Let’s not get too excited about the new UC open access policy

It was announced today that systemwide Academic Senate representing the 10 campuses of the University of California system had passed an “open access” policy.

The policy will work like this. Before assigning copyright to publishers, all UC faculty will grant the university a non-exclusive license to make the works freely available, provide the university with a copy of the work, and select a creative commons license under which is will be made freely available in UC’s eScholarship archive.

A lot of work went into passing this, and its backers – especially UCLA’s Chris Kielty – are to be commended for the cat herding process required to get it though UC’s faculty governance process.

I’m already seeing lots of people celebrating this step as a great advance for open access. But color me skeptical. This policy has a major, major hole – an optional faculty opt-out. This is there because enough faculty wanted the right to publish their works in ways that were incompatible with the policy that the policy would not have passed without the provision.

Unfortunately, this means that the policy is completely toothless. It provides a ready means for people to make their works available – which is great. And having the default be open is great. But nobody is compelled to do it in any meaningful way – therefore it is little more than a voluntary system.

More importantly, the opt-out provides journals with a way of ensuring that works published in their journals are not subject to the policy. At UCSF and MIT and other places, many large publishers, especially in biomedicine, are requiring that authors at institutions with policies like the UC policy opt-out of the system as a condition of publishing. At MIT, these publishers include AAAS, Nature, PNAS, Elsevier and many others.

We can expect more and more publishers to demand opt-outs as the number of institutions with open/public access policies grows. In the early days of such “green” open access, publishers were pretty open about allowing authors to post manuscript versions of their papers in university archives. They were open because there was no cost to them. Nobody was going to cancel a subscription because they could get a tiny fraction of the articles in a journal for free somewhere on the internet.

However, as more universities – especially big ones like UC – move towards institutional archiving policies, an increasing fraction of the papers published in subscription journals could end up in archives – which WOULD threaten their business models. So, of course (and as I and others predicted a decade ago), subscription publishers are now doing their best to prevent these articles from becoming available.

So long as the incentives in academia push people to publish in journals of high prestige, authors are going to do whatever the journal wants with respect to voluntary policies at their universities. And so, we’re really back to where we were before. Faculty can make their work freely available if they want to – and now have an extra way to do it. But if they don’t want to, they don’t have to.

The only way this is going to change is if universities create mandatory open access policies – with no opt-outs or exceptions. But this would likely require actions from university administrators who have, for decades, completely ignored this issue.

So don’t get me wrong. I’m happy the faculty senate at UC did something, and I think the eScholarship repository will likely become an important source of scholarly papers in many fields, and the use of CC licenses is great. And maybe the opt out will be eliminated as the policy is reviewed (I doubt it). But, because of the opt out, this is a largely symbolic gesture – a minor event in the history of open access, not the watershed event that some people are making it out to be.

This entry was posted in open access, public access, University of California. Bookmark the permalink. Both comments and trackbacks are currently closed.


  1. Posted August 3, 2013 at 1:47 pm | Permalink

    Simple Way To Make UC OA Mandate Work

    Aside from the default copyright-reservation mandate with opt-out, always add an immediate-deposit clause without opt-out.

    The deposit need not be immediately made OA, but it needs to be deposited in the institutional repository immediately upon acceptance for publication. While access to the deposit is embargoed, the repository can implement the eprint-request Button with which users can request and authors can provide the sprint with one click each.

  2. Christopher Kelty
    Posted August 3, 2013 at 3:28 pm | Permalink


    An open access policy without an opt-out is impossible. It is also a totally uninteresting argument.

    You know, when I argue with my father-in-law, who is an astronomer, he will sometimes explain to me that something is impossible. By this he means in his endearing but backhandedly condescending way that I don’t understand enough about physics or chemistry or engineering to know that something I’ve proposed violates various laws or regularities in nature. I never question this, because as a diverted scientist (I won’t say failed–other things were more interesting), I respect that I might not fully get it.

    You seem not to have this respect when people explain that it is impossible to have a policy without an opt out clause. And I mean the word the way my father-in-law does. I do not mean that there is some political concession at work, or that in order to get something done, principles have to be compromised, or that it is possible but really, really hard. I mean it is *impossible*.

