Who's Guilty?
Who’s Guilty? (Photo credit: Wikipedia)

About a year ago, while attending a publishing strategy meeting focused on open access (OA), a couple of OA publishers told me that eLife articles had just been published on PubMed Central (PMC) and were available nowhere else. Given PMC’s protestations that it “is not a publisher,” this situation seemed wrong. More aggravating to these publishers was the fact that eLife apparently received special treatment, avoiding stated requirements before their content was not only included in PMC but actually published originally there. On top of it all, eLife had not yet demonstrated any independent publishing capabilities, since the eLife site had not yet launched.

I wrote about how this looked from the outside and included notes from my initial interview with the Director of the National Center for Biotechnology Information (NCBI) at the NLM, David Lipman (who is also Executive Secretary for the PMC National Advisory Committee). Lipman represented during our conversation last year that eLife had submitted a PMC application using “the regular procedure” while also asserting that he knew nothing about any conversations that might have gone on leading up to the launch of eLife on PMC.

After what struck me as unsatisfactory answers and aggressive stone-walling, I filed a Freedom of Information Act (FOIA) request about a week later. I’ve been receiving documents based on this and other FOIA requests ever since and have published numerous posts on what I’ve received. The final packet responsive to my initial request arrived earlier this month — another 330+ pages of emails and internal documents to pore over. Many documents were familiar, but some expanded on items I’d seen before, while some were brand new and a bit startling.

As an aside, I never anticipated that inquiring about eLife and PMC would yield 859 pages of documentation. Now that the entire request has been fulfilled, the sheer scope of what went on can be appreciated. It wasn’t nine pages or 59 pages — it was 859 pages.

A few things come into sharper focus thanks to these additional documents:

  • Lipman was intimately involved in discussions around the inclusion of eLife and its special treatment, from the very beginning through the moment when eLife articles debuted on PMC
  • When Lipman disavowed knowledge of what happened leading to eLife publishing its first content on PMC, he was not telling the truth
  • Plans to publish eLife articles on PMC were fairly concrete even before eLife secured its ISSN or was able to put up any sort of basic Web presence
  • In order to pull off the strange publication process that eLife required, the NLM had to resort to “lying” to its own systems and processes
  • Those involved were very clear that this was special treatment for a “friend,” and that they wanted to conceal this special treatment from the community at large

This is a long post in which long email threads are recreated, typos and all.

Let the Games Begin — Again
Last February, I published a post detailing the copies of relevant emails I’d received at the time. As a refresher, on February 24, 2012, Mark Patterson at eLife (then writing from a Wellcome Trust account) pitches the idea of publishing eLife articles on PMC in an email to Ed Sequeira at the NLM, with copies to Lipman and Jo McEntyre of UK PMC.

That same day, in a newly revealed email, Lipman responds to Sequeira:

Ed –

I’m not sure what Mark has in mind. I don’t see what a preview eLife article is “accepted for publication” that is not simply a published article in eLife. What do you think? They should just start the journal in “preview” mode.

djl

Exactly what “‘preview’ mode” represents isn’t clear. As you’ll see, Lipman has a habit of using quote marks around vague ideas he’s trying to push. Sequeira responds to Lipman about 70 minutes later:

It sounds like they don’t want to launch the journal officially until oct or nov, but they still want to publish articles officially through PMC. Otherwise they could release these on their own site — unless they aren’t ready technically. Sounds weird. I’ll respond and suggest we talk on the phone.

Ed

Lipman responds five minutes later:

Maybe if possible I should be on this.

A minute later, Sequeira responds:

If you’d rather respond to him directly, that’s fine with me. Let me know. Otherwise, I’ll copy you on my response so that we can coordinate on a time for a call.

Five minutes later, the thread ends with Lipman replying:

I will respond and cc you.

As noted in the post from earlier this year detailing the first batch of emails about these interactions, Lipman wrote an intentionally vague email responding to Patterson and McEntyre that same day, cc’ing Sequeira. We now have additional emails showing more conversations branching from this thread and showing Lipman’s intimate involvement in setting up eLife on PMC. Three days later (February 27, 2012), Lipman wrote to Sequeira, copying Bart Trawick:

Ed –

I spoke with Mark Patterson and he had an interesting idea for doing these “preview articles”.

Bart – Mark Patterson is now the Managing Editor for eLIFE (HHMI, Wellcome Trust, & Max Planck Institutes upcoming OA journal). Patterson is interested in having some accepted articles publicly available prior to the formal opening of the journal.

He wanted to simply have peer-reviewed & copyedited articles submitted through the NIHMS system. Then, when the journal is up and running, they would send the OA articles for PMC and for their site. It’s kind of a cool idea.

I realize there are issues for this – for example, there won’t be any pre-existing pubmed records. But can we think of a way to handle this?

djl

NIHMS is the National Institutes of Health’s Manuscript Submission system, which was designed to facilitate the submission of final peer-reviewed manuscripts — you know, the kind that cost NIH $47-50 each to convert into XML and QA. Here we see Patterson trying to create a back door into PMC for eLife papers, and Lipman pushing staff to find a way to make it happen.

Sequeira responds the same day:

It’s a nice idea, but it runs into the common problem we have here: if we do it for one group we have to be willing to do it for others as well and then we have to find ways to shoo away the riff raff. I don’t see how the NIHMS pathway being a factor here one way or the other — unless we’re talking solely about articles funded by our participating agencies: NIH, Wellcome, HHMI.

With the funder qualification, we could create a mini-repository of pre-publication papers. But then the question becomes: how long will this last; will anyone contribute to it after eLife is born; will this be short lived like RRN?

A better option may be to have Wellcome advertise these articles via UKPMC. They already offer content that’s not in PMC. They could just put up abstracts and PDFs and since it’s a Wellcome journal, there’s less of a problem with creating a precedent for every other publisher that wants publicity. If this seems viable, we could think through the details.

Ed

Note that Sequeira is clear that eLife’s request is about publicity. Trawick responds to the thread about a half-hour later:

The technical issue from the NIHMS side is that manuscripts are matched to journal records before processing. If the journal record does not exist in the NLM Catalog (not unusual), we kick off a process with LO where they investigate into the legitimacy of the journal, gather the ISSN info, etc, and create the record. Then we match the manuscript to this and processing can begin.

