The speed that things are happening at the moment is astonishing.

Whenever we talk about the economics of open access — when I argue that it costs the community eight times as much to publish a paywalled article with Elsevier as it does to publish it as open access with PLoS ONE — I always hear the same argument in response.  And it’s a good argument.  It goes like this:

Yes, the total cost to libraries around the world of an Elsevier article may be eight times the cost to the author of publishing an open-access article that is free to read.  But you can only expect to save that money if libraries cancel their Elsevier subscriptions and plough that money into funding open-access publications instead.  And no library will ever do that, because the researchers that they serve need the subscriptions.

Well, it turns out — somewhat to my own surprise, I’ll admit — that libraries will cancel their Elsevier subscriptions.  The Department of Mathematics at the Technical University of Munich has just voted to do exactly that:

Because of unsustainable subscription prices and conditions, the board of directors of the mathematics department has voted to cancel all of its subscriptions to Elsevier journals by 2013.

So what does this mean?  A lot of things.

1. This is no idle far-in-the-future threat: 2013 is only one year away!  So this is an actual policy.  Something that they’re going to do.

2. Universities are not messing about.  When Harvard say they can’t afford subscriptions, they probably mean it — it’s not just a negotiating tactic.

3. Where one university department leads, others will probably follow.  Maybe initially it will be mostly maths departments in other universities; maybe it will be other departments of the Technical University of Munich; maybe it will be all of Harvard.

4. So far, this announcement is only about cancelling subscriptions and says nothing about open access.  If that’s all they do, it will be a mere cost-saving exercise and a missed opportunity.  To be truly transformational, the department needs to channel a significant chunk of its subscription savings into funding Gold OA publications.

5. Publishers who are paying attention will surely start to realise that they have pushed their exploitative prices too far, and that they don’t hold libraries in a steely grip any more.  I wonder how this will play into investment advice regarding Elsevier?

This isn’t the kind of problem that can be fixed by hiring a PR person.  I’ve argued this before, but if Elsevier are going to survive, they’ll need to be much clearer in the their communications, eliminate practices that alienate authors, and ultimately change their business model entirely.

For more on the Munich development (interesting more for the comments they may generate than for carrying much additional information):

• Google+ page by John Baez
• Blog post by Tim Gowers

These are worrying days for barrier-based publishers.  In the few days since I posted part 2 of this series, we have yet another major development in the Open Access world: UK Science Minister David Willetts’ announcement that “we will make publicly funded research accessible free of charge to readers”.  Exactly what form this will take is not yet clear, but the signs seem to point to an FRPAA-like universal Green-OA mandate for all research funded by the government.

What does this mean for Elsevier and their competitors?  It looks to me like a swift and sudden pulling of the rug from under their feet.  When Green OA becomes ubiqitous enough that researchers start looking at repositories as a first line of attack rather than the if-all-else-fails option, the need for journal subscriptions will fall away precipitously — and with it, publishers’ revenue streams.

In the face of that prospect, we now come to the third and final part of the How Elsevier Can Save Itself series.  (Final, that is, apart form the forthcoming coda on how other publishers should react.)  Previously we talked about easy measures and medium measures.  These ones are hard.  They require vision, courage, and determination.

1. Convert to Gold open access

It sounds radical; it sounds like a Won’t Ever Happen.  But, really, what is the alternative?  Subscription revenue is on the way out.  The signs are everywhere — not just in blogs and on Twitter, but in legislation in the US, in the UK and elsewhere.  Publishers reliant on subscription revenue must find another source of income or they will crash.

This may seem so obvious that it doesn’t need saying.  Unfortunately for Elsevier, a there’s a steady stream of no-we’re-not-sinking denialism issuing from venues such as the Scholarly Kitchen, and because that’s what publishers want to hear, they might fall into the trap of believing it because they want to.  (I have a vision of a world ten years hence when open access is all but universal; somewhere in the mid-west, Kent Anderson and David Crotty(*) have locked themselves into a underground bunker with a supply of tinned food and bottled water, and spend their days writing blog-posts to each other about how awesome the subscription-and-gatekeeper model is, and how there’s no need for change and how naive all these OA zealots are.)

