Well that was quick. Less than a month after the Finch working group published its recommendations on the future of open access, UK science minister David Willetts has responded, saying in effect “Let’s go for it.” The government has taken essentially all of the recommendations on board and has committed the country to making all its publicly-funded research available for free online by 2014.
Except that it’s not quite that simple. There are weak points in the government’s response, but in other areas the policy implementation has actually gone beyond what Finch recommended. Let me deal with the weaknesses first.
The point-by-point response (4-page PDF) by David Willetts to the ten recommendations, which cover a range of issues, makes it clear that there will be no new money to lubricate the transition to open access. As a result, the implementation of several of the recommendations remains decidedly aspirational. It is yet to be seen how the health sector or businesses will secure access to the research literature, how university libraries will negotiate with publishers to ensure that subscription prices reflect the increase in open access content, or how scholarly monographs will be paid for (a major concern for he social sciences and humanities). These matters are now left for the relevant institutions and stakeholders to figure out — but with no timescale for resolution imposed.
On the plus side, the proposals that relate most directly to the work of publicly-funded researchers are laid out much more clearly because they have already incorporated into the new open access policy of Research Councils UK (RCUK), which was also timed for release today. The policy appears to have retained all the muscle that was evident in a draft document that was circulated back in March and even shows signs of having worked out. It’s a strong statement that surpasses the Finch recommendations.
For starters, from 1st April 2013:
Peer reviewed research papers which result from research that is wholly or partially funded by the Research Councils:
1. must be published in journals which are compliant with Research Council policy on Open Access
2. must include details of the funding that supported the research, and a statement on how the underlying research materials – such as data, samples or models – can be accessed.
The document makes it clear that a compliant journal is one that permits either immediate free access to readers on payment of an Article Processing Charge (APC) (gold OA) or, if that option is not made available by the publisher, allows the author to deposit their final peer-reviewed version in a repository (green OA) no more than 6 months after publication (12 months for AHRC and ESRC funded work in the humanities, economics and social sciences).
What is more, if an APC is paid, the article must be accorded a CC-BY Creative Commons licence, which allows extensive rights to copy and distribute the content, even by commercial organisations. If no APC is paid, the deposited copy must still be made available “without restrictions on non-commercial re-use” (I think this may mean CC-BY-NC but would welcome any correction). These are robust conditions.
What is even more is that both routes to open access must allow “unrestricted use of manual and automated text and data mining tools”, a condition that will facilitate deeper and broader analyses of the research literature.
And that’s not all — condition 2 above places an obligation on authors to ensure that the data that their conclusions are based on are also made available and should be clearly sign-posted within the paper. This is an important dimension to open access — it is good for transparency and for research integrity. I half suspect this a response to criticisms of climate scientists who, in the wake of climategate were criticised for not making their data easily accessible; but perhaps it simply reflects a growing trend.
What about the money, and the vexed issue of costs, which has excited much of the negative response to the Finch report? Here again there is bad and good. On the bad side, there is no new money to help the excess costs of transition (when OA and subscription costs have to be borne), but that is hardly a surprise in these austere times and may even help to exert some downward pressure on the level of APCs.
The good thing, however, is that the Research Councils have finally adopted a flexible model for funding of publications that is similar to the one adopted by the Wellcome Trust. What will now happen is that money for publication costs, rather than being included in time-limited funding awards to individual investigators, will be paid as a block grant to the host institution, which will be required to establish a open access fund. This nicely addresses a criticism I made back in February, but what isn’t yet clear is exactly how these awards will be calculated. I imagine there is some concern among universities as to whether the Research Councils will get their sums right. Nevertheless establishment of such funds (already in existence in some universities, including my own), also creates a possible mechanism for the transfer of journal subscription funds, currently provided via HEFCE, as the publication model shifts from subscriptions to APC-supported open access.
There is still much to be done. The UK government deserves credit for staking out its position so boldly but this is a risky stratagem and it it to be hoped that its ambition will set an example for other countries to emulate. We can have some reassurance in the fact that that there is already momentum towards open access in the US and the EU.
