I have seen several journals switch from closed to open access models in the past. Is that simply a decision of the editorial board? Do they themselves have to convince the publisher (let's assume it is a commercial publisher)? And if they have to convince the publisher, what arguments/proof would convince the publisher?