It is with great pleasure that I acknowledge that Canada's
Social Sciences and Humanities Research Council has selected my proposal Sustaining the knowledge commons (open access scholarship) for funding in the amount of $71,000 Canadian for the period 2014 - 2016. This suite of research projects involves an open research approach; to facilitate sharing of knowledge in this area (both existing knowledge and new knowledge gained through this suite of research projects) I've created a new project blog,
Sustaining the Knowledge Commons (open access scholarship) which can be found at
sustainingknowledgecommons.org A summary of the research proposal has been posted on this new blog and is copied below for the convenience of readers of IJPE. Watch for more on this as the research unfolds.
Sustaining the
knowledge commons (open access scholarship)
Summary
Open access literature is digital, online, free of charge
and free of most copyright and licensing restrictions (Suber, 2012). Open
access to the scholarly literature is a public good that will “accelerate
research, enrich education, share the learning of the rich with the poor and
the poor with the rich, make this literature as useful as it can be, and lay
the foundation for uniting humanity in a common intellectual conversation and
quest for knowledge” (Budapest Open Access Initiative, 2002). Much has been
accomplished in transitioning scholarly works to open access in the past
decade. The Directory of Open Access
Journals lists about 10 thousand fully open access, peer reviewed scholarly
journals. The relatively new Directory of
Open Access Books lists over 1,500 open access scholarly monographs from
more than 50 publishers. There are thousands of open access archives around the
world, housing millions of documents. Hundreds of research funding agencies and
universities require open access to the results of research they support, as
listed in the Registry of Open Access
Material Archiving Policies. However, much remains to be done, and in
particular there is a need to address the challenge of ensuring sustainable
funding mechanisms for open access publishing, especially high-quality
publishing led by scholars that is free to prioritize the needs of scholars and
the public good (as opposed to the single bottom line of profit).
The proposed research would create a macro-level analysis on
the economics of an open access scholarly journal publishing system through the
examination of three separate but related areas of research: 1. a thorough
in-depth analysis on the article processing fees charged by some open access
journals; 2. an examination of the resources needed by small not-for-profit
scholar-led publishers (e.g. needs for editorial or technical support); and 3. supplementing
the work of Edgar & Willinsky (2010) that found an average revenue of $188
per article for journals using Open Journal Systems, by factoring in infrastructure
costs for journal hosting services. The proposed budget allocates over 80% to
research assistantships as this proposal is my first step towards the creation
of a research centre and the active participation of students is a crucial
first step to the research centre. In addition, the proposed research will help
university libraries to make prudent decisions to transition the underlying
economics of scholarly publishing (currently the vast majority of funding for
the system comes from university library budgets) from a subscriptions /
purchase to an open access basis. An open research approach to knowledge
dissemination will help journals and libraries struggling with the transition
early on in the process through sharing knowledge to date through a dedicated
website. This research falls under the SSHRC priority area Digital Economy. Knowledge gained through this project is expected
to benefit the other newer “open” movements, including open education, open
data and open government through the development of scenarios for transition
that might be worth considering in these other areas.
This proposal builds on research developed for my 2012
doctoral dissertation, Freedom for
scholarship in the internet age (Morrison, 2012) and published in First Monday (Morrison, 2013) that
strongly suggest a potential for transitioning to a scholarly communication
system that is more cost-effective than the current system as well as being
fully open access.
Objectives
The overarching objective is to understand how best to bring
about the transition of published scholarly works to a knowledge commons. The
knowledge commons is conceived of as a system where the world’s scholarly
knowledge is available to everyone, everywhere, to draw from and contribute to,
one that prioritizes the values and needs of scholars, scholarship, and the
public good, and is open to all by default, with exceptions as necessary to
accommodate other social values such as the right to individual privacy.
The specific objective of the proposed research is to address
the question of how to transition the underlying economics of the system from a
subscriptions / purchase model to one that funds works at the production stage.
This research will develop a macro analysis of the economics of scholarly
publishing that demonstrates the potential to transition the global spend of
the world’s university libraries from support for subscriptions / publishing to
support for open access production and to achieve significant cost savings at
the same time. This issue was developed as part of my dissertation research
(Morrison, 2012) and subsequently published in First Monday (Morrison, 2013).
