Pro-commercial publisher arguments almost always include the term "added value" or something similar. In other words, the big publishers add something beyond the raw manuscript and figures that are provided by the authors. I think very few people will dispute this claim, at least at its face*. The publishers:
- facilitate peer review by paying for a manuscript handling system
(either licensing a commercial product or installing an open source
product on servers they pay for) [note that this is not the same as doing the peer review, which is done by volunteer referees and unpaid or minimally-paid editors]
- do some copy editing
- format the manuscripts into a pretty PDF and web page
- provide a veneer of respectability with well-known journal "brands"
- distribute the journals to libraries and interested readers, via subscriptions, web hosting, and proprietary search engines
- and other miscellaneous things
Look, I appreciate the fact that all of this costs money. Somebody needs to be paid to do the formatting into the appropriate medium (whether web page or PDF), technical staff need to make sure the authors submit the files in the right format, it costs money to run a server, programmers don't come cheap, and all of the various functions of a business/journal aren't free (office space, salaries for necessary employees, etc.).
But does it really cost so much that publishers have to charge $37.95 for a single PDF file, or $392 for a personal subscription to a journal?
Maybe the answer is yes (forgetting the 30%+ profits for many major publishers). Maybe it does cost a lot of money to produce an article. Fine. Just do a better job of convincing me that it's worth it. Particularly when some of the most labor-intensive tasks (typesetting and peer review) are provided for free by the authors and their colleagues.
Many large publishers have an established list of things they do that cost money. They've done a decent job of publicizing these talking points, judging by the facts that they show up so often in comment feeds and that I was able to assemble the bullet points above virtually from memory.
However, publishers have performed miserably at convincing us that $37.95 is a reasonable price for a PDF download. Elsevier and company could deflect much criticism if they were to be more honest and transparent about the costs behind a journal article. How much time/money actually goes into formatting? How much does it really cost to serve a file to the internet, over multiple years? What is the honest per-article cost for the manuscript submission system? How many people actually buy articles? Instead we're stuck with the broken record of "oh, this stuff costs money, OA advocates just think it all happens for free. . ."
Finally, here's my most pressing question: If economies of scale apply to publishing, why are the largest publishers providing some of the most expensive services? (in terms of solo journal subscription rates, individual PDF downloads, and open access fees) Wow, would I love the answer to that one!
Post script: It seems that many folks are having similar thoughts. Check out Björn Brembs' round-up here.