This was a question that I received at a recent event where I spoke. Having set out the economic problems of the subscription model and the difficulties of cross-subsidy for learned societies, a questioner piped up:
"We're a small learned society, charging £25 for our journal. We use the funds to give reductions to Ph.D. students and, when people want their articles to be openly available, we let them. We don't have many subscribers and we publish about 10 articles a year. Tell us what we're doing wrong."
The implication behind the "Tell us what we're doing wrong" was, as I read it, a little smug (perhaps it wasn't and it was a genuine question, in which case, apologies). It seemed to have this subtext: "you've spoken about mega-publishers and the above-inflation budgetary crisis. You've also spoken about the problems of learned societies getting their cross-subsidy from library budgets. But we're not part of that problem. We don't need to do anything. Argue your way out of that one".
1. "Publishers" is not a homogeneous category. I always take a great deal of care, in my talks, to point out that university presses are very different to mega-commercial publishers (although the bigger UPs have some similarities to their shareholder-beholding brethren). Small society journals that exist to disseminate knowledge, at relatively low cost, are not doing mega-economic damage. But then, I never implied they were.
2. But, with this model, it's not hard to achieve an economic setup that will enable OA. £25 is not much for one subscriber. But, for ten articles per year, that's £2.50 per article. Again, perhaps not much. Scale this up to the entirety of scholarly communication and it still might be cheaper than the current system (I estimate that some publishers charge ~$20 per institution per article). But it's hardly nothing. The costs of production, though, for an open access equivalent (taking the Ubiquity Press costings) come to about £300 per article. Assuming 10 articles, that's £3000 per year. That's just 120 subscribers paying the £25 and it could be made available to everyone. Why not simply put this on to a membership fee and, perhaps, give early access to members before the full issue is published at the end of each year, or similar?
3. If the goal is to disseminate knowledge, OA can help with your small subscriber base. I won't recap the download figures, or the citation studies on OA articles, all of which are well known. Instead, consider that, if you want the broadest audience, it makes much more sense to underwrite the costs to get to OA and then let anyone see the work. Unless, of course, you're happy with a small, insular set of readers. This, though, comes with a longer term danger of disciplinary invisibility and the commensurate reductions in funding for work that this entails.
In other words: at this scale, the economics to cover the costs of OA are not hard to achieve. OA could boost your readership and the profile of your society. Sure, you're not wrecking library budgets. Just because you're not drowning the planet, though, doesn't mean that you couldn't do things better. And it's doable in these cases.