Ronny Fernandez asks:

You know the result that a per unit tax levied against consumers has the same effect as a per unit tax on producers? Does it imply that fining the victims of crimes should be just as deterring of the happenings of crimes as fining criminals?

This question feels like a great example of what I think of as a “flossing question”. Sometimes I notice that two different concepts are pointers to the same place in my mind. For example, I remember noticing that when I thought of Chicago and Boston I was internally thinking of the same concept; I still kind of feel this way about Tibet and Nepal. A flossing question is one which forces you to really clearly think through the difference, so that you afterward like you’ve separated them out and cleared out conceptual gunk that had built up because of the false equivocation. I love thinking about really basic-sounding questions; this is part of why.

It’s not equivalent to fine perpetrators and victims of crime. It’s true that in both cases, levying a fine will reduce the crime rate. But fining victims of crime generally leads to a less socially efficient outcome, while fining perpetrators leads to a more efficient outcome.

The central difference between the two cases is that both producers and consumers consent to a trade, and so if you want a trade to happen less you can disincentivize either party, but victims of crimes don’t consent to the crime, and so disincentivizing them doesn’t cause the crime to “just not happen”, it causes them to do costly things to avoid the crime.

Another way of putting this is that in the analogy between fines and pigouvian taxes, the “criminals” who are choosing to do something which has negative effects on others are both the producers and the consumers, and the “victims” are everyone else.

When victims of crimes do things to reduce the probability of a crime, this is inefficient. Despite my bad track record on the topic, let’s use the example of bike theft. Imagine that in the absence of policing, cyclists will spend $50 on bike locks to prevent their bikes being stolen. This is a worse outcome than if they could have spent no money on bike locks and had their bikes protected by the ability of the police to punish crime. And if you levy fines on bike theft, then the rate of bike theft will decrease, but this will reduce cyclist welfare, because they will now be spending more on preventing their bikes from getting stolen than they think is worthwhile. This is analogous to the situation where people who are being inconvenienced by pollution: they’ll spend the optimal amount of money to mitigate the harms of the pollution, but this might be worse than if the cost of getting the polluter not to pollute in the first place.

But if it’s cheaper to prevent bike theft with police than bike locks, why did the cyclists buy bike locks instead of buying policing? Partially it’s because policing is the kind of good that makes sense for governments to provide, because it has positive externalities and has good redistributive effects and so on, but a more important answer is that the legal system is allowed to threaten you with taking your stuff and putting you in jail. This is convenient because you can disincentivize people a whole lot by threatening them. And private individuals can’t replicate this because only the legal system is allowed to threaten people like that. (Security guards only have good threats because they can give you to the real police, and because they’re legally advantaged if they get in fights with you while you’re breaking the law.)

See comments on Facebook here.