Believe it or not, we have gotten papers from ‘faux’ journals before, including one that was just some guy’s blog.

-bart

“LO” is “library operations.” I don’t know what “RRN” is.

In a response to Sequeira’s email above, Lipman shows his hand:

Hmmm – I think there’s still a work-around here because we have a “relationship” with HHMI and Wellcome Trust. If either or both of them “sponsor” these papers, the $ side is covered. Furthermore, if either or both “certify” these papers, then we are covered in terms of setting a precedent with any given journal. Finally, in terms of publisher permissions – as these are OA, there’s no problem there.

So the key issue would be to get LO to accept the legitimacy of ELife.

I think this is still a possibility then – let me know if you want to discuss this more.

djl

The phrases in “quotes” are worth pausing over. There is a “relationship” between PMC and HHMI and the Wellcome Trust. Indeed there is — so much so that the Wellcome Trust was allowed to handpick a successor for the PMC NAC as one Wellcome employee rotated off, and they picked their solicitor to take over this de facto seat on the NAC. When the Wellcome Trust’s solicitor’s term was expiring in late 2012, Lipman wrote in an email to a group trying to drum up candidates:

We certainly want to get someone from elife and I’ll look at that group of editors.

Then there is the use of the word “certify.” Every journal “certifies” papers by publishing them — some more plausibly than others. So how the certification of eLife differs from the certification of any other journal isn’t clear. I think the nice way to say it is that Lipman is being artful as he tries to move the eLife cause forward.


A Brief Interlude As We Contemplate the Concepts of “Fairness” and “Honesty”

I’d like to pause a moment so we can regain our bearings and remind ourselves that two issues at the heart of this controversy are fairness and honesty. To put this into context, I’ll use an unrelated email exchange from June 11, 2012, between Chris Kelly and Sequeira regarding Bentham, an OA publisher:

Ed,

FYI, we have a pesky Bentham Open editor back asking for reasons why we aren’t taking Bentham journals any longer. No matter what we say now, I fully expect this will turn into a dust up with Matthew Honan at some point shortly . . . but I’m about to rehash Abby’s original response to Urooj from January and hope that gets this guy off our case for a few minutes . . .

Chris

Sequeira responds:

Chris

I’m in the mood to take a more aggressive / blunt stance here. He says “Obviously, not to be indexed in PubMed will hinder any progress of the journal and as the editor I need to know.”

My response: We do not accept a journal into PMC or PubMed with the idea of helping the journal progress and prosper. We expect the journal to do that on its own. We suggest that you asked Bentham management about their previous communications with PMC. This is not a matter of the quality of a single journal. It is a matter of the publishing practices of the organization overall.

This independent exchange provides some interesting contrast to the eLife and PMC discussions. Entire organizations can get special treatment — special positive treatment or special negative treatment. But most striking, of course, is Sequeira’s claim, contemporaneous with PMC’s moves toward assisting with the launch of eLife, that:

[w]e do not accept a journal into PMC or PubMed with the idea of helping the journal progress and prosper. We expect the journal to do that on its own.

The evidence suggests otherwise, unfortunately. That’s not fair, and that’s not honest.


And Now Back to Our Main Feature . . .

Returning to our email exchanges from February and March 2012 between Lipman, Sequeira, Patterson, et al, we start at a point originally captured in an earlier post. On March 4, 2012, Patterson emailed Sequeira, with a cc to Lipman and Melissa Dodd, head of production operations at eLife (scan of these emails):

Hi Ed

I’ve had further discussion with David about using the PMC author submission system to distribute accepted versions of eLife articles before launch of the journal around November of this year.

To enable this to happen, I understand that we have to have eLife included in the NLM catalogue, which in turn requires an ISSN. I’ve applied for an ISSN and will let you know as soon as this is available. Meanwhile, can you let me know of any other information that you’ll need for the purposes of the catalogue?

. . .

Best wishes, and many thanks in advance for your help with this initiative.

Mark

Now, we get into a new branch on the same tree. Lipman replies to Sequeira the same day (Sunday) in a separate thread, copying Dennis Benson at the NIH:

Ed –

I told Mark that if either Wellcome Trust or HHMI are willing to take responsibility for these documents – i.e., as if they funded them, then we can handle the tagging costs effectively drawing from funds they’ve previously provided for tagging. According to Mark, there will only be a dozen or so.

THe notion would be that once the paper is peer reviewed and copy edited, they would use e.g. the HHMI route to upload the manuscript and we’d tag this and return. One eLife is fully in production, they could replace these documents – Jim mentioned that we have some workflow to supercede an author manuscript version with the published version.

Let me know if we need to discuss.

David

Sequeira responds the next day:

David

As Mark mentions, the first step is to get an NLM catalog record for eLife. LO would normally require some physical evidence of a journal beyond an ISSN, e.g., at least a skeletal journal site. So we’ll probably have to pull some strings to get the catalog record. Have you spoken with Sheldon or anyone else there about this or do you want us to?

Tagging itself is not an issue. Wellcome could even process these on its end and deposit them via the UKMSS pipeline. In that case the NIHMS doesn’t have to be involved at all.

Normally it isn’t a problem to replace a manuscript with a published version, because the manuscript isn’t released in PMC until the article has been published and we have a proper citation for it. Citation matching can be done automatically.

These articles are quirks. The official citation will be sometime in the future. So at the very least we’ll have to handle the replacements manually. We’ll also have to figure out some creative way of citing them for now — unless eLife is willing to assign them permanent citations now.

Ed

A week later, on March 12, 2012, Sequeira emails Lipman with copies to Chris Kelly, Jeff Beck, and Sergey Krasnov:

David

LO has created a skeletal catalog record for eLife, http://www.ncbi.nlm.nih.gov/nlmcatalog/101579614. We can flesh it out once Mark has an ISSN and the journal information site that he expects to be ready in a couple of months. I’m talking to him later this week about official publication dates and other details for the ‘pre-launch’ articles.