Let’s assume for the moment that someone at Elsevier can see the oncoming revenue evaporation, and is making plans that go beyond “sell all my stock and get a job in another business”.  What revenue streams could be brought online to replace the lost subscriptions?

Well, what are publishers good at?  Managing the editorial process; proof-reading and copy-editing; page layout; consistent formatting across a brand; reliable large-scale web-hosting.  Those remain valuable services which publishers could charge for.

Guess what?  Charging authors for those services is Gold Open Access.  It’s the only business model that might still work when Green is pervasive.  (Only “might”?  Yes; see below.)

2. Support the FRPAA

Wait, what?  Didn’t I just say that Green-OA mandates like the FRPAA are going to Eat Elsevier’s Lunch?  Yes I did; and yes they are.  The thing is, there is nothing Elsevier can do to stop that.  The RWA debacle should have taught them that, if they doubted it before.

So what do you do when an unstoppable enemy is charging at you?  The judo move is to use that momentum to your own advantage.  Or, to switch metaphors, if a train is heading down the track towards you, it’s better to hop on board than to try to stop it.

I can see that this is going to be a hard lesson for a company that’s been used to having its own way since forever.  There’s a good chance that it believes in its own invincibility — its manifest destiny, if you will.  You know, like Blockbuster, Borders and the News of the World.

But if Elsevier has the agility, there’s an opportunity right now to radically reposition itself as the progressive, author-friendly publisher that listens to what academics want.  Given the appalling behaviour that’s got Elsevier into the current situation, it’ll take something dramatic to do that: supporting the FRPAA is the best chance they have, maybe the only chance.

3. Do it now

Here’s why I said that Gold OA is the only business model that might still work when Green is pervasive.  Once Green really takes hold, there’s a danger that scholars will lose interest in journals.  We’ve already seen this start to happen in maths and physics, where pre-prints on arXiv are ubiquitous.  In those fields, it’s still considered valuable to get a paper into a “good journal”; but what counts most is getting your work on the pre-print server where people can see and cite it.

This danger is all the more real because of the rise of Altmetrics.  A few years back when arXiv was establishing itself, journal impact factor was about the only way of getting a quick-and-dirty measurement of how “good” a paper was.  That’s changing; and as it changes, it’s destroying one of the main reasons for the continuing dominance of the journal system as the primary way to share science.  I don’t know how quickly Altmetrics are going to grow in influence, how well they can be designed to avoid gaming, or indeed whether the whole initiative will crash and burn.  But I do know that I’m more likely to tell people that our neck-posture paper has been cited 37 times than that it was in a journal with impact factor 1.949.  Because what a paper actually does is more important than where it hangs out.

So the spectre hanging over Elsevier isn’t just the destruction of their subscription model, but the possibility that by the time they get into gear and switch to Gold OA, it’ll all be over and the world will have lost interest in journals.  Certainly if university libraries start redirecting their saved subscription money towards in-house publishing efforts, the chances of that will increase dramatically.

This might not happen.  Inertia surrounding the journal system, and bean-counters’ ingrained love for impact factors, might be enough to keep journals alive and important, with Green OA and in-house publishing remaining second-tier.  But it might.  If I was Mister Elsevier, I wouldn’t be betting the farm on it.

What won’t help

A quick word on two things Elsevier could do that I have not mentioned in this series — because I don’t think they will make any difference in the long term:

• Reducing subscription prices.
• Unbundling subscriptions from their “Big Deals”.

Yes, I know that these moves would address two of the three complaints of the Cost of Knowledge boycott.  But in the end I think they’re red herrings.  All they do is make the current system of barriers marginally less unpalatable.  And that’s like patching holes in the Titanic, because the ship is going down.

The challenge is to re-tool for a world where barriers won’t make money.  Because that is not a strategy that can keep working.  It just isn’t.

Can they do it?

