Recently, Stuart Shieber, Director of the Office for Scholarly Communication at Harvard published the edited text of a talk he gave at the annual meeting of the Linguistic Society of America, in which he praised open access (OA) as a better system for learned societies. This is an important topic, especially now because  Parliamentary hearings are going on in the UK exploring how RCUK mandates may affect learned societies.

Judging from the text, Shieber’s talk argued points his facts directly undercut, argued facts he didn’t understand, and asserted realities that don’t exist, yet he failed to realize any of this.

Shieber is a relentless advocate of OA publishing, and has been a force at Harvard driving their policies in this regard. In his recent post, Shieber argues that not-for-profit publishers are more efficient than commercial publishers because they command a lower price in the market for their goods, don’t engage nearly as much in bundling, and have smaller margins. This interpretation of the facts has four obvious problems:

• Commercial publishers often publish on behalf of not-for-profit societies.
• Commanding a higher price in the market is a sign of efficiency and effectiveness.
• Having larger margins is a sign of efficiency.
• Bundling is a sign of scale, which can only occur if there is efficiency.

In short, Shieber is looking through the wrong end of the binoculars. Of course, his main argument hinges on a related piece of equally spurious logic:

. . . the reason that scholarly societies benefit from playing in the open-access APC market rather than the closed-access subscription fee market is the difference in the goods being sold. When the good is a journal bundle, the companies with the biggest bundles, the large commercial publishers, win. When the good is publisher services for an individual article, the publishers that can deliver those services for an individual article most efficiently, the non-profit publishers, win. Sure, there are economies of scale, but empirical evidence shows that the scholarly societies are already far better able to efficiently deliver services despite any scale disadvantage.

As we’ll see, Shieber’s “empirical evidence” is inadequate. He hasn’t parsed the market correctly. And the superior profitability and price advantages of publications going through a commercial publisher — whether those publications are owned by a non-profit society or the commercial publisher — make a mockery of this line of reasoning. In fact, OA will probably be done more efficiently by large commercial publishers — viz, the acquisition of BioMed Central by Springer.

Basic business tenets also work against his argument. Every business has a core set of functions — IT, legal, HR, finance, facilities, insurance, and so forth. If the organization is small, those functions take up a large proportion of revenues and staff. The organization is comparatively inefficient. Larger organizations can use systems, time zones, currency hedges, and many other techniques to increase efficiency, making the money they spend go farther. They also tend to have lower overheads — less of their revenues devoted to supporting core functions.

The assertion that commercial publishers are less efficient than non-profit publishers is just wrong. Laughably wrong. Their margins are better, their market penetration is better, and their sales forces are better. That’s why so many non-profit societies sign contracts with commercial publishers. They want the benefits these organizations can and do deliver. And there is a reason these organizations can beat the status quo of running publications in-house — they are more efficient and effective.

Of course, nothing can dissuade the true believer from the assertion that OA is superior. So, we get a litany of facts and reasoning. Let’s examine Shieber’s facts and reason a little more closely to see how much they have to be elided to fit the argument.

To start, Shieber uses the economic concept of a “complement” incorrectly, asserting that two journals complement each other in the same way hot dogs and hot dog buns complement each other. That is, when hot dog sales fall, hot dog bun sales fall. That kind of relationship does not exist between journals. It exists between audiences and journals — if there are no more druids, druidic journals disappear, for example. But Journal A’s usage doesn’t drive Journal’s B usage in any appreciable and direct way. His use of the term smacks of sophistry.

Cherry-picked facts come next. Shieber compares subscription prices between commercial and not-for-profit journals. The data he uses are a decade old, from 2002. Unfortunately, the comparison doesn’t represent reality in any year — commercial publishers publish not-for-profit society journals. A potentially enlightening comparison would be to compare three different cohorts:

1. Journals owned by for-profit publishers with no not-for-profit involved
2. Society-owned journals published by commercial publishers
3. Society-owned journals published independently

Despite using economic concepts incorrectly, having the wrong framework, and relying on outdated facts, Shieber argues on. His main observation is that price differentials are a clear sign of market failure. He’s right — in a commodity-based market. His example is Coke vs. Pepsi. These are cola commodities. While it pains me to admit, I will settle for a Pepsi when Coke isn’t available. They are, to some degree, interchangeable. But in a market that has differentiated goods, prices diverge. Think about the difference in price between a Scion and a Lexus — both made by the same company, both automobiles, but brand and features differentiate their prices significantly. Price divergence is not a sign of a malfunctioning market of differentiated, non-commodity goods. Think First Class vs. coach. Think Godiva vs. Nestlé. Think Nordstrom vs. Wal-Mart. Shieber’s economic analysis is simply wrong.

