Open
access: reflections from the
Based on a paper presented at the UKSG seminar
‘Scientific Publications: Free for all?’
The Geological Society,
ANN OKERSON
Abstract
This article offers a perspective about open
access and the current 'churn' in scholarly journal publication from a
collections development and scholarly communications specialist librarian in
the
For
the purposes of this article I am going to use the metaphor of The Good, the
Bad and the Ugly, which is a classic 1966 film about the search for hidden
treasure. Even the ‘Good’ is badly
flawed person. All three are
inter-dependent yet do not trust each other.
See if the search for ever-improving access to information bears any
resemblance to this tale.
In
the
Even
the largest state (public) institutions receive very little funding from their
states. Let me give you the example of
the
We
have 50 states with their own state university systems, scores of library
consortia, and 60,000 librarians who are members of the American Library
Association. Many have very different views about all kinds of library
issues. However, one of the things that
Americans do seem to agree on is that we do not want the government involved
very much in our lives.
Open
access can be a highly emotional topic, and yet it turns out that no one is
opposed to it. In fact, almost everyone
is sure that they are contributing towards open access in some way or another,
whatever the method or the reason. One
reason might be that they are the Good; another is that open access is at least
to some degree a natural consequence of the ease of information access on the
Internet: the way in which such information ‘leaks.’ Another reason is that there has been a fair
amount of marketplace push-back by librarians and others against the traditional
journal subscription models, particularly of the higher-priced publishers, and
this pushback has made an impact. These
days, we see more competition in the publishing arena, which leads to more
rapid change than we have been used to.
Still, the theoretical agreement about the benefits of wide access
disintegrates rapidly in meetings as people opine about who and what is Good,
or Bad, or Ugly, as if effective scholarly communications did not need the
interdependence of scholars, publishers, librarians, and other players in the
communications chain.
Librarians
think of themselves as Good -- indeed, that is why I became a librarian, to
become part of the mission of bringing knowledge and information to users. Librarians have a long tradition of providing
on-site access to information. We
circulate materials widely and we do this not only in our own institutions but
also broadly for users we don’t even know, in completely different locations,
through inter-library loan and document delivery, whether through high tech or
standard, traditional means. In the
Librarians
have also embraced and transformed our services using electronic technologies
and the Internet. The library community
has been supporting alternative publishing ventures and models, in order to
test their effectiveness and sustainability in a time when technology has
transformed so many publishing-related activities.
I
also believe that publishers can be called Good. Many journal publishers have long traditions
of distributing free or cheap print subscriptions to institutions and countries
that cannot afford to pay in
Publishers
have had to shift their journals from print to electronic publishing mechanisms
and platforms at considerable expense, with extensive planning and
re-tooling. Many publishers have been
willing to participate and experiment with new business models, whether print,
electronic bundles or consortial, and we are now seeing open access
experiments, along with library and consortial payment models to support
them. This experimentation is
extraordinarily beneficial.
Most
publishers have loosened up a great deal from their earlier fears that library
site licenses would lead to numerous copyright violations. A report on RoMEO (Rights Metadata for Open
archiving) here in the UK stated that approximately 90% of publishers do not
require exclusive copyright transfer as a condition for article publication and
allow authors to licence their articles or give extensive re-use rights such as
the right to mount their articles on institutional repositories and/or on their
own home pages. That is an exceptional
change in less than ten years. This is
also Good.
Today's
government interest in STM publishing is also Good. Agencies and funding bodies have generally
been very detached from scholarly publication, what it costs, and who has
access, in the belief that a hands-off policy expresses proper neutrality about
the doing of science. That is changing
with various sectors now asking for agency and government involvement from the
UK House of Commons and the US National Institutes of Health (NIH). We are
suddenly awash in studies, recommendations and suggestions for policy change.
However,
none of the foundations or government agencies that have explored open access
and recommended change have proposed to alter their own dissemination
practices. For example, the US National
Science Foundation should publicly identify all the proposals it has funded,
linked to any progress reports and final reports from the scientists who do the
research. Likewise, the Wellcome Trust,
the NIH and others should do the same.