I said at the start that, while the changes outlined in earlier posts are no-brainers, or at least only moderately difficult, these ones will require vision, courage, and determination.  I think Elsevier has people with those qualities.  Unfortunately, it also has a massive amount of institutional inertia.  It’s hard to steer a supertanker.  Even when it’s heading straight for the rocks, there are going to be plenty of people whose attitude is: heck, we’re a supertanker — we don’t need to be worried about a bunch of stinkin’ rocks.

So I’d like to think they can do it.  But my head says probably not.  And as Clay Shirky points out, “You’re gonna miss us when we’re gone!” is not much of a business model.

(*) My mistake: I was thinking of David Wojick here — not David Crotty, who is much more reasonable. I’m not fixing this now because it would change the interpretation of some of the comments below.

It seems the world is conveniently arranging itself for the benefit of this occasional series.  Every time I am about to post an installment, something apposite happens out there.  Just as I was preparing part 0, Bernstein Research’s investment report Is Elsevier Heading for a Political Train-Wreck? came out; just before part 1, Elsevier decided that the solution to their problems was to hire a PR guy; and now, as I prepare part 2, America’s richest university says publicly that it can’t afford spiralling subscription fees any more.

It’s not my intention to gloat, but the recent cascade of events must surely be giving Elsevier and other big barrier-based publishers pause for thought.  A trickle of outrage has swiftly grown into a tide.  That may become a tidal wave.  Real change is needed to avert disaster.

Last time we looked at things that Elsevier should do right now, at no cost to itself: be explicit about terms of “sponsored articles” and non-sponsored articles; make it trivially easy to find sponsored articles; stop lying about copyright transfer; root out and destroy stupid conditions.

This time, we’ll go on to measures that probably will cost Elsevier something (though most likely not as much as they fear).

1. Change the “sponsored article” licence to CC-BY

Springer’s Open Choice is their equivalent of Elsevier’s “sponsored articles“.  Except it’s not, because Springer’s version is true Budapest-compliant open access.  Not only that, it’s trivially easy to tell that it’s the case, because the page simply says:

Copyright and Open Access License

If authors choose open access in the Springer Open Choice program, they will not be required to transfer their copyright; the copyright remains with the author.

NEW! Springer now permits commercial use for Open Choice, as all Open Choice articles are published under the Creative Commons Attribution (CC BY) license from January 16, 2012 onwards.

This is the right way to do it: by using a standard licence instead of writing their own terms, Springer tap into accumulated understanding of what CC BY means.  And by using that particular licence, they remove all doubt about what you are and aren’t able to do with such articles.  They’re freed to be used in any reasonable way, for the benefit of science.

If someone pays Elsevier $3000 to sponsor an article, it’s been paid for.  There is no legitimate reason for Elsevier to take the copyright under such circumstances, and no reasonable expectation that they should be able to make more money by charging for any use of such an article.  Worse, this kind of action makes Elsevier look mercenary and works directly against the partnering-with-authors impression that they otherwise try to nurture.

Special bonus: if Elsevier switch to the BOAI-compliant CC BY licence for such articles, they can stop weaseling around calling their scheme “sponsored articles”, and truthfully call the results open access.  And right now, Elsevier really, really needs to be able to say that it’s supporting open access.

What does Elsevier stand to lose by doing this?  A tiny revenue stream arising from people paying to actually use sponsored articles.  Of course I am not privy to Elsevier’s internal accounting, so I don’t know what that amount is, but I would be amazed if it’s more than 0.01% of their revenue and pretty surprised if it was more than 0.001%.  Not worth it.

2. Stop being obstructive about text-mining

When you read a published paper — say, because your library has a subscription that allows you to — you process the information that it contains and use that in your own work.  Later on, you publish that work, and what you publish has benefitted from the facts that were in the papers you read.  This is called research, or sometimes “standing on the shoulders of giants”, and it’s how science is made.

It’s how we’ve been doing science, in fact, since 1665.  But now it’s the 21st century and we have millions of papers to read instead of a handful.  Happily we have computers to do the reading for us, and in many fields they can extract important information such as use of taxonomic terms or chemical reactions.  This is called text-mining.  Just like when we read papers ourselves, text-mining extracts facts rather than their particular expression; and facts are not subject to copyright.  So it follows that anyone who has access to papers has the right to text-mine them.