Moreover, as will be clear from this synopsis, many details of the process remain to be sorted. Some commentators have bemoaned the fact that the government’s endorsement panders too readily to publishers’ interests and current prices (see Stevan Harnad’s comments in the Guardian). However, it is important to see today’s announcement not as an end-point but as a beginning. If we, the community that generates and reviews the research literature, want a publication system that is accessible and effective and that represents good value for money, we have to agitate for it (and the attendant culture change, part of which will involve breaking free from impact factors). We have to get involved in the establishment of the publication funds at our universities and shape the ways that they are run. There is plenty of scope to get this right; the government has made a good start but it is down to us to finish the job.
If you would like to hear me being interviewed briefly about today’s announcement on the BBC World Service’s NewsHour, start at 44:00 (not sure how long this link will work).
The simple take on all of this (i.e., the BBC summary) is that the cost to the tax payer, and the loss to the research budget, will be £50M. I don’t know how accurate this figure is, but any erosion of the UK’s modest science budget is discomforting. Furthermore, if my reading of these changes is correct, there is nothing here to suggest that the financial hit will be absorbed in any way by the massive profits of the big publishers. Perhaps as you say, it is now all up to us, the academics, to forget impact factors and choose the free or low cost publication options.
The $50m headline figure is very much a crude estimate. It could be higher but it could also — depending on a lot of factors — be much less. Even zero. The actions of other countries and any renewed competition in the publishing industry could affect the transitional costs a great deal.
I think the RCUK’s policy will be causing publishers some discomfort — they certainly reacted badly when the draft policy came out and the published policy is, if anything, even stronger. But is it down to us to shape this market by our actions. Already I see emerging fears about rationalisation of the numbers of publications that people will be allowed by their institutional fund and I’ll bet the research-intensive universities are concerned about the possibility of having to shoulder a greater burden than at present.
But if we think OA is a good idea — and I think most of us do! — we have to find a way to make it work. The current system is unsustainable in its present form — simply too expensive — so we are duty bound to find a cheaper system.
Slightly belated, but …
I think the £50-£60M transition cost is a huge over-estimate based on what publishers would like to have paid to them. For a more detailed analysis see my recent Guardian article.
Great as usual Stephen…
I too am confused on the CC-BY-NC issue. Many publishers in my field have close-to CC-BY-NC but not quite. For many, commercial use requires explicit permission. My understanding is that the Wellcome Trust’s point is about free access e.g. for text mining and so requires CC-BY-NC. I am still confused on their approach to Science/Nature etc though. It seems that there could be an exception since these publishers do not offer an OA model.
See:
http://www.wellcome.ac.uk/stellent/groups/corporatesite/@policy_communications/documents/web_document/WTVM055715.pdf
Specifically on point 12:
“Note: in relation to publishers who do not offer a compliant author-pays option, but are willing for content to be made freely available via PMC/UKPMC within six months of publication (e.g. Nature, Science, NEJM, JAMA etc.), although we encourage these publishers to attach a CC-BY licence to the self-archived, author manuscripts, we will not require this.”
As for cost – who knows, my own RC grants already include publication costs for a modest number of papers (calculated by us). We base costs largely on OA charges expecting no further page charges (e.g. Journal of Cell Science) but we could blow our entire 3 year budget in one go by publishing in Cell Reports or similar!
Whatever these policies emerge as they will need to be absolutely explicit. I don’t think it fair to expect individual researchers to plough through these license agreements and retain a good understanding of what is and isn’t compliant.
Perhaps RCUK will simply produce a list of “acceptable” journals/publishers without all the caveats. SHERPA/ROMEO for example doesn’t seem to address the -NC component… yet.
This arrangement sounds much like the one NIH has with medical researchers…except there is a 12 month moving-wall…JSTOR users also experience a similar moving-wall.
http://publicaccess.nih.gov/
http://about.jstor.org/content-collections/moving-wall
Thanks for the links Clayton. Out of interest — are there any significant journals that don’t comply with the NIH requirements?
The cost depends on whether the government capitulates to publishers who want to charge something like ten times the actual cost of web-only publishing (I fear they might).
