Open‐access journals — nice idea, shame about the numbers?

Years ago, I prepared a presentation including a slide entitled ‘The Strategy Gap’. ‘Strategy’ was represented by a woolly cloud at the top of the slide, while ‘Finance’ was represented by a stark rectangle at the bottom. In the middle was ‘The Gap’. The point was that the strategic discussion in organizations often occurs in a different room, at a different time, between different people from the discussion of finances. The consequences may be familiar to you. Grand strategic designs offering an exciting future are abandoned midway through their execution because they are not compatible with the pressure for short-term results. Plans are undertaken which, it becomes apparent, could not be afforded or could never have produced an adequate return on investment. In looking recently at the future for journals proposed by the Budapest Open Access Initiative, I suspect that here is an example of such a disconnect. Looked at strategically, the proposals are impressively well thought out. The approach is not directly confrontational, as is that of, say, the Public Library of Science campaign. The message is not ‘You, commercial publishers, must do this or else . . .’ Rather it is ‘We intend to establish a new model. You, commercial publishers, can follow us or see your business wither away.’ This is the classic ‘art of Japanese strategy’ – avoid confrontation with people bigger than you, move the battleground to where their strengths become weaknesses and your weaknesses become strengths. The compelling economic argument for the superiority of the new model is that traditional publishers have revenues of around $4,000 per paper published, while alternative models have costs in the range of $300–1,000. At this point I would like to step back into the persona which I used to inhabit, that of the cynical finance director, and pick away at these figures. To assist me I have data from a benchmarking project which I recently completed on behalf of ALPSP. First, the figure of $4,000 revenue per paper. This is borne out by the benchmarking study, which has most participants in the range of £2,000–2,500. Next, the cost per article of the alternative models, $300–1,000. Six out of eight of the participants in the study had first-copy costs in this range. There has to be some serious inconsistency here. The $4,000 per paper is based on publishers’ revenues, which cover a multitude of things apart from the direct costs of processing papers. These include not merely print and physical distribution (which the Budapest model dispenses with, being onlineonly) and subscription management (which the Budapest model dispenses with, being open access) but also sales and marketing expenditure, indirect overheads, profit and the cost of development. About 20% of journal publishers’ direct costs are accounted for by print and physical distribution. Subscription management accounts for 7% of costs, so what the figures actually show is that an online-only openaccess journal would have costs of very roughly one-third less than a traditional journal. There is still a very large gap between the revenue per article and the cost per article or first-copy cost, even after taking out these elements. What accounts for this? Sales and marketing is one thing, accounting for around 13% of costs. One could argue that it will be redundant with an open-access product, but I doubt it. My experience in business has been that nothing happens without someone selling in some form. The cost of selling to libraries will be replaced by the cost of ‘selling’ to authors or their employers. The costs may be described as ‘author service’ but costs there will be. Next, organizations do not run themselves. They need support functions such as finance, legal, human resources and IT. They also need senior managers who Personal View 75