The proposed research will make it possible to develop more
accurate scenarios for the potential for overall transition of the scholarly
publishing system by collecting and analyzing data on key components of the
system: the current costs per article charged by existing open access
publishers; the necessary future costs based on the actual resource needs of
scholar-led publishers; and infrastructure costs for support services such as
university library and university press journal hosting services. A more
accurate economic analysis will assist university libraries with making the
case to collaborate to transition support for scholarly publishing from
subscriptions / purchase to support for production so that works can be open
access.
Context
The context section consists of several sub-sections,
covering the history and present of scholarly publishing, the theoretical
framework and major research to date on the three sub-projects.
Until the Second World War, scholarly societies published
nearly all scholarly journals (Mabe, 2003, 2011). Journals were published in
print and distributed to society members and subscribers, many of which were
university libraries. After the Second World War, the commercial sector became
involved in scholarly journal publishing, particularly in the area of science,
technology and medicine (STM), areas of scholarship that the commercial sector
in general had become interested in as technology was viewed as having the
potential to open up new areas for commercial exploitation (Price, 1963). In
the next few decades a “serials crisis” developed, as documented by the
Association of Research Libraries (1989). Even the largest university libraries
were no longer able to purchase all of the scholarly journals. Average journal
prices increased at rates above inflation year after year. During the same
period, scholarly monograph purchases by university libraries declined from
about 3 – 5,000 copies per monograph to about 300 – 500 copies per monograph,
causing a different kind of crisis in scholarly monograph publishing (Thompson,
2005). Brown (2007) describes a university press system in crisis.
This situation illustrates what I call irrational rationality (Morrison, 2012). Universities, through
their libraries, fund a system where a small number of very large commercial
publishers enjoy exceptional profits in the 30-40% range while in the same time
frame they reduce or eliminate the modest subsidies traditionally provided to
university presses. Every element of this system is rational. For-profit
corporations are expected to return maximum profits to their shareholders. In
tight financial times, it makes sense that universities cut services like
university presses that serve the whole system but are not essential to their
own operations. However, all of these rational elements add up to a system that
funds extraordinary profits for a few scholarly publishers while threatening
the existence of other scholarly publishers and even the careers of scholars
who need to publish monographs and find it increasingly difficult to do so
(Harley et al, 2010). The concept of irrational rationality builds on the
intellectual tradition initiated by Weber (1968) in the nineteenth century, Lukács
(1967) and Marcuse (1964) who articulated the difference between rationality
based on values and goal-oriented or instrumental rationality. A real world society-wide
example of this in modern society is the contrast between the common human
value of having an ongoing ecosystem capable of sustaining a high quality of
human life for ourselves and our children and our inaction on climate change,
identified by the World Economic Forum (2013) as one of the ten top global
trends for 2014.
Rather than analyze irrational rationality I propose a holistic
or systemic approach. This is illustrated by my macro level analysis of the
global spend on scholarly journal articles by university libraries
(approximately $5.6 billion annually) and the global production of articles
(approximately 1.5 million annually) (Morrison, 2013). In recent years the
commons has emerged or re-emerged as one potential alternative. A number of
scholars have written about the potential of information technologies to
facilitate enclosure of intellectual property as an emerging stage of
capitalism. In 1989, Mosco described this as the pay-per society in which emerges usage charges for things that used
to be free or charged for on a blanket basis as having the potential to
radically change society; Lessig (1999) makes similar arguments in Code: and other laws of cyberspace. Enzsenberger
(1974) described the potential of new media to facilitate a new more democratic
form of communication, while warning that social action would be necessary in
order for the technology to fulfill this function. Ostrom’s (2000)
groundbreaking Governing the commons
effectively debunks arguments against the impossibility of collaborative
approaches such as Hardin’s notion of the “tragedy of the market” and provides
substantive evidence of highly effective commons-based approaches. Heller
(1998) warns about the tragedy of the anti-commons. Bollier (2007), Boyle
(2003), Lessig (2004), Vaidhyanathan, and Hess & Ostrom (2007) discuss the
enclosure movement and the potential of the commons in terms of culture, knowledge
and information. Caffentzis (2012) focuses on the knowledge commons. In seeking
alternatives it may be wise to consider the perspectives of other societies
such as the first nations approach to intellectual property as articulated by
Young-Ing (2006) and the idea of the gift in traditional societies as
explicated by anthropologist Mauss (2002).
