re: In practice it often isn't a good idea, and in principle I'm generally against it except in cases where there is a problem with an externality.
Yes, but all taxes penalize and reward. Taxes are by nature social engineering.
re: Do you mean a tax based on the estimated fuel consumption of a vehicle, or a tax on fuel?
Without really thinking it through, I would say both. Current gas taxes obviously favor drivers that get high gas milage; they pay less in tax. But that incentive isn't working very well.
My inclination is towards a sales tax based on EPA milage ratings. Figure a mean, and give increasing credits for cars below the mean; increasing taxes above the mean. People respond better to chunks of money than they do to incremental amounts at the pump.
re: For a particular grade or quality of crude there is pretty much just one world price. If we buy oil from Mexico and Italy is buying from Iran...
Yes, but I can't imagine Iran saying to the US "You've been bad, so we are going to cut off Italy's oil supply". Also it's politically more confortable not having to do business directly with your enemy.
You didn't respond to this part: Presumably gas purchases would go down, feeding new money into the economy. Also, with reduced demand, the commodity price of oil should go down, compounding the effect. I think I read that oil represents ~25% of our trade deficit, so reduced consumption helps in that regard.
Honestly, I don't see a downside... and I see a lot of benefits. The auto companies would certainly have to hustle and increase spending on R&D... and they are not doing great right now. The oil companies would certainly whine like crazy, but they are currently hoarding obscene profits. From a social benefit perspective (without even talking about pollution), I think it's a win win win proposition.
John |