University of California
University of California (Photo credit: Wikipedia)

The new open access policy recently announced by the University of California (UC) seems to have engendered a lot of confusion, at least in the press. Physics Today even has an article on the confusion, noting that the press coverage is largely focused on the OA political debate rather than what UC is actually doing. In that regard I think that two features of the new policy are worth noting. First it is as much a project as a policy and second that in an important sense the publishers are calling the shots.

The project aspect is reflected in the policy’s implementation timeline, which extends until June 2015, or roughly two years from now. The project only begins on November 1, 2013, when faculty at three UC campuses must start participating. After that various things happen, including two policy reviews, adding the rest of the campuses, and most interestingly, an attempt at automated manuscript harvesting. It sounds like this harvesting will be from the journal websites but I could find no further information on this aspect of the project.

Also interesting is the lack of any specific compliance mechanism, nor any mention of penalties for non-compliance. There is an extensive FAQ page but I can find no mention of compliance. Instead the FAQ Q&A wanders off into a discussion of OA per se. But the Policy Statement itself ends with this intriguing sentence:

“The Faculty calls upon the Academic Senate and the University of California to develop and monitor mechanisms that would render implementation and compliance with the policy as convenient for the Faculty as possible.”

This suggests that that the issue of compliance has been left to a later date, perhaps after the first review, in May 2014.

The publisher’s role is only apparent when one looks at the draft entry form which authors use to specify the status of their articles, including asking for embargoes and waivers. Most of the UC documentation and much of the press coverage makes it sound like the author can specify an embargo period or get a waiver for their article as they like. But the entry form defines embargos and waivers as follows:

Embargo: Request verification that your publication will not be displayed in UC’s eScholarship repository until your publisher’s required embargo period has expired.”

Waiver: If required by your publisher, request a waiver to opt-out of the Open Access license for a single publication.”

So basically the UC policy works like this. If the publisher allows it, then the article will be posted by the repository immediately. If there is a publisher-specified embargo period then it will be honored. If there is no such period then the article will not be posted by the repository. In short the UC repository is simply doing whatever the publisher allows. How this is a political victory for OA is beyond me. Perhaps the confusion stems from the fact that none of this is mentioned in the UC press release or the FAQ. Or maybe I too have missed something in this complex project. Speaking of complexity it is worth noting that the authors have a lot to do for each article, as does the library. Given that in many cases that article will be available free from the journal one wonders if it is worth it?

I did enjoy this FAQ title:

My publisher is offering me Open Access for $(absurd amount). Should I pay for this?

It suggests that UC is not a big fan of Gold OA.


39 Thoughts on "Some University of California Open Access Policy Confusions"

Not sure what you mean, Albert. The authors have to abide by the publisher’s rules, which puts the publishers in control of the posting of their articles.

The authors choose the publishers, not the other way ’round. Publishers have competed for authors ever since I can remember.

It is not that simple, Albert. Journals have high rejection rates so in an important sense they choose the authors.

But a journal can’t reject an article unless it’s submitted. So yes, it’s true that both journals and authors engage in some degree of mutual selection — but it’s also true (and significant) that the process doesn’t begin until the author makes a choice between competing journals.

SOME journals have rejection rates, most likely in the Humanities and social sciences. The others, particularly in the sciences, exhibit incredible growth in page counts and prices that rise ever faster than tuitions and fees.

Reputable journals have rejection rates. And while I have seen no comprehensive studies, several of the science journals I work with have both increased their page counts AND increased their rejection rates. There is simply more research being done than ever before, more papers submitted, and more quality work available.

And since you mention the rise in tuition rates:

Moody’s released statistics showing tuition and fees rising 300 percent versus the Consumer Price Index between 1990 and 2011.

Historically, the numbers of journal articles has doubled every 15 years or so for more than 300 years. Prices, adjusted for inflation, follow numbers of articles and pages. More interestingly, Federal spending on academic R&D, adjusted for inflation, has also doubled every 15 years or so since Vannever Bush set the policy of government support for research.

But this doesn’t address David’s point, which is that growth in page counts doesn’t imply less discrimination. Saying that science journals don’t have rejection rates because they show constantly-increasing page counts is, in an environment of explosive research growth, a non-sequitur.

It would seem that UC is saying, “we encourage OA as long as it doesn’t cost us anything.” That’s the poison pill that assures the failure of this project. Refusing to recognize that there are resource requirements to publishing, even in the digital era, assures that this project will not end well. That is, unless this project doesn’t invoke new and more efficient ways to address those costs. Shifting costs as in Gold OA really doesn’t address the inefficiencies that make OA so difficult to implement fairly. They seem to understand that at least.

“we encourage OA as long as it doesn’t cost us anything.”

