Mr Komanoff --

I'm a big fan of your work -- and I agree that people are price sensitive, and that rising prices are dampening demand a bit.  Still, I think your spreadsheet substantially overestimates the short-term price elasticity of gasoline, perhaps by a factor of 3.

To check one of your key assumptions -- that a 10% increase in real income corresponds to a 10% increase in gas consumption -- I checked federal figures for total personal income in Washington State, from 1969 through 2002, and compared it with federal figures for total gas consumption in Washington state.

The result -- on average, a 10% increase in total personal income corresponded to a 4.8% increase in gasoline consumption.  Which means that the elasticity of gas consumption, relative to income, is about 0.48, rather than 1.

When I plug that lower figure into your spreadsheet, I get a much lower short-term price elasticity for gasoline -- 6%, rather than 20%.

It would probably be worthwhile comparing GDP, total US personal income, and total US gas consumption; and also do the same thing for each state.  That might give you a more accurate estimate for the income elasticity of gas consumption.  But my guess is that no matter where you look real income has risen much faster than gas consumption over the long-term -- which means that gas price elasticity is lower than 20%.

I'm pretty sure that I've done the income-gas consumption elasticity calculations wrong, now that I look at them again.  I'm sure that it's less than 100%, by a substantial margin (from 1969 to 2002, total personal income went up by 250% in constant dollars, and gas consumption went up 87%).  But I'm not sure the elasticity is really .48; perhaps it's more, perhaps less.  

Either way, it seems like it would be wise to take a look at the issue...