Instead these agencies are targeting the downstream articles that arise
out of their funded research, articles to which publishers have added value. Why is that?
There
seems to be agreement about only one thing in the open access discussions
(apart from the fact that everyone supports the concept): no matter how it is to be achieved, open
access is not going to be free. Someone
somewhere somehow will need to pay for the process of managing, reviewing,
editing, producing, electronically distributing and hosting the journals or
articles -- as well as delivering numerous value-adding features. After that, the disagreement begins. Open access players, whether pro, con or
neutral, are at odds about many things.
Some of the many areas of disagreement devolve around exactly what open
access is:
§
Does open access require all articles to be free to all
readers from the moment of publication?
§
What version would be satisfactory to open access advocates?
§
Does delayed free access, such as free access after six
months, still count as open access?
§
Or is it a pathetically weak substitute, a kind of sop to
the readers?
§
Do the numerous developing nations’ initiatives count as
open access?
§
How about hybrids in which articles prepaid up-front by
authors are open access and articles that cannot be paid for up-front are not
open access?
Many
open access proponents criticize the ‘imperfect’ compromises, and many
librarians are hyper-critical about a large amount of wonderful change and
progress in a very short time.
Another
set of questions in play these days relates to cost of open access:
§
Just how much does or should open access cost?
§
Who is going to pay for it?
§
How will we sustain this access?
§
Will open access require entirely new business models such
as ‘author pays’/’funder pays’?
§
Will it be done through institutional repositories, and what
is the role of government, if any?
§
Will libraries save money under open access? (I am somewhat skeptical as there is very
little hard data to answer most of these questions, and there is a great deal
of speculation.)
Sometimes
called ‘bundling’, the ‘big deal’ in the
Think
about national and multi-national site licences. The Canadian National Site Licensing Project
has brought huge amounts of content to all institutions of higher learning in
the country. The eIFL project of the
Open Society Institute has resulted in licenses for free access to huge numbers
of journals by readers in dozens of developing nations in the world. Some of the users who participate in those
projects have told me that this looks and feels exactly like open access and
they have said, "You rich westerners should go away and solve your rich
people’s problems. We are now starting,
thanks to the publishers and the web, to get the access we need."
Institutional
repositories (IRs) are sometimes thought of in the same breath with open
access, because so far they have been conceived and set up as the freely
accessible and searchable sites of not-for-profit organizations such as
universities. IRs are intended to
provide opportunities for institutions to care for and disseminate their
authors' creations and allow a wide range of faculty and staff to gain
experience in using information technologies.
The question often arises: will
such institutional repositories provide access to works that have so far been
organized and distributed by publishers?
Will IRs displace or even replace publishers, especially STM journal
publishers as we have known them? Will
IRs be the or at least a path for achieving the open access
ideal?
In
September of 2004, prior to speaking at an STM meeting at the Frankfurt Book
Fair, I surveyed informally 40 of my colleagues in the largest research
libraries in North America and received 16 responses. Of those 16, six have in place some kind of
institutional repository, and the other 10 plan to have one within the next
five years. Frequently expressed
concerns were that:
§
Institutional repositories are costly to create.
§
Institutional repositories and placement of materials in
them are a hard sell on my campus.
§
Long-term sustainability of the IR is a daunting issue.
§
So is a compatible infrastructure, cross-searchable with
other IRs around the world.
§
One librarian wrote: "Personally I am intimidated by
the burden of preservation of archive materials, considering the wide variety
of formats that are likely to be deposited."
§
Several said that the different levels of rights and access
issues loomed complicated and very large.
At
a SPARC (Scholarly Publishing and Academic Resources Coalition) and
SPARC-Europe sponsored institutional repository symposium in Washington DC, one
panel addressed the topic of IRs and rights management. One of the biggest concerns academic authors
have about university IRs is that access be provided to their materials in ways
that are comfortable or appropriate for these authors. Clearly, levels of rights can be complex, and
that those who are responsible for these repositories are spending a great deal
of time handling rights issues, agreements, templates and the campus discourse
about rights and copyrights (much as publishers do)!
