Interesting data, poor inferences

Apparently, the UK higher education system is a ‘knock-out’ performer.  Although the UK system is ranked 8th in the world — not bad at all — what is holding it back on that measure is central investment. Per capita, that is, higher education is not well funded in the UK, and the system as a whole pays the price. However, we can compensate for that, correcting for the relative poverty of the country (compared to, say, the US or Switzerland) and thus its relatively low level of public investment. This yields a different ranking, with the UK way up in second place. (Behind Serbia, interestingly.)

All very interesting. It makes you wonder from where all the calls for the UK system to be over-hauled in one way or another are coming (how marking and degrees are scaled being an old chestnut). If it ain’t broke, don’t fix it. On the other hand, is also makes you wonder what would happen if there was some higher level of investment…

The THE article first cited though, goes one step further. The reason why the UK system punches above its weight in this way is the Russell group — i.e. the concentration of funding into roughly 15% of the institutions. (We should also include a few other Universities who are not technically in the RG, but who enjoy big funding and strong research reputations, such as Leicester or Durham.) The article, or rather the chap being quoted, cites the recent REF results as evidence. Evidence for what, exactly? That research is concentrated, to be sure; that funding is concentrated, definitely. But it is certainly not evidence that concentration delivers ‘bang for the buck’. On the contrary, the fact that an army of researchers, working at the other 85% of institutions, are still doing international quality research with bread-crumbs of funding, much less infrastructure, and no endowments to speak of… THAT is bang for the buck. The cost per international-standard research publication at a ‘new’ university is a fraction that of the Russell Group, and I can prove it.

For example, let’s take two random Universities: Newcastle University and Northumbria. Now, Newcastle did pretty well in REF2014, staying in the top twenty nation wide. They submitted a total of 2806 outputs rated 3 and 4*. Northumbria are quite a way down the REF table, but comfortably in the top 100. They submitted 823 outputs at that quality rating. Now, clearly, there is a concentration of research in Newcastle, who produced three and a half times as many quality outputs. However, Newcastle received 35.6 million pounds in recurrent research funding in 2009-10, while Northumbria received … 3.26 million. In other words, with ten times the funding, they only managed three and a half times the number of quality outputs. It’s the same story when we compare Teeside and, say, York. The former required £6440 to produce an international quality paper, while York required £11,900.

Now, I know this is not the whole story. There is a difference between 3 and 4*, and also we may be comparing inherently expensive forms of research (laboratory sciences) with cheaper ones. On the other hand, in the above calculations, I haven’t included other big sources of funding, which are also highly concentrated, such as Research Council grants, European funds, and so forth. In the absence of more study, the evidence clearly goes against the ‘funding concentration’ hypothesis. (Hold on, who wrote this report that the THE is cheering about? — Universitas21 — and who are their UK members: all Russell Group. I’m seeing a pattern emerging.)


Leave a Reply

Your email address will not be published. Required fields are marked *