Re: The Green and Gold Roads to Open Access

From: Stevan Harnad <>
Date: Thu, 6 Nov 2003 04:52:29 +0000

[This is the reply to a science writer for a forthcoming article on
open access.]

On Wed, 5 Nov 2003, [identity deleted] wrote:

> 1) What do you see as the most important reason to allow open access to
> journals?

There are a number of non-reasons, or side-reasons:

But there is only one decisive, incontestable reason: to maximize
research impact:

If you want to know why research impact is the be-all and end-all of
research for researchers, their institutions and their funders:

Researchers do research (and their research institutions employ them to
do it, and their research-funders fund them to do it, and the tax-payers
pay their funding) in order that the research results should be read,
used, and applied, to the benefit of all of us. That is research impact,
and that is why research is done, and supported. Anything that blocks
access to those research findings is blocking research impact, hence going
against the interests of research, researchers, their employers, their
funders, and the tax-payers that fund the funders.

> 2) Do you see any problems that might result from open access. If yes,
> what, and how might these problems be addressed.

No problems whatsoever. The only problem is how to get there from here.
Right now, most of the planet's annual research output (across all
fields of science and scholarship) -- about 2,500,000 articles per year
-- is published in the planet's 24,000 peer-reviewed research journals.
Of those 24,000 journals, only about 500 (<< 5%) are currently open-access
journals ("gold" journals, in the terminology I will explain in a

Hence only <<5% of yearly research output appears in open-access
journals. The remaining >>95% appears in toll-access journals. The
problem is how to make it all open-access, so all that daily, weekly,
monthly, and yearly research impact stops being needlessly lost because
of access-denial to all those would-be users whose institutions cannot
afford the access-tolls (subscriptions or license fees):

Waiting for 23,500 toll-access journals to convert to open-access or to
be replaced by 23,500 competing open-access journals would be a long
(and perhaps endless) wait, but there is another road to open access
besides the "golden" road of open-access journal-publishing, namely, the
"green" road of open-access self-archiving (by authors, self-archiving
their own toll-access-journal articles in their own institutional
open-access eprint archives):

The "green" strategy is BOAI-1, the first of the two open-access
strategies of the Budapest Open Access Initiative (BOAI). The "gold"
strategy is BOAI-2:

To solve the problem of getting *there* (100% open access) from *here*
(under 10% open access) all that is needed is that researchers either
publish in a suitable "gold" journal, if one exists in their field
(<<5%) or they publish in a suitable "green" journal (one that supports
author self-archiving -- at least 55% of journals already do) and also
self-archive the article.

With the help of new open-access "publish-or-perish" policies on the part
of research institutions and researcher funders, as recommended in last
week's Berlin Declaration, researchers will be choosing gold and green
journals to publish in, which will encourage the remaining "white"
journals to go "green" (i.e., self-archiving-friendly).

The result will be (1) that research and researchers have open access
(though all or even most journals won't yet be "gold," only "green").
Then, if and when -- and no one knows if and when this will happen,
because self-archiving is gradual and anarchic, so it will not be clear at
what point 100% of the contents of any particular one of the 23,500 green
journals are openly accessible, and hence when it would be safe for an
institution to cancel its toll access subscriptions -- but if and when
toll-access cancellations do start occurring because of competition to
the toll-access versions from the open-access versions, this will start
generating institutional windfall toll-savings as it also generates
journal toll-revenue loses.

First journals will cut costs in order to keep making ends meet, and
they will drop products and services that are no longer necessary. These
will include the creation and distribution of the paper version, and
eventually even the creation and distribution of the text will be left
to the self-archiving authors and their institutions. The only essential
service that peer-reviewed journals will always have to be the ones to
continue providing is peer-review itself (and perhaps some editing). The
cost of that is about $500 per article. But what the planet is paying
right now per article (summed up from the tolls paid by those
institutions that can afford that particular journal) is about $1500 per

So it is already clear why there is no problem about any eventual
transition from toll-access journals to open-access journals: By the
time such a transition ever becomes necessary, the research institutions
(i.e., mostly the universities) will have three times the windfall savings
from cancelling the tolls for buying *in* one another's research output
to now redirect one third of their savings toward paying the peer-review
costs for their own research output (already all being self-archived).

The only thing that looks like a problem *now* is how to fund the
existing *gold* journals, before the windfall toll-savings have started
happening. But the solution for that is already falling into place:
Research funders are committing themselves to paying the costs of
publishing in open-access journals such as the BiomedCentral and Public
Library of Science journals (and the Berlin Declaration, as well as the
Bethesda Statement and the Wellcome Trust Statement, support this too).

> 3) If the system stays as is, do you think scientists now have adequate
> access, or do you see the problems as being in the future as fees for
> access increase?

Researchers definitely do not have adequate access in the status quo, but
the system need not change radically: only research-institution and
research-funder policy -- and hence researcher practise -- needs to
change. According to an open-access policy for research publication
(as in the Berlin Declaration), all peer-reviewed research must be made
open access, via either the gold or the green route.

To repeat: It was spiralling journal costs that helped *alert* us to the
access problem, but the access problem is not the size of the tolls but
the access-denial to those would-be users worldwide whose institutions
cannot afford the tolls to the particular journal in which any
particular piece of research appears: There are 24,000 journals, and no
institution can afford toll-access to more than a small portion of them.
So the open-access version is for those would-be users whose
institutions can't afford the toll-access version. But even if all
24,000 journals were charging tolls *at cost*, with no profit margin,
that *still* would not solve the access problem. Access would still have
to be provided to would-be users whose institutions could not afford
access. Otherwise all their potential contributions to research impact
would be lost. That is what open access -- via either the gold or green
route -- is meant to remedy.

> 4) Please add any other comments you may have that would help my
> clinician audience understand your viewpoint.

Clinicians should understand that research publication is about research
impact -- the uptake, usage, application and citation of research. That
is why, for example, citation-impact measures -- as well as other, new,
online measures of impact -- are so important to researchers.
Researchers will want to maximize the impact of their work by
maximizing access to it -- via the gold and green strategies.

Journals need not go gold for the sake of open access, but they do
need to go green, for trying to oppose (by remaining "white") what
is obviously so overwhelmingly in the best interests of research,
researchers, research institutions, research-funders and tax-payers
would not be in the interests of journals, and would definitely risk
losing their authors to other journals that are more responsible and
responsive to the needs of research, and to the powerful new possibilities
that have been made available to them by the online era.

Stevan Harnad
Received on Thu Nov 06 2003 - 04:52:29 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:08 GMT