Re: The Green and Gold Roads to Open Access

From: Stevan Harnad <>
Date: Tue, 23 Mar 2004 17:27:20 +0000

On Tue, 23 Mar 2004, Waaijers, Leo wrote:

> As someone must bear the costs, this someone must be the author's
> institution then. However, in many a case this is the same institution as
> the reader's. So, at the end of the day the financial effects of both
> approaches (toll gate, or 'open submission' as Declan Butler calls it
> elegantly, and 'open access') meet at the table of the financial manager of
> the institution.

Yes, both Toll-Access (TA) Journal-Publishing and Open-Access
(OA) Journal-Publishing are paid for by the institution -- the
reader/institution in the one case and the author-institution in the

But it must be noted that OA provision via author self-archiving is
orthogonal to this; it is done in parallel. It neither decreases nor
increases an institution's expenses (the annual expense per paper
self-archived is truly trivial). It merely increases an institution's
access to the research output of other institutions -- and increases
the impact of an institution's own research output. The latter in turn
usually means more research funding for the institution. But not for the
library. The library spends neither more nor less, and receives neither
more nor less, with institutional self-archiving.

It is this that needs to be borne in mind in reckoning the costs and
benefits of institutional self-archiving, not its implications for the
institutional library budget!

> And, whether you like it or not, (s)he wants to compare.
> When promoting open access in the Netherlands, I am confronted with
> questions about the underlying business model of open access.

We are once again back to OA journal-publishing, its business model,
and its implications for the library journals expenditure!

This has (almost) *nothing* to do with the benefits of self-archiving.
To reckon it this way is to force both self-archiving and OA itself into
a Procrustean bed, to try to shape it according to this arbitrary
(and almost unrelated) metric.

(I say "almost" because in fact there is the hypothetical possibility that
the growth of OA via self-archiving might eventually save libraries money.
But that is hypothetical, whereas the research benefits of OA are
immediate and objective, and have nothing whatsoever to do with library
budgets one way or the other!)

> For open access journals, the gold road, this is not too difficult. I can
> easily demonstrate that the scientific community pays Elsevier $ 8000 for
> having an article refereed, published and made accesible to a minority of
> that same community, where BMC asks $525 and PLoS $1500 for refereeing,
> publishing and making the article accesible to everybody. The subsequent
> discussion is then reduced to the question whether BMC is too cheap or PLoS
> too expensive. I allways answer that, contrary to the subscription world,
> the open access world operates in a market situation and that will keep
> prices competetive.

That's all fine, but again irrelevant, because you are again comparing
OA *publishing* and its costs with TA publishing and its costs, whereas
I am talking about OA *provision* (through self-archiving).

Although the calculations look more concrete, I believe that the
probability of eventual overall library savings arising from a conversion
to OA journal publishing is even smaller and more remote than the
probability of eventual library savings arising from OA self-archiving --
the "almost" I spoke about earlier -- because the first of these
probabilities is based
not on what it would cost or save per journal but on the probability
that many, most, or all journals will convert to the OA publishing model,
and when!

Your projected savings look attractive enough, but they are based on a
hypothetical large-scale conversion that has neither taken place nor shows
signs of taking place. The growth of the number of OA journals reported
in the Directory of Open Access Journals (DOAJ) during the past year (798
OA out of about 24,000 peer-reviewed journals in all at the moment)
was merely a growth in the *reporting* rate (because no such directory
existed previously). Many of those journals have been OA for years now.
Once the reporting catches up, we will be able to track the absolute
number of OA journals and the number of OA articles they publish,
their proportion of total journals and articles -- under 5% today --
and, most important, their rate of growth.

Without realistic indications of significant growth I would suggest
that your budgetary calculations are rather beside the point.

Besides, this has nothing to do with OA self-archiving, for which we
*do* have current size estimates (there are at least 3-5 times as many
TA articles per year being made OA through self-archiving by their authors
today than through being published in OA journals). OA Self-archiving is
also growing:

Though until stable data are also available for the growth in the number
of articles in OA Journals, it will not be possible to compare the
growth rates.

One thing can already be said with certainty, though: The growth rate of
self-archiving could easily be made much faster! All it requires is an
official OA provision policy on the part of universities:

At least 55% of publishers are already "green" -- i.e., they have given their
official green light to author self-archiving.

So whereas self-archiving is right now around 20%
all it would take would be official institutional OA Provision
Policies to immediately
raise that to at the very least 55%
but in fact much more likely 100%

Meanwhile, however, your own calculations are all focussed exclusively
on "gold": the 5% of journals that are OA:

What you need to ask yourself -- in terms of probability and practicality
-- is which is more likely: That the 5% gold will rise to 55% or 100% gold
by creating or converting more OA Journals? or that the 20% green will
rise to 55% or 100% green by creating more Institutional Eprint Archives
and Institutions making up their minds to adopt policies to fill them!