    What’s more, the opt out was not included for any of the reasons you suggest. It was included because it *must be there* for any policy of this size and scope. Anything else requires a dictatorship or a gazillion dollars for lawyers. We are not Saudi Arabia, we’re the University of California. It’s a subtle but important difference 🙂

    I’m also sad that this is the only argument you can mount, because it is also boring. It’s boring because there is a much easier problem to solve: how do you get the 5% of people who *do* opt out to stop doing it (because we know that the number is around 5%)? What are the incentives, the deals, the tricks, the promises, the alternatives that would make it undeniably more advantageous for any individual faculty member to deposit an article (perhaps with a CC-by license, and with as short an embargo as possible) than to ask for a waiver? Rather than trying to concoct a draconian policy that, like exceeding the speed of light, is impossible, why not propose a path towards encouraging that 5% to do something other than opt out? Then we might get somewhere, and you might be able to stop harping on this point once and for all.

    • Posted August 7, 2013 at 11:46 am | Permalink

      Why is it “impossible” for the university to require that the output of its faculty be publicly available? There are plenty of precedents for the university limiting the freedom of its faculty to operate in the marketplace. They require that I sign over any patent rights I have in my work. They require that they own the copyright on software I produce.

      And as for the opt-out percentages – you know the 5% is a crap figure. The proper way to measure opt-outs is the number of actual opt-outs plus the number of people who just ignore the whole thing and don’t even bother to ask for a waiver. I suspect (and my quick and dirty analyses of Harvard’s archive support) the notion that the real problem is non-compliance.

      • Christopher Kelty
        Posted August 8, 2013 at 3:55 pm | Permalink

        Can I just say, since people seem bamboozled by your argument here, that there are only two options: opt-in or opt out. There is no third option in open access policies passed by faculty senates. The triumph of any open access policy is *switching the default* from opt in to opt out–switching from powerless individuals doing everything on their own to the collective power of a prior license grant. That’s all. There is no such thing as a policy without an opt out, unless you mean no policy at all.

        Why? Let’s take a simple example. Anyone who has to use copyrighted materials in their work–humanities or sciences– has to negotiate licenses for that work. Let’s assume they do so without fully thinking about the implications or the kind of license they need, and the don’t negotiate rights to make the work openly available. I’m sure you would call them idiots, and we can berate them for failing to stand up for themselves, but it will happen, with 100% certainty, often. If we had a policy that had no opt-out of any kind, the university would be forcing those people to violate a contract or copyright law by forcing them to make their work open access. Lawsuits, grievances and ill will ensue. With 8000 faculty, even if only 1% do this that’s 80 lawsuits a year, and a whole lot of anti-open access sentiment. And that’s just one example.

        So you need an opt-out for those people. I suppose you could have a very *difficult* opt-out procedure, in which faculty have to put together a big file documenting why they will opt out, submit it to an office that does not yet exist at UC, and then let some bureaucrat decide who gets to opt out and who doesn’t. That way you solve the lawsuit problem, but you make the ill-will problem a thousand times worse and a hundred times more expensive. But some publishers would game even a complicated opt-out policy, that too is inevitable. And that’s just one example.

        And to be clear the 5% is not crap– it is a very clear and reliable number. But you are right, what it does not capture is how many people *deposit* their articles. Perhaps what you think we need is a required *deposit*? Well guess what, the policy has a requirement that people deposit, and it requires that of everyone who does not opt out (i.e. 5% of people). The fact that people shirk their responsibilities is hardly something you can correct with a policy, and it is not the publishers who are causing some percentage of the other 95% to not deposit. If you have a good idea to improve the rate of deposit, lot’s of people want to hear it. Deposit is the issue, not opt-out.

        • Posted August 8, 2013 at 4:53 pm | Permalink

          Why do you say there’s no such thing as a policy with no opt-out?

          I understand that faculty seemed reluctant to pass one, and so an opt out was included. But that is not the same as saying a provision with no opt out is can’t exist. It can. The university has plenty of policies that we can not opt out of – he most relevant being the intellectual property ones. I don’t have the right to say “Nah, I’d rather not sign over my IP to the university, I’d rather sell it to a biotech company and pocket all the money myself.”

          All it would take to have a policy that required faculty to make ALL of their works freely available is the will on the part of the administration and faculty to do it. This is what I am criticizing – the fact that there is this huge disconnect between the rhetoric around this policy “UC requires OA” or “UC embraces OA” and the reality, which is that the opt out provision exists because the faculty and university as a whole do NOT embrace OA, and I am not going to let them get away with pretending they do.

          And we disagree about the significance of the opt out. 5% may be a “hard” number, but I did a back of the envelope calculation based on information about publisher policies at MIT, which show that at least 50% of articles in the sciences are published in journals that require opt outs. I don’t know what the number is in the humanities, but I suspect it’s similar. So the 5% number is just a mirage – the problem is way bigger, and is only going to get worse.

          I’ve said it before and I’ll say it again. This policy is a start. But it’s a mistake to pat ourselves on the back and pretend that we’ve solved the access problem when this policy BY DESIGN does nothing to attack the reasons that we have the access problem in the first place – the billions of dollars of subsidies from universities given to publishers who restrict access to their contents, and the prestige economy in academia that has turned many academics into vassals of high profile journals in their field.

          • Christopher Kelty
            Posted August 8, 2013 at 6:37 pm | Permalink

            So perhaps there is more to explain here. The policy we passed is a Faculty Senate Policy– not yet a University administrative policy. The Faculty recommended this policy to UCOP and they will turn it into a university policy.

            Despite the fact that we have a strong faculty governance policy, we can’t change administrative policy, we can only recommend changes to the Academic Personnel Manual, and then it is up to the University to determine what those changes ultimately will look like.

            The only other thing we have control over, by virtue of the 1992 UC Copyright policy, is that we own our copyrights in the work we produce (and that should cover software as well, per your claim above, so something fishy is going on), ergo, we can pass a policy agreeing to license our work, which is what we’ve done.

            Now, it is true that the University could pass a policy that would a) expropriate our copyrights and b) force us to make our work available. I can’t really imagine the world in which they would have the *legitimate* power to do so, but they could expend the last drop of political capital they have doing so. But they would never have faculty support in doing so, regardless of how unanimously everyone supported Open Access– the anger at draconian, unjust policy making will outweigh ideological principle every time.

            The situation with the patent policy is a case in point– the university has in fact expropriated intellectual property under the theory that any patentable invention done with UC resources is something they control, but to do this, they have incurred massive suspicion, massive expense and with a few blockbuster exceptions (which is what keeps people chasing this mythical beast), have really done very little with what they have taken. Three years ago when UC was forced to change the patent policy in response to a court ruling, they then had to *force* every member of the university to sign the new policy. Ask around, I’m sure a few people are still angry about that. People might support more patents and beneficial UC funding in general, but they still HATE the UC patent policy. And as a result people DO opt out of the patent policy all the time– by gaming it. They conceal results, start companies on the side and move research into the public sector in order to keep the money for themselves– bad for everyone.

            The point is that there is not a world in which the political will you speak of could exist, and it is not because people don’t support open access, it’s because they value justice and fairness far more. This is, in my opinion a Good Thing. I obviously think we should have more open access too, and I think more people should know more about it and support it more– why else would I have spent three years trying to do this?

            We should have a better concept of “politically impossible” that allows us to respect the messy but good solution, in favor of the perfect solution that compromises everyone’s freedoms. We’ll never move forward otherwise, and 20 years from know we’ll still be trying to get every last paper in every last discipline into some perfect state of openness.

            And I think you are wrong about the calculation. The number is 5% of published papers– a number that is admittedly hard to estimate because no one keeps good data– not 5% of the deposited articles.

  3. Ian Holmes
    Posted August 5, 2013 at 6:17 pm | Permalink

    Michael, few have done more to improve access than you. It’s hugely encouraging to see your activism on the scene. And yet I disagree with some premises of your argument. As I’ve noted before, I think that attempts to impose “Open Access” from the top down are unworkable & misguided (hence the gaping loopholes in this policy, which as you observe, were necessary to pass it) and, furthermore, rhetorical attempts to turn Open Access publishing into a “moral” issue are extremely narrow & ultimately incorrect.