In
response to my survey question about the kind of materials that would be placed
in their IRs, the highest priority for many were faculty and student scholarly
electronic works and databases, as well as many electronic research and
teaching/learning projects on campus, projects that have no other permanent
home. Other categories included ‘grey
literature’ (informal publications) created on campus and content the library
staff digitises or creates, such as images (which might not be able to be made
freely accessible to the world). Less
than half of the respondents (7 of 16) will target articles that their faculty
submit to journals, especially in STM fields.
So, will IRs in the US be in a position to replace journal
publishers? At the moment, it seems that
while repositories will expand access to a number of articles, the coverage
will be far from complete or comprehensive, as compared to articles that
continue to be produced in more formal publishing outlets.
My
take-away from the survey, as well as from US discussions in general, is that
we will indeed invest significant resources in IRs on our campuses. We will focus less on STM journal articles
than on the urgent need to care for our institutional creations, such as
faculty and student databases, data sets, teaching materials and research
projects since those do not have any other outlets, whereas journal articles
do. It is unlikely that under this kind
of scenario in the US, scattered local versions of STM articles would compete
effectively with the completeness or the value that the publishing community
adds.
The
costs of journal publishing are not all that well understood, and pricing for
open access is understood even less well.
Open access as advocated these days requires that all costs associated
with initial article/journal creation and long-term hosting, value-adding,
and migration need to be collected ‘up front’, as articles are unlikely to
yield any additional revenues in the future.
(See the BioMed Central e-mail signoff, ‘All Use is Fair Use.’) Happily, several important publishers, among
them the US National Academy of Sciences and Oxford University Press, have
begun open access experiments, from which we all can learn a great deal. Let me briefly describe a couple of examples
that have me far from sanguine that the costs to my library will decrease.
Oxford
University Press's Nucleic Acids Research. The current subscription price for this
important journal is $2,855 for paper plus an online edition. In 2005, this
journal will become open access and the library subscription will remain at the
current level, while faculty will also contribute $500 per submitted
article. We have estimated that about 22
articles would be published in NAR by Yale authors in 2004. If exactly the same numbers hold true in
2005, Yale University would pay approximately $11,000 for faculty
submissions. (If we did not have a
library subscription, our authors would pay $1,500 dollars an article.) Thus, a crude estimation of the price that
Yale University will be paying for worldwide open access to NAR next
year will be just short of $14,000. That
is not an increase for the library, but it is a significant institutional
increase and something to think about.
The
National Academy of Science's journal PNAS (Publications of the National
Academy of Sciences). Our library's
current electronic subscription costs approximately $3,000. As of 2005 this journal will convert to open
access. Authors will be charged an
optional $1,000 dollars per article, because a survey of those authors said
they were not willing or able to pay any more at this time. That $1,000 is certainly well less than the
cost of publishing that article, and will be charged in addition to the
library's subscription cost.
In
short, at this time no reduction in library subscription prices is expected,
and there will be increased costs university-wide. Of course, at some future time, if OUP and
NAS secure sufficient author revenue, they may adjust the library subscription
price downward. Nearly a year ago, I
made some back-of-the-envelope calculations, aimed at assessing whether the
Library would be ahead or behind in an entirely open access publishing
environment. We made some rough
approximations of Yale author output, using PubMed Central and ISI's Web of
Science, from which we estimated that the University's authors had
published as many as 4,000 STM articles in the previous year. We had expended in that year somewhere
between $3.6M - $4M for STM journals. Had
we no library subscriptions for STM journals, and if our
authors each paid $1,000 per published article, we would more or less spend as
much money in the open access environment as we do today on our
subscriptions. However, note that most
publishers estimate the cost of article production as somewhere between $800 to
$5000 with many clustering at the $2500-$3000 level. So, a per-article cost to Yale of only $1,000
does not seem realistic. Nor does it
take into account the many other kinds of subscriptions and licenses we
purchase.