Unfortunately, Elsevier (like many other publishers) has a history of being obstructive about mining.  Even when they set out to help — and to give credit, they are making an effort — it’s of the form “We are keen to arrange a teleconference with you all to discuss ways to enable text mining for academics at Cambridge University“.  Which is a waste of everyone’s time: first, we want to be doing science, not making conference calls; second, it’s dumb to limit this negotiation process to one text-mining project at a time.  (Even Elsevier’s own Director of Universal Access, has commented that she worries that “she might be overwhelmed by requests from others who also want text mining access“.

So this is dumb.  Since other publishers (with the obvious exceptions) are equally obstructive, Elsevier has an opportunity to lead the way.  They should demonstrate their commitment to open science by laying down this simple principle: if it’s been paid for (by subscription or OA), you can let computers as well as people read the information.

Again, what will this cost?  Maybe a little: perhaps Elsevier have managed to negotiate contracts with some customers where they pay extra for the privilege of mining, and they’d have to forego that revenue.  On the other hand, they’d be able to save a lot of money by not having to hold conference calls that involve six people at a time, including a vice president and three directors.  So overall this could even be a net financial win for Elsevier.

3. Dump the “you can self-archive unless mandated to” rule

Elsevier’s Article Posting Policies state:

Elsevier believes that individual authors should be able to distribute their AAMs for their personal voluntary needs and interests, e.g. posting to their websites or their institution’s repository, e-mailing to colleagues.

(AAMs are “accepted author manuscripts”, defined as “the author’s version of the manuscript of an article that has been accepted for publication and which may include any author-incorporated changes suggested through the processes of submission processing, peer review, and editor-author communications”.)

So far, so good: Elsevier supports self-archiving.  But it goes on:

Deposit in, or posting to, subject-oriented or centralized repositories (such as PubMed Central), or institutional repositories with systematic posting mandates is permitted only under specific agreements between Elsevier and the repository, agency or institution, and only consistent with the publisher’s policies concerning such repositories.

The page then clarifies that you’re not allowed to self-archive in response to:

Institutional, funding body or government manuscript posting policies or mandates that aim to aggregate and openly distribute the work by its researchers or funded researchers.

In other words, you’re allowed to self-archive so long as you’re not under a mandate to do so.  If you’re mandated to self-archive then you’re not allowed to.

This is not just obstructive, but absurd.  I’m not sure whether it’s more Kafka or Borges, but either way it makes Elsevier look dumb.  The analogy that springs to mind is a stroppy toddler who likes going to the playground, but arbitrarily refuses when a parent suggests it.

It’s been pointed out many times that there is no evidence that self-archiving (“Green open access”) harms publishers in any way.  Elsevier’s current policy is baseless, and makes them look both obstructive and absurd.  They should fix it.

One more quote from the policy page:

We routinely analyse and modify our policies to ensure we are responding to authors’ needs and concerns, and the concerns in general of the research and scholarly communities.

That is encouraging.  Let’s hope it’s true.

Probable cost to Elsevier if they fix this?  Hard to evaluate, but at the moment the sum of all available evidence (i.e. none whatsoever) doesn’t give any reason to think there will be a cost at all.

4. Withdraw opposition to the FRPAA

This one is surely a no-brainer.  Elsevier’s support for the RWA, both financial and rhetorical, catalysed a level of fury among researchers that’s like nothing they’ve seen before.  The Cost of Knowledge boycott, which has now surpassed ten thousand signatories and is going strong, is only the tip of the iceberg.  Elsevier recognised that supporting the RWA was an appalling tactical misstep, and publicly withdrew their support, resulting shortly thereafter in the RWA’s unlamented death.

So it was with something of a facepalm that, five minutes later, I saw Elsevier listed as one of the signatories on the Association of American Publishers’ letter campaigning against the FRPAA.

