The same logic also says that the income tax, VAT, etc. should be set at zero, as they don’t have specific negative externalities. If that approach gives enough revenue for all government functions, then great, problem solved.

If not, then a tax will have to be increased beyond the optimum Pigovian point. At that point, I’d say it still makes more sense to increase resource based taxes such as fuel duty, as that will continue to reduce a negative externality and therefore be less damaging than increasing income or sales based taxes.

Not if you’re a plant it ain’t.

Either I travel to work or I don’t. If I do, how does an increased tax (or even the present level of tax) optimise the social consequences of my action?

Tim adds: Marikets don’t adjust immediately. In fact, there’s at least one serious economist who thinks that it takes a century to reach a new equilibrium after an exogenous shock….and a change in tax rates can be considered such. I think a century is probably too long but that’s just opinion. The increased tax will alter your decisions over the entire timespan that you take such decisions. Which job you take next, whether you commute, home work, drive, walk, cycle….all are, over time affect by that tax change.

Tim adds: No. The point is “socially optimal”. There are costs and benefits to everything. We want behaviour to change to the point that we are fully accounting for the costs of the externalities in our calculation of costs and benefits. We don’t want behaviour to change to the point where the externalities no longer exist….for that is failing to take account of the benefits of the actions that we’ve just taxed out of existence.

For example, some water buffalo farting does indeed increase the chance that a Bangladeshi farmer will be drowned along with boiling flipper in 2200 or so. That’s a bad thing, a cost. But that water buffalo farting also means that some rice farmer can plough his paddy this year and next. That ‘s a good thing, a benefit. What we want is the socially optimal balance between people eating this year and people and flipper not drowning/boiling in 2200. Thus, we discourage the use of the water buffalo today by taxing the amount of damage that such use will cause in the future (I’m ignoring discount rates and lots of other complexity). We don’t ban wbs, we just make sure the use covers the bad things.

Is it? Not being contrary here, but isn’t the idea to correct an externality by providing an incentive to alter behaviour. Wiki suggests this: “A tax shifts the marginal private cost curve up by the amount of the tax. Faced with this cost increase, the producers have an incentive to reduce output to the socially optimum level by reducing the marginal externality to the marginal tax.”

I’m not in favour of emissions taxing, but surely the logic would be to raise them until behaviour changes.