of two-tier networks significantly impact the effectiveness of knowledge transfer and innovation performance (Su & Yan, 2023).

In addition, centrality metrics highlight how entities active in multiple project scopes (regional, national, European) can amplify the impact of external collaborations. When a node scores high on degree or eigenvector centrality within the European network but maintains similar prominence locally, it serves as a powerful conduit, systematically feeding back newly acquired technological insights, as suggested by previous analysis (Kochenkova et al., 2016; Vermeulen & Pyka, 2018). This interplay is especially significant for addressing the persistent challenge of underperformance in certain strategic domains, where local clusters may lack direct global connections. By identifying the specific nodes and tie strengths that bridge these gaps, policymakers can design targeted initiatives—such as specialized cluster calls or co-funding schemes—ensuring that frontier knowledge does not remain concentrated among a few top-tier organizations but instead circulates more evenly throughout the Aragonese innovation landscape.

Hence, SNA not only quantifies the presence of cross-level collaborations but also pinpoints where organizational roles and tie strengths either facilitate or hamper knowledge transfer. This is in line of previous research works indicating that centrality analysis can be applied to classify patterns in knowledge transfer process (Lee & Su, 2021). This granularity offers a robust, evidence-based platform for refining Aragón’s innovation strategies, better aligning with RIS3 objectives around capacity building, stakeholder coordination, and inclusive growth across all levels of the innovation network.

Another central criticism of the RIS3 Aragón evaluation was the difficulty in correlating financial investments with tangible outcomes, especially given the abstract indicators previously employed. This study’s use of weighted networks addresses that gap by associating tie strength with funding amounts, providing a more transparent view of how financial resources are distributed and exploited. By integrating funding data into the network analysis, it becomes possible to identify whether an entity’s considerable budget truly translates into broader connectivity, higher coordination roles, or increased knowledge dissemination, as analysed in previous papers (Kreye & Perunovic, 2020).

For instance, entities with high funding but low degree or eigenvector centrality may command substantial resources without creating robust links to other actors, reducing the collective benefits of such investment. Conversely, organizations with lower budgets yet intensive collaborative ties could demonstrate a higher impact on innovation diffusion, embodying the principle that effective interlinkages and coordination can amplify relatively modest resource inputs (Centeno, 2021; Wang et al., 2019). This alignment between funding data and centrality metrics thus moves beyond conventional expenditure-focused evaluations, enabling policymakers to discern where financial commitment genuinely elevates the region’s innovation capacity—and where it falls short.

Moreover, comparing weighted ties across the regional, national, and European levels clarifies how resource intensity differs by project scope. Some entities excel at securing large sums in national programs yet remain peripheral in European consortia, a mismatch that could hinder them from scaling to globally significant innovations. Conversely, others exhibit moderate funding inflows but maintain cross-level connectivity, suggesting a more strategic use of resources that facilitates knowledge spillovers. These conclusions are in line with previous research papers emphasizing the need to account for the intermediary role of entities (e.g., universities, research centres) in transforming public funding into tangible innovation outcomes (Stergiou et al., 2024).

These insights respond directly to the evaluation’s concern about abstract and insufficiently outcome-oriented measures. Thus, SNA-based funding weights permit a more evidence-driven examination of whether large investments meaningfully advance RIS3 objectives, such as stronger collaborative networks and long-term ecosystem resilience.

By applying SNA to Aragón’s multi-level competitive projects, this study tackles key gaps in the RIS3 2014–2020 assessment by correlating financial inputs with tangible collaboration metrics. Our findings highlight bridging institutions that promote knowledge transfer but also expose network imbalances and under-engaged sectors, mirroring the issues raised in the original evaluation. Weighting ties by funding addresses the difficulty of linking financial investments to concrete outcomes, showing that well-funded entities do not always occupy central network positions. Overall, these insights underscore the necessity of adopting a robust, network-oriented framework to better align policy objectives with real-world innovation dynamics. Additionally, by answering the first two research questions, this study offers evidence-based guidance for future regional innovation strategies in Aragón and beyond, thus fulfilling the third research question.