Ed

Lipman replies less than an hour later to cheerlead for eLife:

Sounds great Ed – I hope they get overwhelmed with papers. . .

djl


It Becomes Clear That eLife Is Driving the Bus

Three days later (March 15, 2012), Sequeira sends a long and specific email to the team of Kelly, Beck, and Krasnov, adding Sarah Post Calhoun:

For the record, my conversation with Mark today. We agreed to mull over questions about final published version, what’s released in PMC initially, etc. and discuss again in a week or two.

Expect to have a journal info site up in 3 weeks. Will include details about editorial board and editorial / review policies. Online submission site in place by the end of May.

About to apply for an issn. (eLife was incorporated as a non-profit corp in the US but Mark is operating out of the UK, so there was a question of where they get the issn.)

Talking to members of the editorial board about submitting ‘seed’ articles for publication in eLife in the next few months, instead of sending them to Nature, Cell, etc. Focusing on high impact papers; quality over quantity. They may get about a dozen of these in the next 6 to 8 months, but don’t really know at this point. Hoping to have about 50 articles, total, published by year end.

Authors of most of the seed articles will be funded by HHMI, WT, or Max Planck, but there could be some who aren’t funded by any of these three. So there could be some articles that, technically, aren’t HHMI or UKPMC manuscripts.

Need to differentiate between the official start date of the journal — which has to be no later than the date the first article is released in PMC — and the launch of a fancy journal site (with a party, etc.), which they’re hoping will be in nov/dec. Hope to have the first seed articles ready for PMC by july/aug.

Aim to give us the seed articles in final, copy-edited form (emphasis on content; PDF formatting, etc. may not be finalized till later). Still, it’s unlikely that we’ll really have the final, final content in PMC for the seed articles at initial deposit. So I’m thinking we treat the initial deposits as manuscripts in PMC, which can be replaced later by the official published version (even though both will have the same publication / release date).

The official date of publication for a seed article will/must be the day on which it is released in PMC. This means that if we decide to load them as manuscripts in PMC, the manuscripts will have citations in the form “eLife. Author manuscript; available in PMC 2012 July 15. Published in final edited form as: eLife. 2012; 1: e12345. Published online 2012 July 15.”

The final published article will be an illusion at this stage. The physical object won’t exist until later. But citing it will establish the official date of publication and allow us to rationally replace a manuscript with the final published version in PMC eventually. (The illusion becomes reality — or, we’re no longer lying.)

eLife will list the seed articles as published, on its temporary site, when they’re released in PMC. Article citations on the eLife site will take you to the version that’s available in PMC.

Makes sense for them to publish one volume per year with no issues. At the least, it makes it easier to create the proper (final) citation for the seed articles. They won’t need some irregular, sparsely populated, issues. (Mark thought that PMC requires journals to have issues.)

Once the journal is in full production (yr-end 2012) they’re considering immediate release of an accepted manuscript, which will be replaced later by the final, edited version. Other than this, they don’t expect to have versioning.

This email bears some analysis. First, this concrete plan sounds like it was largely dictated to PMC by Patterson, who seems to have been emboldened by his conversations with Lipman. Second, the lengths to which PMC is going to accommodate eLife are extraordinary — basically, fooling their systems across the board in order to help eLife launch sooner and gain a market advantage. The use of words like “illusion” and “lying” is striking. Third, eLife is getting all of this attention at no cost — basically, a custom platform build months in the making is being completed on the backs of US taxpayers.

The attitude of entitlement at eLife doesn’t stop there, unfortunately. As we move from March 2012 to September 2012, things get a little frosty between eLife and PMC, as eLife’s preferred launch date approaches. This email dated September 19, 2012, from Sequeira to Lipman, with copies to Kelly, Krasnov, Jim Ostell, and Peter Cooper, indicates the emerging tensions:

David

Details on the eLife situation.

Jeff and Chris have been working with the people there for the past two months on tagging of their content for PMC. So far we’ve only seen concocted ‘articles’ intended to test the various constructs they expect to see in real papers.

Yesterday they told us that they would get us files for 8 real articles on Friday (9/21) and that they’d like to have the journal go live in PMC on Oct. 3. When Jeff told Mark Patterson last week that we can’t guarantee processing times until we’ve seen real articles, Mark apparently got a bit testy. Back in March, Mark expected to have the first articles ready to publish in July/Aug and their full-fledged journal site live in Nov/Dec, so he’s probably under pressure to get something out now.

We’d like to have two weeks after getting clean files to get the journal up in PMC. They’ve said that the first batch of articles will consist of four research articles, three insight articles and an editorial. I think we need alll of them at a minimum in order to make this look like a ‘real’ journal and not just a PR effort we’re doing on behalf of our friends.

I’d also like to make sure that whatever they say in their press release makes it clear that these articles in PMC are the first officially published articles for the journal, i.e., this is the launch of the journal even if their own journal site isn’t going to be ready for some months. A note I have from a phone conversation with Mark in March says “eLife will list the seed articles as published, on its temporary site, when they’re released in PMC. Article citations on the eLife site will take you to the version that’s available in PMC.”

Incidentally, Jennifer McLellan asked Neil last week whether we’d like to co-sponsor their press release. I said no, we don’t want to make it look like we’re doing something special for eLife. (message attached)

Let me know if you need more info.

Ed

About 45 minutes later, James Ostell, who was included on Sequeira’s email to Lipman and who is Chief of the Information Engineering Branch of the NCBI, replies to the entire chain:

They are giving us less than two weeks to set up a new journal in PMC based on content we have never seen. I think that is pretty cheeky, especially since they are behind in their deadlines and asking us to go early on ours.

There are more emails in this thread, alternating between disbelief at how “cheeky” eLife is being and confidence that they can get it done nonetheless. It’s interesting to watch this UK-based publisher pushing US government employees around so blithely, spending US tax dollars with impunity.

