Self-Archiving Mandates

---

SUMMARY: The publisher lobby can defeat the Federal Research Public Access Act (FRPAA) (to mandate self-archiving of all federally funded research so as to make it freely accessible online) if FRPAA is promoted merely or mainly as a means of providing public (student, practitioner, patient, general public) access to publicly funded research. Publishers can and will argue (and already are arguing) that the public does not really want or need access to most of this specialized peer-reviewed journal literature, across all fields (not just clinically relevant medicine), written by and for specialized researchers, and that it is hence not justified to put publishers' subscription revenues at potential risk by mandating that this entire literature must be made freely accessible online. Instead, publishers will propose special arrangements in which they themselves would make the tiny fraction of what they publish that is of potential public interest freely accessible online. The right response to this by FRPAA proponents is to make it very explicit that the primary purpose of the bill is not public READS, but "CURES" -- i.e., the public benefits that come from applying and building upon research findings in further research and practical applications, for the sake of which the research was publicly funded in the first place. And CURES come from researcher access and usage -- researchers applying and building upon current research in further research and applications -- not from public access and usage. Because no researcher can currently afford access to anywhere near all the research they might need to read and use, researcher self-archiving substantially accelerates and increases research usage and impact, which is the measure of speed and progress toward CURES. And substantially accelerating and increasing progress toward CURES -- unlike providing public READS -- does outweigh any hypothetical risk to publisher revenues (although there is as yet absolutely no evidence that self-archiving reduces subscription revenues.) Moreover, with free online access (Open Access) to self-archived research, the public (students, practitioners, patients, tax-payers) get full access too, as a secondary benefit, but not because that is the primary benefit from or justification for mandating Open Access self-archiving.

---

ANON: " Your arguments are totally logical. However, a factor you are not taking into account: if researchers are focused on their research- impact - politicians are focused on their own image and reelection potential. It is the politicians who need to vote in FRPAA."

And it is the publisher lobby that will be pressuring them not to. SPPP (Student/Practitioner/Patient/Public) access is a good intro, to get the politicians' and voters' attention, but then you need a follow-through that can hold up against the publisher lobby -- and SPPP-access has no follow-through when publishers inevitably say, as they will (and are already):

"You want to mandate that our business revenue should be put at risk for the sake of SPPP-access, yet there is no evidence that the SPPP reads (or has the slightest wish to read) most of the highly specialized research that we publish! Why not just make a side-deal that we make publicly accessible that tiny fraction of (mostly clinical-medical) research that is likely to be of SPPP interest, and leave the rest of it -- which is the overwhelming majority of it -- alone, rather than putting all of our revenues at risk for no objective reason?

(And denigrate logic all you like, in the end, the pro-mandate argument has to make sense, otherwise the publisher lobby wins and the OA self-archiving mandate -- and the best interests of research and the public that funds it -- lose.)

The requisite follow-through is CURES, not SPPP-access. Students, practitioners, patients and the public do not produce CURES, researchers do. And the reason researcher usage and impact is so important is not because it produces money and prizes for researchers, but because it generates CURES. In fact, that is what research is funded for, not to produce reading matter for the SPPP.

("CURES" is of course over-simplified too, and medically biassed, but it will do, as long as it is put in scare-quotes or CAPs: more generally, it means applications of research, including technology; even more generally, it means pure research progress itself, which might eventually lead to applications; and when it comes to social science and especially the humanities, which rarely has any applications at all, it means the production of specialized scholarship, which we presumably fund because we think it is a social benefit to promote scholarship, not because the general public or even students actually need or wish to read the peer-reviewed journal articles reporting the research the public funds, written by specialists for specialists, but because the public wants to promote scholarly progress, which may eventually trickle down into education.)

ANON: " Is there evidence that FRPAA will result in the kind of citations that politicians care about - photo ops and positive pieces in the news, funding support and votes so that they can be re-elected?"

CURES produce photo-ops, and for researchers to produce CURES, researchers (not SPPP) need to have access to the ongoing research, in order to use it and build on it.

