From: Rick Anderson <[log in to unmask]> Date: Mon, 23 Sep 2013 14:14:23 +0000 >My problems with Rick's suggestion come in what little we know of the >nature of green OA provision. Phil Davis has indicated that >researchers cherry pick their publications, putting their best >articles out as Green OA. See his article in BMJ 2008;337:a568 that >notes: " prestigious articles are more likely to be made freely >accessible." http://www.bmj.com/content/337/bmj.a568 If that is >still true, cancelling on the basis of green OA would mean, yes, the >most cited articles have a fair chance of being readable to Univ. of >Utah researchers. . For a research project, article, review, grant >application etc. it's not just the top of the pile that gets sorted >through, its the whole thing. So one potential outcome for the U of U >is that Utah will pay MORE for the less cited literature ie. in ILL >and document delivery fees. or researchers will miss it all together. I apologize, but I'm having a hard time following the thread of this paragraph. I think the upshot is that canceling a Green journal might lead the library to pay more in per-article transaction fees than it would have for the subscription -- is that correct? If so, then in such a case the library should probably reinstate the subscription. >And even the green OA article has potential issues. What researcher >will be satisfied with the OA version? That's a good point -- but unfortunately, librarians have never been (and never will be) able to make all researchers equally satisfied. We have to treat some of them preferentially. How that treatment is distributed will (or should) be determined by the priorities of the library's sponsoring institution. So at my university, for example, we have to give physicists more support than we give to researchers studying Latin American literature. This means that if Latin American literature journals and high-energy physics journals go Green at similar rates (and if our budgets remain flat relative to STM price increases), we'll probably start canceling Latin American literature journals in order to maintain access to the physics journals, despite the dissatisfaction this causes among our Latin American lit scholars. And as the budget/price trends continue, we'll eventually start canceling some of the Green physics journals in order to maintain subscriptions to the non-Green ones, despite the dissatisfaction of the physics researchers. We don't have the option of making everyone equally happy. >They need to cite the final >form of publication, probably why arXiv has not had an appreciable >impact on subscription rates to High Energy Physics journals. Correct me if I'm wrong, but I don't think what usually goes to the arXiv is the same thing as what gets deposited under most Green mandates. As I understand it, Green OA is generally about authors' final versions. The arXiv generally deals in more preliminary versions, which I believe would largely account for the arXiv's relatively low impact on subscription rates. Even more significant, though, is the centrality of high-energy physics journals to the research mission of so many universities. As I've said before, journals that are centrally important are relatively unlikely to be canceled even if they go 100% Green. It's the ones on the margins that would be under greater threat. But that means lots and lots of journals. >I know of no evidence that green OA is a threat to subscription based >journals, and in fact as "free " advertising, probably pushes the >researcher to the original. Well, it's notoriously difficult to collect data about things that haven't happened yet. Until we see truly pervasive adoption of Green policies, there won't be any direct evidence of the impact of pervasive adoption of Green policies. In the meantime, we do have common sense, which says that it's pretty hard to give away content for free and continue selling it at the same time. (Bear in mind that giving away your content is not the same thing as advertising it. And giving away free samples only enhances sales when the samples are small and access to them is limited.) The other thing we do have is this study from ALPSP last year: http://bit.ly/16TPU7L. It's not the same thing as direct evidence, but I think it's pretty strongly suggestive. --- Rick Anderson Assoc. Dean for Scholarly Resources & Collections Marriott Library, University of Utah [log in to unmask]