How a Narrow Metric Took Over the Meaning of International Research Collaboration

International research collaboration has become one of the most coveted currencies in modern academia. Universities boast of cross-border partnerships, governments fund them to boost national standing, and global rankings increasingly treat internationally co-authored papers as proof of institutional excellence. But according to a new systematic review published in the journal Higher Education, the way this collaboration is measured bears little resemblance to what it actually means — and the gap may be quietly reshaping who gets credit for the world’s scientific work.

The study, conducted by Xinzi Ow Yong and Yusuf Oldac of The Education University of Hong Kong, is the first to systematically interrogate how international research collaboration is both defined and measured across the scholarly literature. The researchers combed through more than 10,000 records from Scopus, Web of Science, and ProQuest Social Science, ultimately analyzing 723 peer-reviewed studies published between 1991 and 2026 in both English and Chinese. Their verdict is stark: the field suffers from a profound mismatch between the rich, plural ways collaboration is conceptualised and the narrow, publication-obsessed ways it is operationalised.

At the heart of the problem lies a single, dominant proxy: international co-authorship. When researchers want to measure collaboration across borders, they almost invariably turn to bibliometric databases and count papers whose authors list affiliations in different countries. The numbers are striking. Across the 723 studies reviewed, a staggering 97.2 percent incorporated at least one bibliometric element, and 57.7 percent operationalised collaboration specifically through co-authored publications or author collaboration networks. Meanwhile, approaches that treat collaboration as a process (just 3.5 percent of studies), as a researcher orientation (0.6 percent), or as an object of explicit conceptualisation (2.2 percent) remain vanishingly rare.

This concentration is not accidental. The authors argue that international research collaboration has become what they call a metricised governance object — a phenomenon transformed into a number that institutions can chase, rankings can rank, and funders can reward. Because co-authorship data is scalable, standardised, and readily available through commercial databases, it has become the default lens through which the entire concept is viewed. The result is a kind of circular logic: measurable collaboration is rewarded partly because it is visible, and it becomes more visible because institutions reward it. Collaboration that does not culminate in a co-authored paper — capacity building, co-supervision, mentoring, informal intellectual exchange, long-term partnership development — simply vanishes from the record.

The conceptual landscape, by contrast, has grown far richer. Early work in the 1990s, notably by Katz and Martin, treated collaboration as a social process and warned that co-authorship offers only a partial trace of collaborative activity. Since then, the literature has expanded to encompass collaboration as a relationship built on trust and social capital, as a self-organising network shaped by preferential attachment, as a researcher-level orientation shaped by mobility and international experience, and as a governance instrument embedded in institutional and geopolitical structures. Yet despite this conceptual broadening, the operational toolkit has remained remarkably static. The authors describe this as asymmetric development: the vocabulary of collaboration has widened far faster than the means of observing it.

The consequences of this gap are not merely academic. The review identifies three distinct forms of slippage between concept and measure. First, collaboration as process gets reduced to collaboration as output: a jointly authored paper captures the trace of joint work while leaving the negotiation, coordination, and exchange that produced it unexamined. Second, collaboration as relationship gets reduced to an affiliation trace: two authors linked to institutions in different countries may register as international collaborators even if the depth, symmetry, duration, or quality of their working relationship is entirely unknown. Third, formative drivers such as researcher mobility, institutional governance, trust, and informal exchange disappear entirely unless they happen to culminate in an internationally co-authored publication.

The metrics can misfire in both directions. In one study cited by the review, more than two-thirds of internationally co-authored papers from a Russian dataset involved diaspora authors, and roughly 20 percent only gained their international status through a second affiliation — cases where the co-authorship label overstates genuine cross-border engagement. Conversely, in European survey data, 63.8 percent of academics reported engaging in international research collaboration, but only 50.8 percent had produced an internationally co-authored publication — a case where the metric dramatically undercounts real collaborative activity. The same bibliometric classification, in other words, can represent fundamentally different underlying social processes.

These distortions carry serious equity implications. The review argues that dominant metrics reproduce Western-centric assumptions about visibility, impact, and legitimate participation, marginalising scholars and knowledge systems that fall outside prevailing standards. Researchers and institutions in resource-constrained or peripheral contexts may participate substantially in international research in ways poorly captured by publication data, while well-resourced actors are far better positioned to convert collaboration into high-visibility outputs. North-South partnerships, even when framed as developmental or humanitarian, can remain constrained by existing power relations. As global science becomes increasingly multipolar, status hierarchies within collaboration networks remain stubbornly resilient.

The timing of the review is significant. The corpus shows explosive growth: only 19.4 percent of the included studies appeared by 2015, while 80.6 percent were published between 2016 and 2026, with the period from 2021 onward alone accounting for 56.4 percent of the total. This acceleration coincided with mounting policy and evaluative pressure to demonstrate international research performance — yet the quantitative expansion has produced no comparable diversification in how collaboration is evidenced. Ironically, the disciplinary analysis reveals that education research, the field with arguably the greatest stake in how collaboration is evaluated, is among the least represented in the evidence base informing its own policy and practice.

Looking ahead, the authors anticipate that geopolitical reconfiguration will make cross-border research relationships increasingly uncertain and contested. They call for more reflexive approaches to evaluating international collaboration: researchers should complement co-authorship data with surveys, interviews, mobility records, funding information, and longitudinal network data, while policymakers and institutional leaders should broaden what universities recognise as valuable collaboration. Change will likely be gradual, given how deeply publication, funding, and ranking infrastructures are institutionalised. But the review’s central warning stands: as long as one narrow proxy stands in for the concept as a whole, the scientific community will keep measuring only the most visible sliver of the collaborative enterprise — and rewarding it at the expense of everything else.

Subject of Research: The gap between conceptualisation and measurement of international research collaboration in higher education research evaluation

Article Title: What counts as international research collaboration? The gap between conceptualisation and operationalisation in higher education research evaluation

Article References: What counts as international research collaboration? The gap between conceptualisation and operationalisation in higher education research evaluation. (n.d.). https://doi.org/10.1007/s10734-026-01788-x

Image Credits: AI Generated

DOI: 10.1007/s10734-026-01788-x

Keywords: international research collaboration, higher education, research evaluation, bibliometrics, co-authorship, scientific metrics, internationalisation, research policy, global science, systematic review, research governance, Global South