The ARL has published a very helpful report for universities and colleges that receive NIH funding regarding their options for facilitating their authors' compliance with the requirements of the new NIH Open Access Mandate: Complying with the NIH Public Access Policy - Copyright Considerations and Options (SPARC).
I would note that even without this mandate, many publishers had already established policies that permitted public access posting (about 60% according to stats available from SherpaRomeo). The interaction between these policies, typical publisher contracts, and the new Mandate's requirements that authors retain sufficient rights to grant the NIH the public access rights necessary for grant compliance would make an interesting addendum to this report. The report 's author does not explore this option, perhaps believing that to leave it up to authors to wade through such contract/policy/regulation interaction subtleties is probably not the best risk management strategy. Or, he may not realize that a large percentage of publishers have these policies. I certainly think it's worth having a look at. If the combination of the policy and the contract properly referencing the policy were sufficient to meet the requirements of the mandate, it might reduce the administrative burden in many, though not all cases. But that might be the rub right there: depending on the institution's size, the strategy that requires the least case-by-case might be a better choice even if it isn't technically necessary in all cases.
