Collaborations between universities and cultural institutions: achieving transformative research impact 1

Summary

This blog post has been written following an NCACE Evidence Café event held in November 2025 on the theme of this blog’s title. It uses data from REF 2021 Impact Case Studies to identify formal collaborations between universities and cultural institutions, discussing the dynamics of how knowledge is produced and what drives transformative collaborations.

It identifies two key modes of knowledge production – traditional knowledge production and context driven knowledge production - and positions cultural institutions as active partners, not just recipients, in knowledge creation. It also sets out some brief implications for universities, cultural institutions and policy makers.

Importance of collaborations between universities and cultural institutions

Cultural institutions — such as museums, libraries, archives, collections, theatres, heritage sites — are vital hubs of knowledge preservation, creation and diffusion. They preserve and share culture and heritage while engaging diverse audiences. Their public mission closely aligns with universities’ educational and civic roles, providing rich opportunities for collaboration (Comunian & Faggian, 2014)[1]. By combining expertise and resources, universities and cultural institutions can turn academic research into benefits for society (Wilson et al., 2021[2]; Rossi et al., 2025[3]). This aligns with the ‘research impact’ agenda, according to which universities are increasingly expected to show that their work produces value beyond academia – generating impacts that are economic, cultural, social, environmental, political, and so on.

However, most of the evidence collected about research impact still focuses on collaborations around science and technology, where outcomes are easier to measure (Hessels & Van Lente, 2008)[4]. We still do not have a large amount of evidence about how universities and cultural institutions work together to produce and share knowledge, and what kinds of impact they achieve.

Evidence of collaborations between universities and cultural institutions

Using data from the REF 2021 Impact Case Studies database[5], our work aims to fill this gap by investigating collaborations between universities and cultural institutions, particularly trying to identify which collaborations have the most transformative impact.

The study draws on 6,361 REF 2021 impact case studies, focusing on 3,018 that involved formal partners. Among these, 286 cases included a cultural institution. Each case provided information about the type of institution, its partners, research area, and the kind of impact achieved—whether cultural, economic, environmental, social, political, legal or technological.

Figure 1. Cultural institutions involved as formal partners in REF 2021 impact case studies

Collaborations involving cultural institutions were typically smaller, more diverse, and more likely to have received external funding (but from a lower number of external funders) than collaborations that did not involve cultural institutions. The former focused less on traditional academic publication outcomes and more on creative knowledge, such as performances, exhibitions, or digital archives.

Table 1. Comparison between formal collaborations with and without cultural institutions

Different types of collaborations between universities and cultural institutions

In order to identify patterns and distinguish between different types of collaborations, we build on ideas proposed by Gibbons et al. (1994)[6], to identify two main ways knowledge is produced:

• Mode 1 – Traditional academic research within universities. This tends to focuses on disciplinary expertise, peer review, and theoretical work, but usually involves limited interaction with outside partners. This approach strengthens specialist knowledge but can be detached from real-world needs.
• Mode 2 – Collaborative, problem-focused research. This tends to bring together universities, industry, government, and community partners to co-create knowledge that addresses practical issues. It values diverse perspectives and produces more socially relevant results.

Table 2. Comparison between Mode 1 and Mode 2 knowledge production

Knowledge dynamics and relationships

Mode 1 knowledge production processes are more likely to result in cumulative knowledge dynamics, that is, in the production of new knowledge that is close to and building upon the already similar knowledge bases of the participants (who are mainly academics, with external stakeholders playing the role of knowledge recipients). Therefore, we expect Mode 1 knowledge production processes to generate impacts on areas of knowledge that are closely related to the knowledge base of the collaborators. Additionally, in Mode 1 knowledge production, research teams are more homogeneous and hail from similar disciplinary and institutional backgrounds. In these contexts, bonding forms of social capital play a central role by reinforcing trust, stability, and shared interpretive frameworks that sustain incremental learning and path-dependent knowledge cumulation (Coleman, 1988)[7]. Therefore, we expect Mode 1 knowledge production processes to generate impact to specific social groups closely connected with the participants in the process.

Mode 2 knowledge production processes are more likely to result in combinatorial knowledge dynamics, that is, in the production of new knowledge that integrates the already quite different knowledge bases of the participants, and which can be quite far from the initial knowledge bases that the participants possessed. Therefore, we expect Mode 2 knowledge production processes to generate impacts on areas of knowledge that are not closely related to the knowledge base of the collaborators. Additionally, in Mode 2 knowledge production actors span organisational, institutional, and sectoral divides. To bridge these interfaces and build a common knowledge base, actors must actively invest in establishing mutual understanding. In terms of impact, we expect Mode 2 knowledge production processes to generate impact to broader social groups not closely connected with the participants in the process.

What drives transformative collaborations

We find that cultural impact is by far the most common outcome. Museums, galleries, and heritage sites lead in this area, while science-based cultural institutions like zoos, observatories and botanical gardens tend to contribute more to environmental and global impacts. Smaller, focused research teams often achieve deeper cultural engagement, while larger teams tend to spread their impact across sectors.

We also find that collaborations can be broadly classify into a Mode 1 vs Mode 2 pattern:

• Mode 1 Collaborations – Small, focused collaborations between academics and cultural institutions, mostly in the arts and humanities. They rely on traditional research methods and, as we expected, produce strong cultural outcomes with local reach.
• Mode 2 Collaborations – Larger, more diverse collaborations that also include universities, charities, and community groups. They are more experimental and practice-based, and, as we expected, they generate wider social, technological, or environmental impacts, often internationally. They are more likely to address contemporary Grand Challenges such as climate change, health, technology for social good, and placemaking.

Together, these modes form a balanced ecosystem: Mode 1 deepens cultural and disciplinary expertise, while Mode 2 drives broader societal change.

Figure 2. Types of impact of Mode 1 and Mode 2 collaborations

Figure 3. Geographical span of impact of Mode 1 and Mode 2 collaborations