Moreover, the politicians are not just responsive to votes, as you know, but also to money and lobbying, especially from big business, and to what fosters or threatens business revenue flows. Yes, "public access to publicly funded research" sounds like a good vote-getter on the surface, even if it doesn't amount to much research the public would actually want to access and. But the publishing lobby is another matter, and they are the ones to contend with now.

It's not the vote-getting power of the OA principle that has been blocking the RCUK policy for two years and that has watered down the ,a href="http://publicaccess.nih.gov/policy.htm">NIH public access policy to near-nothingness: it's the publisher lobby; and this time FRPAA has to come forearmed: If it tries to coast on the public-access-to-publicly-funded-research slogan alone, or primarily, it will be defeated, no matter how sexy it may sound as a vote-getter. (And, by the way, most individual citizens don't read research and couldn't care less about this issue, one way or the other.)

Publishers will float doomsday scenarios about ruinous risks to their ability to make ends meet if self-archiving is mandated (not based on any evidence, but sounding ominous just the same). These doomsday scenarios need a more convincing answer than that "we are doing it so the public can read the research it funds" -- because then the publishers will simply adduce the abundant evidence that the public is not reading most of the peer-reviewed research they publish, and would not and could not have the slightest interest in ever reading it. So the revenue-risk is completely unjustified.

Not so if the rationale is CURES rather than SPPP READS, for research progress and the possibility of cures is the very reason we fund research in the first place. CURES -- but not READS -- offset publishers' hypothetical doomsday scenario very effectively.

ANON: " To put it another way: is there research showing that politicians care about researcher-impact at all, never mind enough to stand up to the publisher anti-OA lobbying?"

Politicians care about CURES, and "cures" is the simple (simplistic) encapsulation of research uptake, usage, application, productivity and progress. And that in turn is something that can only come from researchers using and applying research, not from the public, reading research. And it is for CURES that the public is funding research in the first place, not for its own READing delectation.

So the right public issue politicians need to focus on is CURES, not SPPP-access; and CURES means research usage and impact, which comes from researcher-use, not from SPPP-reading.

ANON: " Arguments focussed on students, patients, and the public are much more likely to persuade politicians than arguments based exclusively on benefits for researchers. The two streams of arguments complement each other. It is not necessary, or desirable, to limit pro-OA arguments."

OA is not about benefits to researchers! It is about CURES. Researcher access means more progress and momentum toward CURES.

Moreover, it is now no longer just about persuading politicians but about resisting the publisher lobby, which is trying to dissuade politicians. Answers to their objections are needed too; and SPPP-access is not the answer, CURES is; and that means researcher-access, not SPPP-access. (Yet, let us not forget, SPPP-access can and will come too, with the OA territory: So it's fine to mention both benefits, but essential to make it clear that CURES is the primary rationale for mandating self-archiving, and READS merely a secondary benefit.

ANON: " The politician who cares about patients but thinks the researcher-arguments are abstract, will support a patient-based OA argument. It is unlikely that a person with this viewpoint would support a research-only focused argument."

The focus is on CURES, not on abstract researcher-arguments: Everyone knows that CURES come from researchers, not from students, practitioners, patients or the general public. I think that is a concrete matter that politicians and voters are quite capable of understanding. And it has the virtue of trumping the publishers' arguments about hypothetical revenue risks: progress toward actual CURES (monitored in the form of research impact) trumps hypothetical revenue risks; SPPP-READS do not.