The night of the same day, Lipman emails Patterson:

Mark —

I was chatting with Ed Sequeira about the first few eLife articles and it sounds like we should touch base by phone to discuss this. As I understand things, you want the articles to go live by Oct. 3 however we still don’t have the articles. It’s difficult for us to promise that kind of turn around so we should chat briefly ASAP. If we have the articles e.g. by tomorrow and there are no issues, then it may be possible to have them ready by October 3. So let’s chat ASAP.

David

Patterson replies the next morning:

Hi David

I’m in Budapest at the moment, so it’s going to be tricky for me to speak until I get home. I could definitely chat on Saturday if you’re around/don’t mind. I’ll touch base with the folks in Cambridge in the meantime.

I completely understand if we haven’t built in enough time in the process, and we should err on the side of caution to make sure it all looks good. And let me say too, that Ian and Melissa have really appreciated the help and flexibility that your guys at PMC have offered to us – it’s been great!

I might be able to step out of the conference in the afternoon today or tomorrow to skype, but the other problem is that the internet connection here is s**t.

Speak soon

Mark

Lipman forwarded this thread back to Sequeira, Kelly, Krasnov, and Beck:

So Mark seems more calm now. Keep me posted and let me know if you want me to do anything.

In case there was any doubt that Lipman was involved throughout this process, from the initial discussions through to hand-holding eLife during the final steps, we can now lay those doubts to rest — despite Lipman’s initial and demonstrably false assertions that he didn’t know anything about the details of what happened to cause eLife articles to appear on PMC early.

And if there was any doubt as to which group was catering to which, it’s clear PMC was catering to eLife at this time, taking favoritism to some pretty extreme lengths.


More About PR Efforts

One of the most memorable quotes from the earlier batch of eLife and PMC emails came from Sequeira, who wrote in an email about edits to the eLife press release:

Thanks for letting me have a look at this. . . . My reaction to the editorial: Now that we know for sure that you love us, we’d like you to pretend that we’re complete strangers in public.

In an email from the day before these memorable lines were written, Sequeira writes to his internal staff (Kelly, Krasnov, Beck, and Cooper):

At Jeff at least knows, I’m concerned about how eLife presents its ‘launch’ in PMC. The wrong slant provides fodder for everyone from the mystery journals we reject to people like [redacted in source document]. I mentioned this concern when I talked to David the other day and he wasn’t too concerned about it.

Let me know what you think. Am I over-reacting? I’ve made a bunch of edits to their editorial. I’m not expecting them to use my exact words, just trying to convey a change of tone. The differences may be too subtle to matter, so I’d like your opinions before I respond to Mark or say anything further to David.

Thanks.

Ed

Krasnov’s reply starts another interesting exchange, and shows the general lack of comfort with the approach to eLife that permeates many exchanges:

Ed, I think your version is the best what we can do.

Do we have an answer on the question – who could this journal pass PMC acceptance criteria?

How could it be accepted for  NLM collection (was it?) with zero articles published?

Can we explain it?

We are going to tighten up PMC acceptance criteria but this one is a bad precedent for US government agency. Am I missing something?

Sequeira responds a few hours later:

Sergey

This is definitely an exception, but I think we can justify the selection for the NLM collection prior to any articles being published. See attached correspondence with LO. NLM has done this sort of early selection with other high profile journals. David’s influenced helped but it’s a lot different from what we did for BMC. I was never comfortable with how we handled them. I’m fine with eLife as long as they don’t present the release in PMC as some kind of pre-launch. That was my agreement with Mark.

Ed

This reference to “what we did for BMC” is hinted at again in another email chain starting on October 16, 2012, just a day after the eLife articles went live on PMC. The thread has Sequeira seeking advice about developing a statement to respond to complaints. Judith Eannarino writes the last message on the thread:

Ed: Minus the last sentence, the response looks fine.

Working group: Just to review — in general, NLM only rarely makes special arrangements to accept forthcoming content; indeed, the only other title that comes to mind is PLOS Currents. In addition, NLM has given BMC journals preferential treatment in a number of cases, accepting several titles that do not yet meet the current 6 month/30-paper criterion.

-Judith

We’ve seen Eannarino before — she is the LO individual who gave eLife a “what the crap” record number in order to grease the skids for it.

There are a few points to pull out of this message. First, special accommodations are made supposedly based on the credentials of the publisher — but eLife had no credentials at this time other than funder credentials (which are not publisher credentials), while PLOS and BMC credentials certainly pale compared to those of major society or commercial publishers with 100+-year track records, many of whom were at the time waiting to get their OA content into PMC and thereby into PubMed. Second, if BMC (which was founded by Lipman’s friend Vitek Tracz, the man who also hired Lipman to edit a journal at BMC) received some form of special treatment sufficiently beyond the pale to leave an uncomfortable memory in employees minds, and then F1000 Research, another Tracz product, received its own special treatment this year, what are we to think about conflicts of interest and fair play?


Conclusion

While this appears to be “more of the same,” it is both a lot more of the same — creating a more comprehensive picture of the duplicity and favoritism practiced at PMC around eLife and potentially other publications — and it is in some ways new. The new information includes:

  • insights into how uncomfortable NLM staff was with the way eLife was being handled
  • documentation of how involved Lipman was end-to-end and at a detailed level in launching eLife on PMC
  • evidence of how much PMC catered to eLife’s whims
  • emails showing how funders like Wellcome and HHMI are viewed by government employees working on PMC
  • clues that the conflicts of interest between Lipman and Tracz may go deeper (and further back) than we’ve previously thought

Overall, these emails provide yet another eye-opening, embarrassing, and disappointing portrait of PubMed Central management.

Kent Anderson

Kent Anderson

Kent Anderson is the CEO of RedLink and RedLink Network, a past-President of SSP, and the founder of the Scholarly Kitchen. He has worked as Publisher at AAAS/Science, CEO/Publisher of JBJS, Inc., a publishing executive at the Massachusetts Medical Society, Publishing Director of the New England Journal of Medicine, and Director of Medical Journals at the American Academy of Pediatrics. Opinions on social media or blogs are his own.

