The Long-Tail in Libraries

Cross-posted from my new blog, Libraries are for Use....

I know, I know...the Long-Tail is not exactly a new concept.  It's been a buzz-word since Chris Anderson's article in Wired in 2004 (I remember that article).  Indeed, it is merely an extension of power-laws that have been known for centuries.  Even librarians have been familiar with these distributions for decades, looking at all those circulation and journal use studies.

This article, however, is pretty intriguing because it extends that concept a little farther than I had seen before (request through ILL if you don't have access):

Petros A. Kostagiolas, Nikolaos Korfiatis, Marios Poulos
A long-tail inspired measure to assess resource use in information services
Library & Information Science Research, Volume 34, Issue 4, October 2012, Pages 317–323
http://dx.doi.org/10.1016/j.lisr.2012.05.002

What is most interesting is how the authors applied measures of income disparities between countries to disparities of book usage.  I had learned about these "macroeconometric" measures, the Lorenz curve and the Gini coefficient, in my readings of The Economist, so I was quite intrigued by this application.

So, I've been reviewing the application of "long-tail" distributions in library & information science, reading these articles to get back up to speed:

• The Long Tail (book by Anderson)
• Libraries and the Long Tail: Some Thoughts about Libraries in a Network Age (D-Lib article by Lorcan Dempsey, 2006)
• The Long Tail, Copyright and Libraries (article in LIBER Quarterly by Julian Van Borm, 2009)
• More, much more...

The problem that I see with applying the long-tail concept to book circulation is that books are physical items, although Netflix DVDs are often included in long-tail discussions.  But libraries have not yet successfully implemented a delivery model that rivals Netflix's.  I would, however, like to look more carefully at applying these models and concepts to journal usage, particularly given the growth of discovery systems.  Has the long tail extended?  Anderson started his inquiry into long-tail distributions when he was told that 98% of all songs available by a particular "digital jukebox" provider were accessed at least once a quarter.  98%!  Book circulation studies have shown much, much shorter tails - the best I've seen was 50% of titles circulated at least once in 5 years.  Can discovery systems lengthen this tail?  Even better, can it "thicken" the tail (getting more usage of our articles)?