I think this nicely sums up a lot of the NIMBY ( thinking that has gone into university OA policies. As noted in a recent post, the universities’ commitment to the open distribution of research results stops just at the point where it might reduce technology transfer revenue streams:

Thank you. I keep reading this policy over and over and I still can’t quite figure it out. I do find the tone incredibly hostile toward publishers, as you noted above, and so I start to get defensive about it. Anyway, all the licensing stuff confuses me too. If they deposit a paper and grant UC a CC-BY or CC-BY-NC license, can the author then sign a traditional copyright transfer form to the publisher? It also bothers me that I can’t find the CTA amendment they want authors to send us. You need to have certain article metadata to generate the form. Guess I’ll have to wait until one shows up.

Good point Angela! UC gets a license to post the article in the archive but I have no idea what the licensing terms of the posted article are. Perhaps someone can tell us, or maybe that has not been decided, or maybe there are none, which would be very interesting indeed (think Napster).

When faculty post in eScholarship, they will be asked to choose one of the Creative Commons licenses–most will choose either CC-by or CC-NC. This is actually an important point, because many OA repositories don’t license articles this way, but simple post them with copyright intact for people to read. Any other uses–in class, in a reader, in data mining etc. are therefore not specified, or not allowed. The UC approach requires that the license terms be clear and that faculty authors are the person who determines what rights are allowed for any given article.

I think this raises some questions though–for example arXiv has a similar structure but points out the following caveat:

Note that if you intend to submit, or have submitted, your article to a journal then you should verify that the license you intend to select does not conflict with the journal license or copyright transfer agreement. Many journal agreements permit submission to arXiv with the non-exclusive license to distribute which arXiv has used since 2004. The Creative Commons Attribution license in particular, permits commercial reuse and thus conflicts with many journal agreements.

This requirement may result in more authors opting for waivers as the various licensing schemes and requirements may be limiting, and are certainly somewhat confusing. It’s also probably worth noting that when given a choice, researchers seem to prefer more restrictive licenses that give them more control over what others can do with their written work.

David, what ArXiv says is true *after* you have signed a publisher agreement. OA policies, however, are about reserving a non-exclusive right *before* you sign the agreement. All publishers will get a notice from UC about this reserved non-exclusive right, and faculty may choose to include a standard addendum alerting them. If publishers object to the terms, then they must ask the author to opt out completely, or embargo the article.

And absolutely many faculty authors will choose restrictive licenses. This is a good reason why publishers shouldn’t be too anxious and it is also consistent with academic freedom. I and others routinely urge colleagues to go with the CC-by route, but many academics value the integrity of their work above all.

I want to thank you for coming in on this conversation. It is good to have someone to ask these questions of. My question is how often you think this happens where a paper is written and not submitted to a journal. Why would an author write a paper and post it in the repository before submitting to a journal? This goes against what I hear from our researchers all the time. Also, if the publisher has a policy of not reviewing papers that are already readily available online, does the author who may have posted a paper in your system then need to have it yanked?

I believe in physics it is actually quite normal to submit your papers to ArXiv for informal peer review, get feedback and share it with colleagues, revise it, maybe go through a couple rounds of this and then submit it to a physics journal for formal publication. This has been going on since 1991. Thousands and thousands of papers have been submitted to ArXiv and it seems to work quite well in that field.

My understanding (and this is from a distance) is that use of arXiv varies tremendously among disciplines in physics. Some fields use it frequently, others not at all. And from what I’m told, if one thinks one’s paper is likely to be published in one of the absolute top journals (Nature, Science, etc) then it generally doesn’t go up on arXiv.

I think this program costs UC quite a lot. The library has to look up each embargoed article’s embargo period and then program it correctly. Under the US OSTP OA guidance we may see what I call “green hybrid” journals, which means different embargo periods for different articles, depending on which agency funded the work. If so then getting the embargo period right will take some serious work on the library’s part.

Baumol’s Cost Disease strikes again. UC staff making work for UC staff. Max Weber would be gratified.

I do not see OA repositories as “make work” projects. Some OA advocates claim great benefits form such repositories.

One example: at my library, the IR has become the new home for the “stacks copies” of theses and dissertations — resulting in significant savings (printing, binding, processing, shelving), hugely enhanced discoverability, and far greater access to the papers’ content. (AHA will be pleased to know that our authors have complete freedom to embargo their dissertations for as long as they wish.) If it did nothing else, our IR would be a worthwhile investment for this reason.

David, this is a very keen reading of the policy and its difficulties. I wish there were 10 of you for every faculty member who objected to sky-falling nonsense about communism (I kid you not). But there are definitely some unknowns in the policy and its implementation. They are not, however, unknown unknowns and nor, I believe, are they specific to the UC policy, but would plague any University who passes such policies.