Discussion

22 Thoughts on "PubMed Central and eLife — New Documents Reveal More Evidence of Impropriety and Bias"

Thanks for taking the initiative to uncover these emails describing the backdoor deals between eLife and PMC. Frankly, the special treatment eLife received is ridiculous and totally unfair to other new publishers, such as ourselves. I would argue it is because of such backdoor operations that innovation in publishing is severely hindered. Different models from “mystery” (aka NEW!) journals will, as you show, never receive such treatment and accordingly their growth will be hindered.

I hope this post gets the attention it deserves! Maybe you could write PMC to help with that, they seem to be good at it…

Joshua Nicholson
CEO & Co-founder of The Winnower

The lack of transparency is truly frightening, especially where the government is the primary publisher. The first amendment makes it almost impossible to block something from being published in a private journal, but nothing REQUIRES the government to publish anything. PubMed Central can effectively censor by just hitting a delete button. Scary! Where are the adults in charge of this process?

Another outstanding post, Kent. Keep up the pressure. PMC should not be in the business of picking winners and losers.

This is government in action. So many people want more government. You got it.

Kent, thanks for your efforts to uncover the duplicity at PMC and the efforts of Lipman to help his friends. I wish I knew how best to get Lipman reprimanded for his actions. Perhaps it is time for you or one of us to ask Francis Collins for an explanation. Oops, can’t do so. The NIH and the rest of the government have been furloughed. Maybe we can find a way to permanently furlough Lipman.

Kent,

You posted [http://scholarlykitchen.sspnet.org/2012/10/22/somethings-rotten-in-bethesda-the-troubling-tale-of-pubmed-central-pubmed-and-elife/] that PMC was the primary publisher for not only eLife but two other journals….

Is there a FOIA request pending for the Journal of the Medical Library Association or the Journal of Biomolecular Techniques? Are these the only other two you could suss out?

I wonder what behind the scenes connections between Dr. Lipman and his publishing friends that may be lurking – could this be more endemic than 3 journals?

This seems like it could be deep.

Thank you again for your diligence in uncovering their mysterious processes.

Tony

I’ve only seen those two journals you mentioned as ones PMC is publishing. There are emails in the recent batch about these journals. I might try to pull something together at some point.

It is strange to have these two journals benefiting from US taxpayer funding in this way. Ultimately, they don’t have to pay for designing or implementing their own journal sites, so they get to skip a lot of expenses most publishers have to balance.

Again, if someone could explain this without forcing me or someone else to resort to a FOIA request, that would be great!

Both are journals of “nonprofit” societies – so I guess they get the benefit of double dipping on the taxpayer, although it seems to me they have sufficient revenue to support a basic website:

The 2011 Form 990 Medical Library Association has about $2.7m in net assets and had taxable unrelated business income from its “publications” of about $168,000 – enough to build a basic $50k Atypon site for sure! (Anecdotally, its a no-embargo Open Access journal, but the society lists it as a membership benefit).

The 2011 Form 990 for the Association of Biomolecular Resource Facilities has about $1.04 in net assets – no breakout for journal income which considering its “free”

I’m not sure if there is an explanation beyond the NLM believing it can engage in this decision-making without consequence – either because it is within their authority, or if not, nobody will stand up to them.

Registration Required:
MLA 990: http://www.guidestar.org/FinDocuments/2011/360/540/2011-360540525-08e4f4b7-9.pdf
ABRF 990: http://www.guidestar.org/FinDocuments/2011/561/659/2011-561659510-08c6dc65-9.pdf

Kent – I can save you the trouble of another FOIA request since I was the editor of the JMLA (then the Bulletin of the Medical Library Association) when we started depositing content into PMC. This goes back to the very early days of PMC – the fall of 1999. (Long before the NIH Public Access Policy was even a glimmer in Heather Joseph’s eye). I had just been appointed to succeed Michael Homan as editor (I served until 2006), and there was much discussion among MLA Board members & headquarters staff about how to make the journal available electronically, preferably open access. The various options that had been identified so far were all in the $20K plus range annually (which, despite T Anthony’s superficial analysis of the association’s budget would have been a significant cost to an organization that has always operated on a slightly frayed shoestring).

In November of that year Lipman did a presentation about PMC at the annual AAMC/AAHSL meeting. NLM was trying to encourage publishers to deposit their content and as Lipman described the conditions of participation it occurred to me that we probably met those conditions. I introduced myself to Lipman afterwards and he asked me to follow up by email, which I did on December 30, 1999. He then put me in touch with his colleague Liz Pope for followup. I see that it was six months (June 22, 2000) before I got back in touch with Pope. Much of this delay was caused by the need to get consensus from the MLA Board that this what we actually wanted to do (in those antediluvian days, the board only did business when they met in person three times a year – how quaint!) Between working through the technical issues with NLM and sorting through the policy issues with the MLA Board, it was another year before the first issues of the BMLA/JMLA appeared.

What may be relevant to the current discussion is the Draft PubMed Central Policy that Pope sent me in July of 2000, describing it as “a first draft of a PMC policy for working with publishers.” In that document, the “Minimum Requirements for Participation” are given:

“Pubmed Central will accept only peer-reviewed material from journals in the life sciences that satisfy one of the following criteria:

“• The journal must be archived in one of the major abstracting and indexing services such as EMBASE, Biosis, MEDLINE, Science Citation Index, Agricola, PsychINFO or Chemical Abstracts.

“• If it is a new journal, at least three members of the journal’s editorial board must currently be principal investigators on research grants from major funding agencies (e.g., NIH, NSF, DOE, NASA, or HHMI in the U.S., or equivalent organizations abroad).”

Note that there was no requirement that the journal have its own digital presence. The BMLA clearly met the conditions, so there was nothing untoward or exceptional about our participation or NLM’s willingness to host the content. Any journal, print or otherwise, could have participated under those same conditions.

The JMLA Editorial Board (on which I am currently serving a 3-year term) does take up the issue from time to time of creating a fully featured website for the journal, but for the time being the current arrangement meets our needs. Marty Frank (who I consider a good friend) and I have sparred occasionally about what he perceives as the “unfairness” of this, but I can’t get too worked up about it. I have a hard time figuring out who is harmed by it.

