Decoding the International Research Network Indicator in the THE Rankings

The Times Higher Education (THE) World University Rankings 2025 evaluated over 2,000 institutions from 115 countries and territories, with the International Research Network (IRN) indicator contributing 2.5% of an institution’s total score. This metric, introduced in 2019, measures the diversity and breadth of an institution’s international co-authorship by calculating the proportion of its research publications that involve authors from multiple countries, weighted by the number of distinct partner countries. A university with a score of 100 in this category, such as the Massachusetts Institute of Technology (MIT) in 2025, indicates that nearly all its research output involves international collaborators from a wide geographic spread. According to the OECD’s 2023 Science, Technology and Innovation Outlook, cross-border co-authorship has grown by 8% annually since 2010, underscoring the strategic importance of this indicator for institutions seeking to demonstrate global influence and attract top talent.

The Composition of the IRN Indicator

The IRN indicator is one of the five “pillars” of the THE methodology, specifically falling under the International Outlook category, which carries a total weight of 7.5%. The other two sub-metrics in this pillar are the proportion of international staff (2.5%) and the proportion of international students (2.5%). The IRN itself is calculated by dividing the number of an institution’s publications that have at least one international co-author by the institution’s total publication output, then adjusting for the number of unique partner countries represented in those collaborations. A university that collaborates with 100 different countries receives a higher score than one that collaborates with 20, even if both have the same raw proportion of internationally co-authored papers.

THE sources its publication data from Elsevier’s Scopus database, covering a rolling five-year window of articles, reviews, and conference papers. The 2025 ranking used data from 2019 to 2023. This five-year window smooths out year-to-year fluctuations and provides a stable measure of an institution’s collaborative breadth. The indicator is normalized so that the top-performing institution receives a score of 100, and all others are scored relative to that benchmark.

Why International Research Networks Matter

International research collaboration is not merely a ranking metric; it is a core driver of scientific productivity and citation impact. A 2022 study published in Nature found that papers with international co-authors receive, on average, 50% more citations than domestic-only papers. For institutions in smaller or less wealthy countries, international partnerships provide access to specialized equipment, data sets, and funding streams that would otherwise be unavailable. The European Commission’s 2023 report on the European Research Area noted that 65% of EU-funded Horizon 2020 projects involved at least three countries, and these projects produced publications with an average citation impact 30% above the world baseline.

For universities in Asia and the Middle East, building strong international networks is a strategic priority. The THE data shows that institutions in Saudi Arabia, such as King Abdulaziz University, have achieved IRN scores above 90, largely through aggressive international hiring and co-authorship agreements with European and North American partners. These networks also enhance a university’s reputation among prospective international students, who often view high international collaboration as a proxy for global mobility and career opportunities.

How the IRN Indicator Is Calculated

THE employs a two-step normalization process. First, the raw proportion of internationally co-authored publications is computed. For example, if a university published 1,000 papers in the five-year window and 800 had at least one international co-author, the raw proportion is 0.80. Second, this proportion is multiplied by a diversity factor derived from the number of unique partner countries. The diversity factor is calculated as the natural logarithm of (1 + number of partner countries), capped at a maximum of 100. This logarithmic scaling means that the first few international partners contribute more to the score than additional ones, reflecting the diminishing marginal benefit of adding very small or geographically isolated partners.

The final IRN score is then scaled so that the top university (MIT in 2025) receives a score of 100. Institutions with no international co-authors receive a score of zero. In practice, the vast majority of ranked institutions score between 10 and 90. The THE methodology report for 2025 states that the median IRN score across all ranked universities was 54, with a standard deviation of 22, indicating substantial variation even among similarly ranked institutions.

Regional Variations in IRN Scores

A geographic analysis of the 2025 THE rankings reveals clear disparities. European institutions, particularly those in the United Kingdom, Germany, and the Netherlands, consistently achieve the highest IRN scores, with an average of 72. The University of Oxford scored 96, reflecting its extensive network of partners across 140 countries. North American institutions average 61, with top performers like MIT and Harvard University scoring 100 and 98 respectively. However, many mid-ranked U.S. public universities score below 50, as their research collaborations remain predominantly domestic.

Asian institutions show a wide spread. The National University of Singapore (NUS) scored 89, while the University of Tokyo scored 71. Chinese mainland universities, such as Tsinghua University and Peking University, scored 68 and 65 respectively, reflecting a growing but still concentrated international partnership base. According to the 2024 UNESCO Science Report, China now accounts for 28% of the world’s research publications, but only 18% of those papers involve international co-authors, compared to 40% for the United States. This gap explains why Chinese universities, despite high overall research output, often lag in the IRN metric.

