REF 2014: study ranks subjects by competitiveness

I am surprised that the benchmark used for research excellence is the number departments IN THE UK. I would have thought by now that it would be more standard practice to use international benchmarks for this. A fairly common indicator for international research excellence is the proportion of a group's scientific outout that is among the top 10% most highly cited IN THAT FIELD. The combination of proportionality plus subject specficity plus internationality excludes the kind of bias that prof. Kelly aims to address. Kees Kouwenaar Senior advisor international strategy Vrije Universiteit Amsterdam

We should take this opportunity not to beat each other up within universities, but to be very upbeat about how good the research coming out of British universities is. Of course, it may be an uncomfortable truth that the RAE/REF is part of this, and that it has driven research quality upwards. (The problem of Kees Kouwenaar's arguement is that there is not a common indicator or measure of 'international research excellence'. Not least the measures that there are only for work in English. Moreover, the REF categories do stipulate international standing as a criterion.)