The research excellence framework is seen by most academics as a necessary evil. Necessary because submission is crucial for future research income. Evil because it is stressful and time-consuming.
Added to this are three ethical problems posed by the exercise, which will be used to distribute about £1.6 billion in annual quality-related research income from 2014.
First of all, the REF only concerns research-active staff, which can lead to the marginalisation of teaching-only academics. On the one hand, teaching-only staff are at least spared all the paperwork involved. But on the other, the fact that their research-active colleagues have to spend so much time on the REF can leave those who are not involved shouldering even more of the teaching burden, leading to further alienation.
This disaffection will only be exacerbated by the increasing emphasis in annual review forms on REF-related activity, which can lead to teaching-only staff having to leave large swathes of the forms entirely blank.
A second ethical problem is that the added value attributed to collaboration and interdisciplinarity by the REF incentivises the dishonest attribution of authorship. If your boss asked you to add someone's name to a paper because otherwise they wouldn't be entered into the REF, it could be hard to refuse.
Let's imagine that Trevor has a host of publications on which he is the rightful first author. His boss, Miranda, asks him to put Joanne, a young researcher, as a named author on some of his papers to boost the department's REF score for collaboration, despite the papers in question being Trevor's work alone.
This puts Trevor in an invidious position: he knows he ought to refuse as guest authorship is unethical, but he wants to please Miranda and help his colleague and department - and the paper in question isn't part of his REF submission anyway. This leads us to the next issue.
A more serious problem is posed by the fact that the REF only allows a maximum of four submissions over a six-year period.
Let's imagine that Arthur and Bob are rival researchers at different institutions. Each submits four absolutely outstanding pieces of work to the REF and is justly awarded a great deal of cash - they each receive the same amount. This all seems perfectly fair.
But Bob submitted almost all of the research he did from 2008 until 2013 - he only had one other quite high-quality paper that could have been submitted, and a few third-rate ones. Arthur, on the other hand, was spoilt for choice, as he had 20 publications of the highest quality and no inferior ones.
Why is the REF completely insensitive to quantity in this sense?
This omission is all the more important given the assessment's emphasis on impact (20 per cent, against 15 per cent for environment and 65 per cent for output).
It will judge the impact of all of Bob's top-class research against only a small proportion of Arthur's. Arthur's research might well have led to five times the impact of Bob's, but because he can submit only four papers, he might as well have not done the work as far as the REF is concerned.
(The author of this article is a case in point. I have had 25 publications in leading journals since 2008, and will probably have 40 by the end of 2013. Apart from this, I have not experienced any of these ethical problems first-hand.)
If impact judgements are based on case studies rather than outputs, the same problem applies, as the number of submissible case studies is limited, too.
As currently described, the REF looks at quality but not the quantity of quality, which is clearly wrong. Only if Bob lets other people take the credit for his "spare" research (see the second ethical problem mentioned above) will it be included in the REF.
These three ethical issues deserve attention. As it stands, the REF is exacerbating a schism between research and teaching staff, encouraging deceptive attribution of authorship and failing to give due credit (and cash) to deserving research. As such, it needs some serious refereeing.