APPENDIX 74
Memorandum from the Institute of Physics
The Institute of Physics is a leading international
professional body and learned society, with over 37,000 members.
The Institute's primary purpose is to promote the advancement
and application of physics: through support to education, research,
industry and business and engagement with the public and key stakeholders.
The Institute welcomes the opportunity to respond to this Inquiry,
which is particularly relevant to our extensive learned society
publishing activities. The attached annex summarises the Institute's
views.
The Institute has published academic journals
continuously since its foundation in 1874. Today the Institute's
publishing is carried out through a wholly owned subsidiary company,
Institute of Physics Publishing Ltd (IOP Publishing). The company
employs more than 200 people in Bristol and has offices in the
USA, Germany, Russia, China and Japan. IOP publishes 40 journals,
six technical magazines and between 40 and 50 new book titles
each year, with journals being by far the largest area of activity.
IOP journals are international in terms of content
and circulation. Around 92 per cent of submitted papers and 94
per cent of subscription income come from outside the UK. This
performance has been recognised by the award of the Queen's Award
for Export Achievement in 1990, 1995 and 2000. International sales
of IOP journals generate a surplus that is transferred annually
by Gift Aid to the Institute. Income from publishing forms the
largest element of the Institute's total income for its charitable
activities in support of physics in the UK and Ireland.
ACCESS TO
IOP PUBLICATIONS
The notice of this Inquiry says, inter alia,
that the Committee "will be asking what measures are being
taken [by] the publishing industry...to ensure that researchers,
teachers and students have access to the publications they need
in order to carry out their work effectively".
Most users of IOP journals are in universities,
some in government laboratories and other research institutions,
and a small number in industry and commerce. Access to IOP journals
has been dramatically improved by the internet, both in the UK
and around the world. It is our intention that every researcher
who wishes to read an IOP journal article is able to do so. We
believe we have accomplished this within the existing model of
journal subscriptions. The following have been the main influences.
Electronic access: IOP was the first science
publisher in the world to place all its journals on the internet,
in January 1996. Any institution that subscribes to an IOP journal
receives the print copy. In addition, every person in the institution
can read the journal electronically, by accessing the IOP website
from any computer on the institution's network. The number of
IOP articles read on the internet has almost doubled each year,
from about 50,000 full-text electronic downloads in 1996 to more
than three million in 2003. These figures far exceed the readership
that was possible in paper alone.
Consortium arrangements: in a typical consortium,
a group of university libraries pools resources to buy journals
collectively. We supply all our journals electronically to all
members of the consortium, with print copies available if required.
We currently have 42 consortium arrangements in 22 countries,
covering more than 1,000 institutions. In the USA over 200 universities
are covered by contracts with 13 regional consortia. We have national
consortium contracts in Canada, Denmark, India, Iran, Russia,
South Korea and Switzerland, and major contracts in China, France,
Germany, Greece, Italy, Sweden, Taiwan and Turkey. Overall, about
a third of our subscriptions are now in the form of consortium
arrangements. As a result the effective distribution of our journals
has increased fourfold, without a concomitant increase in cost
to the research community. We have further arrangements to permit
very low cost, in some cases free, access to less developed countries.
Free for 30 days: every paper published in an
IOP journal is accessible free to all for its first 30 days on
our website. This means that all researchers, whether or not their
organisations subscribe to any of our journals, can read and download
the latest research papers published by IOP.
Document delivery: finally, it is always possible
to buy single articles from an IOP journal through any one of
several document suppliers.
UK: we make special arrangements to ensure access
to researchers in the British Isles. Our first national arrangement
was in the UK between 1996 and 1998, when HEFCE (Higher Education
Funding Council for England) made national "site-licence"
arrangements with IOP and two other publishers. When this national
site licence ended, we continued to supply all our journals electronically
without charge to home universities who maintained their subscription
to any IOP journal(s).
Five years on, we now have a choice of options
for home universities with physics departments. Some choose to
subscribe to all our journals both in print and electronically.
A significant group is in a consortium receiving all IOP journals
electronically. Others opt on an individual basis to receive some
journals in print and the rest electronically for a small additional
cost. Finally, a small number buy only the few titles in which
they have an active research interest. In addition, individuals
who are members of IOP can buy the journals, for personal use,
at very low cost.
INTEGRITY AND
THE SCIENTIFIC
PROCESS
The notice of this Inquiry also says, inter
alia, that: "The inquiry will examine the impact that the
current trend towards e-publishing may have on the integrity of
journals and the scientific process."