As you say, success in achieving open access will depend, in part, on breaking free of the impact factor nonsense. I can think of nothing that would benefit science more than doing that. It would avoid the sort of nonsense that we’ve recently seen at Queen Mary, University of London (who in their right mind would want to work there?)
I agree though I fear putting the publication funds in the hands of universities to allocate may even exacerbate the publishing bias towards high IF journals, which of course gives their publishers the whip hand. As a community, we really do need to cure ourselves of this addiction. Even Philip Campbell thinks its unhealthy, I gather.
You ‘gather’? It’s a little stronger than that: Nature has been saying this publicly for at least 7 years (editorial: http://www.nature.com/nature/journal/v435/n7045/full/4351003b.html ) and probably longer.
I was referring to a tweet that I remembered appearing last week but couldn’t find. Hence the circumspection. 😉
But for sure, it’s an old line (I cited it in this post) but Nature still takes out full page adverts when new impact factors come out. Why is that?
Might Nature be persuaded also to publish (in the same advert) the distribution from which the IF is calculated to show the full-range of citation stats for all papers from the previous 2 years? howp about it, in the interests of rigour? Of course such a distribution would show why it is inadequate to fixate on an inappropriately calculated arithmetic mean.
But I don’t blame journals for acting in their own best interests; I agree with Philip Campbell that the responsibility for the mis-use of IFs lies with researchers (and university research ‘managers’).
Why aren’t IFs calculated properly? For instance, you could throw out the top and bottom 25% of the data points and calculate a more realistic mean. This kind of practices are widely used in setting bank interest rates, scoring athletes, etc.
It will be interesting to see if an IF calculated this way will be closer to the current IF values for top journals, or pehaps the “smaller” lower IF journals. If someone has done these analyses I will be really interested to see the data.
I don’t think it matters how IFs are calculated (though your suggestion may result in a more accurate assessment of ‘average’ impact). The problem is that they are applied to individuals and papers when they should only be applied to determine the relative merits of journals. Journal level metrics, we have discovered, lead to bad habits. We need article level metrics to come to the fore to stop people chasing fool’s gold.
However they are calculated, IFs are unsuitable for assessing the worth of individuals. Everyone (apart from Queen Mary) has known that for years.
But I like the idea put forward by Elton Zeqiraj because it would narrow considerably the gap between journals. At the very least the median number of citations should be used, not the mean. The median is far lower than the mean. It’s statistically illiterate to use the mean to characterise the highly skewed distribution of citation numbers.
Anything that helps to destroy the hegemony of the "top" five journals would help. It is the obsession for publishing in Nature that is the biggest obstacle to open access.
I like the idea of branding anyone who crows about IFs as ‘statistically illiterate’. Time to go on the offensive…! BTW – do you know if anyone has done an analysis of journal citations using medians rather than means?
I see two potential problems with the gold access model that will most likely result in academics paying more in the short and medium term. The first one is the time that will take libraries to break free from subscription charges. How much will it cost us to read content published before open access? Similarly, how long will it take the rest of the world to make the same commitment to open access? It will not work if only british scientist make their articles publicly available. The second one is, who will set the price for APC and will journals increase their page charges?
We appear to be so dependent on the current publishing system because of impact factors, that journals can pretty much charge anything they want. How did we allow this to happen? Personally, I favour the green access model with current academic publishers playing a smaller role. Eventually everything will stabilise.
The excess transition cost is a concern that was recognised in the Finch Report. The trouble is, much depends on what happens elsewhere in the world, though there are already encouraging noises from the US and the EU (as noted in my post), and no-one can predict how long that will last.
And again, IFs remain troublesome. You can expect publishers to play them up, since for them it is part of ‘added value’, but as a measure of individual achievement IFs are woefully inadequate (see David C’s comment above).
I hope at least that these recent announcements will continue to stir the pot of discussion on open access; it’s already been bubbling since the Elsevier boycott but has yet to engage all hearts and minds.
The market.
And that alone is a huge reason why Gold OA is a better model than subscriptions.
Stephen wrote:
I too wish that the otherwise admirable RCUK document had been more explicit on this point. Is there someone we can contact for clarification?
I assume that “I think this may mean CC-BY-NC” means that you think the NC clause is prohibited here? The wording is not quite clear.