If your argument can be resolved by the scientific method, then why even bother arguing?
To come up with an argument which can be tested experimentally. At least in physics, which is the subject area here.
You're often not that lucky.
Alas.
Should the government be bigger or smaller? In principle, you could randomly sort the 50 states into three groups where you increase the budget of one, decrease the budget of another, and leave the last as a control, but nobody's going to do that.
Ah, but there are other ways. There are 200 other countries in the world, and some come very close to the experiments you want to perform, now or in the past. Does gun control reduce crime? Tested. Does universal healthcare reduce medical expenditures? Tested. Does changing attitudes toward women in STEM improve the fraction of exceptional women scientists? Tested. Does decriminalizing marijuana lead to more drug use? Tested.
You can argue about these issues with all the Bayesian might until cows come home, it is no substitute for a good experiment.
There are 200 other countries in the world, and some come very close to the experiments you want to perform, now or in the past.
They let you look for correlations. They don't let you run a good experiment. People can come up with theories that explain the facts, just as MWI and the Copenhagen Interpretation explain the facts. Science won't save us here.
Sean Carroll, physicist and proponent of Everettian Quantum Mechanics, has just posted a new article going over some of the common objections to EQM and why they are false. Of particular interest to us as rationalists:
Very reminiscent of the quantum physics sequence here! I find that this distinction between number of entities and number of postulates is something that I need to remind people of all the time.
META: This is my first post; if I have done anything wrong, or could have done something better, please tell me!