1) authors pay the costs of publishing, open access or not.
2) societies pay the costs, which are often supported in part by member dues, ethically questionable advertising, authors fees, and, yes, charges on gated articles.
3) funders, often the government, pay the costs, open access or not.
All of these have strengths and weaknesses, but being a well-known MMT sympathizer and recognizing that, in the U.S. anyway, much research is funded directly by the federal government–which as a currency issuer can’t run out of money, in principle, I’m going with open access publishing with grants covering publishing costs. In reality, of course, this would require NIH, which has a fixed budget, to make the case for slightly higher allocations to publishing expenses or to get more money from Congress**.
Not much chance on the latter, so are we ok with re-routing 1-2% of a modular NIH grant (and slightly higher for many NSF grants) to publishing?
*I’ve left off universities and ‘rich dudes’, because rich dudes aren’t widespread on this issue, and university administrators would sell their own mothers these days.
**The idea that open access publishing fees could lead to inflation is ridiculous.