Stevan Harnad
American Scientist Open Access Forum

---

SUMMARY: "Public Access to Publicly Funded Research" is a good way to draw voter and elected-official attention to the need for Open Access, but then it has to be made clear that research is publicly funded primarily for the sake of the benefits it brings to the public: so its findings can be used and built upon in the form of further research progress and applications ("CURES") and not primarily because the public has a great desire or need to read peer-reviewed research articles ("READS"), written by and for researchers (and research students). Hence it is primarily for the sake of making publicly funded research accessible to its intended users -- researchers -- that Open Access is so important and urgent for the public that funds it: because, right now, researchers do not have access to all or most of the research they need in order to maximize research progress and applications (CURES), to the benefit of the public that funded the research. Public access READs, however, come as a secondary side-benefit of Open Access anyway (for everyone, including students, practitioners and patients); yet the reason it is so important not to argue that public READs are the primary rationale for Open Access is that that rationale can be challenged (and perhaps even defeated) by the publisher lobby, arguing that it is unjustified to put their subscription revenues at potential risk by mandating author self-archiving (even though there exists no objective evidence that author self-archiving causes subscription decline) merely for the sake of public READS when the public (including students, practitioners and patients) in fact has little interest in reading peer-reviewed research journal articles apart from a minority of health-relevant articles, for which the publishers can instead make special arrangements to make them -- and only them -- Open Access. In contrast, the true primary rationale for Open Access, "CURES" (i.e., research progress and applications) for the public that funded the research, does over-ride hypothetical risks to publisher subscription revenue (for which no actual evidence even exists).

---

Below is a reply to an anonymized query on an often-confused issue concerning Open Access (OA), the rationale for providing OA, and the rationale for mandating the provision of OA (by mandating self-archiving, as the RCUK in the UK, the FRPAA in the US and the European Commission (EC) are each proposing to do):

(1) OA is about Open Access to research: about 2.5 million articles per year, published in about 24,000 peer-reviewed research journals and congress proceedings in all disciplines, from maths, physics and engineering to biology and medical sciences, to the social sciences and the humanities.

(2) The only ones who can provide access to these 2.5 million annual researchers articles are their authors, the researchers: either by publishing them in an OA journal, or by publishing them in a conventional journal and also self-archiving them in their institutional repository (or a central one).

(3) Author surveys have shown that although only about 15% of authors self-archive spontaneously, 95% will comply if mandated to self-archive by their research funders and/or their institutions. The half-dozen self-archiving mandates that have already been adopted (including those of the Wellcome Trust, CERN, and several universities) have since confirmed this high compliance rate.

(4) The FRPAA, RCUK and EU mandate proposals are facing stout opposition from the publisher lobby, even though 94% of journals have already given their green light for immediate author self-archiving.

(5) The two main bases for the publishers' objections are that OA is unnecessary and that mandating self-archiving would put their subscription revenues at risk.

(6) There is to date no evidence at all that self-archiving reduces publisher subscription revenues, but even if it were ever to do so, the question is whether the benefits (to research, researchers, and the public that funds them) outweigh the risks (to publishers).

(7) The chief evidence that the benefits (to research, researchers, and the public that funds them) outweigh the risks (to publishers) is that OA substantially accelerates and increases research usage and impact, hence research progress and productivity (as measured by download and citation counts): This is the primary motivation for mandating the provision of OA, through self-archiving: to maximize research usage and impact, hence research productivity and progress ("CURES").

(8) A side-benefit is that OA increases access and usage for practitioners, patients and the general public too -- in those fields in which there is practitioner, patient, and public (PPP) interest in the research articles. (It has to be noted, though, that there is as yet no systematic quantitative measure of this PPP interest and that PPP interest is almost certainly limited to only a small fraction of the annual 2.5 million articles published across all research fields.)

(9) Hence the primary rationale for mandating OA self-archiving has to be the objectively measurable and demonstrated benefits it provides for research, researchers, and the public that funds them -- in terms of usage and citations as measures of progress toward CURES -- rather than merely PPP interest.

(10) OA of course also provides access for PPP use, but the important strategic point to understand is that PPP-needs (being limited to only a fraction of the OA fields and being difficult as yet to measure and document) cannot be adduced as the primary reason for putting publisher subscription revenues at even hypothetical risk -- otherwise there is a strong actual risk that the proposals to mandate self-archiving will be successfully defeated by the publisher lobby. The rationale for OA is primarily to provide access to research by researchers for researchers -- all for the benefit of the public that funded the research. PPP access is merely a side-benefit of OA. It would come nowhere near justifying an OA mandate just on its own.