A key finding of this analysis is the potential for ongoing SNA to serve as a real-time evaluation method for RIS3, enabling policymakers to detect collaboration bottlenecks—such as network fragmentation or low local engagement—before they undermine strategic goals. Because the effects of subsidization can take up to two years to manifest (Mulier & Samarin, 2021), tracking shifts in cohesion and centrality metrics provides an early-warning system. This data-driven approach also complements or replaces the abstract indicators critiqued in the RIS3 Aragón 2014–2020 evaluation, creating a dynamic feedback loop between observed network structures and the refinement of policy interventions. It can strengthen accountability, encourage diversified leadership in both public and private sectors, and keep initiatives aligned with the actual rather than theoretical composition of the innovation network.

Beyond continuous monitoring, reinforcing governance and strategic coordination is crucial for addressing the lingering “lack of correlation” noted in the RIS3 evaluation. By systematically tracking regional, national, and European-level collaborations, policymakers can align strategic lines more effectively with project selection, ensuring that local clusters and companies join large-scale initiatives. Enhancing multi-sector engagement and knowledge-retention mechanisms at every call (Whetsell et al., 2020; Zarei et al., 2024) would diminish over-dependence on a handful of gatekeepers and broaden leadership roles—both public and private. In tandem, SNA-inspired metrics and comprehensive data integration can ensure that resource allocation reflects genuine performance instead of superficial criteria. As shown in previous research (Arranz & Arroyabe, 2023), any failure to complement European strategies with regional policies can undercut private-sector participation, underscoring the need for stronger policy alignment across multiple funding layers.

Taken together, these recommendations highlight a path forward for policymakers in any European region struggling with RIS3 evaluation: adopt continuous SNA monitoring, advance data-driven evaluations with weighted network analyses, and calibrate support measures to fortify both existing and emerging collaborators.

In Aragón’s context, strengthening technology transfer from major regional participants to SMEs stands out as a top priority, ensuring that advanced knowledge gained through European consortia benefits local businesses (Hervás-Oliver et al., 2021). Creating denser, more cohesive networks can be facilitated by calls that support larger, multi-actor consortia—spanning companies, SMEs, research centers, and clusters—thus addressing underrepresentation issues noted in the RIS3 2014–2020 evaluation (Caloffi et al., 2019; European Commission, 2023). Encouraging Aragonese partners to join EU projects through co-funding or recognition programs would further integrate smaller stakeholders into high-level initiatives (European Commission, 2023). Additionally, improving local coordination of European calls by offering capacity-building in consortium management can tackle longstanding governance critiques and increase the share of initiatives led by local actors (Querejeta et al., 2018).

Implementing such measures in tandem with SNA-based insights—such as identifying bridging institutions and weighting collaboration by funding—allows policymakers to reduce network inefficiencies, generate new synergies between local and international consortia, and embed advanced capabilities more firmly within the regional economy. Taken together, these targeted actions move beyond purely qualitative metrics in RIS3 evaluations by offering a quantifiable roadmap for inclusive and sustained growth. By focusing on knowledge transfer, diverse consortia formation, and integrated local-global participation, Aragón can better align its strategic objectives with concrete innovation outcomes.

5. Conclusions

This study assesses the structure and functioning of Aragón’s innovation ecosystem by applying Social Network Analysis to a comprehensive set of competitive public projects across regional, national, and European levels. In this sense, it aims to evaluate the extent to which these multilevel collaborations and funding channels have supported the objectives of the region’s RIS3 (2014–2020), complementing the evaluation made by the regional authorities. Through an in-depth examination of factors such as participation patterns, network cohesion, and the strategic position of key actors, the analysis provides a perspective of how effectively the innovation system has fostered knowledge exchange, mobilized resources, and catalysed collaboration among diverse stakeholders.

The results reveal a complex, multilayered innovation network spanning more than 2400 projects and 8200 entities, with a striking predominance of European funding (86% of total ecosystem funding) alongside varying degrees of regional and national participation. Network cohesion metrics underscored the larger size, higher closure, and stronger integration of European calls, while local/regional networks exhibited tighter density but lower overall connectedness. Centrality analyses showed key Aragonese entities—particularly the UNIZAR, ITA and CIRCE—acting as bridging institutions; however, they also highlighted uneven participation by SMEs and large companies, as well as potentially underutilized roles for certain actors in strategic domains. These insights confirm that, despite gains under RIS3, the region’s innovation ecosystem remains fragmented and reliant on a small set of gatekeepers or coordinators.

