University Rankings

This is a guest post by my colleague John O’Hagan of TCD Economics Department:

The recent ranking of world universities was given considerable prominence in the Irish Times. Seán Flynn in his piece of September 11th (“Why Irish universities should take world rankings seriously”) argued that the fall of Ireland’s top universities in the world rankings should not be dismissed?.

I do not wish necessarily to disagree with this argument or to open up here the debate about the merits of rankings. As Mr Flynn states they do matter, if only because they influence the choice of academics/researchers concerning where to work, and the choice of students, graduate students in particular, with regard to study location. In the absence of other more reliable information, upon what else can such matters be decided?

It appears that the QS World University Rankings are now the most reliable guide to the overall performance of a university, although Times Higher Education (THS), with even more recent rankings and which used to work jointly with QS, might challenge this. I do not desire to get into this discussion either, but wish to point out some important other issues that must be borne in mind when considering the rankings. I will make these points in relation to the recent QS rankings.

The first issue is the variation in the ?scores? attached to each university. Cambridge, the highest ranked university in the 2010 rankings, was given a score of 100. The tenth-ranked university, Princeton, was assigned a score of around 96, that is, just four points below Cambridge. Trinity, ranked at 52, had a score of 75. This figure for Trinity is well below that of Princeton, but not much below the university say ranked 39th, namely Brown University in the US (score 80.5).

But it is not much higher either than the score of the university ranked 100th (Oslo with a score of around 64). Thus very little can sometimes divide the universities separated in rank by 30 to 50 places. As such, it might be best to use three/five-year averages rather than the score for one year. Bearing this in mind, UCD has moved very considerably up the rankings in the last five years (over 100 places between 2006 and 2010), despite the slippage in 2010.

The second point to make is that a university ranking can vary significantly by academic area. Taking UCD as an example again, it ranked 89th in Arts and Humanities but 261st in Natural Sciences. The variation in Trinity was less marked: 52nd in Arts and Humanities to 99th in Engineering and IT. Harvard, the second ranked university overall, had a rank of 22nd in Engineering and IT but 1st in the Social Sciences and Life Sciences/Biomedicine. It could be argued that it is these discipline rankings, and not the overall rankings, which really matter to academics/researchers and graduate students (and possibly even more disaggregated rankings, like for economics or sociology say within the Social Sciences).

The third point to make is to advocate caution with regard to explanations for variation in rankings. There may for example be too much of a readiness to attribute a fall in ranking to a drop in public funding. There is in fact little evidence as far as I know linking public funding and the ranking of a university. If so, how then can the very low ranking of some French and German universities be explained? How can we, in the Irish context, explain the 9 and 25-places fall in the Trinity and UCD rankings in 2010 and the 23-place rise in the UCC ranking? Universities should perhaps also look at their own performance and operation (e.g. salary levels and structures, promotions, governance), as well as highlighting any funding deficiencies, in responding to changes in ranking.

In this regard, one must also remember that in an Irish context it is GNI (Gross National Income) and not GDP that is the more appropriate benchmark to use in assessing total state expenditure on education. When using GDP, Ireland comes out below the OECD and EU averages, but is about average when GNI is used.

The fourth and final point I wish to make is what appears to be a strong bias in the rankings towards the English-speaking world. Of the top 30 universities using the GS ranking, 26 are in the English-speaking world. Is it really believable that the UK has eight universities in the top 30, whereas developed countries of similar size and maturity such as France and Germany have none? The caution in this regard is borne out when looking at other, more Europe-based rankings, such as the Leiden Rankings. While their rankings are based on research alone, the listings are more in line with what one would expect. On one of their rankings, the University of Göttingen in Germany is ranked first in Europe (including the UK), with Lausanne (Switzerland) ranked second, and generally their top-50 rankings include a large number of Continental European universities.

It might then be more appropriate to compare Irish universities to UK universities in the GS rankings; on this basis, Trinity ranks 9th in the British Isles and UCD ranks 20th. These are not poor rankings, as for example in the case of Trinity they put the College ahead of universities such as Warwick, York, Durham and St Andrews, all very reputable institutions. Trinity though ranks well behind Bristol and Edinburgh universities and does not come near the scores assigned to the top four UK universities, namely Cambridge, University College London (UCL), Imperial College London, and Oxford (all with scores of around 98 or higher). So while the slippage in the rankings this year may be a legitimate cause for some concern, the more pressing question perhaps is can an Irish university ever get into the Top 5 rankings in these islands , and if so how?