Cornell
University Library staff conducted a similar study -- only far more
meticulously -- with surprisingly similar results. My conclusion from these two examples
(admittedly limited in number) is that savings to large research libraries
under an open access model are unlikely, unless substantial production cost
reductions can be realised by many categories of publisher. I am not sure where those would come from;
but one hopes there is potential for at least some reduction. Also, I have described (in the NAR and
PNAS examples) potential and significant cost shifts away from the
library to other parts of the university, which raises a number of
institutional policy issues. One could
ask: if there is such a shift, and if other
parts of the university are paying for STM publications, could not budgets
for libraries be commensurately reduced?
The
NIH recommended in July 2004 that PubMed Central should host all articles that
result from NIH-funded research, after a maximum lag of six months after journal
publication. The NIH and any funding
agencies are well within their rights to ask this as part of the condition of
providing research grants. Yet, it is
not clear to what extent access will be improved thereby. The NIH currently funds about 20 - 25% of US
biomedical research, which though high is far from
100%. HighWire Press asserts that it
already offers in its free content (its publishers generally exercise an
embargo period or ‘moving wall’ of their choice, anywhere from six months to
two years) about three times as many free articles as PubMed Central.
Publishers
are asserting that there may be better ways than the NIH recommendation to
improve access, and that different versions of the same article between journal
and PMC are likely to cause confusion for readers. Many have asked that the NIH instead permit
links to publishers' own journal sites and have requested as well to choose the
time after which publication will be available for free via PMC. We will know in December 2004 what the NIH has
decided, and only time is going to tell us what the impact of that initiative
will be upon formal journal publications.
[Author's note: as of press time,
2/1/2005, the future of this recommendation continues to be uncertain.]
Our
first commitment is to our users, to meeting their research, teaching, and
learning needs. Next, our commitment is
to describing the scholarly communication system to them, engaging in clear
discourse, so that librarians and users understand each others' needs. As librarians, we need to engage with
university administrations and other partners on campus to share the likelihood
of significant cost shifts within our universities, costs that are beyond the
purview of the library to mandate or pay for.
Finally, we have a commitment to the scholarly communication system
itself. We should treasure its
achievements, which have evolved over several hundred years. It isn’t a bad system, yet it could surely be
better. Thus, we librarians want to be
well informed, to support our colleague publishers’ experiments, to be curious
and open-minded, and to examine evidence that emerges from those
experiments. We must continue to push
back in our e-negotiations with publishers, in order to make clear what both our
users and we need, and can afford to pay. One of my mantras during these times of churn
and change is: "Expect no
windfalls, expect no miracles, expect no dazzling budget relief; be civil and
turn down some of the rhetorical temperature."
My
final point is neither good, nor bad nor ugly.
Let me draw our attention to a couple of recent important articles. Several authors suggested last fall that open
access might be misplaced as the total centre of our attention. As I read their arguments, they resonated
with me a great deal. In the September
2004 issue of D-Lib Magazine on
the Web, Herbert von de Sompel and colleagues published ‘Re-thinking the
Scholarly Communication.' Their thesis
is that open availability is only one small dimension of scholarly
communication. We must look, they
reason, at the rapidly changing nature and pace of research, think about all
the new media that are coming before us, re-think models totally, and develop
goals that hasten discovery and enrich information for our readers. It is apparent that these authors do not
think that the scholarly journal is the centre of all scholarly communications
activity: it is one piece. In October, the Charleston Advisor
published an editorial by the UK's David Worlock, called, ‘Open Access -- Free
for All?’ He asserted, perceptively,
that today's dialogue needs to be about many things: about broken economic models, the
transformation of scientific research, the journal, access, and the power of
the web.
Let
me conclude thus: let us work hard for
the freest possible access to information, while recognizing that as
librarians, publishers, and other key players in the scholarly environment, we
have other, even larger and more encompassing challenges to meet.
Article © Ann Okerson
Ann Okerson
Associate University Librarian
Collections & International Programs
Yale University
e-mail: ann.okerson@yale.edu