In the experience of IOP, e-publishing does
not in itself threaten the integrity of learned journals. Indeed,
the greater access to and use of journals through the internet
strengthens their role in the research community. Both established
and new journals attract papers only if they create a powerful
academic reputation. Electronic publication is an advantage equally
available to all, and a disadvantage to none.
However it is important to recognise that the
availability of information on the web is changing our culture
in many subtle, as well as unsubtle, ways. There are signs that
people brought up with information so freely available via search
engines and key words may not have the same tacit understanding
of information ownership, and hence the need to cite sources,
as those who sought references in paper journals. This is an issue
that should be addressed in the training of scientists and engineers.
In physics, an electronic "pre-print server"
(sometimes called an "e-print server") has existed since
1990, originally based at Los Alamos National Laboratory and now
at Cornell University. Physicists can deposit their papers electronically
for the world to read, free of charge, before the paper has been
peer reviewed. This enables physicists to be aware of their colleagues'
work long before it appears in a formal journal. At the same time,
scientists continue to require formal publication in a respected,
refereed journal, both for dissemination of their work and for
career progression. Many physics papers exist in these two parallel
forms.
In relation to the scientific process, the wide
availability of journals electronically is a huge benefit. Traditionally
the researcher had to walk to the library to browse and read current
journals and to seek earlier references in bound volumes. Today,
the researcher at his or her own desktop can rapidly locate papers
of interest by entering relevant search terms, by electronic browsing,
or by e-mail notification. Full papers are then available at a
mouse-click. References to papers in other journals and even from
historic archives appear at another mouse-click or two. IOP was
one of the first publishers to introduce electronic reference-linking.
We have also digitised and made available electronically our entire
journal archive back to 1874.
Electronic publishing further enhances the scientific
process by permitting more and different kinds of content. Colour
illustrations were formerly rare in journals or had to be paid
for by the author; on the internet they are ubiquitous and free.
Moving images are also common in our electronic journals. Large
data sets can be included or reached by a simple link.
What impact do publishers' current policies on
pricing and provision of scientific journals, particularly "big
deal schemes", have on libraries and the teaching and research
communities they serve?
Without doubt the price increases of journals
in the 1990s led to subscription cancellations. Our members in
university departments were acutely aware of this trend. However,
it has now eased, and at the same time electronic publishing has
led to increased access. IOP's philosophy aims to ensure that
our journals are available to our members and all researchers
in the UK and around the world, as outlined earlier.
The "big deal" describes the offer
by a publisher of a large numbers of journals electronically at
a large discount but still typically at a total cost of several
hundred thousand pounds. Librarians often commit the first large
tranche of their budget to buying, for example, the big deal pack
from a large publisher and a journals database supplier. There
may remain only a small amount of budget to buy journals from
learned society publishers such as IOP and other relatively small
companies. Whilst it is a concern for smaller publishers, we would
not wish to overstate the importance of this issue, particularly
in the UK.
IOP itself offers an all-journal pack, but not
on the big deal scale of the main commercial publishers (the discounted
pack price for all 40 IOP journals is £25,894), and there
is no obligation to buy all IOP journals. A librarian can choose
this pack or various other packs of 3, 5, 8, 12 or 15 journals,
or can simply take individual journals from the catalogue.
What action should Government, academic institutions
and publishers be taking to promote a competitive market in scientific
publications?
The market in scientific publications is highly
competitive today. Journals compete for submissions and for sales.
We cannot readily envisage any action by the
UK Government that would further promote competition. The UK represents
less than 10 per cent of the market for scientific publications
and hence any influence on the economic model would be marginal.
In any event, our academic researchers want their papers to be
seen by their international peers, which means publishing in internationally
recognised journals. The overall funding of universities is clearly
influenced by Government and bears directly upon library budgets
but is beyond the scope of our evidence.
Academics can certainly influence the market,
individually and to some extent through their institutions. Individual
scientists may agree or decline to serve on a journal's Editorial
Board, to referee its papers, and to submit their work to it.
On the whole their choices are determined by their opinion of
the journal's scope and prestige but cases have occurred where
a group of academics has transferred its commitment from one journal
to another because the first journal was considered too expensive.
The academic institution, as employer, could
in theory direct its staff to co-operate with specified journals
and not with others, but few institutions would contemplate such
a policy. The institution can support its library with adequate
funding, and the library in turn can influence the market by its
journal selections.
