Three modes of horizontal networking
Excerpt from forthcoming Working Paper of Slovenian Evaluation Society, XVIII/1(2025) (Bojan Radej, Barbara Simić, Jasna Šinigoj, EVALUATION OF NETWORKED RESEARCH, The Case of ERA-NET’s CoFund Action ‘GeoERA’)
Horizontal networking arises in ERA-NET through three networking modes: cross-country cooperation, cross-cutting collaboration, and cross-thematic integration (Naujokaitytė et al., 2023; Gøtke et al., 2016), each contributing distinct yet interdependent values to the overall network effectiveness.[1]
Cross-country networking in this paper refers to cooperation between national or regional level public authorities or agencies, acting as central bodies in their domain, often with delegated responsibility for allocating national resources, shaping national research agendas, or representing priorities in transnational initiatives. Cross-country networking provides ERA-NET Actions with political relevance and institutional validity. Its evaluation is carried out through an administrative lens. Administrative indicators measure the mechanics and logistics of cooperation, not its substance: procedural and upward accountability, whether the project is in line with predefined legal frameworks, and project milestones. Indicators of success are largely formal and output oriented, such as the number of joint calls organised, the number of co-funded projects, or the share of states represented.
Cross-cutting networking refers to cooperation that embeds normative or functional values — such as gender equality, ethics, or standardisation (denoted as ‘hub’ in Figure 1) serving as universal common denominators across independent thematic fields (OECD, 2014). This mode of networking operationalises horizontal objectives through vertical implementation: cross-cutting issues apply uniformly across all network members. It creates consistency by boundary-setting mechanisms defined at the macro level. Its logic of change is often linear: normative standards or strategic directives trigger mainstreaming processes that, in turn, enhance design quality and generate cumulative system-wide gains (OECD, 2014). Evaluation of cross-cutting networks is essentially a managerial undertaking, often checkbox-driven, focused on adherence to central guidance. It uses managerial, performance-oriented outcome metrics, such as participation rates across nations, documented transnational coordination processes, and quantifiable pan-European achievements.
Figure 1: Three modes of networking, a hypothetical network of 8 members
The concept of cross-thematic networking in research adopts different logic of cooperation and its evaluation. It pursues horizontal objectives through horizontal means, making it the most authentic mode of networking. Cross-thematic networking creates collaborative structures that integrate interdependent thematic domains to co-produce solutions that are unattainable within siloed approaches. Unlike hierarchical approaches, its rationale emerges from the members’ pursuit of synergies. Its theory of change is transformative: networks of relations evolve into patterns that correlate at meta-level. This is integrative problem space that enables holistic knowledge flows beyond established national or disciplinary boundaries without imposing hierarchy or enforcing relativism.
The effects of cross-thematic networking cannot be captured by standard measures. Whereas cross-country and cross-cutting networking lend themselves to conventional assessment through tangible, uniformly observable effects, cross-thematic networking requires more indirect approach. Its effects emerge indirectly. Evaluation must capture the strength of inter-thematic connectivity through synergies arising from boundary-crossing exchanges. Such evaluation must be comparative, process-sensitive, and capable of capturing non-linear pathways of integration (Radej, 2021a).
Cross-thematic networking must be evaluated at the meso-level. This view has already been validated in geosciences by Tinker (2015) through his operational concept of the ‘radical middle’. This framework highlights the crucial role of analysing overlaps and interactions between distinct thematic domains.
((The original study was recognised as a success story and best practice in the European Commission publication “Success Stories from the Research Management Community, A catalogue of best practices and achievements” by the European Research Area Platform for demonstrating how evaluation can strengthen the programming, governance of public-public research networks across Europe.))
[1] In the literature, various classifications of networking can be found. Heller-Schuh et al. (2011) distinguish between small-world networks, distributed cluster networks, and networked communities — each arising from distinct types of sub-programmes. While this typology provides insights into variability of forms in network formation, it remains limited in scope since it cannot account for trade-offs and compatibility among the three networking forms.
