It is not often the case that our own analyses of publisher journal price increases in liblicense-l or lis-e-resources trigger a comment of the analysts at Exane BNP Paribas, one of the top brokers for European Equities.
On July 6 we published our analyses on both the Springer 2010 price list (and SCOAP3) and the Elsevier 2010 subscription price list, which got quite some feedback on Twitter, especially our added message that “#Elsevier Medical titles increase by 7.5 % on average for 2010, some by as much as 25%.” ( http://twitter.com/bckaemper/status/3171066523 ), a day later BNP Paribas submitted their comment to liblicense-l. As it was distributed only now (3 hrs ago) via liblicense-l and will appear on the liblicense-l web archive only next week, we quote it here in full:
From: Sami Kassab [firstname.lastname@example.org]
Sent: Friday, August 07, 2009 3:49 AM
Subject: Reed Elsevier (-): Announces higher journal price
increases than expected
Hello, please find enclosed a comment on Reed Elsevier (-)
regarding its 2010 science journal pricing announcement:
*Reed Elsevier has announced its 2010 journal price increases
Elsevier has just published its 2010 print journal price list. We
estimate the median price increases at 5% with key titles such as
The Lancet up 9.5% and Cell Press up 4%. While these price
increases only pertain to print-only subscriptions (c 10% of
divisional revenues), we believe they suggest that Elsevier has
maintained an aggressive pricing policy despite current library
budget constraints. Its main competitor, Springer, also announced
a 5% price increase for 2010.
* More aggressive pricing than expected
Given library budget pressure, we have assumed 2010 net price
increases of 3% in our model (0% for 2010 subscription renewals
and 5% for existing contracts), resulting in -1% Elsevier’s
organic revenue decline in 2010. As is the case elsewhere in
Media, listed prices increases can vary considerably from net
price increases due to significant discounts. We expect pressure
on Elsevier’s discounts and renewal rates to be under pressure in
the coming months. We maintain our 2010 organic revenue growth
forecasts for Elsevier but they now look conservative.
* Open Access is putting pressure on pricing
We see Elsevier’s decision to cuts prices for its nuclear physics
journals by 20% as a reflection of the progress of Open Access,
which prevails in this discipline. Should Open Access reach
critical mass in other disciplines, similar pricing pressure is
likely. With the FRPAA act likely to pass in the US, Open Access
will continue to grow.
* Cautious view maintained
We continue to believe the stock could remain ‘dead money’ until
further clarification of the new investment budget and 2010
top-line trends. While we would close short positions, we
continue to prefer Pearson (+) and Wolters Kluwer (-) in
Exane BNP Paribas
Office: + 44 20 7039 9448
Mobile:+ 44 7795 528 365
What BNP Paribas fails to emphasize is that it is not open access per se that is threatening Elsevier (High Energy Physics since long have had almost 100% open access uptake, so critical mass has long been reached in this special field, quite different from other physics areas), but that they are loosing the battle for authors, possibly due to their reluctance to support SCOAP3. As I wrote, they publish now considerably less than 4 years ago (update: about 30%, for more accurate updated estimates cf. the comment below) and seem to have lost in submissions from authors for their HEP journals. With such a reduction in size, prices also had to come down. In the new open access scholarly publishing market, journals will compete for authors even more than now. SCOAP3 certainly raised the awareness for both the scientific community’s expectation to fully convert these journals to OA and the unsustainable prices that had risen to absurd record prices. It is clear that subscriptions are now under even more pressure because of the global economic crisis that especially hit american libraries very hard.
For another illustration of the growing impact of hybrid open access programs on online journals prices, this time at Oxford University Press, cf. my newest posting to liblicense-l and lis-e-resources, Disturbing spread of dyscalculia in recent publisher price lists and announcements (Aug 7, 2009)
Bernd-Christoph Kämper (@bckaemper), Stuttgart University Library
2 Gedanken zu „Reed Elsevier: Announces higher journal price increases than expected, BNP Paribas: Open Access is putting pressure on pricing“
Very interesting, Klaus! Do you have more detail on the figures on decreasing author submissions? Thanks for adding this to the Open Access Tracking Project.
The evidence for decreasing author submissions is only indirect, as we do not have access to those data. We really looked at journal output in pages/year.* An alternative explanation could be that Elsevier drastically increased rejection rates by beiing more selective, but there is no indication that they did so. Looking at ISI impact factors and article counts, we find:
Nuclear Physics B:
2008: 4.158 (350), 2007: 4.645 (313), 2006: 5.199 (405), 2005: 5,522 (524), 2004: 5.819 (476), 2003: 5.297 (534)
A SCOAP3 core journal. Impact factor has decreased by 30% since 2004, size by 1/3.
