Harvard Open Access Policy Benchmark Needed

It is important to calculate what percentage of the total annual refereed journal article output of Harvard (participating Faculties) is represented by the c. 6457 deposits to date in Harvard's DASH Repository since adoption of Harvard's OA Policy? That is the objective measure of the success of an OA policy, and hence of whether it provides a model ready for other universities to emulate -- or whether it still needs some tweaks (e.g., to make it more like the U. Liege ID/OA policy, which (1) requires immediate deposit with no waiver, (2) only requests (but does not require) that the deposit be made immediately OA, (3) designates repository deposit as the sole means of submitting journal articles for research performance review, and has generated 67,631 deposits to date). The global baseline rate of making articles OA (without any OA policy) is about 20% (varying by discipline). The target is of course 100%. And about 60% is a benchmark, because that is the percentage of journals that already endorse immediate OA deposit (hence do not require Harvard-style rights retention in order to make deposits OA immediately). It is extremely important to get a clear idea of exactly how well Harvard's policy is doing after nearly 4 years: If the deposit rate is near 100%, it is doing as well as or better than all other kinds of OA mandates. If it is close to 60%, that's still good, but it's not clear whether its rights-retention clause is the cause, or its deposit clause. And if it's closer to 20%, then Harvard's deposit clause is not working and needs upgrading to ID/OA. This is all the more important since it is the Harvard model that other universities are likely to follow, come what may.