John Hudson asks "what is wrong with (the) academic 'transfer market'" induced by recent research assessment exercises (THES, July 31)? I would like to offer two related answers. The first is that it is not a transfer market like that used by football clubs to buy players. Then the selling club receives millions of pounds for surrendering its claim on the player whose skills it has nurtured. This nurture is not irrelevant to the present case. The money for university departments riding on their RAE ratings is presumably designed to provide them with an incentive to provide an academic environment conducive to fruitful research.
If this is the case it would seem natural that credit for the research should go to the institution providing the facilities - and ambience - in which it was done, although this presents practical difficulties.
My second argument relates to the conditions under which it would not matter whether the Higher Education Funding Council for England rewarded departments buying in recently research-productive scholars or those in which the research was done.
Suppose young researchers knew what they could achieve in departments A, B or C and that as a result they would subsequently be offered rewards R, S or T by developments D, Y or Z. The most rewarding career might start by taking an ill-paid job at research-friendly A in the knowledge that the largest reward (T) would eventually be earned from entrepreneurial Z. Spending time and money to make a department research-friendly is then made viable by the willingness of productive researchers to work there for a while for less.
Thus even the transfer market- inducing form of RAE might indirectly provide appropriate incentives for running research-friendly (as opposed to proven researcher-friendly) departments.
For the result that the structure of the institutional incentives makes no difference to hold, a number of demanding conditions is required:
* researchers have to know their productivity in different places with certainty or be undeterred by the uncertainty
* pay scales have to be flexible and employers have to be able to recognise the talent of those not demanding high salaries
* there has to be a good capital market so that the scholar taking the low pay at research-friendly A in expectation of an ultimate financial reward from entrepreneurial Z (and their partner) can live in the meantime as well as they could as hack teacher(s) not doing research at dull and unstimulating C (or even Z itself).
Since these conditions do not hold, RAE methodology conducive to a transfer market is demonstrably less effective than the hypothetical alternative in providing institutions with incentives to foster fruitful research by rewarding them directly for its achievement.
John Flemming Warden, Wadham College University of Oxford