Ben, the main arguments for taxation to overcome internalities are laid out in the first couple of sections of this paper. Here is the abstract of a paper that purports to show increased happiness as a result of excise taxes.

The problem with internalities is that they prevent people from maximising their lifetime utility. We can tell that we’re seeing it if forecast decisions don’t match observed decisions when all else has remained constant. Obviously, that’s not easy to identify, which is why laboratory experiments have been done in controlled conditions to confirm the existence of inconsistency.

It’s possible that inconsistency would cause underconsumption of some goods; however, that wouldn’t happen with addictive goods unless our understanding of them changed dramatically. So subsidising alcohol or cigarettes is unlikely to be justifiable using internalities. Maybe you’d subsidise gym visits or something else that people want to do in future but never get around to.

The usual counter-argument to taxation is that we’re judging lifetime welfare from one point in time: maximising it will reduce instantaneous welfare from the point of view of every future period. That’s a fair point, but concepts of intertemporal welfare have to use some reference frame and using the current period’s seems better than any other. Of course, if you prefer the reference frame of the evening drinker to the same person waking up in the morning, then you might tax at different levels but you’d still choose to tax. It’s just a precommitment device enforced by a third party and they’re always useful for people who act inconsistently.

You point out that people learn about their inconsistencies and try to correct them. That’s certainly true and is the reason why we observe so many types of precommitment behaviour in everyday life. Unfortunately, precommitment is not always possible without the help of a third party who has enforcement power. That’s where government action might be useful.

First, I think it is a red herring in the alcohol debate. Let’s say our drinker decides in advance to drink until he passes out, but when the drinking starts he’s not in the mood. Should the government step in and correct this time inconsistency? Time inconsistency is always presented as working against the individual, but why should it? And where it doesn’t, is anybody going to cite time inconsistency as a reason to put more drinks in front of the now-unwilling? Red herring, me thinks.

Second, how do you distinguish time inconsistency from mistakes or from the effect of environmental factors? Our hypothetical drinker planning just two drinks unexpectedly bumps into old friends and has five.

Third, what, actually, is wrong with time inconsistency? In what general sense are the preferences of the person in the morning thinking about drinking superior to those of the person with a different view in the evening? What is it about internalities that tells you not to write rules that convince the drinker planning two drinks to actually go ahead and have five? Nothing, as far as I can tell. That asymmetry is comes from elsewhere – the policymaker’s own preferences, I presume?

Fourth, how can we ever develop any general confidence that somebody external can ever be a better judge of the welfare of person A – letalone of the population? That requires knowing preferences – simply observing changing views about what’s optimal for one’s self is not sufficient to establish time inconsistency. For every genuine case of avoided regret, I would expect 100 cases of avoided enjoyment. Why? Because on average I do not believe people repeatedly beat themselves over the head in a way not justified by benefits. There is learning. The economist who sees behaviour that looks suboptimal is probably missing things from her spreadsheet.

To me the internalities argument is a fairly transparent attempt at a technical justification for paternalism. I think it is a manifestly untestable hypothesis. It appears incoherent. Any suggestion that policy should follow from it badly underestimates the difficulties officials have in doing anything, let alone writing rules that correct for internal biases. Internalities looks to me like a technical expression of intolerance. A sham, in other words.

Tell me what I’m missing, Rauparaha. I will listen and read if you provide links.

Definition? I don’t see a definition … are you making fun of my ranting 😛