Some have proposed that each university should
create a local electronic archive where its academics would deposit
all their papers, both from the past and currently, as many individual
authors do already on their personal websites. The papers would
then be available free of charge to anyone who accesses the university's
website. Papers would continue to be published in formal journals
in the traditional way. Where IOP is the publisher, our copyright
practice permits such deposit.
What are the consequences of increasing numbers
of open-access journals, for example for the operation of the
Research Assessment Exercise and other selection processes? Should
the government support such a trend, and if so how?
Journal publishing from the receipt of articles
to their formal publication involves costs, including the cost
of the critical refereeing process which every credible journal
must undertake. The open access model assumes that authors pay
the cost of publishing, and the journal is then made available
to all electronically, free of charge.
IOP created one of the first open access journals
in 1998: the New Journal of Physics (NJP). The term open access
was unknown at the time. NJP was the result of an experiment in
web-enabled alternative publishing models, and the project is
now a collaboration with the German Physical Society (DPG). NJP
has levied an author charge of $500 per paper, this year just
increased to $560. We have found authors reluctant to pay. Even
if all authors paid, we would need a far higher page charge in
order for the journal to cover its costs. From the start, IOP
and DPG have subsidised the experiment by about £100,000
a year and continue to do so.
It is worth noting here that the electronic
readership of NJP is not significantly higher than our regular
subscription journals. In 2003 the number of articles downloaded
from NJP on our website was 97,000, while the average across all
our journals was 81,000 per journal. The average includes some
journals of comparable size and scope to NJP with many more downloads,
and some specialist niche journals with low readership. Free open
access to NJP is certainly a selling point to attract authors
who must pay to publish there, but the statistics do not suggest
that it enjoys a level of readership denied to authors in our
traditional subscription journals.
For open access to become the norm requires
some fundamental changes, for example to the way research and
university infrastructure funding are distributed. Money currently
in the university system for library budgets would need in future
to be delivered through departmental or individual investigator
research budgets. Authors would face a barrier to publication.
At present the dominant criterion for research publication in
any of the world's science journals is approval by the journal's
peer review process (as even those journals which have "page
charges" are likely to waive these for high quality articles
from authors who state that they cannot pay). Under open access
where author charges (ie page charges) were the only way that
the administrative costs of the publication process were met,
ability to pay would be a further criterion.
Publishers in turn would be influenced to accept
papers that came with money attached. Journals would benefit financially
by publishing more papers, potentially increasing the number of
weaker papers in the system—a contrast to the existing model
where the economics favour tough peer review and selection on
merit. Publishers might also be forced to reduce the attention
given to papers, in the processes of peer review, editing, mark-up,
electronic coding, and facilities on the electronic journal website
such as searching and reference linking. Publishers who reacted
in this way, publishing large numbers of papers with little attention,
would be able to charge the lowest author fee and hence attract
more submissions. Such an outcome would destabilise the current
social structure of the academic community where publication in
respected journals is regarded as an important measure of achievement.
On current experience, open access journals
would require external subsidy in addition to author charges.
As far as we are aware, all current open access journals are subsidised
by their publishers, by philanthropic grants or by lump sum subventions
from supporting institutions. Economics suggests this is a less
reliable model than the market system currently in place.
Moreover, IOP does not believe the overall cost
would be significantly less. There is a cost to publishing a paper,
variously estimated between $1,000 and $3,000. The only net money
removed from the system is supposed to be the publisher's profit
margin. However, if open access publishers do operate without
profit then their journals will have an uncertain future.
In relation to the Research Assessment Exercise
(RAE), IOP believes the consequences of open access would create
new difficulties. The last RAE physics panel was chaired by Sir
John Enderby, a vice-president of the Royal Society and President-Elect
of IOP. We asked Sir John for his observations, which are as follows:
"RAE panels were required to define international,
national and sub-national research. Peer recognition was deemed
to be of considerable significance. In physics the panel paid
attention to the overall quality of the cited publications and
undertook to read at least a quarter of those submitted. However,
it was clearly impossible to consider all the remaining papers
in detail so the panel relied upon their perception of the rigour
with which papers were refereed by the publishers.
"Without the collective experience of the
quality controls used by publishers, the task facing the panels
would change dramatically. There are real difficulties in ensuring
that open access and high quality can actually coexist within
a realistic business model. Assessment panels would have to consider
each submitted article, which means that either the panels are
given much longer to do their work or that the methodology of
the RAE is changed.