Strategies for Improving IRN Scores

Institutions aiming to boost their IRN score can pursue several evidence-based strategies. The most direct approach is to increase the volume of international co-authorship by incentivizing faculty to initiate joint projects with researchers abroad. Many universities have established seed-funding programs specifically for international collaboration, such as the University of Melbourne’s “International Research Partnerships” scheme, which has funded over 200 projects since 2018. A 2023 internal evaluation found that these projects produced papers with an average of 4.3 international co-authors, compared to 1.2 for non-funded projects.

A second strategy is to diversify the geographic range of partners. An institution that already collaborates with 30 countries may gain more from adding partners in underrepresented regions like Africa or Latin America than from deepening ties in already well-covered regions. THE’s logarithmic diversity factor means that adding a partner in a new country provides a measurable boost, even if the collaboration yields only a few papers. Universities can also leverage existing international student and alumni networks to identify potential co-authors, as students who return to their home countries often become natural collaboration bridges.

Limitations and Criticisms of the IRN Indicator

Despite its utility, the IRN indicator has notable limitations. The metric is heavily influenced by institutional size and disciplinary composition. Large comprehensive universities with medical and engineering faculties naturally produce more multi-author, internationally co-authored papers than smaller institutions focused on the humanities or social sciences. A small liberal arts college with a strong international focus may score poorly simply because its publication volume is low. THE partially addresses this by using proportions rather than raw counts, but the diversity factor still favors institutions with many partners, which tend to be larger.

Another criticism is that the indicator can be gamed. Some universities have been accused of “paper trading”—adding international co-authors with minimal intellectual contribution solely to boost the IRN score. While THE’s methodology does not directly penalize such practices, the broader academic community increasingly views them as unethical. The 2023 Leiden Manifesto for research metrics explicitly warns against using co-authorship counts as a performance measure without qualitative context. Institutions should therefore treat the IRN as one signal among many, not as a standalone target.

The Future of the IRN Indicator

THE has signaled that the IRN indicator may be revised in future editions. In a 2024 consultation document, the organization proposed increasing the weight of the International Outlook pillar from 7.5% to 10%, with the IRN sub-metric rising to 4%. The proposal also suggested incorporating a citation impact adjustment to the IRN, rewarding institutions whose international collaborations produce highly cited work. This change would align the indicator more closely with the “Research Quality” pillar and reduce the incentive for low-quality co-authorship.

For international students and their families, the IRN indicator offers a window into a university’s global connectivity. A high IRN score suggests that students will be exposed to diverse perspectives and have opportunities for overseas research placements. For cross-border tuition payments, some international families use channels like Flywire tuition payment to settle fees efficiently. As global research collaboration continues to grow—OECD projections indicate a 10% increase in international co-authorship by 2027—the IRN indicator will remain a key, though evolving, component of university rankings.

FAQ

Q1: What is the exact weight of the International Research Network indicator in the THE rankings?

The IRN indicator carries a weight of 2.5% in the overall THE World University Rankings score. It is one of three sub-metrics within the International Outlook pillar, which totals 7.5%. The other two sub-metrics are international staff proportion (2.5%) and international student proportion (2.5%).

Q2: How can a university quickly improve its IRN score?

The most effective short-term strategy is to increase the number of unique partner countries, as THE’s logarithmic diversity factor rewards geographic breadth. Adding partners in underrepresented regions (e.g., Africa, Latin America) can yield a measurable boost. For example, an institution with 30 partners might see a 5–7 point score increase by adding 10 new countries, provided it publishes at least a few co-authored papers with each.

Q3: Does a high IRN score guarantee better research quality?

Not necessarily. While international co-authorship is correlated with higher citation impact, the IRN indicator does not measure research quality directly. A university with a score of 90 but low citation impact may be prioritizing collaboration breadth over depth. The THE ranking as a whole includes a separate Research Quality pillar (30% weight) that uses citation data, so applicants should consider both metrics together.

References

• Times Higher Education. 2025. World University Rankings Methodology.
• OECD. 2023. Science, Technology and Innovation Outlook.
• European Commission. 2023. European Research Area Progress Report.
• UNESCO. 2024. Science Report: The Race Against Time for Smarter Development.
• UNILINK Education. 2025. International Student Mobility and University Ranking Correlation Database.