Tag: user behavior

What Library Circulation Data Shows

Unless current patterns change, by 2020 university libraries will no longer have circulation desks. This claim may seem hyperbolic if you’ve been observing your library, or even if you’ve been glancing over ACRL or National Center for Education Statistics data. If you have been looking at the data, you might be familiar with a pattern that looks like this:

This chart shows total circulation for academic libraries, and while there’s a decline it certainly doesn’t look like it will hit zero anytime soon, definitely not in just 8 years. But there is a problem with this data and this perspective on library statistics.  When we talk about “total circulation” we’re talking about a property of the library, we’re not really thinking about users.

Here’s another set of data that you need to look at to really understand circulation:
Academic enrollment has been rising rapidly.  This means more students, which in turns means greater circulation.  So if total circulation has been dropping despite an increase in users then something else must be going on.  So rather than asking the question “How many items does my library circulate?” we need to alter that to “How many items does the average student checkout?”

Here is that data:

This chart shows the upper/lower quartiles and median for circulation per FTE student.  As you can see this data shows a much more dramatic drop in the circulation of library materials. Rising student populations hide this fact.

[source: http://xkcd.com/605/]

But 2020? Can I be serious?  The simple linear regression model in the charts is probably a good predictor of 2012, but not necessarily 2020. Hitting zero without flattening out seems pretty unlikely. However, it is worth noting the circulation per user in the lower quartile for less than 4 year colleges reached 1.1 in 2010. If you’re averaging around 1 item per user, every user that takes out 2 items means there’s another who has checked out 0.

What’s Happening Here?

Rather than waste too much time trying to predict a future we’ll live in in less than a decade, let’s explore the more interesting question: “What’s happening here?”

By far the number one hypothesis I get when I show people this data is “Clearly this is just because of the rise of e-journals and e-books”. This hypothesis is reasonable: What has happened is simply that users have switched from print to electronic. This data represents a shift in media, nothing more.

But there are 2 very large problems with this hypothesis.

First, print journal circulation is not universal among academic libraries. In the cases where there is no print journal circulation the effect of e-journals would not be present in circulation data. However, I don’t have information to point out exactly how many academic libraries did circulate print journals. Maybe the effect of e-journals on just the libraries that do circulate serials could effect the data for everyone. The data we have already shown resolves this issue. Libraries that did circulate serials would have higher circulation per user than those that did not. By showing different quartiles we can address this discrepancy in the data between libraries that did and did not circulate journals. If you look at the data you’ll see that indeed the upper quartile does seem to have a higher rate of decline, but not enough to validate this hypothesis. The median and lower quartiles also experience this shift, so something else must be at work.

Second, e-books were not largely adopted until the mid 2000s, yet the decline preceding 2000 is at least as steep as after. If you look at the chart below you’ll notice that ebook acquisition rates did not exceed print until 2010:

Ebooks, of course, do have an effect on usage, but they’re not the primary factor in this change.

So clearly we must reject the hypothesis that this is merely a media shift. Certainly the shift from print to electronic has had some effect, but it is not the sole cause. If it’s not a shift in media, the most reasonable explanation is that it’s a shift in user behavior.  Students are simply not using books (in any format) as much as they used to.

What is Causing this Shift in User Behavior?

The next question is what is the cause of this shift.

I think the most simple answer is the web. 1996 is the first data point showing a drop in circulation. Of course the web was quite small then, but AOL and Yahoo! were already around, and the Internet Archive had been founded.  If you think back to a pre-web time, pretty much anything you needed to know more about required a trip to the library and checking out a book.

The most important thing to take away is that, regardless of cause, user behavior has changed and by all data points is still changing.  In the end, the greatest question is how will academic libraries adapt?  It is clear that the answer is not as simple as a transition to a new media. To survive, librarians must find the answer before we have enough data to prove these predictions.

If you enjoyed exploring this data please check out Library Data and follow @librarydata on twitter.

Data Source:

• All library data referenced in this post comes from the Library Statistics Program (National Center for Education Statistics) nces.ed.gov/pubsearch/getpubcats.asp?sid=041#
• Data regarding fall enrollments is from “Fast Facts” (National Center for Education Statistics) http://nces.ed.gov/fastfacts/display.asp?id=98

About our guest author: Will Kurt is a software engineer at Articulate Global, pursuing his masters in computer science at the University of Nevada, Reno and is a former librarian. He holds an MLIS from Simmons College and has worked in various roles in public, private and special libraries at organizations such as: MIT, BBN Technologies and the University of Nevada, Reno. He has written and presented on a range of topics including: play, user interfaces, functional programming and data
visualization.