This post is an updated version of an attempt last fall to assess the penetration of open source in the U.S. ILS market. There were two posts in LISNews, one on October 15, 2007 and the second on October 19. In summary, they reported that about 1% of U.S. Public Libraries used open source ILSs, depending on the variable one used. The data gathering for this reprise of that study took place on March 3, 2008 and I find about the same percentages now as then.
The point of this exercise was to establish a baseline to measure the growth of open source solutions in libraries. I use U.S. public libraries because there is a universe, national-level data series that gives us a denominator with which to calculate percentages. It is convenient that both Evergreen and Koha, the two major open-source ILSs used in the U.S. have their homes with public libraries and are just beginning to be adopted by academic libraries so the percentages are lower. I used ILSs because they are relatively easy to discover by using Marshall Breeding’s lib-web-cats. The latest national-level data are for fiscal year 2005 so we will be using these data to make estimates of the current state of ILSs. And at this point, the vagaries of data become complex so I have moved the Methodological discussion to a separate page.
Table 1 gives summary data and has two sets of summaries. Back to Table 1 in just a bit
Table 2 has summary statistics for public library systems actually running Koha as of March 3 (there were 15).
Table 3 has summary statistics for library systems running Evergreen (48).
Table 4 has summary statistics for another estimate of public library systems running or planning on running Koha at some time in the indefinite future, according to lib-web-cats. There were 81, including the 15 in Table 1. Note the column to the far right. It has the ILSs actually being run by these Koha libraries on March 3.
Thus,the first two lines of Table 1 are based on summary data from a total of 63 (48+15) systems and the second two lines on a similar total of 129 (48+81) systems.
In these latest national figures, we have a total of 9,198 public library systems. There are about 17,000 actual “outlets” (central libraries, branches, and bookmobiles) in these systems but the detailed data here are only available for systems.I have picked data from these libraries that are indicative and provided totals of those data.
Looking at the top two lines in Table 1, as of March 3, 63 were running either Evergreen or Koha.That is .7% of the total number of systems. However, if you look at population served to measure penetration, it is 1.7%; circulations, 1.0%; if you prefer total operating expenditures, these 63 libraries in FY 2005 were .9% of the U.S. total.There are other summary figures there, but we can say that these numbers hover around 1%.
Last October, there were 62 libraries running either Evergreen or Koha and, using FY 2004 data, I found these were .67% of systems and the other figures were roughly similar to the March figures.
A New Wind Blows
Recently MassCat and WALDO announced that they will be working with LibLime implementing Koha. Also recently, the Indiana Open Source ILS Initiative and the Michigan Evergreen Project committed to work on Evergreen implementations with Equinox Software. All libraries in these consortia will not be implementing these plans tomorrow and some libraries in them may not implement either Koha or Evergreen ever. For instance, WALDO’s implementation of Koha is apparently to be with a subset of its total membership. 24 academic members with each maintaining its independence according to my understanding of what John Stromquist, Executive Director of WALDO, told attendees at the VALE Symposium. Given the state of my hearing, I would like to see the film when it is posted on that site.
These announcements plus other work in the pipeline indicate the low numbers seen so far are forming a foundation of faster growth ahead