As to the situation with eLife, I’ve refrained from commenting, in part because my own long association with NLM certainly puts my objectivity in question. (I did a post-graduate fellowship there in 83/84 and stayed on for several years as a Technical Information Specialist. I’ve served on numerous panels and task forces in the decades since, and consider many of the current and past NLM administrators to be valued colleagues, mentors, and friends – although I suppose I should mention that I don’t know Lipman personally and that brief conversation and follow-up email from 14 years ago is the only direct contact I’ve had with him).

But I will point out that NLM has always taken an aggressive posture in pursuit of their statutory responsibility to preserve and make available the world’s biomedical literature. They fought back hard in the 60s & 70s when they were sued by Williams & Wilkins for their interlibrary loan practices, resulting in a landmark Supreme Court ruling that laid the groundwork for how Fair Use is interpreted in current ILL practice. When I was there in the mid-80s they were under attack from database vendors who wanted them to stop providing access to MEDLINE at rates cheaper than the vendors could. Then, as now, the cry was unfair competition, but NLM continued their developments and eventually was able to make MEDLINE available for free via PubMed — something which is generally agreed to be a very good thing.

They have always been willing to bend their rules and processes in the service of what they believe is greater fealty to their mission. I’ve not spoken to anyone at NLM directly about the eLife situation and I’m not going to try to parse the rightness or wrongness of elements of the email trail that you’ve laid out, but I suspect that many people there greeted the initial announcement with enthusiasm, and when people like Schekman and Patterson were brought on board it would have been apparent that this was an experiment that had some muscle behind it. I can well imagine them thinking that it was experiment worth bending the rules to support.

Believe me, I can certainly sympathize with the angst of those publishers who feel they’ve been treated unfairly, but I can’t get too worked up about it. It does not seem to me to rise to the levels of impropriety that you claim. But, like I say, I’m probably not an objective observer.

Thanks, Scott. The unfairness is multi-faceted, as I’ve pointed out before. PMC is allowing JMLA to publish at taxpayers’ expense while keeping it off the market, and keeping it from paying a vendor or some sort the usual fees. This gives JMLA an unfair advantage in the market, deprives companies of normal and customary revenues, and creates overhead the US government has to support.

The other unfairness is cronyism. That is, why is the “muscle” of eLife more significant than the OA muscle of SAGE or Wolters-Kluwer or Elsevier or the American College of [fill in the blank]? Why only BMC (where Lipman’s entangelments are clear), PLOS (entanglements again), and eLife (more entanglements)? It gives the appearance of cronyism, and the emails confirm this perception.

The basic question is why did PMC hide this from its own advisory committee, want to hide it from the community, and evade direct questions posed more than a year ago? If they were proud of what they did, why not say so? And why hide it from their own National Advisory Committee?

JMLA is getting special treatment. It’s not generalizable, and it hurts the market, albeit slightly.

How is JMLA dealing with the government shutdown and loading of new content?

Thanks for saving me the chore of another FOIA request!

You’re welcome — it was fun to go back through those old emails and remember how it all came together.

Like I said, I’m not going to get into debating the issues of the eLife situation, but with regard to the JMLA, had we decided to set up an independent site (or if MLA decides to do so in the future), we’d still participate in PMC, so the cost to the taxpayer would be the same — I don’t see how the decision not to set up a separate site results in any additional taxpayer cost. As far as getting special treatment, as I hope I pointed out, the option we took would have been available to any other journal that met the very broad criteria at the time.

Ah, but you’re right about that $20K a year that we avoided sending to Highwire or Allen Press or that we might have invested in setting up the operation in-house — clearly a crippling factor in the demise of the scholarly publishing industry.

As to how the journal is handling the shutdown? We took the cut-rate option 14 years ago and you get what you pay for. After more than a century of publication, I imagine we’ll survive.

Not paying $20-50K for work is using government infrastructure and costing money. That’s the salary of an FTE at some level, with related taxes and so forth. The model is not generalizable. JMLA is lucky to have gotten in during the days when this was allowed, but I think the board of JMLA should reconsider and pay full-freight for its publishing operations. It might make the medical librarians more sympathetic to the technology costs publishers bear. As it is, it may be easy for the editors and board at JMLA to consider technology as free.

You’re right, JMLA will survive. But quality of life matters. Freeloading off the government is not a great way to live. Sorry, but that’s how it looks.

Mr. Plutchak,

It seems that JMLA and ABRF are the first to be pointed out – perhaps there is still a FOIA request looming – how many others are enjoying the “framed in freebie” of PMC?

Thank you for taking the time to lay out the historical review of the JMLA’s entrance into PubMed Central. I sincerely found it interesting – particularly when compared with the current criteria, of which there seems to be little except perhaps reliance on the CDM when it comes to rejections, and “bending” the rules when it comes to accepting those with muscle as you described poignantly. These are serious issues, and while MLA enjoys a grandfathered benefit, it certainly can pay its own way.

Please accept my apologies for my public performance of a superficial review of MLA’s budgets and I in turn will accept your invitation to dive deeper into the last 3 years of Federal Tax Forms 990 of the MLA. Unfortunately, on this I believe we only have in common our desire to avoid using our given first names.

I’ll dispense with the nitty-gritty details of society management expenses and ways to “afford” a publishing site on a “shoestring” budget, including the largess of salaries, office rent for in the art deco building on the 19th floor above Morton’s steakhouse, etc – the MLA reports spending the following on its website annually:

2009 – $267,229 (including public relations)
2010 – $197,930
2011 – $193,372

I’ve developed fully functional websites for much less and I’m not buying your “shoestring” budget statement.

Noticeably, I did not find a line item for the “$20,000” (or today’s equivalent) for the in-kind taxable value of the service our Federal government is donating to the MLA.

I read with great interest your TScott blog entry of 2/22/12 regarding FRPAA that supports Open Access policies that “encourage subscription publishers to develop fully OA business models, while injecting a level of market competition that could help to hold publication fees down.” Would part of that market competition include a injecting free government websites for publishers who choose not to have one?

Respectfully, T. Anthony Howell

It’s also worth noting that MLA publishes (behind access controls) two other publications — MLA News and MLA FOCUS. MLA does not apparently use government infrastructure to publish these, so clearly they are capable of executing publishing events on their own.

