Right, that could have been more clear. It referred to money one often has to pay to make a work available as open access when it's in one of the traditional "prestigious" journals, i.e. the non-subscription ones.

However, in that case, too, the money is not the researcher's. That said, they are the ones that spend it.

In the case of subscription journals, the argument is similar, except that in this case, it's academic libraries who will have to pay whatever amount of public money a publisher asks of them just to be able to access that research. If the researcher could obtain their credentials elsewhere, they would be able to make their work available somewhere where the libraries and the general public would not have to pay to access it.