This research builds on macro analysis of the economics of
scholarly publishing conducted by the International Association of Scientific,
Technical, and Medical Publishers (STM) designed for the purpose of business
planning for STM members (Ware & Mabe, 2012, 2009). The United Kingdom has
been a leader in conducting in-depth economic analysis of the scholarly
communication system at a national level. The Research Information Network
(2008) released the report Activities,
costs and funding flows in the scholarly communications system in the UK. Houghton
and colleagues have conducted major macro-economic analysis of the potential
for transition to an open access system at a national level. The most in-depth
research, conducted in the UK, found that cost savings could be achieved with a
full switch to open access by the UK for its own research with 3 different
methods, with the smallest savings resulting from a full switch from
subscriptions-based to open access publishing, greater savings with open access
archiving, and the greatest potential savings through a more transformative
approach, using open access archives with a peer review overlay (Houghton et
al, 2009a, 2009b). This research also draws on global best estimates of the
world’s scholarly peer-reviewed journal production (Björk et al, 2008, 2010).
The third project in the overall proposal deals with research
on economic models for supporting open access, including Crow’s (2006) work on
publishing cooperatives, and the Scholarly Publishing and Academic Resources
Coalition’s Income models for open
access: an overview of current practice (Crow, 2009), and an overview of innovative models for support for open access
focusing on collaborative support by libraries that I wrote as the literature
review for a national survey on library and university press publishing
(Taylor, Morrison, Owen, Vézina and Waller, 2013). One result of this survey
was a finding that Canadian university libraries would be willing to support a
number of different approaches to funding open access, with collaborative
approaches being the one option that all libraries would support to some
extent, and none would (a priori) oppose. Examples of collaborative models for
support including the ongoing work of the Stanford Encyclopedia of Philosophy
to use library funding (similar to subscriptions) to build an endowment to fund
ongoing open access for this scholar-led encyclopedia (Sanville, 2005). Another
example: as of January 2014, the Sponsoring Consortium for Particle Physics
Publishing (2014) is implementing a new fully open access publishing system for
all articles in high energy physics,
a remarkable accomplishment involving global cooperation among libraries and
negotiations with all publishers involved in this area.
A key element for the macro analysis of the economics of
scholarly publishing is the average cost per peer-reviewed journal article.
Previous research indicates a wide range of actual costs, with Willinsky
reporting a range from 0 to $20,000 per article (Willinsky, 2006). The low end
of the range is made possible by the large percentage of work contributed by academics
on a voluntary basis by scholars. Harvard’s Shieber (2012) explains how the
high-impact peer-reviewed Journal of
Machine Learning manages on $10 per article. My macro analysis (Morrison,
2013) illustrates the importance of this key element. If the average cost per
article in an open access environment were $1,350 (the current cost per article
for publishing in PLoS ONE), then libraries could fund a global system at one
third the current global library spend. Even at an average cost of $2,000 -
$3,000, substantial savings are possible. However, at an average of $5,000, the
system would cost more than at present.
Edgar and Willinsky’s (2010) survey of over 900 journals
using the open source software Open Journal Systems found an average spend of
$188 per article. This suggests that in addition to prioritizing scholarship
this “renaissance of scholar-led publishing” described by the authors might be
a great deal more affordable than the current system as well. The proposed
research builds on the work of Edgar and Willinsky, aiming for a more accurate
costing taking into account such elements as the cost of universities’ hosting
and support of journals, as well as whether these new journals have factored in
the necessary resources for
sustainable open access publishing in the long term. For example, do these
journals reporting such low costs have all of the resources needed to sustain
their journals in the long term, or is there an over-reliance on volunteer
labour with the potential for burn-out? If this is the case, what resources could
be provided (copyediting, teaching relief, technical support and training,
etc.) that would prevent this from happening? Currently I am conducting
preliminary interviews designed to develop one or more focus groups and survey
research and several interviewees to date have expressed their perspective that
this research is very much needed.
