Have you ever wondered about the pecking order when it comes to university presses? What counts as a best seller for university presses? These questions have crossed my mind, but I have never seen a study that answers them. The fact that sales data for books is unusually difficult to obtain hasn't made it any easier to analyze these kinds of patterns and trends. Today I had an idea for figuring this stuff out.

I first went to the Worldcat database. When I did my searching there, I used the word"history" in the subject field and the word"university" in the publisher field. The cast a fairly wide net and yielded about three times as many books as reported in the history category by Bookwire, but since it was the best measure I could come up with I stuck with it. I limited my search to the last half decade, from 2002 to 2006. I also used five hundred libraries as an easy measuring point. Some history books are sold to just a few libraries in WorldCat, while every once in a while an academic history book is sold to 2,500 or more libraries.

Here is a quick summary of what I found in WorldCat:

• Half of all the academic history books that are sold to any number of libraries are published by just eight university presses: Oxford, Cambridge, Yale, Harvard, California, Chicago, Columbia, and Princeton.
  

• A little less than six percent of history books published by university presses are sold to five hundred or more libraries.
  

• Half of all the academic history books that are sold to five hundred or more libraries are published by just five university presses: Oxford, Yale, Cambridge, Harvard, and Princeton.
  

• When it comes to which university presses sell the highest proportion of their academic history books to five hundred or more libraries, a very different ranking emerges:
  

I was kind of surprised that Oxford and Cambridge fell to the bottom half of the above table when I sorted it by percentage. They have an incredibly high number of history books that are purchased by five hundred or more libraries, but the rest of their history books may not receive the same kind of marketing and promotion. This data suggests to me that medium-sized university presses may be better able to sell a higher proportion of their history titles since they can give books more personalized attention. I was also impressed when I saw that one out of every five history books published by the University Press of Kansas gets sold to five hundred or more libraries. Maybe this is a consequence of their political history series. This data also has me wondering how low we should set the bar for best-sellers in our field. Is being purchased by five hundred libraries (as measured by WorldCat) enough to qualify? What do the rest of you think of this data?