The UC Faculty Survey results are summarized in a somewhat misleading
way:
"There is limited but significant use of alternative forms of
scholarship, with 21% of faculty having published in open-access
journals, and 14% having posted peer-reviewed articles in
institutional repositories or disciplinary repositories."
(1) The practise in question is making published articles open access
(not "alternative forms of scholarship").
(2) 21% of UC Faculty published articles in OA journals and 14% posted
published postprints in repositories.
(3) But 31% posted posted postprints on personal or departmental
websites (and 29% posted preprints).
So the comparison between OA publishing and OA self-archiving is not 21%
vs. 14%. It's 21% vs. either 31% or anything up to 74% (if the 3 forms
of self-archiving were additive).
UC should correct these summary figures. Otherwise it is giving a very
misleading picture of the actual proportions at UC between the two ways
of providing OA. This is important, because it is OA self-archiving that
has the greatest scope for growth and acceleration, as OA publishing
cannot be mandated, but OA self-archiving can (and should be).