Atrios says a gasoline tax isn’t such a good idea because demand for gasoline is highly inelastic. I’m pretty sure that’s not right. I don’t have the paper handy, but as I recall studies of the 1973 crunch showed that the short-run price-elasticity of demand for gasoline was about -0.1. But over five years, as people bought smaller cars (today, of course, they could by hybrids), and either moved closer to work or found jobs closer to home, the elasticity was rougly 1.0: a 10% increase in gasoline prices decreased consumption by about 10%.
Much more importantly, that assumes that all of the tax increase is paid by consumers. That’s almost certainly wrong. U.S. gasoline consumption is a big piece of world oil consumption. As prices go up and the quantity consumed falls, the equilibrium price of a barrel of crude falls, too. So some of those taxes get paid by owners of oil-wells in the U.S., and some of them get paid by, just for example, Saudi Arabia, Iran, and Russia.
Gasoline, or even petroleum, is probably the wrong thing to tax; it would be better to tax fossil fuels generally by putting an effluent fee on carbon emissions.