Fortunately, this is not even an either/or question: We can and should be doing
both, in parallel. But this will not happen if we keep ignoring or holding
back or failing to promote green because of calculations we do on gold!

It is also a good idea to put oneself in the shoes of one's opponent:
If you were a journal publisher today -- not a rapacious journal
publisher, just one stably making ends meet via the TA cost-recovery
model -- would you be inclined to make the sacrifice and take the risk
of converting to the OA (gold) cost-recovery model today, because the research
community wants and needs OA, as demonstrated by the dramatic data on the
impact-enhancing effects of OA?

You are not rapacious, so you don't want to oppose what the research
community wants and needs in the face of its demonstrable power to enhance
research impact. But you don't dare take the risk of converting to gold.
Is it not a sensible -- and fairer -- compromise to convert to green,
and let the research community self-archive for itself (at no sacrifice or
risk to itself),
if it wants and needs OA so much, rather than to take all the risk
and sacrifice on yourself? Would you not perhaps even begin to doubt whether
the research community really wants and needs OA, if it does not even make the
effort to provide it for itself when given the official green light?

*This* is why we need official institutional OA provision policies,
right now! Otherwse even the credibility of our clamour for OA is at risk of
being discounted as not being serious or sincere.

> But the business case for the green road is far more difficult to explain.
> First we have to pay Elsevier $8000 for the publication of the article, then
> we have to beg permission for self archiving it and then some institution
> has to put it in its institutional repository (again costs!) to make it
> worldwide accessible.

You misdescribe the status quo as well as the premises:

(1) We are already paying Elsevier the $8000. Whether or not we elect
to continue to do so is a journal-budget matter, but it is completely
orthogonal to OA now!

(2) No one needs to beg Elsevier for anything! There is already
blanket permission from all 1700+ Elsevier journals to self-archive the
preprint. Please first go ahead and at least get that done, and then we can
continue this conversation!

(3) As to creating the Eprint Archive: The software is free, the start-up is
a few days sysad work, the upkeep is minimal, the cost is negligible, and
over 100 institutions already have them worldwide:
All that's needed is a policy of *filling* them -- and that will not come from
contemplating hypothetical
library journal savings from a hypothetical transition to gold!

> To hesitant looks my defense is: "It's better than
> nothing", but I always have the feeling that I am not very convincing. I
> could parafrase you and try: "Money is irrelevant". Still not convincing, I
> am afraid.

You are not very convincing because it appears that you still have not
sorted out the logic of it, nor where its benefits lie!
It is not about saving the institutional
library money. It is about maximising the impact of institutional research

> May be that's why for articles the green road is less succesful than we al
> wish it to be.

Please let us get the relative success story in order:

Gold is currently at 5% and growing at a not-yet known rate. Its obstacles
are: creating or converting 23,000 TA journals, finding the money to
fund them, and convincing authors to publish in them.

Green is currently at 20%, growing well, but not nearly fast enough
yet. Its obstacles are not creating Institutional OA Archives (that is
cheap and easy) but just convincing authors to self-archive in them. To
convince them to do that, even the green light from 55% of publishers
has not proved to be enough. It requires an institutional OA provision

*That* is what you should be telling your university administration
and researchers. Not "It's better than nothing"!

And yes, money is irrelevant! Self-archiving will not make your
institutional library's journal expenses more or less. That's not
what it's for! It's for maximising your institution's research impact,
and pari passu, for maximising your institution's access to the research
output of other institutions.

If you want to be convincing, you have to be convinced. And to be
convinced, you first have to understand!

> Also because the permission to self archive it, is not always
> given or, when it is given, is limited to access within the author's
> institution.

That is simply incorrect. Please consult
so you can give accurate advice on this point.

> For the time being it is better than nothing, but it is not a
> sustainable solution in my opinion.

Well, even characterized as better than nothing, it is indeed better
than nothing. But better compared to what? OA Publishing alone?

And a sustainable solution for what? Library journal budget problems,
or Open Access?

Stevan Harnad

NOTE: A complete archive of the ongoing discussion of providing open
access to the peer-reviewed research literature online (1998-2004)
is available at the American Scientist Open Access Forum:
        To join the Forum:
        Post discussion to:
        Hypermail Archive:

Unified Dual Open-Access-Provision Policy:
    BOAI-2 ("gold"): Publish your article in a suitable open-access
            journal whenever one exists.
    BOAI-1 ("green"): Otherwise, publish your article in a suitable
            toll-access journal and also self-archive it.
Received on Tue Mar 23 2004 - 17:27:20 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:25 GMT