"A further point is that the present system
clearly "date-stamps" the submitted article thereby
establishing priority and ensuring content stability at the time
of publication. Open access may lead to less than honest practice
against which the panels would need to be on their guard, thus
further increasing their workload."
Should the Government support a trend to open
access? We know the advantages and disadvantages of the present
model. IOP experience with open access has not demonstrated significant
advantages in terms of access to and dissemination of research
results over our conventional journals, and shows the financial
difficulties of moving to open access without a significant change
in the way universities are funded. To replace the present model
with a new model carries both known and unknown risks, without
it being clear that there is a major benefit. For the Government
to fund such a change would mean the Government assuming a large
part of the risk, potentially for a considerable time. That is
not to say that academics and publishers should not continue to
experiment with alternative, web-enabled, publishing models.
How effectively are the Legal Deposit Libraries
making available non-print publications to the research community,
and what steps should they be taking in this respect?
Copies of all IOP's 36 print journals are made
available to the Deposit Libraries in keeping with statutory requirements.
However, the deposit of electronic-only titles is not yet a legal
requirement and most deposit libraries have shown no interest
in acquiring our non-print publications.
Two exceptions are the British Library and Cambridge
University Library: both have online access to all 40 IOP journals
including the four that are electronic-only. Oxford University
has online access to 15 titles by way of subscriptions from the
Clarendon Laboratory, including three of the four journals that
are electronic-only. The National Library of Wales, the National
Library of Scotland and Trinity College, Dublin, do not have online
access to the four electronic-only journals.
IOP would support a statutory requirement for
the deposit of non-print publications including electronic journals,
and we understand that the Legal Deposit Libraries Bill, which
recently came into force, enables the introduction of such a requirement.
However, we wish to urge that access to electronic
journals in Deposit Libraries be available only to authorised
users sitting at computers in the library building. The legal
requirement for deposit is founded on the desire for national
archive and preservation. It also happens to provide access to
deposited publications by library visitors. Historically, publishers
have been happy to co-operate, knowing that any loss of normal
sales consequent upon the free copy in, say, the British Library,
was trivial. That would not be the case in the internet age if
the Deposit Library were to put preservation items on its open
website. The entire existence of journals would be threatened,
and with them the primary means of academic communication.
As noted earlier, IOP has produced its own complete
electronic archive of all its journals back to 1874. We maintain
this archive ourselves and we have also sold copies to many universities
around the world. There are no technical limitations to our offering
a full electronic archive to Deposit Libraries.
What impact will trends in academic journal publishing
have on the risks of
scientific fraud and malpractice?
In October 2003 IOP hosted a workshop organised
by the International Union of Pure and Applied Physics (IUPAP)
on scientific misconduct and the role of physics journals in its
investigation and prevention. It concluded that while physics
has appeared relatively free of misconduct—compared with
biomedicine, for example—scrutiny of recent cases suggests
that it may be more widespread than previously admitted, if "misconduct"
is interpreted broadly.
For example, there appears to be a growth in
plagiarism, probably the most common type of misconduct thanks
to the ease of cutting and pasting text electronically. Other
types of misconduct discussed by the IUPAP group included duplicate
submission or publication; the improper inclusion of an author's
name and conversely the improper exclusion of one; improper manipulation
of data; the reporting of only good results; conflicts of interest
or other bias leading to distorted claims; and referee misconduct.
Partly in recognition of these issues, IOP has
established an Ethics Committee with external membership to set
clear guidelines for the conduct of physicists and to develop
appropriate educational material.
In relation to trends in publishing, it seems
possible that more diverse and widespread means of publishing
could permit more misconduct. In particular any informal electronic
publishing, unaccompanied by supervision, creates such a risk.
Conversely the traditional process of peer review protects against
fraud in two ways: first, the referees and editor may detect the
misconduct; secondly, fearing discovery, the author is less likely
to attempt it. In future availability of electronic tools to identify,
for example, common or closely similar passages of text or figures
will provide a further means to check for plagiarism which can
readily be incorporated into the review process.
The peer review process, probably assisted by
new on-line tools, is likely to remain a critical step in maintaining
quality and integrity in scientific publishing. Any effective
publishing model must therefore be able to deliver a peer review
process which is respected and timely.
February 2004
| 
