Question: How would a gas tax reduce consumption by enough to make it worthwhile?
Answer: The full question asked by Garth Brazelton was: "How would a gas tax reduce gas consumption by a degree that would warrant such a tax increase on Pigovian grounds, given the evidence of such a small elasticity in gas demand? It seems the evidence supports the fact that a gas tax would be great for revenue, not for externality offsetting."
Any rise in the gas tax would be warranted on Pigovian grounds, so long as the tax increase was not excessive (and given political realities, overshooting the optimal gas tax rate seems unlikely). I don't see the issue here, unless you believe that unless we hit the exact optimal tax rate, any change in the tax rate would be unwarranted. An improvement is an improvement, even if we miss the "ideal" point.
As we saw in How would a gas tax reduce consumption?, increases in gasoline prices do reduce consumption. The exact amount that consumption will fall will depend on a number of factors, such as the jursidction the tax increase in being placed on, but we do have available historical averages.
An aside: Does anyone know if the price-elasticity of gasoline rise or fall, depending on the initial price? That is, would a 20% rise in the gasoline tax reduce consumption twice as much as a 10% rise? Less than twice as much? More than twice as much? I suspect it's the latter - but I'm not sure. Anyone know any good studies?
These things are always easier to think about with concrete figures. So we will use some:
Right now in London, Ontario, the price of gasoline, including all taxes, is approximately 90 cents (Canadian) a litre. There are already a few taxes on gasoline, such as a 10 cent a litre federal tax (as discussed in the article The Gas Tax - Answering NoPigou's Questions.
Let's raise the federal tax from 10 cents to 19 cents a litre. Close to the entire amount should be passed along to the consumer, so the final price of gasoline should go from 90 to 99 cents a litre.
(In fact, that's not entirely accurate, since the 6% GST is applied to the price of gasoline inclusive of all taxes. So consumers get taxed on the tax the pay. Seriously. Anyhow, the final tax rise should be 9.54 cents. But for the purposes of this exercise, we will ignore the GST.)
With the 9 cent increase in the excise tax in gasoline, we have a 10% rise in prices. That 10% rise in taxes should lead to an immediate decline in gasoline consumption of 2-3%. Over the space of a few years, gasoline consumption should decline by 5-6%, all else being equal.
According to this article from CTV, Canadians consume approximately 40 billion litres of gasoline. If we enacted our tax, we should see gasoline consumption fall by between 0.8-1.2 billion litres immediately and 2.0-2.4 billion litres over the span of a few years. Since burning a litre of gasoline produces about 2.4 kg of CO2 (Source: Government of Canada), the tax would reduce CO2 emissions by about 5.3 million tonnes a year. That is not an insignificant amount.
As far as government revenue goes, a 10 cent tax on 40 billion gallons brings in about 4 billion dollars in revenue to the Canadian government. A 19 cent tax on 37.8 billion gallons brings in about 7.18 billion dollars in revenue. So Garth is correct - a rise in gas taxes would increase government revenue significantly.
In the next section, we examine how government could use this money to reduce other taxes.
Next: How can we be sure the government would use the additional revenue to offset other taxes?
Previous: How would a gas tax reduce road congestion?
Main: Gas Tax and Carbon Tax FAQ
