Here are some thoughts that came up during a private conversation about Elsevier a few weeks ago... The question is, why is open access publishing so expensive

Here are some thoughts that came up during a private conversation about Elsevier a few weeks ago... The question is, why is open access publishing so expensive in many journals? From what I understand, it's because of the need to pay the technical staff (technical editors, sysadmins, etc.) - costs that in traditional publishing are paid by your library (i.e. spread out as a tax across all faculty's grant overheads+taxpayer money). These costs can be driven down by consolidating the journal publishing platforms (i.e. providing a service for people to organize their own journals and a market for technical editors to provide their services to the content editors). And critically, there needs to be a way to incentivize authors to drive down the total cost of publishing. A big part of the reason big old publishers continue to exist now is that there is an open loop in the incentives: the authors don't bear the costs that they incur to the universities, and there is no direct feedback from the audience that is shut out by this cost. If authors are given an idea of what it really costs to read their article, and the breakdown of who bears that cost, the journals will have to compete on that total cost. But this will not be very effective until university libraries start breaking down the cost of subscriptions to individual users. So one way to approach this is by providing university libraries with a system to estimate the individual cost of each article view (i.e. crudely, if a subscription costs X and Y users view Z articles, then each user will have incurred some cost that is X/Y on average). Then start discussing what to do about who carries the cost burden from the users who are outliers on top. This is the first step toward closing the incentives loop. I agree that there are too many journals, but unfortunately there will need to be even more before the ones with the right combinations of cost and editorial structure will win out. Going back to the common platform for publishing and editorial services that I mentioned: the real kicker will come when the platform is integrated with a social network (where nodes can be organizations as well as people). In other words, once you provide an easy and trusted way for editors to write down their web of trust, display that web to others, and act based on it, you will have put online the informal web of trust (reputation) that underlies publishing. This opens really powerful possibilities. In the meantime, traditional publishers are holding on for dear life and using increasingly desperate tactics (like buying legislation, in Elsevier's case) to hold on to their money funnel. This is holding back science by artificially restricting access to publicly funded research. (And don't get me started on the textbook business... although good things are coming to that because Apple will disrupt it from the bottom.)