This article observes the results and mobility of the São Paulo State HE institutions performance in the THE World University Ranking 2020.
The main highlights are the steady increase of citation performance of USP, the impact of recent cuts on the Federal institutions, and a highlight on a particularity of THE, which favors small institutions over the large ones, and is strongly connected to public financing opportunities.
2017-19 results of USP Unesp and Unicamp in the THE World universities ranking
|Institution||Year||Group||Overall||Teaching||Research||Citations||Ind. income||Intl. Outlook|
Analysis of Results:
USP has remained in the same group since 2017. Despite significant financial hardship, which we would expect to see reflected in a fall in teaching score, where university budget constitutes 7.5% of the total, the scores have actually improved, reflecting a continuing presence in the reputation score from USP’s exceptional international reputation. Citation score continues to grow in steadily, reflecting greater insertion into international research networks. The research indicator score has fallen, mainly because of the reduction in federal research funding over the past few years. It should be noted that these scores are based on the previous year (2017-2018), and therefore the dual crises of Capes and CNPq will not be reflected in this indicator until the 2021 edition, where it would be reasonable to expect a large drop in “research”.
Unicamp has fallen a group into 601-800 this year, although this fact should be tempered with the fact that it has only actually lost points in the teaching category. This is a measure of staff-student ratios and total core budget, combined with a reputation survey. The decline in core budget, as well as expansion of student numbers probably accounts for this difference. Despite having dropped a group, the university’s performance has not particularly changed from the previous year, with other universities growing more quickly than Unicamp is able to.
Unesp’s model is unlikely to be favoured by the model the Times Higher employs; it is large, much of its output and mission is dedicated to local impact and its resourcesare spread thinly across a wide number of areas. That said, it has improved significantly in this edition in all dimensions, although it has not regained the citation score that it lost in 2018. Unifesp recorded an almost identical increase in the same year, suggesting that this was an attribution issue rather than a real drop (see table in Annex). However, given that Unifesp dropped 29 points in this indicator this year, it is not clear where these citations have now been attributed, given that neither Unesp or USP have advanced as much – from our own investigation of Scopus, we know that Unesp has a lower field normalised citation rate than her neighbours, but it is not as dramatically lower as this table would suggest. This indicator is the key for Unesp to progress in this ranking, and occupy a higher position than it currently does.
Who does this ranking favour?
The Times Higher is very heavily normalised by size – all of the indicators apart from reputation are size-independent (68% of the total). This means that smaller universities are more successful than larger ones – there is no university larger than USP that is ranked higher, and there are just 8 universities over 60,000 students ranked the same or higher than USP, and just one (University of Toronto) in the top 100.
This also means that extremely well-financed universities are also favoured. British universities, who are declining in this ranking in the ranking spent an average of $494,000 per member of staff, compared with Germany who now spend $1.2 million per member of staff (a result of the excellence initiative, which increased this spend by 38% in five years) and China, where $1.47 million spent (57.1% increase). It is clear that success in this ranking is heavily tied to public financial resources at nations’ disposal. Any government initiative that aims to improve in this measure must consider this experience from other countries who have improved (Germany has increased from 3 to 23 in the top 200, China from 5 to 7, with two in the top 30).
Recommendations for improvement in Times Higher for the São Paulo State universities
To be considered among the Life Sciences Challengers, USP needs to attend to its overall average citation count. It should study the policies of those universities of similar profile who are generating greater impact. A score of 55 would elevate it into the top 150 universities, ceteris paribus. Given its growth in this indicator in the past few years, it would be not entirely unrealistic to expect this target to be reached in the next five years. What will hold the university back is that this will not be fast enough to attain this position because of the growth of other institutions in countries with more stable funding systems. The methodological analysis suggests that progress in this ranking is tied to intensiveness of financing, and therefore progress will be difficult without increasing this.
Unicamp should first aim to reach the citation impact levels of Technion, University of Twente and Ruhr Bochum, similar universities (around 55) in order to climb significantly in this ranking. Given that growth in this indicator has been slow over the past three years, the university should look into strategies to encourage higher risk and higher reward research. Its industrial income score is compatible with being considered a technology challenger, although it could be improved.
The main obstacle for Unesp, keeping it within the category that Times Higher describe as a “Core Competences” university is its international reputation. Because the teaching and research reputations are heavily interlinked, improving this score has an outsized effect on the ranking. The deep instability in citation score (combined with the seeming exchange of citation scores with Unifesp), means that there is an ongoing issue of institutional attribution of citations caused by the university’s foreign language naming practices. On Scopus, Unesp has an FNCI of 0.92, compared to 1.06 for USP and 1.56 for Free University of Berlin. The considerable methodological uncertainty for Unesp needs to be dealt with before concrete conclusions can be drawn from its performance.