Not all pennies are the same...  

A truly curious economic research finding: we react differently to a gas price increase when it is caused by taxes than when it is simply supply/demand driven. In fact, we cut consumption more for a penny increase in taxes compared to a penny in intrinsic gas price.

That's from a new NBER working paper by Shanjun Li, Joshua Linn, and Erich Meuhlegger. As the authors note, this has some interesting implications. It suggests, first, that estimations of the revenue that can be raised from petrol tax increases that are based on elasticities with respect to petrol prices will overstate assumed revenue gains. On the other hand, it means that reductions in consumption driven by tax changes should be less painful than those driven by movements in the price of oil. If America is interested in cutting its dependence on oil, then weaning consumers off petrol via tax rises will be easier on the economy than simply letting market-price variation do the work.
The paper suggests that more work is needed to understand the causation, but they point toward one logical factor: consumers may be more likely to read tax changes as permanent. A household that observes what looks like a permanent increase in petrol costs due to tax rises will quickly adjust its behaviour to minimise the burden—by driving less or purchasing more efficient vehicles. The household may delay such action when market movements send prices up as it waits to see how persistent the change will be. That delay represents more profit for producers and more of a hit to other household consumption than we'd get with a straightforward tax hike. [More]

Regardless, I think we are learning we will cut consumption when the cost increases. In fact, between Boomers slowing down, young people driving less and later, and better cars we may surprise ourselves how low we can go.

On the other hand, that's not good news for volume based gas taxes, unless we raise them commensurately.