Friday, February 03, 2006

Journal Pricing

In a previous entry I wrote some about the nuts and bolts of academic journals. I also promised that I'd write a series of entries on this topic. Alas, I'm far behind the schedule I imagined for myself when I made that first entry. So, to try and get myself back on track, here are some thoughts on journal pricing.

The question I ended with last time, was: How much do journals cost, and how much should they cost? Most journals charge different rates for libraries and individuals. I'll focus on library subscription rates. Let's start with two journals that I've been thinking about lately.

  1. Physical Review E (PRE), a physics journal that covers statistical, nonlinear, and biological and soft-matter physics. I've published a few articles in here.

  2. The Journal of Theoretical Biology (JTB), which is, as the name suggests, mainly about theoretical biology. I've never published there, but I was recently asked to referee and article for them.


The table below shows some various data for each journal. At first blush, both journals seem to be cost about the same. However, PRE publishes many more pages than JTB. If one calculates the cost per page, PRE turns out to be a relative bargain. JTB costs three times more per page than PRE.



Journal NamePhysical Review EJ. Theoretical Biology
Institutional Cost per year$4,315$4,085
Total number of pages published (year)19,097 (2004)2,724 (2005)
Price per page$0.226$0.667

(Note: PRE charges different rates to different sized institution. The rate I've shown above is the most expensive rate corresponding to the largest institution. Their least expensive rate is $2,795.)

One question that immediately jumps to mind when looking at this table is: Why are the two journals priced so differently? One journal costs three times more than the other. This is pretty remarkable; imagine going to the store and finding that one brand of milk costs three times as much as another brand. This would be strange. It might make you suspect that there must be something fundamentally better about the three-times-as-expensive milk.

So, is there something fundamentally better about JTB? Nope. Not so far as I can tell. The average article quality might be slightly higher in JTB than PRE, but I doubt by much. Both use essentially the same refereeing and editing and typesetting processes. JTB might be on a slightly higher quality paper. However, most people read journal articles by gaining access to an electronic copy and printing it out on their own printer.

So, why is JTB three times as expensive as PRE? The answer is simple: PRE is a non-profit entity. It's published by the American Physical Society (APS), which is the (non-profit) professional society for physicists in the U.S. JTB, on the other hand,is published by Elsevier, a for-profit publishing company based in Amsterdam. PRE's primary goal is to generate sufficient revenue to offset costs, while simultaneously maximizing access to its journal. JTB's primary goal, however, is to maximize profit. JTB thus charges what the market will bear; PRE charges what the need to cover expenses.

Does JTB really make a profit, or is this just conjecture on my part? It's hard to know about JTB in particular, but Elsevier, who publishes around 1700 journals, is clearly doing very, very well. Reed-Elsevier, the company that holds Elsevier, reported a 2004 adjusted pre-tax profit of 1.51 Billion Euros. Elsevier itself, the academic, medical, and technical publishing and communications arm of Reed-Elsevier, reported revenues of 2 Billion Euros and an adjusted operating profit of 676 Million Euros. (You can access their most recent financial statement here, and an interesting report by Morgan Stanley here.)

The situation illustrated above for JTB and PRE is quite typical of the difference in pricing for for-profit and non-profit journals. A number of researchers have done careful studies of this issue. Some results of these studies are collected in Carl T. Bergstrom and Theodore C. Bergstrom, "The costs and benefits of library site licenses to academic journals," PNAS. 101:897-902. Here is a table, using data cited in Bergstrom and Bergstrom, of mean price per page (for institutional subscriptions) for for-profit and non-profit journals, for several academic disciplines:



DisciplineNon-profitFor-profit
Ecology0.191.01
Economics0.170.83
Physics0.190.63

All costs are in US dollars per page. See Bergstrom and Bergstrom for further details. (For a thorough and lucid discussion of journal pricing, I strongly recommend Theodore C. Bergstrom, "Free Labor for Costly Journals?" Journal of Economic Perspectives. 15:183-198. Although the focus is on economics journals, what he has to say applies to all other fields, as well. Both of the papers by Bergstrom are available on Ted Bergstrom's website here.)

In any event, the trend is very clear: for-profit journals are between 3 and 5 times more expensive than their non-profit counterparts. But perhaps its the case that the for-profit journals are better. Journal quality is often measured by the number of citations to articles published in it. If lots of other articles cite the articles in a particular journal, that's an indicator that the journal is pretty good. This isn't a perfect measure of journal quality, but it seems a good place to start. Bergstrom and Bergstrom also compile data for the cost per citation to a journal. I.e., they don't worry about the number of pages in a journal, but rather the number of times articles in that journal are cited. They find that price differentials between non-profit and for-profit journals are even larger for cost per citation than they are for cost per page. So much for the better quality theory.

Questions for next time: Where does this money for institutional subscriptions come from? Where does it go? And why does it matter how much journals cost, anyway?

No comments: