I’m fine with the study’s definition if we’re willing to be consistent about it. For example, what’s the cost of all violent/property crime, including jail and prosecution?
Are you willing to say that any portion of that not paid for by criminals is a “subsidy to criminals”? If so, then you’re at least using your terms in good faith. If not, you’re applying the label unfairly.
The ad absurdum there is that you have to say that seizing all assets from convicted criminals and forcing them to work off their debts is merely “ending the crime subsidy”. Most people would feel a little dissonance at that kind of label.
I think the silliness comes from trying to apply idea of subsidies to people, the concept is meant for companies operating in a market. It's meant to address fairness of one company competing with another.
It's not arbitrary, companies and countries sue each other in WTO to address different level of subsidies. I never heard of criminals doing the same. I think we are not looking to optimise rates of car theft with market forces. Do you?
Companies are different entities entirely, we don't talk about profit when we educate schoolkids, and I can't write off expenses before paying income tax, while companies can.
Right, there’s sort of a gentlemen’s agreement not to be consistent regarding the equation of subsidies and “unrecovered negative externalities”. That just means that if you have enough political power, you can ignore demands to be consistent. It doesn’t mean, as you seem to think, that there’s an actual substantive difference between:
a) “Letting polluters get away with not compensating victims of pollution is a subsidy.” vs
b) “Letting muggers get away with not compensating their victims is a subsidy.”
Are you willing to say that any portion of that not paid for by criminals is a “subsidy to criminals”? If so, then you’re at least using your terms in good faith. If not, you’re applying the label unfairly.