Sin taxes and liberty

I have discussed quite a few times sin taxes that are instituted to redress some individual behavior. Indeed, people make choices that harm others or themselves directly or indirectly. They create congestion and pollution by driving a care, they increase tax- or insurance-financed health care cost by smoking or becoming obese, they also become public hazards by being drunk. This is one reason why we often tax automotive fuels, unhealthy foods, tobacco and alcohol more than other goods.

But some people object to such sin taxes because they infringe on personal liberties. One of them is Gilles Saint-Paul, whose work I have several times discussed here in a positive light (I, II, III), but this time I have to disagree. His view is that the state is too paternalistic when it intervenes in otherwise free markets with sin taxes, and this has become worse since behavioral economics has highlighted choice patterns that deviate from standard utilitarianism. Well, this is exactly the point. Behavioral economics has brought forward that there are situations were people take actions that they later regret. This is precisely when they would appreciate (at least later) some paternalism in the sense that the state can provide them with a commitment device.

So why does Saint-Paul object to this? His argument is that one should not object to personal choices, and that people should only blame themselves for poor choices. But what if one can help them? Should this not happen only because it is the state? He complains that economists have abandoned utilitarism, which maximizes the sum of individual utilities. I do not think that is correct, but he seems to completely ignore that there are externalities out there, that there is regret, that there are temptations, and that there is lack of commitment. and all this should not be myopically ignored when computing utilities. He is going as far as comparing this supposed abandonment of utilitarianism to eugenics. In other words, he sees excessive government intervention. I agree that there is potential for this, but I see no demonstration that this is happening, and the mere fact that there is intervention is not sufficient, as Saint-Paul seems to imply in a rather puzzling paper.

PS: Robert Wiblin at Overcoming Bias has recently made a similar argument to libertarians in favor of paternalism and also finds it "incredibly obvious." Yet, it needs to be made.