Physics Letters B:
2008: 4.034 (929), 2007: 4.189 (840), 2006: 5.043 (999), 2005: 5.301 (955), 2004: 4.619 (1038), 2003: 4.066 (968)
A SCOAP3 core journal. Impact factor has decreased by 25% to 30% since 2004/05 for both journals. This means that also the better papers have gone to other journals.
We have to revise our first estimate of a 50% reduction in output for PLB, as the picture is complicated by changes in frequency and discrepancies between scheduled volumes for a subscription year and the printing of indices in former volume. We find now that the reduction in output in pages was similar to NPB, about 30%-40%. The reduction in the number of papers published for PLB was less, around 10%-20%
Nuclear Instruments and Methods in Physics Research A (Elsevier):
2008: 1.019 (884), 2007: 1.114 (1888), 2006: 1.185 (1650), 2005: 1.244 (1359), 2004: 1.349 (1692), 2003: 1.166 (1551)
A SCOAP3 “broadband journal” with 25% HEP content. The sharp decrease in article numbers for this title comes about because Elsevier published much less conference proceedings in NIMA from 2008 on. No of pages published decreased from 10792 to 6340 (40%), whereas the price was lowered only by 15% (from 2007 to 2008). Impact factor has decreased by 25% over the last 4 years.
For comparison the other SCOAP3 core journals:
Journal of High Energy Physics (SISSA):
2008: 5.375 (1284), 2007: 5.659 (1247), 2006: 5.393 (1027), 2005: 5.944 (859), 2004: 6.503 (885), 2003: 6.057 (809)
JHEP has grown by 45% in the last 4 years, but impact factor has decreased by 20% probably because it captured a larger part of the mss. lost by Elsevier.
Physical Review D (APS):
2008: 5.050 (2863), 2007: 4.696 (2268), 2006: 4.896 (2375), 2005: 4.852 (2247), 2004: 5.156 (2277), 2003: 4.599 (1964)
PRD is also going strong (25% more published than 4 years ago) but has maintained its high impact factor.
European Physical Journal C (Springer):
2008: 2.805 (323), 2007: 3.255 (369), 2006: 3.251 (214), 2005: 3.209 (306), 2004: 3.486 (302), 2003: 3.580 (287)
Springer’s EPJ C has also lost 30% in Impact factor, while its size has not changed substantially.
To set this in perspective: The aggregate impact factors for the ISI category: Physics, Particles & Fields were (Year: aggregate IF (median IF, #articles):
2008: 3.165 (2.015, 9178), 2007: 3.034 (1.894, 9720), 2006: 3.058 (1.564, 8209), 2005: 2.988 (1.659, 9471), 2004: 3.137 (1.589, 8751)
This is only a preliminary analysis. Caution is also necessary because it is unclear how much of the real citations are captured by ISI if many early citations on HEP articles are on their preprint version at arxive instead of the final published article. This could considerably distort the picture. (I still have to check what has been reported in the literature on this, and I’m glad if people direct me to such studies.)
Elsevier’s analysis of the same data (Research Trends, August 2008) is highly misleading and selfserving, as they compare total citations to a journal in a given year over all articles (the citations can be to any previous year, but largely since 1996, as Scopus has only incomplete data before this) to the journal’s article output in that year and call this the “average journal citation per article” [=(cumulated) cites to all articles divided by number of current articles]. It is therefore ridiculous if Publishing Director David Clark maintains “Nuclear Physics B has consistently maintained its high standards despite the reduction in the number of papers being published in particle physics”, emphasizing the upward trend and the fact that Nuclear Physics B has the highest “average journal citation per article”, 77.15 in 2007. The sure recipe to drive this factor even higher is to rapidly reduce the output in papers [the denominator, number of current articles] even further because it is well known that the citation half life of these journals is around 10 years and it will take some time for the cumulated citations to come down. In fact, the trendline for average journal citation per article would go through the roof if Elsevier stopped publishing altogether in HEP. The impact factor, on the other hand, measures the average citation rate with much less latency, namely through a citation window of only the previous two years and takes into account the citations and number of articles published in both years.
Bernd-Christoph Kämper (@bckaemper), Stuttgart University Library