Here is my detailed reply to a well-meaning (anonymous) query concerning PPP interests:

ANON: " When I read your 8-point agenda I believe that the clinical faculty would feel that they were not being embraced in it."

I think you are not quite understanding the OA problem, hence its solution: The objective is to provide free online access (OA) for all would-be users (whether they be researchers or practitioners, patients and public [PPP]).

The problem, however, is that the providers of the research, namely, the researchers who wrote the research articles, are not yet providing OA to their articles spontaneously.

The solution is to mandate that they must provide OA, for the benefit of the public that funds their research -- by self-archiving their own final, refereed, accepted drafts of their own articles free online in institutional or central repositories.

In order to get that solution (mandate) adopted, it is necessary to persuade those who are in a position to mandate self-archiving -- namely the researchers' own funders and institutions -- to mandate it. In order to persuade them to mandate it, it is necessary to persuade them that there is a need to mandate OA -- especially because the publishers are trying to prevent self-archiving mandates, or, failing that, to embargo them, because they fear they could reduce their subscription revenues (even though there is no evidence of this, even after 15 years of self-archiving, some of it at or near 100% for years now in some subfields).

Now comes the critical point: To persuade researchers and their funders and employers that there is indeed a strong need to mandate self-archiving despite the publishers' objections that there is no need for OA and that it might put their subscription revenues at risk, you have to make it clear exactly what the need for OA is, and how and why it is to researchers' advantage to self-archive their research:

The chief need for OA is on the part of those who are in the position to use and apply the research, for the benefit of the public that funded it, namely, the researchers by and for whom the research articles were written. And the objective measure of their need is download and citation counts: It has been demonstrated that self-archiving accelerates and increases downloads and citations substantially (meaning that without it, many potential users are denied access). Citation counts mean salary and funding for researchers, and overheads for their institutions, and both citations and downloads mean a return on the funder's investment of tax-payer money in funding the research, in terms of research productivity, applications and progress ("CURES"), in all fields.

So the way to solve the problem of how to persuade researchers to provide OA is to persuade funders and institutions to mandate self-archiving. And the way to persuade them to mandate self-archiving is to persuade them that OA is to the advantage of research and researchers (and their institutions and funders and especially the tax-payers that fund the funders) because it both accelerates and increases research citations and downloads (i.e., research impact and progress: "CURES").

Downloads are not as yet being systematically measured and compiled (although they will be eventually), but citations are already being systematically measured and compiled -- and, moreover, they are correlated with downloads.

So the simple, straightforward argument for mandating self-archiving, the one that is immune to publishers' objections that OA is unneeded or that it might ruin their business, is that self-archiving is optimal for research progress itself ("CURES"), because it substantially increases research citations, which indicates that the research is being taken up, used, applied and built upon.

If we could add download counts to the argument, and downloads in particular by practitioners, patients and public (PPP), we would, but there are no such download counts yet, so we cannot add them directly and empirically to the usage/impact argument. It is not necessary, however, because free access for researchers also means free access for everyone else too, including PPP. So there is no need to adduce specific evidence that there is substantial PPP demand and need for access (especially because in most specialized fields there is unlikely to be!).

We cannot, however, say that the primary reason we need OA is because of PPP needs, because (1) we have no data on PPP use yet and (2) PPP use applies to only a small fraction of the research literature -- 2.5 million articles a year, across all fields, in 24,000 journals. Hence this is not a valid argument for OA self-archiving in general, and, if put up front as the main reason for seeking OA mandates, would lead to debate, delay and defeat after years of haggling, with publisher offers of "special deals," with the publishers making only a select subset of their articles OA -- those that might have some PPP interest -- rather than all articles, which would put all of the research journal needlessly at (hypothetical) risk, for no compelling reason.

