Opening the black box of university rankings

Opening the black box of university rankings

Today two university rankings were released, one by CWTS and one by Times Higher Education. Ludo Waltman and Nees Jan van Eck, responsible for the CWTS Leiden Ranking, use this occasion to discuss the importance of transparency of rankings.

The Times Higher Education (THE) World University Rankings 2025, published today, includes 12 Dutch universities, of which five are ranked in the top 100 worldwide. One prominent Dutch university, Utrecht University, is not included in the ranking. Last year this university decided to discontinue submitting data to the THE World University Rankings. Henk Kummeling, Rector Magnificus of Utrecht University, pointed out that rankings like those published by THE are not consistent with open science. Instead, Kummeling argued, rankings must be “transparent, verifiable and relevant” to the ambitions of an institution.

At CWTS we also released a ranking today: the CWTS Leiden Ranking Open Edition 2024. Unlike the THE World University Rankings, the Leiden Ranking Open Edition is fully transparent. The data underlying the ranking, obtained from the OpenAlex database, is completely open, and so are the source codes of the algorithms used to calculate the indicators included in the ranking.

How many Dutch universities are included in the top 100 worldwide according to the Leiden Ranking Open Edition? The ranking on purpose does not provide a single definitive answer. The answer depends on the indicator you are interested in. For instance, suppose you are interested in the overall publication output of a university. The ranking then tells you that three Dutch universities belong to the top 100 worldwide. However, if you are interested in the share of the publications of a university that are open access, the ranking shows that 12 of the 13 Dutch universities belong to the top 100. In other words, Dutch universities may not be among the biggest scientific powerhouses worldwide, but they do have a leading position in terms of the level of openness of the knowledge they produce.

The need for full transparency

The lack of transparency of the major commercial university rankings has been strongly criticized, for instance in a report of an expert group of Universities of the Netherlands (of which one of us was a member) and in a recent report of the European University Association. Instead of making the data underlying its World University Rankings openly available, THE sells the data to academic institutions, claiming the data will help them “remain globally competitive”. There is a painful irony here: most of the data THE sells to institutions has been produced by the institutions themselves. Partly the data was prepared by institutions and then submitted to THE, and partly the data results from surveys completed by the academic staff of institutions. To open the black box of the THE World University Rankings, academic institutions need to pay THE to get access to data they produced themselves!

The opaque nature of many university rankings violates basic principles of open science and is fundamentally incompatible with initiatives promoting openness of research information, such as the Barcelona Declaration on Open Research Information. Fortunately, this is increasingly being recognized, for instance by the universities in the Coimbra Group. Moreover, universities are starting to take concrete actions. In a recent statement, Université de Lorraine emphasized the need for a transition to “open and transparent” rankings, referring to the Leiden Ranking Open Edition as an example of such a ranking. Moreover, “to contribute to a transition towards open rankings, which provide institutions with the opportunity to regain control over research and teaching management indicators”, Université de Lorraine announced that “it chooses to withdraw from the commercial THE and QS rankings by no longer providing the necessary data”. After Utrecht University and University of Zurich, Université de Lorraine is the third European University within a year making the decision to discontinue its participation in the THE World University Rankings.

Advancing the transition to transparent rankings

Henk Kummeling is completely right that university rankings must be “transparent, verifiable and relevant”. The Leiden Ranking Open Edition is indeed fully transparent and verifiable, and by refraining from providing a ranking of the ‘overall performance’ of universities, it encourages users to focus on the indicators that are most relevant to them.

Making the Leiden Ranking fully transparent took a substantial effort from our team at CWTS as well as the external partners that we collaborated with. To support a broader transition to transparent rankings, we would be happy to share the experience we gained in this process with the providers of other rankings, for instance to help them move from closed proprietary bibliometric data sources to open alternatives.

To advance the transition to transparent rankings, universities also need to take action. Ideally universities would follow the example of Utrecht, Zurich, and Lorraine and discontinue their participation in non-transparent rankings. If this is considered too big of a step, there are other smaller steps universities can take, such as changing the way they communicate about rankings, making the data they submit to rankings openly available, and joining the More Than Our Rank initiative. Leiden University is currently working on these steps.

University rankings create complex dilemmas for universities and their leaders. Changing how universities deal with rankings requires leaders that do not run away from their responsibility. Last year Chris Brink aptly posed the question all university leaders should think about: “As a leader, you will be conscious of your legacy … if there is a chance that peddling rankings may come to be viewed somewhat like smoking or digging coal or selling arms, do you really still wish to be seen in the company of the rankers?”

Header image by Jonathan Farber on Unsplash
DOI: 10.59350/8frt3-50m98 (export/download/cite this blog post)

0 Comments

Add a comment