From: Stevan Harnad <[log in to unmask]>
Date: Wed, 28 Jun 2017 11:24:21 -0400

On Jun 28, 2017, at 9:57 AM, Andrew Odlyzko <[log in to unmask]> wrote:

we could operate an adequate scholarly publishing business, with the
> current level of peer review, at $300 per article, or 10% what it costs
> Elsevier.  The main obstacle is inertia.


"I think that the true figure for peer-review implementation alone
across all refereed journals probably averages closer to $200 per article,
or even lower. Hence, quality-control costs account for only 10% of the
collective tolls actually being paid per article.”

*Nature* *410*, 1024-1025 (*26 April 2001*) | doi:10.1038/35074210

https://www.nature.com/nature/journal/v410/n6832/full/4101024a0.html


Inertia indeed, on the part of the publishing industry, predictably, but on
the part of the research community, deplorably…

*Stevan Harnad*

On Wed, Jun 28, 2017 at 9:57 AM, Andrew Odlyzko <[log in to unmask]> wrote:

> Perhaps a Kazhakstani graduate student can provide simple distribution
> of files at a very low cost.  But once you get into providing anything
> resembling serious curation, and even more when you get into peer review,
> costs do mount up.  For example, arXiv costs about $10 per preprint
> submitted (if we divide the annual cost of the arXiv by the number of
> new submissions, and so don't worry about the accounting niceties of
> splitting the costs between handling new and old papers).  For a few
> million papers per year for all of scholarly publishing, this gets
> beyond the capability of a Kazhakstani graduate student.
>
>
> This rough estimate of $10 per preprint for arXiv, and others to be quoted,
> are all from the paper "Open Access, library and publisher competition, and
> the evolution of general commerce," Evaluation Review, vol. 39, no. 1,
> Feb. 2015, pp. 130-163,
>
>     http://dx.doi.org/10.1177/0193841X13514751
>
> and (for those who can't get inside the paywall), a preprint is at
>
>     http://www.dtc.umn.edu/~odlyzko/doc/libpubcomp.pdf
>
> Going beyond preprint distribution (and the very light level of screening
> by volunteer editors, which does exist at arXiv, at no monetary cost),
> Elsevier collects about $5,000 in total on average for each article they
> publish.  About $2,000 is their profit, and the remaining $3,000 covers
> what they claim are necessary costs.  As many (including your truly) have
> been arguing for a couple of decades, the necessity of those costs (leaving
> the profit question aside) is extremely questionable, and we now have lots
> of examples of lower cost journals.  It seems clear (some estimates and
> references in the paper cited above) that we could operate an adequate
> scholarly publishing business, with the current level of peer review,
> at $300 per article, or 10% what it costs Elsevier.  The main obstacle
> is inertia.
>
> Andrew