Open access, again

“I had not intended to come back to this so soon. But a government-sponsored report of a committee on the subject, chaired by Dame Janet Finch CBE and including probabilist and statistician Peter Donnelly, appeared earlier this month, and it doesn’t reassure me... Press reporting of this suggested that it concentrated almost exclusively on the ‘gold’ option (author pays); I took a look to see whether this was actually the case. Indeed it was. The report mentions three publishing routes: subscription-based journals, open-access journals, and repositories. [No mention is made of the proven concept of high-quality free journals such as Electronic Journal of Combinatorics or Theory of Computing.] It concentrates on the second option, and discusses in some detail how this should be paid for. The basic argument is: Author-pays open access will not cost more in the long run, since Universities will save by not having to pay for journal subscriptions. But there will be a transition period of indefinite length during which extra money will need to be found. They estimate an extra £50-60 million per year for the higher education sector, made up of £38 million for paying charges for open-access journals, £10 million for extensions to licences, £3-5 million for repositories, and £5 million transition costs. The reason this worries me is that I am not sure they have their figures right. Nature says that the Wellcome Trust paid an average of £1422 per paper last year, but that some journals would charge much more:Nature itself would charge £6500 per article. Now £38 million for articles costing around £1500-2000 each will allow the publication of somewhere between 20000 and 25000 articles by British academics and health researchers. A quick estimate based on RAE returns and my own department’s publication figures suggests that British pure mathematicians alone publish over 1000 papers per annum; and pure mathematicians are far from the most prolific members of the academy. I suspect that the money they have suggested will not be anywhere near enough to pay to publish all research. Then there is the problem of whether, in these austere times, the Government will be prepared to pay £50 million a year for an indefinite period. It seems just as likely that they will say “We are .all in this together – academics will just have to take a bit of a cut, and publish fewer papers, or publish in cheaper journals”... Now a university, faced with not having enough money to pay for all its researchers to publish their results, will have to prioritize. Some researchers will get the support (no doubt, those who hold big grants), others will not. Pure mathematics is certain to lose out. Others who will lose out are categories such as temporary research staff (post-docs) and emeritus staff (me, soon!). Last time I wrote about this I predicted that we would have a two-speed research system. That looks even more likely now. I do not see my subject as among the winners here. The bottom line is that the people in power are not going to help us: we have to help ourselves, somehow.”