Holly Mercer reports that the self-archiving rate in library and information science is nearly 50% among librarians (and double the 20% global baseline even among nonlibrarians). Nevertheless, not even all articles for which immediate OA self-archiving has been endorsed by their publishers (c. 58-68%) are yet being self-archived even in library and information science, let alone the over 90% after embargo (or the 100% that can be deposited immediately in Closed Access allowing the semi-automatic eprint-request Button to provide Almost-OA during any embargo).
Among the potential solutions, the most important and effective one is for institutions and funders to mandate self-archiving. (Several library faculties have already taken the intiaitive of doing this.) It is also important to make institutional repository deposit the official mechanism for submitting publications for institutional and national performance review (see Liège model).
One slight correction: Alma Swan’s reported rate of 49% self-archiving was not for total articles; it was just the percentage of authors who said they had self-archived at least once. (And both Alma’s studies and those of others have found that authors are often not sure what they mean when they say they have self-archived!) This too will be self-corrected as self-archiving mandates, with their links to research assessment, grow.