Sheiber’s thinking doesn’t even make internal sense. He notes that APCs vary greatly, ranging in his expert opinion from $0 to $3,000 (in reality, there are higher APCs on the market already). If OA is selling a commodity (peer-review, copy editing, and publication), how could prices diverge like this? And wasn’t he just claiming that price divergence is by definition a sign of a dysfunctional market? Therefore, is the OA market already broken?

He points to more out-of-date data to assert that most OA journals don’t charge APCs. It’s from a 2009 post of his examining 2007 data. The data set is no longer available via the link Shieber provides, but it’s probably irrelevant anyhow. Things have changed significantly. How much have things changed? Well, 2007 is when PLoS ONE launched. In 2008, BioMed Central was acquired by Springer. The NIH Public Access policy went into effect in 2008. And so forth. These data are from a different era.

Sheiber also brags about how Harvard was the first university to resist the “Big Deal,” but then goes on to explain how disaggregating the Big Deal landed them back at the same place, but with less to show for it. That is, they ended up paying as much as they’d paid before, but for 30 journals rather than the 130 they’d had through bundling. This hurts two of his points. Apparently, bundling is an efficient way to sell and buy journals, proving that commercial publishers are more efficient in the market. Also, it’s apparent that the best journals in the bundle is what Harvard was paying for, but in the bundle they also received some strong second- and third-tier journals, many of which probably came from non-profit societies using Elsevier as their publisher. This is what bundles do — they help send revenues across more titles, many of which come from small societies. Bundles help smaller societies. Therefore, bundling is a boon to non-profit societies using commercial publishers.

But, of course, Shieber’s goal is to convince us that OA is better for not-for-profit societies. To find a more current source of information, let’s look at an example that emerged from the Parliamentary hearings on the same day that Shieber’s post was published. In this example, we’re dealing with a UK non-profit society (the Tavistock Institute) that gleans $1,633,565 per year in revenues by publishing 60 articles per year in the journal Human Relations. Their publisher is SAGE, a commercial publisher. If the Tavistock Institute were to go to a complete OA model with attendant CC-BY licenses, the Institute would make $90,000 per year at Shieber’s proposed rate of $1,500 per article. In other words, their journal would lose $1,543,565 in revenues by shifting to OA. How does this help the Tavistock Institute?

This kind of trade-off isn’t uncommon, and it’s why societies are so concerned about unthinking mandates and policy shifts. You can see this example and many more in the public evidence available online for the UK’s Parliamentary proceedings.

Shieber throws accusations with abandon. Does the Big Deal violate anti-trust regulations? He points to a 2004 paper — one single paper — that suggested it may. What has happened in the last eight years? Based on my online searches, the answer is, “Nothing.” There was one speculative paper, and then crickets. Shieber uses the empty rhetorical trick of playing organ music to evoke anxiety.

Finally, Shieber notes that 600 scholarly societies publish OA journals. However, when you begin clicking on links in the list he points to, a 404 error or being sent to a society home page is a very likely result. It seems many of these journals have gone by the boards — journals that seem to be identified with the designation “Transfer to publisher” or “Transfer to society.” Others I looked at are publishing an article every week or two, hardly enough to sustain a robust journal’s infrastructure.

But back to the fundamental question: Is OA better for non-profit societies? Judging from what Shieber is inadvertently telling us, I’d be very concerned if I were running a not-for-profit learned society, especially in the UK. Not only are facts being twisted by OA advocates to suit a narrative, but once those facts are placed in a sensible tableau, the picture that emerges is one full of risk and penury.