The n-Category Café

Depends exactly how the data is calculated, but if you can tell me that then I can tell you the mathematics results. For example, going from the AMS journal data (which is what Ulf Rehman bases his webpage on) then there are 38 Elsevier titles (including Academic Press/Elsevier) out of 276. The mean price per journal for Elsevier is 2638 dollars, compared with 918 for non-Elsevier titles and 1205 overall.

$$\array{ Publisher & Number & Mean Price & Mean & Elsevierian \\ & of Titles & per Title & of Complement & Ratio \\ Elsevier &38 & 2638 & 918 & 1 \\ American Math. Society & 6 & 693 & 1152 & 3.8 \\ Cambridge Univ. Press & 7 & 693 & 1154 &3.8 \\ International Press & 5 & 515 & 1154 & 5.1 \\ Oxford Univ. Press & 8 &729 & 1154 &3.6 \\ Soc. for Indust. Appl. Math & 9 & 620 & 1159 & 4.3 \\ Societe Math. de France & 4 & 328 & 1154 & 8 \\ Springer & 86 & 1380 & 1043 &1.9 \\ Walter de Guyter & 2& 1600 & 1140 & 1.6 \\ Wiley & 9 & 2168& 1110 &1.2 \\ Overall & 276 & 1205 & 0/0 & 2.2 }$$

I’m not sure whether or not the figures for Springer are fair since there are journals listed with ‘Springer’, ‘Springer/Birkhauser’, and ‘Springer-Verlag’. I’ve lumped all of these together.

I don’t know if it’s of any use, but there’s an interesting per-page journal price comparison here:

http://www.mathematik.uni-bielefeld.de/~rehmann/BIB/AMS/Price_per_Page.html

By the way, there’s an awful lot of data out there on maths journals and it can be tricky to track down and gather into a unified document. I’ve put up some instructions on how to get the information that I’m aware of into a usable form on the web here. If anyone knows of any other sources of useful information either let me know or (as it’s a Wiki) add the information yourself.

I don’t think the case against Elsevier should be made based on price. Price is an issue, but it’s not the only or even the most important issue. Elsevier’s prices are high, but not uniquely high.

Elsevier has an extensive history of poor behavior (arms fairs; fake medical journals; Chaos, Solitons, and Fractals; probably other examples we don’t know of yet). This behavior totally disqualifies them as stewards of the mathematical literature. Even if their prices were rock bottom, they have proven themselves unworthy of our trust, and publishing papers with them is a matter of trust (namely, trusting that they will make the papers available in the future at a reasonable price).

Of course, high journal prices alone are already a terrible thing, but that’s a much trickier issue. The problem is where to draw the line. If we insist on very low prices, we might damage quality or keep learned societies from using journal revenues to subsidize other valuable activities. If we allow moderately but not extremely high prices, then the best we can hope for is to end up with a ton of moderately high-priced commercial journals. (That would certainly be an improvement on the current situation, but it wouldn’t solve our problems.) It’s genuinely unclear how to pull this off.

On the other hand, we’ve already got plenty of other reasons to boycott Elsevier. This would not only remove a cancer from the mathematical community, but would also put their competitors on notice that good behavior is required. So, in making a public case against Elsevier, I’d mention high prices but make it clear that the problem is much deeper than just overcharging.

Once Elsevier has been dealt with, we can move on to the question of journal pricing in general. Am I the only one who thinks this is the right approach?

I found out what the University of California pays for various electronic journals. All disciplines are included in this spreadsheet.

The American Chemical Society is the worst in one sense: in 2008 their journals cost an average of $13,767 per title, compared to $6,598 per title for Elsevier and $1,816 per title for Springer–Kluwer.

On the other hand, the average cost per use was $0.43 for American Chemical Society journals, $1.30 for Elsevier and $2.70 for Springer–Kluwer.

(There are also others listed, but I’m too lazy to type them all in.)

You’ll forgive me for being pseudo-anonymous. My university library is having its budget cut, and yesterday it announced that it would like to cancel a lot of mathematics journals: including Crelle’s journal, Mathematische Annalen, Proc. LMS, GAFA, Inventiones and Duke, amongst others.

No Elsevier journals are up for cancellation, I believe because they are in bundles, and so cancelling a single journal is impossible (or simply wouldn’t save any money). Can anyone shed any light on how such bundles work? For example, large numbers were banded around in relation to Chaos, Solitons and Fractals (see [1] below). But how does this number relate to the cost of a bundle? Does that question even make sense?

The library claims to have usage stats for electronic accesses, and if these are accurate, then the journals listed above aren’t be used much at all (which I find hard to believe). However, it is a little hard to argue against the financial logic. But the net effect is that independent publishers, and learned societies (e.g. LMS, AMS, Australian Math. Soc, Canadian Math. Soc) are the big losers. I think we should be supporting these organisations (the profits go back into mathematics) and not the likes of Elsevier.

[1]: I notice that the latest issue of Chaos, Solitons and Fractals is back to publishing what looks like proper, non-linear dynamics stuff…

anonymous wrote:
I notice that the latest issue of Chaos, Solitons and Fractals is back to publishing what looks like proper, non-linear dynamics stuff…

I don’t think so. At least if one does not think of that crackpot stuff to be proper:

Chaos solitons and fractals, volume 41, issue 3 (15 August 2009)

http://www.sciencedirect.com/science/journal/09600779

BPS states, dualities and determining the mass of elementary particles, Pages 1263-1265, M.S. El Naschie

A simple direct connection between superstrings and E8 unification, Pages 1329-1330, M.S. El Naschie

E-eight exceptional Lie groups, Fibonacci lattices and the standard model, Pages 1340-1343, M.S. El Naschie

Apparently that journal continues to contain this crackpot stuff and Elsevier has no shame to publish it.