The first thing to say though is that the main point of these policies is simply to switch the default: rather than every individual having to negotiate individually about every article, this is a collective and forceful statement about what we think the *default* should be. It reserves rights for faculty if they want them, and if they don’t, it’s their choice. That includes the choice to publish with someone who forces them to give up those rights for the privilege of publishing in their journal.

But some clarifications:

1) The harvesting component is indeed a great experiment, and one that will cost some money. I, and a lot of other faculty members, are putting faith in the California Digital Library that they can pull it off. It’s one reason for the timeline and the stepwise implementation– to monitor it and report on its success. Hopefully CDL and UC will do that publicly so lots of people can learn from it.

2) The Embargo/Waiver distinction is confusing, but you are not quite right about it I think. Faculty can either choose to waive permanently (not grant the license to a particular article) or embargo application of the license for a particular period (and this is generally what happens with publishers force faculty to opt out). But if no period is specified, that doesn’t necessarily amount to a permanent waiver. It could mean that publishers are asking for permanent opt out (but I haven’t heard of cases of this); more likely it means that publishers are leaving it deliberately vague in an attempt to manipulate authors into opting out. A related issue is that several large publishers *threaten* that faculty must obtain a waiver, but they don’t *require* the waiver to be included with the publishing agreement. It’s not clear what they think they are doing, but it amounts to a form of bullying, because authors don’t know what to do. Some will waive out of fear, others will take the risk and post anyways, unclear on whether they are breaking a contract or law. Needless to say, we’d like to see that stop. The larger point is the one I started with: the default is switched: rather than authors begging for open access, publishers now have to beg for exclusive rights from authors.

3) Compliance. Ah, compliance. Using the words ‘compliance’ and ‘faculty member’ in the same sentence usually elicits something between a snigger and a hearty dismissive grunt. I don’t know if you have worked in a university, but there is nothing so impossible to imagine as a policy governing academic research that requires faculty to do anything that isn’t already required by state or federal law. There is a lot of detail to how the UC system works that is not really worth going into detail here, but the short answer is: you are right, the compliance mechanism has been “left to a later date.” Most of our discussions, however, have been about carrots and not sticks. Carrots are cheaper and better, and don’t usually elicit the same dismissive laughter. We’ll see how creative CDL and UC can be about that.

4) Lastly, “UC is not a big fan of Gold OA.” Spot on. I was actually quite surprised to hear the amount of faculty concern, anger and downright hostility to the idea– from across the sciences and humanities. Much of this is probably driven by the $(absurd amount) which is the only point of contact many faculty have with the idea. And it is almost always an absurd amount, especially if you are a humanities scholar with no research account. So if that’s what “Gold OA” means (i.e. expensive Hybrid Gold OA), then the UC system is definitely opposed. But if Gold OA means real experiments with funding models that include the university, the funders, foundations, and that are done with respect for what open access means, and the rights faculty want to retain, then there is interest, and probably support. Passing everything onto faculty members however, is a total non-starter.

“The larger point is the one I started with: the default is switched: rather than authors begging for open access, publishers now have to beg for exclusive rights from authors.”

Huh? Publishers get more content than they know what to do with and they don’t particularly care from whom the content comes–with few exceptions. If UC researchers want to post their work on the UC site exclusively, great for them. If they want a journal to supply peer review for validation, that’s a differenmt story.

Journals do not “supply peer review”. Academics do that, and they do it for free. What journals do is manage the peer review process, and that is valuable, but it is not as valuable as many publishers seem to think it is. Since I have run and edited journals I know that it involves a lot of emailing, and some good project management skills. I’ll pay for that service, but it shouldn’t be confused with the work of peer reviewing, which I consider to be more difficult and more time consuming.

But the point is not that authors will post their work instead of publishing in a journal– it’s that they reserve the right to do *both* (i.e. publish it in a journal AND make it freely available) unless a publisher forces them not to.

Thanks for the input Chris, very useful. For those who do not know Chris Kelty is the principal architect of the UC OA policy. He even has a video on the UC Policy website at:

Since you ask, I was on the faculty of Carnegie Mellon about 2.37 eons ago, but I got over it. You are quite right about not telling faculty what to do, which is what makes the UC flip-flop so interesting. Your default condition is universal faculty action and we wait to see how that goes. (At the moment I have a cat bumping my legs and purring but I would never be so foolish as to demand that it do so.)

I am not comfortable with you use of the terms begging and bullying, either way.

My real question is why anyone would want to search on just those articles authored by UC faculty, unless they happened to be studying UC? Surely the journal is a better source of scientific information.

Begging, bullying. You guys are so *sensitive* here 🙂 I’m kidding. But seriously, I think that a lot of what goes on in the course of publishing an article is not about personal feelings, it’s about frustration with the impersonal nature of it–forms to fill out, legalese, confusion, perceived lack of respect (going both ways). Almost all publishers do the right thing by faculty and vice versa. But that doesn’t mean there are not some very powerful, and very strategic, decisions being made by a handful of influential publishers.