I’m genuinely curious about something here.

I’m seeing a whole lot of fuss being made about “freeloading off the government” when a journal allows the government to host its full-text articles, but not a word is spoken about journals allowing the government to host their abstracts and provide a search interface via PubMed. Does this government-provided service not also have some real monetary value, reducing technology development costs (and providing valuable exposure and inbound links) for all publishers with PubMed-indexed journals?

Surely a journal could pay to develop their own search interface, APIs, and meta tagging for their own published papers, and work to build a larger website userbase. But nobody does this, because this would be expensive, and the US government is providing this service for free, as it’s a cheap way to benefit a large number of medical researchers.

Am I missing some fundamental difference here between the two situations? What is it about full-text hosting, as opposed to abstract hosting, that all of a sudden turns leveraging a centralized government database into something that’s awful and entirely anti-free-market?

I think you are missing something fundamental here. Pretty much every journal publisher on earth develops its own search interface, API’s and metadata. Everyone does this, it is part of the services a publisher provides. If a publisher relied solely on PubMed as its only source of traffic, then most, if not all, would see only a tiny fraction of their current traffic. For example, on the medical journals I work with, PubMed refers around 1/10th the traffic that Google does.

It is a question worth asking though–in the age of Google, is maintaining a specialized search engine a good use of NIH funds that could instead go to other projects? At one point, PubMed did offer an important service, a quality check, where only journals deemed worthy of MedLine indexing were listed. This has since devolved, given that PubMed search results now also include all journals deposited in PubMed Central, which has much lower (and apparently inconsistent) requirements for inclusion.

David and Kent, thanks for your comments.

Note that I purposely didn’t confront the issues of fairness, cronyism, or favoritism in my original comment — just the point about freeloading off the government, which in particular felt like an odd line of reasoning directed towards the JMLA. I think Kent somewhat missed the focus of my question in his response, and I don’t really have much to add there.

David’s response raised a good point: that publishers actually already do fund search and database development. (Though I would suspect that only the largest publishers develop full-fledged APIs or search interfaces used by a noticeable number of visitors.)

Now, I probably have a slightly different perspective on the utility of PubMed, as a developer of consumer tools for academics. (I co-develop Paperpile, a web-based reference manager.)

Without PubMed, our service would be vastly less useful for researchers in the life sciences. Many journals provide haphazard metadata and/or APIs at best, and little to nothing at worst. PubMed, both the website itself and the metadata that’s made accessible via NCBI’s well-supported eUtils APIs, has become a vital informatics infrastructure that enables much of the search and discovery going on in life science reference tools. If PubMed funds were diverted elsewhere, not only would an entire ecosystem of tools and databases be damaged, but I expect all publishers would stand to lose much more than just the minority of referrals that PubMed provides.

(Yes, I am aware that the publisher-supported nonprofit CrossRef is also a means by which publishers spend money on technology and API development. But I’m intimately familiar with both databases, and I can confidently say that PubMed far exceeds CrossRef in the accuracy, consistency and usability of its data. From my perspective, the US government’s investment in PubMed is still very well justified.)

But this is getting far afield from my original comment and your responses.

To quickly reach my point: I appreciate your claim that publishers *do* spend money on metadata curation and tech development, but I still haven’t heard an argument that PubMed is providing *no* taxpayer-subsidized value to publishers (both nationally and abroad).

So, assuming PubMed is providing some value and that its existence isn’t altogether controversial, it feels like an odd cherry-picking of idealism to then see derisive comments to the effect that JMLA is a government freeloader if it doesn’t cough up the $20-50k in tech development saved by leveraging PMC for hosting fulltext articles.

Why doesn’t every publisher, large and small, pony up and pay for a PubMed-quality search interface and API? Because it would be an unnecessary and wasteful duplication of effort! And to me, the JMLA situation feels much the same. But maybe that’s where I differ from some of the experienced publishers in this forum.

You’re conflating PubMed with PubMed Central. JMLA is being published in full-text by the NLM/NIH via PubMed Central. That’s far beyond being indexed in PubMed. This series is not about PubMed (the index of abstracts in biomedicine) but about PubMed Central (the full-text platform).

PubMed’s history is an interesting one. It started in 1964. Well, MEDLARS started around then. Index Medicus (a printed tome) started in 1879. It migrated to CD-ROMs and other similar formats, and then went online (the MEDLINE index) as PubMed in the 1990s. The indexing of articles comports well with the NLM’s core mission, and their choice to make their index freely available online in 1997 was part of Al Gore’s well-placed enthusiasm for the Internet. But the index is very different from a redundant full-text repository, and again very different from being the primary publisher of a journal like JMLA. That’s government-supported publishing.

The government has interests in supporting infrastructure at many levels. I wrote about this in a review of the excellent book, “The Entrepreneurial State.” In its proper role, the government helps the economy grow and does things that, as you point out, businesses have difficulty doing on their own in the midst of their own incentives. The problem with PubMed Central is that it is not doing something unique, but rather is competing with publishers (as you note, PubMed is not competing with publishers — however, not complaining about the government serving its purpose is not the same as being a freeloader; public/private partnerships like this exist throughout the economy, as they should). PMC’s competition with publishers is a source of friction, and hurts the economy and reduces the available tax base for the government. Those are signs the government is not in its proper role.

In addition to competing, PMC is acting as a publisher with JMLA and at least one other journal. This is not a valid role for the government, for the same reasons.

The government has a role in the economy. PubMed is highly aligned with the mission of the NLM, helps publishers, and is not competitive. PubMed Central is competitive with publishers, initiatives like CHORUS show that it is not necessary, and it is being run poorly.

Thanks Kent and David for your further comments.

Note that I tried not to conflate PubMed and PubMed Central — I know the difference between the two, and after double-checking my comment I cannot find a place where I mistook them. I was merely trying to find a parallel between the two systems and their relationship to publishers’ technology development costs.

