Funding Mechanisms for BioMed Central Articles, aka the "End of Biomedical Digital Libraries with BioMed Central"

BUT...

In 2003 and 2004, many academic health sciences libraries purchased full memberships in BioMed Central. This enabled researchers at their institutions, including librarians, to publish articles at no charge. Article processing charges for a single article were around $500 at the time BMC began, and now hover around $1,500 per article. UC San Francisco, my employer, paid for a full membership during this time.

Authors at institutions without memberships authors bore the full cost of publication. The membership model was conceived as the open access equivalent of the traditional journal subscription.

Starting in 2005 BMC increased the cost of its memberships. Many libraries responded by reducing their level of membership to "supporting," which offered researchers a discount of 15% off the publication fee. (Famously, Yale dropped support for BMC altogether in August 2007.) This is the level of support that UC San Francisco offers today.

Theoretically researchers with large grants can pay the cost of an open access article from their grant funds. Librarians generally have minimal research support; our core author base was heavily dependent on full memberships in BioMed Central.

Eventually the relationship grew untenable, because it became impossible for most of our authors to publish with us. BDL and BioMed Central formally ended our relationship in September 2007, three years after the first article appeared.

[EDITORIAL NOTE: BioMed Central was sold to Springer in October 2008. The general blogosphere reaction has been that this is evidence that traditional publishers see a bright future for open access as a business proposition. Time will tell. I believe that BMC's experimentation with membership models was part of the process that led to this sale.]