As for why would anyone search a UC repository–that’s a very good question and a big concern. I don’t think OA advocates generally want a world with a thousand different repositories–certainly we all abhor the idea that to find a bunch of research we would have to independently search each of these repositories. But that’s not the technological reality–all these repositories are very much concerned with how to federate data and publications, expose statistics about citations and downloads and related content, make work visible to Google Scholar and other databases, etc. It’s a hard problem to be sure, but it is not the case that UC or CDL expect the work will only be available if you look at eScholarship.

The other thing to say about this is that repositories are good places for people who want to repurpose content– such as — so unlike journals, who have a stake in driving ALL traffic to their site, IRs can be comfortable being resources for other sites. This makes for a richer ecology in the end.

I’d like to chime in here on the subject of the accessibility of faculty research within institutional repositories. As Chris Kelty correctly points out, any model that required researchers to identify particular repositories and then search within them to find relevant publications would be deeply inefficient and unsustainable. It is our intention and has been for the past decade (and here I speak at the UC eScholarship repository manager) to make sure that all of the articles deposited by faculty are full-text indexed by Google. We also take pains to expose article metadata for indexing early and often. The vast majority of readers of the UC-authored research in eScholarship come via a “known-title” or keyword search within Google. Though we support a robust search mechanism within the repository, we by no means assume that researchers are starting their searches there.

I guess the question is this. If the article is freely available in the journal what is the advantage to bringing people into the repository instead?

In terms of access, there is no advantage. And the UC Policy recognizes that, enabling faculty to provide metadata and a link to their publication if it is freely available elsewhere. In terms of archiving, there is a strong argument for enabling faculty to deposit their publications in the institutional repository, thereby ensuring the future preservation and accessibility of that publication, regardless of the fate (or unanticipated changes in access policies) of particular publishers.

Thanks Catherine, however how I do not see where that metadata is provided by the faculty, but perhaps it will become clearer in November. Or is it something the library will provide?

Nor would I call it enabling, but rather requiring, which is quite different. But if the purpose of this policy is simply to guard against disappearing publishers, then it seems excessive, to say the least.

I’m replying here to David, below. Thread won’t allow a direct reply to his comment. The deposit workflow will change substantially by the Nov 1 deadline. Please check back then. We’re also developed a fully integrated harvesting mechanism that will grab metadata from various sources (including indexes, publishers where available, faculty profile systems, etc.). So we’ll have ample opportunity to collect that data and verify with the faculty. The policy is obviously not simply a guard against disappearing publishers – but it’s certainly an added benefit to be able to ensure future access to UC Faculty publications.

Catherine, are you requiring your faculty with papers in your repository to also post errata, corrections and retractions? What about Discussions and Closures that relate to the paper? If they do are you connecting the metadata? In other words, if a paper is flawed and corrected or retracted, would anyone looking at the paper in your repository know that? My concern with repositories, from talking with librarians, is that there is absolutely no effort put in to making sure that the literature is correct and that the only people adequately creating this paper trail are publishers.

Consider also the steady movement of faculty from institution to institution. If Professor X posts a paper in the repository, then moves far away to a new position at non-affiliated University Y, and then posts a correction or a retraction to the paper, it’s unlikely that Professor X is going to get back in touch with the UC librarians to update the repository from his former employer.

Angela – That’s a great question. We certainly have no desire to muddy the waters of the academic record. We’re in the midst of exploring vendor options for a harvesting solution and have called out the importance of supporting the versioning of publications. A lot will depend on the metadata available from the publisher/indexes – but our aim is to keep the repository up to date with the most correct version of each publication. We will also support manual deposit of new versions, should the faculty wish to alert us to any changes. And finally, we will provide, for each article, a link to the version of record with the publisher. It may not be a perfect solution, but we will make every effort to get as close as possible.

One of my basic problems with this scheme is that it makes the teacher a middle person between the university and the publisher, in the complex world of copyrights. Does the UC library have a staff of IP lawyers to advise the faculty on their choices? Or are the faculty just winging it? The burden of IP learning is potentially quite large here, easily hundreds of thousands of hours, not to mention the bad decisions. What is the benefit?

But this is already the way things are structured: the university currently doesn’t do anything for faculty vis a vis publishing. In fact, I have heard University Counsel say explicitly that they will step up to defend faculty on any issue *except* copyright, because our University policy, like many universities, is to leave copyright ownership with the faculty member. So without an OA policy, faculty members must negotiate on their own with publishers, and are legally and personally responsible for understanding the details. With an OA policy, faculty members have the made the collective choice to reserve their rights–unless they choose as individuals to give up those rights. Again, it’s about changing the default.

The larger point–that IP law and policies are complex and risky is true a priori. It is not something that is caused by having an open access policy.

Comments are closed.