The complaint that publishers add little or no value to the scholarly publishing process is one of the most common soundbites used by the OA partisans. And yet, if this were true, why would the NIH Public Access Policy or FRPAA be structured the way that they are?
Although the phrasing that is typically used is along the lines of "taxpayers fund the research so they are entitled to read the results of that research at no additional charge" that's not exactly what the mandates are actually after. Several commentators have suggested that the public's entitlement could be met by making the research progress reports that are required by all funding agencies available. Some agencies already have systems available to do that.
As Phil Davis points out in The Scholarly Kitchen, recent studies show that fewer than half of NIH-funded clinical trials actually result in published articles within 30 months. So wouldn't access to progress reports be a huge improvement over a focus on published articles? Given the ratio of funded studies to published articles it is clear that neither the NIH Public Access Policy nor FRPAA are or could be very successful in achieving the goal of providing access to all the results of federally funded research.
But that's not actually what those policies are trying to achieve. It's only after a publisher has accepted an article for publication that NIH or FRPAA is interested in getting a version of it. Argue all you want about whether or not the publisher adds value, but these mandates are very explicit that there is something that publishers do that is absolutely essential.
It has something to do with peer review apparently. Although again the partisans are quick to point out that the reviews are done for free and the decision is made by an editor who is getting, at best, a tiny honorarium, and that whatever it is that the publisher contributes, it can't be very much or be very costly.
So why do we need the publishers? Why doesn't the NIH policy or FRPAA establish their own independent peer review process if peer review is so important, but so cheap? Then the publishers couldn't complain that they were providing something of value for which they are not being compensated.
But I haven't seen anything like that suggested anywhere. The NIH policy and FRPAA absolutely depend on publishers contributing something. How can something of so little value be so absolutely essential?