That would be the PPP tail wagging the entire OA research dog: PPP needs are only a tiny (though important) subset of OA needs. And, more important, direct PPP access is definitely not the main way the public benefits from OA! Focussing primarily on PPP access is the wrong strategy for persuading researchers, their institutions and their funders of the need to mandate OA, even though PPP access does undeniably have superficial appeal with voters and politicians; in the end, on its own, or in the lead as the primary rationale for Open Access, PPP access would lead to debate, delay and defeat for a self-archiving mandate.

But using PPP access needs as the primary rationale for OA needs is not necessary. The solution is to put the irrefutable direct needs of researchers for research access (for the sake of the research and application benefits -- "CURES" -- it will provide for patients, practitioners, cures, the public) first, and note that OA will also provide PPP access as a side-benefit wherever wanted or needed.

It is ever so important not to weaken the case for OA -- the case that must be put to the researchers and their institutions and funders, across all fields -- by giving primacy to access by patients and practitioners. They will get access anyway. But they are not the research providers: Researchers are; and most of them don't do clinically relevant research; and even those who do are rewarded for their research impact, and not yet for their practical impact. (They will be rewarded for the latter after OA prevails, but not before, so that cannot be used to induce them or their institutions and funders to self-archive: research impact can, and it gives everyone else access too.)

I hope you understand these issues of logic and practicality better now: Only a small fraction of research is PPP-relevant, so the need for PPP access cannot be made the principle argument for OA or OA will lose.

Now some comments:

ANON: " I don't think that folks understand this distinction well. You and I do but researchers=lab to the more social sciences. We have a large health science program here and our faculty have "divisions" (i.e. research faculty versus clinical faculty). It is from these clinical faculty I have extended my appreciation of the problems in the field. When I read your 8-point agenda I believe that the clinical faculty would feel that they were not being embraced in it."

If the clinical faculty publish research (i.e., if they are OA providers), they are embraced by it. If they merely use research, they are irrelevant to a mandate that addresses research providers. However, since OA means OA for everyone, clinicians (indeed, all of PPP) are embraced by its outcome, which is Open Access to all the research they need.

Please distinguish what concerns research providers from what concerns research users. The OA problem is that of getting the providers to go ahead and provide the OA (and the solution is to mandate providing it). And the users are the beneficiaries (whether researchers, practitioners, patients, or the public). (Moreover, the public benefits incomparably more from the CURES than the READS). Please do not conflate the problem of getting access (the user problem) with the problem of getting providers to provide OA (the mandate problem). The solution to the mandate problem is also the provider solution to the user access problem.

ANON: "As we try to go about courting our disciplines I think that the language is important when we cross over to the professional/social sciences. There are few, if any, practitioners of particle physics. But there are lots of nurses, social workers, educators, and so on who could use the research but they are challenged to get it.... the situation is really grim... once students leave the school and move to "disconnected" areas of which there are many)."

You are mixing up the user problem and the provider problem here: The point is that providers have to be mandated to provide OA. You are also mixing up the (minority) practitioner-relevant OA fields with the vast majority of practitioner-irrelevant OA fields. OA and OA mandates need to cover them all, and the research impact argument is the decisive and universal one, not the practitioner argument, which is a minority special case, and could be strategically manipulated by publishers with special side-deals.

By the way, students could be added to PPP too, making it SPPP, and the same argument applies to them: OA gives them access along with the territory, and eventually their usage will be measured and credited too, through download counts. Moreover, to the extent that students are or become researchers, their usage also translates into citations and more research (and "CURES").

ANON: " I think that all that needs to be added is something along the lines of "research-practitioners benefit [from OA] too" and this is particular important to "isolated", "international" and "less-resourced" communities."

It's fine to add SPPP needs to research needs in the overall rationale for OA wherever possible (though I think it is already covered by "all would-be users"). Eventually, Connotea-style tagging will help quantify SPPP need and its benefits, the way it is already quantified by research citations...

Stevan Harnad
American Scientist Open Access Forum