Success and Failure of New Journals

The last couple weeks had me thinking about new journals. 

I "attended" a webinar from bepress, the company that supplies the software for OpenSIUC, called "Journal Make-Over: Practical Steps to Better Journals." A lot of the webinar was about helping fledgling journals develop. The gist that I got out of it was that editors for new journals need to build up a steady stream of submissions and articles to publish. Approaches that would make the journal look desperate could backfire. The stream has to be built through personal networking with potential authors and editorial board members. 

A few days later, a librarian encouraged other librarians on an email list. He encouraged people to contact Thomson Reuters to include a fairly new journal (started in the last five years) from a scholarly society publisher in Web of Science.

I also saw  "New Journals in Education and Psychology: General Trends, Discoverability, and Ubiquitous Journals of the Decade, 2000–2009" in College & Research Libraries. In it, Bernadette Lear found that new journals in education and in psychology from big for-profit publishers were more likely to receive coverage from indexing and abstracting databases and be listed in library collections than new journals from colleges and universities, small for-profits and societies.

One of Lear's other findings was that more than 83% of the new journals in her study that started in 2000-2005 were still publishing in 2010. To put that in perspective, Lear provides a footnote that over half of new magazines fail in the first five years. Heck, the high rate of new business failure is the stuff of urban legend -- though research puts it at about five ninths failing in the first four years.

To me the success rate defies economic logic. How can so many new journals appear year after year with so few failing? Why don't more of them fail? Where would libraries get the money to pay for the new journals? A few big research libraries can find a way to subscribe to new journals (maybe), but college and university libraries aren't faring that well. With the high inflation rate on existing journals, new subscriptions have to be offset by cost cutting somewhere else. Are the costs of production so low that the commercial publishers can release new ones ad infinitum and turn a profit on them from a handful of subscriptions?