Seriously, Elsevier.  Don’t you get it?  All the researchers who hated the RWA also love the FRPAA.  By publicly opposing it, you instantly undid the good that your RWA withdrawal did.  Straight away, you cast yourself again as researchers’ enemy rather than partner.

Just stop it.

Cost to Elsevier: hard to calculate but probably close to zero.  Cost to Elsevier of continuing to oppose the FRPAA: totally wasting the salaries of whatever PR people they hire.  Because while you directly oppose what we want, no amount of PR will persuade us that you’re on our side.

5. Be open about subscription prices

No-one knows what Elsevier charge for journal subscriptions.

Even people within a university often don’t know what their own library is paying, because the librarians are forbidden to tell them.

It’s hardly surprising that this breeds an atmosphere of secrecy and distrust.

Is that what you want, Elsevier?  Huh?  Huh?

Again we come back to the central issue, which is trust.  Elsevier is a science publisher, which means that most of its customers are scientists.  Science has always been done in the open — it’s the nature of what science is.  And this is becoming more and more important with the rise of open-notebook science, and of services like FigShare, DigiMorph and GenBank.  That’s the atmosphere that science is happening in.

If Elsevier wants to be taken seriously by scientists, it needs to be similarly open.  After all, when you keep secrets, people always assume the worst.

—

What do these measures have in common?  None of them will cost Elsevier much — maybe even save them some money.  These are things that should be done as a priority, as in within a month or so.  They’re not hard, they just need the will to make them happen.  If Elsevier don’t move quickly on these things, the door will slam shut and leave them outside.

If they do move quickly, then they’ll have made real steps towards re-casting themelves as a friend of science, and of scientists.

Last time we looked at the state Elsevier has got itself into, and how it needs to make significant changes to regain the trust of researchers (and librarians for that matter).

By coincidence, literally as I was writing that, Elsevier’s Liz Smith tweeted:

I clicked through and looked at the advert:

Can you help us tell our story? We’re looking for someone who is part community manager, part brand manager and part journalist to be in charge of an exciting new website that will help us communicate more effectively and consistently with the research community.

You’ll establish and implement a content strategy for the site, to make sure that we’re talking about the many positive things happening at Elsevier.

This is worth doing — as Richard Poynder pointed out back in January, Elsevier needs to get out more, and hiring someone whose job is just that can’t hurt.  But I hope no-one at Elsevier thinks that it will be enough to “make sure that we’re talking about the many positive things happening at Elsevier”.

The problems run much deeper than that.

So I’m going to discuss some of the things that Elsevier needs to actually do.  By that I mean, not just talking more effectively about what’s happening already, but changes that need to made.

This post will address the easy ones — things that Elsevier should do right now, without even thinking about it.  We’re talking here about things that will have pretty much no cost, and will start to make Elsevier look like people we can do business with.

1. Be explicit about “sponsored article” terms

I’m starting with this one to give Elsevier a head start, since it seems they’ve fixed this now.  As of a few weeks ago, the sponsored article page has a link to a Sponsored Articles – User Rights page which spells out important details about what you are and are not allowed to do with a sponsored article.

(For the moment I am not interested in what those terms are — we’ll discuss how they are and are not satisfactory in the next article.  I am just interested in whether we can find out what the terms are.)

Why am I mentioning this in How Elsevier Can Save Itself when it’s a problem they’ve already fixed?  Two reasons.

First, because it took an incredibly long time to get them to make these terms clear: see previous articles one, two, three, three and a half, four.

Second, the terms are still not completely clear, in that they don’t say who has the copyright in the article.  (It’s Elsevier, by the way, not the author, but we’ll talk about that in the next part.)

This is in contrast to, for example, Springer’s “Open Choice” page, which is crystal clear about both copyright ownership and access conditions in just 55 words of a page that fits on a single sheet of A4.