This project draws on existing and emerging research on
costs of the production of scholarly journal articles, including a major report
conducted by the Wellcome Trust (2004) on the cost of open access journals, a
body of research based on print journals relying on subscriptions as reported
by Tenopir & King (1998), and the data gathered by the Sherpa RoMEO service
on publisher open access article processing charges (based on publisher
self-reported average prices and with an emphasis on UK-based publishers and
the commercial sector). This project will supplement and extend existing
knowledge about costs of producing scholarly journals articles through sampling
open access article processing fees of journals listed in the Directory of Open
Access Journals, a broader and more international set than that included in the
Sherpa RoMEO list, and by triangulating data from interviews, focus groups, surveys
of scholarly publishers and infrastructure costs from case studies of
university hosting and support services to develop a more accurate range of
scholarly journal production costs.
A mixed methods approach is used
because the macro level analysis of the potential for economic transition to
support open access publishing requires several different types of variables.
Open Access Article Processing Charges
The Directory of Open Access Journals is a vetted list of
about 10 thousand fully open access (articles freely available on publication) peer-reviewed
journals. About a quarter of these journals charge article processing fees,
with a small percentage charging article processing fees on a conditional basis
(about two-thirds do not charge article processing fees). DOAJ notes whether
journals charge article processing fees and provides a URL to look up
information, but does not list the amount. A copy of the DOAJ list of journals
charging article processing fees was made in November 2013 and the process of
looking up the amounts was initiated in December 2013. Preliminary results
indicate a skewed distribution with most of these journals published by
relatively large publishers (20 or more journals charging article processing
fees) or very small publishers (1 to 3 journals, with 1 being the most frequent
number of journals). A stratified random sampling method is used to ensure that
a selection of journals from large, medium and small publishers is included.
Preliminary analysis indicates a much more complex situation than anticipated.
While some journals have a straightforward article processing fee, a large
portion have variable fees depending on such variables as location of author,
membership in a society and article length. There is a wide range of article
processing fees, from about $20 US per article to $5,000 per article. The
purposes of this portion of the overall project are to capture another set of
data for comparative purposes, a step which is advisable as some charges may
change (e.g. one publisher, providing a complete set of article processing
fees, advised that some of the journals’ charges might be changed the following
month, and one rationale for this research is to track the possibility of the
commercial sector raising prices at rates above inflation as happened with
subscriptions), and to complete the in-depth analysis of existing charges.
Examples of questions to be explored include whether some of the journals
listed as charging article processing fees are actually producing both print
and online open access journals, with traditional page charges for the print
version and no open access article processing fee. Preliminary analysis
suggests that this is the case for at least some of the journals.
Resource Requirements for Small Scholar-Led Journals in an Open Access
Environment
This project builds on preliminary research in the form of
informal interviews with editors of small scholarly journals currently
underway. Responses to date for a call for participation sent to select
scholarly publishers’ and open access listservs (Canadian Association of
Learned Journals, Scholarly for Scholarly Publishers, Global Open Access List)
indicate a keen interest among publishers in the results of this research.
Inductive methodology will be used to develop one or more focus groups with
publishers from the qualitative results of this research, to be held in
conjunction with the annual general meetings / conferences of the publishers’
associations, and an online survey to be sent to stratified samples of open
access and non-open-access scholar-led journals (i.e. scholarly society and
independent scholar-led journals).
Infrastructure costs estimate
3-4 case studies of library publishing services will be
conducted, representing different types and sizes of library journal hosting
services. For example, the Ontario Council of University Libraries (OCUL)
provides a collaborative journal hosting service for member libraries, while
many other university libraries in Canada and elsewhere offer services targeted
to their individual faculty members. An attempt will be made to include at
least one larger centralized service and one individual institutional service,
as well as organizations offering a slightly different package of services.
Macro analysis of costs for a global shift to open access
Results of these three research projects, in addition to
other relevant information gleaned through an ongoing review of the scholarly
literature and monitoring of related initiatives, such as the Social Sciences
and Humanities Research Councils’ Aid to
Scholarly Journals Program and the Canadian Association of Research
Libraries’ open access group, will be used to develop cost projections for a
global shift to open access based on the range of needed costs uncovered in the
study.
This issue of the Dramatic Growth of Open Access highlights 30 indicators of open access growth that are beyond this background growth of scholarly works - in many cases far beyond, with a range of percentage growth from 5 - 89%. In some cases, high percentage growth reflects early start-ups (low starting figures), but in other cases there are very high growth rates on resources that were very, very large to begin with (these are the highlighted numbers below). Note that some numbers are rounded for ease of understanding; if precise numbers are required, please download the full dataset from the DGOA dataverse.
