Jan 282011

A contact mentioned to me that a large university press was drawing attention to a journal article by the economist Mark Armstrong to support their contention that non-for-profit (NFP) journal owners would be better off having their journals published by a not-for-profit publisher (i.e. themselves) rather than a commercial publisher.

The article (published ironically enough by a large commercial publisher) is Collection Sales: Good Or Bad For Journals? published in Economic Inquiry 48(1), Jan 2010, and available here: http://onlinelibrary.wiley.com/doi/10.1111/j.1465-7295.2008.00207.x/pdf.

Armstrong uses economic modelling to examine the impact of collection sales (bundling) of journals and concludes that “nonprofit journals may benefit from withdrawing from commercial publishers which distribute their own for-profit journals, and joining together to be distributed by less commercial publishers who set relatively low prices for their collections”.

Many NFP journal owners do indeed prefer to use NFP publishers, often for “soft” reasons such as alignment of mission/objectives and “cultural fit”, so I was interested to see that there might be an economic advantage to them for this.

Unfortunately it is hard to see that Armstrong’s economic modelling can have more than very limited relevance to the real world, because the simplifications he uses lead to descriptions that are more caricature than realistic.

For example, he says that

“all journals prefer to participate in collection sales programs, but the two kinds of journals are distributed by publishers with different pricing strategies: for-profit journals are attracted to (or are owned by) publishers who offer them high revenue but relatively low reach, while nonprofit journals use publishers who market their collections with relatively low bundle prices and higher reach.”

This comes near the start of the article; unfortunately for Armstrong’s thesis, it is the commercial publishers that currently offer the largest reaches (through their very large consortia deals), not the NFP publishers. (Not that the NFP publishers aren’t doing their best to catch up, of course!)

He also asserts that

“While for-profit journals do not care about reach, a nonprofit journal will need to consider both its remuneration and reach when choosing its publisher”

In this fantasy world, therefore, Nature (a for-profit journal) would not care about its reach while its arch-rival Science (a not-for-profit) would. This goes beyond caricature, it’s just plain wrong. All journals compete for authors and citations, and reach is one of the key means to these ends.

I’m not qualified to comment on the technical analysis used in the paper but it’s hard to put any credence in any conclusions that are built on such shaky foundations.

I’m inevitably reminded of my old marketing professor at Cranfield, Malcolm MacDonald, who use to say of economists that they would say on the battlefield as they were being overrun by the enemy: “Let us assume a tank. Ceteris paribus, we win”.

Returning to my friends at the university press, I do warm to the idea of evidence-based publishing – first coined I believe by the inestimable Richard Smith at the BMJ – but those professing to follow this path do need to understand the difference between evidence and cherry-picking. Or perhaps I’m just confusing discourse and rhetoric?