LEADER

Being the best? Yes - but best for what?

John Butler-Adam

The Leader in the last issue of the South African Journal of Science (109 7/8) considered DORA - the San Francisco Declaration on Research Assessment - just a month or so before one of the major ranking agencies, Quacquarelli Symonds (QS), released its 2013/2014 rankings of over 800 universities across the world. QS claims that its list captures the top 4% of world universities. If this figure is correct, it means that an appearance on the list, even at a low rank, puts an institution ahead of 96% of the world's universities. Not a recognition to be ignored - and, as it happens, the release of the information has produced an understandable flurry of 'boast' material on the websites of the South African universities on the list, along with a fair volume of media coverage.

Just two countries in Africa have universities that appear in the QS rankings - Egypt and South Africa. Egypt has six institutions that rank between 348 (American University of Cairo) and 701+ (Alexandria University). Seven South African universities are ranked between 145 (University of Cape Town) and 601-650 (University of Johannesburg). The remaining five South African universities are Witwatersrand (313), Stellenbosch (387), Pretoria (471-480), KwaZulu-Natal (501-550) and Rhodes (551-600).

Five of the seven South African universities improved their ranks (where universities fell into clusters of ranks, the midpoints of the clusters have been used as surrogate measures of their actual positions in the rankings). Cape Town improved by 9 places, Stellenbosch by 38 and Witwatersrand, KwaZulu-Natal and Pretoria by 50. Rhodes and Johannesburg were ranked by QS, on application, for the first time this year, so improvement was not an issue - and they had the additional disadvantage of being ranked on only one submission rather than the three to five annual submissions that are the norm in the QS methodology.

The first, and possibly the most obvious, conclusion to be drawn from the shifts is that the higher an institution's ranking, the more difficult it is to move. The 'top' institutions form a very tight cluster - and the higher up the ranking they are, the tighter the cluster to which they belong. Harvard, Oxford, Stanford and Cambridge, for example, juggle their positions by fractions of points each year - or not at all.

Of course, this Leader focuses on just one ranking system (albeit a system that is increasingly significant amongst institutions that consider ranking to be important). In the top tight huddle, for instance, Oxford appears, in 2011, as 4th in the Times rankings, 5th in the QS ranking and 10th in the Academic Ranking of World Universities. In the following year, Oxford was, respectively, at 2nd (an upward shift of 2 places), 5th (no shift) and 10th (no shift). So while the rankings differ according to the selected criteria and methodologies, the tight clustering remains a common feature of the major ranking systems.

An important question to ask is what these rankings might mean. The obvious conclusions to draw would be, for example, that Cape Town is not just the 'best' university in South Africa but, by far, the best in Africa and that Rhodes and Johannesburg have done pretty well for first-time entrants - and entrants that are the only members of the 'Club of Seven' that do not have medical schools.

But the ranking system assumes that there is just one kind of university, with common criteria for measuring comparative success, while in many countries there are institutions that differ in terms of their markets and purposes in the higher education system - some through the student market and others through state steering. South Africa currently has four official types of universities - traditional, comprehensive, technology and a major, dedicated distance institution. In circumstances that both encourage and need a post-school education system that is seriously diverse, and whose institutions are complementary, it seems strange that institutions would wish to be ranked in systems that assume they are all the same, and that they can be directly compared with one another.

In reality, what happened to the citation indices, intended for scholarly journals, but that were arrogated by universities for their own purposes, has now also happened to at least some of the rankings. QS, intended for students and parents in the first instance (a quick review of the QS website will confirm this), has been assumed by universities as a basis for making claims about how much better they are than 'others'.

In his Welcome to the QS Rankings Report 2013-2014, Ben Sowter, head of the QS Intelligence Unit writes:

...QS World Rankings have captured the imagination of prospective students, ranked and non-ranked institutions, policymakers, governments and the public at large. ... Universities are diverse and complex organisations that are integrated into the very fabric of the nations, societies, communities and economies in which they operate. The distillation of their value into a single integer is, from one point of view, nonsensical; but from another, this very complexity is what makes a simple comparative assessment a necessity.

Well, they may be a necessity or, possibly more realistically, a guide to prospective students and their parents, but are these rankings a necessity for the universities themselves, especially in systems that are, or ought to be, diverse - systems that should have a range of institutions, which each aim to be the best in relation to its role and purpose in the system? In the South African case, for instance, is there a danger that universities of technology or comprehensive universities might strive to have index profiles that approximate those of traditional research institutions so that they can have related rankings? Or should they rather strive to have index profiles that meet the needs of their particular student bodies and the special, essential roles they play in a system aiming for diversity? Are there, perhaps, indices that recognise the efforts that institutions make to support underprepared students, or that support young postgraduate students in their teaching and research skills? Might Harvard, Cambridge or University College London care? Maybe not. But Stellenbosch, Rhodes and Johannesburg, for example, might well do so - although for different reasons.

This Leader is most decidedly not a polemic intended to undermine the goals of excellence in higher education, teaching and research at either an institutional or a systemic level, in whatever way that excellence is defined or assessed (even if by ranking systems). Universities and their corollaries have critical, essential roles to play in social, economic and individual development - roles that are quite tightly bound together. And the better they are, it might reasonably be assumed, the better they will be at fulfilling those roles. The concern expressed here, specifically in the South African context, is whether or not ranking systems might become an 'anti-diversity' pressure when the incentives for diversity are still at an early stage.

The South African Journal of Science warmly invites correspondence and commentary on the implications for this, as yet, unresolved challenge in higher education and scientific research. We will be very pleased to have and to consider your views and opinions.