What's the Point of a Little Gas Tax?

Ryan Avent writes of the gas tax that it's trivial in relation to the normal fluctuations of the market:

. . . any politically tolerable carbon price, enacted by cap-and-trade or tax, would add mere cents to the price of a gallon of gas. I don't remember the exact data point, but I recall George Bush railing against Lieberman-Warner by saying that it would add 40 cents to the price of gasoline by the year 2020 (or something like that). At a time when prices were rising by well over a dollar in a matter of months, this was not a particularly frightening statement.

But people hate the idea of expensive gas, and it could be the case that even a small potential increase in prices would make it more difficult to pass a carbon price. For this reason, some greens (Dave Roberts, for instance) argue that a carbon pricing system should exempt transportation. I disagree -- the idea of pricing is that emission reductions will occur in difficult to predict places, which makes me extremely reluctant to exempt such a large sector of the economy -- but I understand where he's coming from.

But the most important thing to understand about the above is that consumers are far more vulnerable to market-driven swings in the price of gasoline than they are to regulation-driven changes. Any potential government-engineered gas price increase pales in comparison to the spike markets delivered in 2007 and 2008.

There are two things about this that trouble me.  The first is that Ryan and I may conceive of a 40 cent increase in 2020 as trivial.  On the other hand, I don't drive to work every day, and when I do, my commute is three miles each way.  Traffic is a vastly bigger problem for me than the cost of gas; I fill up my tank about once a month.  For households out in the Great Beyond, a permanent 40 cent a gallon increase is quite a lot--particularly if the 2008 prices were a bubble, and the permanent cost of a gallon of gas is more like $2.00.

Second, if Ryan is right, why are we doing this?  It did take a lot of increase to change driving behavior, which is why in Europe, taxes can account for as much as 90% of the price of a liter of gas.   There's a plausible argument that a 40 cent tax won't do much to change driving habits.  But then, why have the 40 cent tax at all?  If you have a Pigovian tax that doesn't alter the production of the externality, cap and trave starts to look a lot less like controlling our carbon emissions, and a lot more like taxing people who aren't Ryan Avent and Megan McArdle to top up government revenue.