    In particular, blaming “the incentives in academia [that] push people to publish in journals of high prestige” misses the point. Forget CV points, tenure committees, and cash bonuses for CNS publications from the Chinese Academy of Sciences. Academics ultimately just want their work to reach the most people and have the highest impact. Once preprint archives are seen to be the place to go for the hottest stuff (as they are in physics, and as they inevitably will be in bio) then these other institutions will follow.

    The “prestige” that you complain about is precisely why CNS journals generally tend to have a more professional peer review process than the high-volume OA journals. (I’m not saying it’s perfect!) The real puzzle, to my mind, lies not in the incentives for *authors*, but rather in the incentives for *reviewers*. The reputation economy that underpins peer review (the reason reviewers respond to requests from editors) is thoroughly broken, especially at the vanity presses where editorial judgment is explicitly removed from the picture (which in turn makes reviewers AND editors feel like unrewarded drones).

    I am far more interested in the vibrant constellation of startups that are innovating and disrupting the mechanisms for peer review, than I am in top-down attempts to bolster support for OA journals. Ultimately, OA journals effectively take a step BACKWARDS in terms of the business model (going from “prestigious scientific proceedings” to “vanity presses”), with the dubious benefit of transferring the economic responsibility for access from readers’ libraries (supported by state taxes) to individual authors (supported by unreliable federal grants), and distorting the peer review market while doing so (because now there are just too many papers to review).

    In the meantime, I think it is highly detrimental to attempt to force authors to use publishers with a particular business model. A lot of interesting stuff happens on the fringes. Books, weird journals with quirky editors, etc. It is impossible to assert with any confidence that ALL such endeavors will gain an increased readership with the very conservative, incremental improvement on 20th-century publishing currently offered by a handful of OA publishers. I’m fine with big federal grant-awarding agencies like NIH imposing OA as a condition of funding, but I’m opposed to the idea of limiting academic freedom by requiring academics to publish in a limited collection of OA journals. Fortunately I think my opposition is entirely superfluous since (as this vote proves) it will be almost impossible to force academics to publish a particular way.

  4. Anonymous
    Posted November 20, 2013 at 3:45 pm | Permalink

    Hi Michael,

    I support in theory that the output of academia should be widely, easily, and cheaply – even freely – available to the public. But I have some problems with the logical underpinnings of your arguments, and those of other OA advocates. I’m hoping you can help me bridge the divide between your goals and your arguments in support of those goals.

    You say, for instance:

    “there’s no reason not to make publicly funded works immediately freely available” – See more at: http://www.michaeleisen.org/blog/?p=1361#sthash.B2xtaJnk.dpuf

    “The government (both state and Federal) should simply declare that all of their grantees need to make any publications immediately freely available, period – no exceptions, no questions asked.” – See more at: http://www.michaeleisen.org/blog/?p=1361#sthash.B2xtaJnk.dpuf

    Of course, the public funds all sorts of things. I’m a UC grad, and I took on pretty considerable debt to get my degree from UC (though thankfully, not as much of a debt burden as current UC students are taking on). Assuming you teach courses (or if not, consider your teaching colleagues), could I walk in and take courses without paying tuition? Because UC of course gets significant public funding. In short, since taxpayers fund UC, shouldn’t a UC education be free? What is the “reason not to make publicly funded [education] immediately freely available?” That’s what taxpayers would prefer, and it’s hard to argue that wouldn’t be a public good.

    And of course there are many more examples of taxpayer funded things that people have to pay to use – toll bridges, the post office, state parks, etc. etc.

    In addition to the moral basis for open access, of course there is a strong business case to be made for open access science publishing, and this business case sure makes the moral high ground easy to take: PLOS has made a lot of content in the sciences available to end users at no charge, but as you know that’s not to say that someone isn’t paying for it, and in fact PLOS is doing quite well financially. But how does that translate outside of the sciences, where there is still (meager) public support? Who’s going to pay those author processing charges? PLOS is not getting into open access humanities and social sciences, for good reason – there’s no money in it. So why then should we expect AHA to jump in where wise men fear to tread?

    At a minimum, if it is a moral imperative that scholarly content be open and freely available to all, shouldn’t we be requiring that taxpayer money funding research is also explicitly funding publication – peer review (e.g., Scholar One Manuscript fees), copy editing, HTML conversion, marketing, (author royalties?), etc. etc?

14 Trackbacks