One element of power in the academic world is association or affiliation. Established researchers have it, and early career academics want it. But interestingly, Universities are the ones who really need it. They go about exerting their powers in order to obtain it: hiring this person (and their research), or creating this program or centre in order to become it, and so on. Affiliation brings funding.
So, I propose an experiment.
If you have produced a piece of research (ie. you have something to show for your work), and that work was conducted mostly in an unsupported way – like this:
Then that research should be affiliated in the following way:
[Researcher’s name] is lecturer in XXX at a University in XXX city/country. The University’s name has been withheld in order to protest the lack of research support available. This piece of research cannot be included in research data calculations for that institution without the author’s consent.
Maybe those who enable such things will start to pay attention to the way this pathological system is organised.