Thanks Kent for sharing a bit on the history of PubMed. It’s interesting to learn how databases that are often taken for granted today as having “always been there” actually have much more colorful, dynamic and varied histories. While a grad student at EMBL-EBI, I learned how the UniProt databases was originally distributed (for 13 years!) as the printed Atlas of Protein Sequence and Structure during the 60s and 70s.

Regarding the idea that PMC is competitive to publishers while PubMed is not, I suspect that PMC’s threat of competition may vary between different publishers and journals: namely, it represents a relatively non-competitive piece of technical infrastructure to open access publishers (where the majority of income is in hand in before readers view the article), whereas it may be a competitive and highly disruptive system for closed access publishers (where more income results from readership). Honestly, it’s fascinating to see how the various battle lines have been drawn since the OSTP memorandum.

And David, I like your sensitive treatment of the ‘freeloader’ term. Indeed, it’s hard to blame anyone for taking advantage of a money-saving opportunity, and it does stink when such opportunities are offered in a very unfair way.

I think the use of the term “freeloading” in this case is fair. The notion that “papers” are government-funded is demonstrably untrue — research is government-funded, and the reports to the government are government-funded, and I’m on record as supporting making these reports free. But they are not papers. Papers are separate animals, and there are costs for a publisher even to reject them (at our journal, it’s about $275 per paper just to register the submission, send them to two editors to review to see if we want the paper at all, and then reject them, which means generating records, letters, and so forth). Publishers don’t freeload, but actually remove risks from authors (fund the review, editing, and composition, hosting, etc., while attending to IP and other legal matters, registering the paper widely so it’s known). Sharing government-funded reports with the public is proper. Asking publishers to share the papers they’ve published, which have to cover the costs of rejection and a host of other processes and personnel, not to mention long-term sustainability, is a huge request. But publishers are not “freeloading” off government grants or anything of the kind.

JMLA is freeloading, which is defined as, “To take advantage of the charity, generosity, or hospitality of others.” If they didn’t want to freeload, they could stop tomorrow. But as it is right now, they are taking advantage.

As for PMC offering a “relatively non-competitive piece of technical infrastructure to open access publishers,” my estimates is that the traffic competition PMC poses to PLOS costs PLOS $75K to $125K per year in lost advertising revenues. That’s fairly competitive. In fact, traffic decrements to PLOS content were greater than those observed to subscription publishers, likely because the OA content is free from Day 1 via PMC, which makes PMC an even better substitute. Subscription content has a year embargo.

The JMLA situation is nearly unique. It’s important to think why that is. Essentially, I believe it’s because it was a mistake to offer it in the first place, PMC leadership realized this, and stopped offering it, and now the onus is on JMLA to do the right thing and stop taking advantage of the situation.

Greg–
Thanks for the further thoughts. It’s helpful to get some different viewpoints and see what sorts of benefits a program like PubMed offers to others beyond just the direct traffic of readers and authors. You’re right that it’s really the big publishers who do a great deal of API and other technical development on their own, but for the smaller and independent publishers, these things are still done, just by outsourcing to a typesetter or an online platform provider. This makes it affordable for the smaller and the not-for-profits to gain these functions at a fraction of the costs of doing it yourself (though the downside is less control over what you’re able to provide).

I’m trying to avoid the question of “freeloading” as well, as who’s a freeloader is really in the eye of the beholder. As Kent notes, one of the important functions of government agencies is to partner with and provide resources to help citizens drive the economy. Even so, in nearly every such situation, someone feels left out: publishers are freeloaders because they won’t make govt funded papers free to everyone, researchers are freeloaders because they patent and get rich from govt funded research results, you’re a freeloader because you use a govt service to fuel your company’s service and yet you charge for it.

The problem here isn’t that someone is getting a free ride, it’s the seemingly capricious and suspect nature of who gets access to these free services. If we assume this economic stimulus is indeed a deliberate act by the NLM, then it is perhaps troubling that only certain parties, particularly those with close business and other ties to NLM leaders, are allowed access to those economic benefits. If you were denied access to the NLM’s API’s and metadata, while your competitor was given full free access because he is a close pal of a director at the NLM, you might be upset. You might feel the same way that many OA publishers feel when they see one particular publisher getting special treatment (and a publisher not based in the US to top it off). The fact that those behind PMC deliberately tried to cover their tracks is pretty telling here.

There is nothing unfair about PubMed indexing — or, at least there wasn’t, until it became a downstream event related to PMC inclusion. Now, the fairness of it is a little more questionable. If every publisher could benefit from PMC publishing infrastructure for full-text, we’d have a different set of facts and incentives to analyze. But that’s not the case. The “fuss” is mainly over favoritism, cronyism, and freeloading off taxpayers. Why are some journals allowed to save money by using government infrastructure while others are not? When NLM/NIH says they “are not a publisher,” we should expect them to stay out of the business of being the primary publisher of full-text content.

Having PMC as the secondary publisher of full-text content creates problems for publishers, siphoning traffic away. While PubMed does create some traffic benefit for publishers, that benefit is diminishing quickly as Google and Google Scholar have eclipsed them, leading to David’s excellent questions about whether PubMed is even needed anymore.

Defenders of the JMLA situation admit that they are taking advantage of an opportunity that was briefly available, and is no longer available, while placing a value of $20K per year on their freeloading. It’s probably more like $50K (they’ve been off the market for so long they’ve lost touch with the costs of publishing, which includes creating a search interface, APIs, and deploying metadata). As it’s been about 10 years since they started using PMC as their publisher, they’ve effectively received $200-500K of benefit by freeloading. This keeps JMLA out of the technology marketplace, and shields them from the true costs of running their operation. And, as noted above, the NIH/NLM are operating outside their own stated roles.

As for eLife, they benefited from launch assistance that government employees tried to hush up, attempts at secrecy that included hiding the behavior from their own governance group, which I believe is very unethical. And eLife was operating as an arm mainly of a UK-based charity, while getting help from US taxpayer-funded employees and infrastructure. At the same time, other publishers with better track records who were following the NLM’s stated rules were being shut out of PMC. It was unfair, not transparent, full of cronyism, and wrong. That’s worth some fuss, I believe.

Comments are closed.