As a young tenured academic I have been systematically refusing to do reviews for non-open-access journals. (I'm certainly not senior enough to create a new journal and be taken seriously.) I hoped that this might give some ideas to my peers but it's seen at best as a weird quirk, at worst as a selfish move (because "the community" needs my help to review papers). Usually the answer is some variation of "yeah the system is not perfect".
The only argument I have seen which motivates researchers to care about open access is when it's mandated by funding agencies (e.g., in Europe, the ERC). Researchers really want to get these grants, so when these agencies talk about open access, they listen very carefully. It both makes me somewhat optimistic and quite cynical that the only way to move towards open-access seems to be via funding agencies (as opposed to researchers caring about the problem). Plus, for many agencies, "open access" here means "gold open access", i.e., continuing to work with the usual publishers, who make the articles available online, but move the "costs" (and huge margin) to the authors, with "article processing charges" of $1000-$2000.
I've been arguing that it's a coordination problem.
Aspiring scientists don't have the pull to change the system (if they move individually they only bear the individual disadvantages without bringing about the big collective benefit), and established scientists generally don't care or are too busy with other things [1].
That's why a concerted ("political") move like this cOAlition S thing is so important, and I find it very disappointing that academics speak out against it. I will read more to understand why.
[1] With notable exceptions, like Don Knuth getting the editorial board of the (Elsevier) Journal of Algorithms to resign and move to a new journal, taking the reputation with them.
I have to get as many publications into as highly reputed international journals as possible if I want to have any chance at getting a permanent position, and the vast majority of all reputable journals in my area are behind private paywalls. There is a maximum of 3 reputable, fully open journals and they are still not as prestigious as the established ones.
If I decided to only publish open access journals from now on, that would be immediate professional suicide. According to our institute's internal regulations, I'd be gone in two years from now, maybe even in one if they interpret their rules more strictly.
You're talking about eliminating, or at least sidelining traditional academic publishing. It's a completely different proposal.
That's right, and that's also why the prestigious researchers and universities, those with an already established reputation, have a responsibility: collectively leave the editorial boards of for-profit publishers, set up alternative venues with the help of university libraries. Share the archival, indexing and discovery effort among universities via peer-to-peer digital library federation.
All the tools are there. The same way that places like Stanford, Berkeley and MIT made MOOCs a thing, they can revolutionize scientific publishing.
There's already such venues such as JAIR for AI research (and that was set up looong ago without all the tech we have today), so it's certainly possible. It just needs to become the norm rather than the exception.