The paper findings point to several theoretical contributions. First, by examining networks at multiple levels—regional, national, and European—and incorporating funding amounts to weight the ties, this study goes beyond traditional proxies like patenting rates or co-location analysis. In doing so, it emphasizes the synergy between institutional impulses—such as formal regulations or funding programs—and social capital, characterized by trust and embeddedness, in shaping the broader innovation ecosystem. This dual lens is particularly relevant for understanding how regions like Aragón balance supranational directives, like EU-funded initiatives, with local development priorities. Consequently, the research adds empirical depth to the literature on regional innovation systems by showing how network dynamics—and not just their aggregate outputs—can reveal hidden inefficiencies or untapped collaborative potential in the wake of an RIS3 strategy.

Second, the emphasis on cohesion, centrality and tie strength enables a clearer picture of how resources circulate and accumulate within the ecosystem. Network cohesion measures—such as density, connectedness, and closure—demonstrate the degree to which stakeholders unite around shared goals; when paired with tie weighting, these same measures illuminate whether high-funded consortia also achieve high levels of collaboration or continue to operate in relative isolation. This methodological refinement offers a more granular tool for detecting structural holes or imbalances, particularly in how SMEs and clusters integrate into R&D consortia. It thereby refines existing notions of social capital, highlighting that trust- and norm-based relationships must be sustained by sufficiently strong financial and project-based linkages to produce genuine innovation spillovers.

In terms of practical contributions, the study outcomes carry immediate relevance for regions striving to refine their innovation agendas. One pragmatic insight lies in the identification of key bridging institutions, which demonstrate high betweenness and closeness centralities while also attracting sizable funds. These institutions can function as conduits for advanced knowledge diffusion, ensuring that European-level research outcomes permeate local industries and SMEs. Furthermore, the weighting of ties exposes situations where larger budgets do not necessarily translate into proportionate benefits for the innovation system, which warns policymakers against allocating resources solely based on an entity’s nominal funding success. By recognizing discrepancies between funding and network influence, authorities can implement more equitable and results-oriented frameworks, refining calls to better incorporate diversified consortia.

Additionally, the paper’s findings urge policymakers to integrate network-based metrics more systematically into RIS3 evaluations. Regions can use such metrics not only to spot gaps in stakeholder participation or structural overdependence on a handful of coordinating bodies but also to target capacity-building measures. For instance, private companies or SMEs that demonstrate a high degree of openness in smaller, local calls may also be encouraged, via tailor-made incentives, to coordinate or join consortia in larger, European-funded projects. This encourages deeper alignment between multilevel policy objectives and fosters an environment where social capital and institutional frameworks mutually reinforce one another. In effect, by applying centrality and cohesion data in ongoing monitoring cycles, governments can recalibrate funding calls, governance processes, and stakeholder engagement, thereby aligning R&D investments more closely with long-term regional innovation goals

Lastly, the research proposes several targeted policy recommendations—such as creating denser local consortia, incentivizing the inclusion of regional partners in European calls, and reinforcing the coordination capacities of Aragonese entities—that can mitigate existing imbalances and enhance strategic coherence in subsequent regional strategies.

As final thought, it is also relevant to mention the paper limitations. One constraint stem from the dataset itself, as it covers only competitive public projects between 2014 and 2023, thus excluding private investment in research and innovation. This focus on the evaluation of the 2014–2020 RIS3 strategy also means that new directives and performance indicators introduced under the S3 Aragón 2021–2027 program are not considered. Consequently, the full contemporary context of Aragón’s innovation evolution may not be fully captured.

Moreover, although SNA provides a powerful tool to map and interpret network structures, the analysis omits the networks specifically shaped by the specialization areas (e.g., agro-food, sustainable mobility) defined in the RIS3. Furthermore, the absence of direct consideration of so-called “European champions” in these specialization areas limits understanding of how stronger transnational ties could reinforce Aragón’s capacity in certain key sectors. Future research might integrate these thematic networks and cross-reference them with the presence of top-performing European partners to produce more customized policy interventions.

In conclusion, the SNA-based evaluation complements existing qualitative and quantitative approaches, clarifying how funding, governance structures, and social capital interact within Aragón’s innovation ecosystem. By offering a precise and data-driven lens, this study provides policymakers with actionable insights into both the achievements and shortcomings of RIS3 2014–2020, laying a stronger foundation for crafting inclusive, resilient, and globally competitive innovation strategies in the years to come.