The big lesson in the decade-long flat-lining of gasoline consumption in the Northwest: prices matter. When prices were low, we used more and more gas each year. But as fuel prices rose, our consumption first stabilized, and then began to fall. The inflection point happened in 1999—when the price of gasoline started to drift upwards, after about a decade of low and stable prices. 
So it might seem that prices are high enough now to have a substantial effect on on our driving habits. But here’s the thing: no matter how high gas prices get, driving is still way too cheap. Whether we realize it or not, we impose subtle costs on our neighbors and community every time we take to the road. Our cars spew climate-warming gases into the atmosphere, and create wisps of pollutants that foul our air. Sometimes when we drive, we impede pedestrians trying to cross the street, or slow down other drivers who are also trying to use the roads at the same time. We create noise; we gradually wear away pavement; and we impose small but measurable collision risks on other road users—drivers, pedestrians, and bicyclists alike. And our petroleum addiction has entangled us in military engagements that have cost both lives and treasure.
Of course, we do pay some of those costs as we drive, in the form of gas taxes. But as high as those taxes might seem—55.9 cents per gallon in Washington, 49.4 cents per gallon in Oregon, and 43.4 cents in Idaho—they don’t even cover the cost of the roads we drive on, let alone all of the other external costs of driving.
Some of those costs are tough to peg. Estimates of the social cost of carbon emissions vary widely, from a few pennies to well over a dollar per gallon of gas. When you add in all the other external costs—including both the costs that accrue per gallon of fuel consumed, and those that accrue per mile driven—the price we pay to drive a mile only hints at the costs we impose on society.
What this means is this: no matter how expensive gas gets, it’s still too cheap! Until we take a serious stab at making the price of gasoline tell the truth about its social cost, we’re getting a free ride—and, as a consequence, driving too much. And if we actually started paying what we should for gas, our fuel consumption would fall even faster than it has over the last decade.
Cheering
Let’s have a $100 a year tax on bicycles. Why? To offset the aggravation.
Lang
This is a really interesting graph. The switch from blue to red occurs when prices are below the 1990 level and about the same as the rest of the previous decade. The big changes in gas price don’t come until 5yrs after consumption switched from “increasing” to “stable/falling,” suggesting there are other factors at play. Given the price/consumption history it seems unlikely that consumption would respond that quickly to a small uptick within the recent historical average range. Which begs the question, what are the other factors at play? Overall domestic economic activity would be one obvious choice, but if the switch occurred in 1999 that was a time of economic expansion. So what else is involved?