2. Be explicit about non-sponsored article terms

I have little idea at the moment exactly what I’m allowed to do with regular non-sponsored articles.  I am affiliated with the University of Bristol and have off-campus access via Shibboleth and an intermittently functional VPN, so I assume I am allowed to download ScienceDirect articles.  But I don’t know whether, for example, I’m allowed to email copies to colleagues who should have access but don’t from off-campus; or whether I’m allowed to text-mine the articles that I have access to; or, if so, what I am allowed to do with the results.

To be fair, there is a fairly hefty document on Authors’ Rights & Responsibilities, but that is addressed much more to the authors of articles than to their users.

Immediate Update.  It turns out I was looking in the wrong place: the information doesn’t seem to be on elsevier.com, but it is on sciencedirect.com, where it can be reached by a Terms and Conditions link on almost every page.  Those terms and conditions are very restrictive, but again we’ll discuss that next time.  The important thing is to make sure they are very easy to find.

3. Make it trivially easy to find sponsored articles

When I was first trying to discover the terms of “sponsored articles”, one approach I took was to go to Cretaceous Research, an Elsevier journal, try to find a sample sponsored article, and see what it said about itself.

Well, I couldn’t do it.  The Advanced Search page has fields for subject, date-range and more, but no “limit to sponsored articles” or similar checkbox.

There are lots of reasons people might want to find sponsored articles, and I can’t think of any good reasons why Elsevier might not want them to.  So this should be made very easy.  (Apart from anything else, it’s about the best way to advertise the journal.)

Not providing a searching option makes it look as though Elsevier want to hide sponsored articles — to stop people from getting full value from them.  That perception, whether accurate or not, needs to be dealt with.

4. Stop lying about copyright transfer

This page in the Author’s Rights area discusses a question that comes up a lot:

Why does Elsevier request transfer of copyright?

The research community needs certainty with respect to the validity of scientific papers, which is normally obtained through the editing and peer review processes. The scientific record must be clear and unambiguous. Elsevier believes that, by obtaining copyright transfer, it will always be clear to researchers that when they access an Elsevier site to review a paper, they are reading a final version of the paper which has been edited, peer-reviewed and accepted for publication in an appropriate journal. This eliminates any ambiguity or uncertainty about Elsevier’s ability to distribute, sub-license and protect the article from unauthorized copying, unauthorized distribution, and plagiarism.

This is flagrant nonsense.  No-one — no-one — evaluates the trustworthiness or validity of a paper on the basis of who owns the copyright.  No-one.

So whatever the true reason for copyright transfer, we know it’s not to make it “clear to researchers that when they access an Elsevier site to review a paper, they are reading a final version of the paper.”

When Elsevier tells us things that we know are not true, how can they expect us to believe anything else they say?

5. Root out and destroy and stupid conditions

You sometimes hear stories like this one: an inter-library loan facility where the librarian is sent a PDF, but publisher restrictions do not not allowed it be forwarded to the patron.  Instead, the librarian has to print the PDF out, destroy the original, scan the printout and send the scan to the patron.

Are these stories true?  I don’t know.  If they are, is Elsevier the publisher concerned?  I don’t know.  But Elsevier needs to make sure of that.  At this point, any such stupidities will be discovered, and trumpeted, and ridiculed.

The same goes for any other equally dumb edge-cases.  If they exist, someone’s going to find them.  Better for Elsevier that it be one of their own people, and that they fix it ASAP.

—

What do these measures have in common?  None of them will cost Elsevier anything.  These are things that should be done as soon as humanly possible, by which I mean “within the next week” rather than “we’ll set up a group to look into it, and report back at the next six-monthly management meeting”.

These measures are about transparency and sanity.  They are the kinds of changes that will start to put some trust back in place.  Being up-front and clear about what the access situation is will start to chip away at the sense that Elsevier has something to hide.  Getting rid of palpable lies about copyright transfer will be a start towards enabling us to believe Elsevier when they tell us other things.  None of this is enough to make an enemy into a friend; but it will at least help us to feel we’re facing an honourable enemy.  (Of course, there is a lot more that researchers need from Elsevier beyond that baseline.  We’ll look into that next time.)

Please do shout in the comments if I’ve missed any zero-cost transparency or sanity measures that should have been listed here.