« The $100 Million Beetle | Main | Graph Theory »

September 20, 2005

Is Gas Elastic?

In 1999 you could buy a gallon of gas in Washington state for less than a buck.  As recently as 3 years ago, gas prices averaged about  $1.20 a gallon.  Right now, though, expect to shell out about $2.85.

Washington_per_capita_gasSo what has a 136% price hike done to gasoline consumption?  As it turns out, not a lot.  In 2002, the average Washington resident went through about 8.4 gallons of gas per week.  Based on data through July 2005, that's now down to about 8.1 gallons per week -- a 4 percent reduction.

Elastic?  Not so much. 

Of course, the trends are a bit confusing. The state economy was in the doldrums in 2002, but has picked up a bit of steam since; if economic conditions had remained constant, the decline in gas consumption might have been a little steeper.

Over the longer term, the trends are a little more promising.  Person for person, gas consumption peaked in 1978, at 9.7 gallons per person per week.  We're down 17% from then.  (Which convinces me that federal CAFE standards--while far from perfect--really did accomplish something useful.)

Remember, though, these are per capita trends.  Population growth has worked at cross purposes to improved fuel efficiency:  all told, Washington state is on track to use the same total amount of gasoline this year as it did in 2002, and possibly a bit more.

Posted by ClarkWD | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Is Gas Elastic?:


My first question would be what the curve looks/looked like for electricity, where it is clear and a fact that there is elasticity of demand. The question really is where does the curve start?

It's inconceivable to me that there is NO elasticity of demand for gas...you mean at $10/gallon we will use 8 gallons/week per capita? At some point people will conserve...so all we can say is that gas going from $2/gal to $3/gal doesn't have much of an impact over the very short run.

Posted by: David Sucher | Sep 20, 2005 7:07:36 PM

Even at $3/gallon, gas still barely registers against the other monthly costs of car ownership: insurance, parking, maintenance, and possibly loan payments.

Posted by: Mars Saxman | Sep 20, 2005 7:13:57 PM

I think one needs to look at demand elasticity for gasoline in two contexts: the short term and the long term. In the short term, demand elasticity for gas is very near zero, because people have already invested in the infrastructure for burning the gas (i.e., their cars and their lifestyle choices, particularly housing location) and can't quickly change.

Over the long term, if (and this is crucial) people are convinced that gas prices will stay high and/or go higher, and won't return to "normal" levels, they will make changes in their infrastructure and their lifestyle (buying a fuel efficient car, buying a house close to transit, telecommuting, etc.) that will make a dramatic difference in the amount of fuel they use for transportation.

On another note (and I would love to see Cascadia Scorecard start a thread on this topic), the cost of heating is going to become a major, major issue this winter for everybody who burns heating oil or natural gas, and I would speculate that the demand elasticity for heating fuel is even lower than it is for gasoline.

Posted by: Roy Smith | Sep 20, 2005 9:46:14 PM

Of course gasoline demand is elastic to some degree. Just, apparently, not very much -- at least in the price ranges we're seeing right now. As Mars notes, gas is still fairly cheap, compared with other costs of driving.

For literature on the subject of gas price elasticity, I defer, as always, to the master: Todd Litman of the Victoria Transport Policy Institute. See here:

In table 8, Todd cites this paper...
which compiles elasticity estimates from 69 different studies. The paper finds that, as a mid-range estimate, a 10% rise in real fuel price will result in a ~ 2.5% decrease in consumption within a year.

But this mid-range estimate has a pretty wide standard deviation, and there are few outliers. And it seems to me that the low-outliers have been far more predictive of recent trends than the mid-range estimate.

It'll be interesting to see what will happen if gas prices remain high. Real elasticity kicks in when prices remain high for a while -- and when the psychology of high prices begins to influence the big decisions that determine how much fuel we use (where we live & work, and what we drive).

Posted by: Clark Williams-Derry | Sep 20, 2005 9:53:10 PM

great points.

I've been thinking about doing something on heating, too. Not too much oil heat in our region, really; and the increases in natural gas prices have been somewhat more restrained here than in, say, the midwest.

But I'll try to look into this -- I think it's a very big deal, too.

Posted by: Clark Williams-Derry | Sep 20, 2005 9:59:14 PM

People with less disposable income are being squeezed, and that will continue.

So Gas expenditures relative to a fixed income go up generally by necessity, and Dinner and a Movie happens less and less.

(And it is my pet peeve that in an inflationary period a night at the Movies is keeping up with inflation, if not exceeding it. I nostradamus, that will change in the not too distant future. You can only charge so much for flickering light.)

Posted by: Jon S. | Sep 21, 2005 12:08:44 AM

I'm afraid that we have ended up with just enough government "end of oil" programs for people to believe in them, but not enough to really accomplish anything.

We are really getting, for better and for worse, a free market adaptation to higher gas prices. I don't believe federal mileage programs (including CAFE) are making any contribution.

If we look back, the SUV and light truck loopholes have made it possible for anyone, in any price segment, to choose between high and low MPG.

You have done well in the Northwest simply because you chose differently. The Subaru became your mascot and not the Ford Explorer (or F150). Unfortunately this wasn't true everywhere. Drive out of Subaru country and bigger bad-weather cars become the norm.

So anyway, congratulations on improving your per-capita gas consumption ... but I think you all did it, not CAFE.

Posted by: odograph | Sep 21, 2005 6:30:38 AM

Yesterday I saw an AP article in the PI about Gregiore accusing the oil companies of price gouging. First, I think this accusation is false, as the oil companies will simply charge what the market will bear, as will any unregulated company. Many politicians have advocated for gasoline price regulation, however this will only lead to gas shortages, and other means of rationing will need to be established in place of the supply/demand model.

However, as counterintuitive as it may seem, I think the high prices could be an excellent opportunity for increasing gas taxes. Assuming that the current prices are dictated by supply/demand (and not based on cost of procurement), more taxes would only push down the profits of the oil companies instead of further increasing consumer prices. Of course, convincing consumers of this may be quite difficult, as the general populace wouldn't understand this concept.

Unfortunately, gas tax funds in Washington State must be spent on roads. Obviously it could also be spent on repairing the infrastructure we have, but only political will can stand in the way of building new capacity. However, I'd be intersted to learn exactly what could be built with gas tax funds. Could it be used for road safety, such as sidewalks, guardrails, traffic calming, pedestrian over/underpasses, or shoulders (that could be used by bikes)? Since gas tax funds support the ferries, could it also support railway infrastructure if the state purchased automobile carriers for trains that run in our state?

Posted by: Rodney Rutherford | Sep 21, 2005 8:27:46 AM

All good points so far, especially the long term/short term arguments. How about this idea, based on no documentation. Could elasticity be non-linear? In he sense that as prices increase, there is an accelerated increase in behavior. Perhaps not a "tipping point" at a paricular price, but just a lot more people joining a bandwagon of cancelling trips, carpooling, combining errands, slowing down, the stuff we should be doing anyway. In other words, if the first 10 percent increase in price results in a 2.5% decrease in the sales, the next 10% would cause a 4 to 5% decrease. What do you think?

Posted by: WayneS | Sep 21, 2005 8:42:37 AM

Haha. Wayne, you and I are totally on the same wavelength (see the "Sic Transit" discussion thread from a couple days ago-- and be sure to ignore my conceptual dyslexia with "price elasticity").

I think if you look at this price elasticity primer (http://www.digitaleconomist.com/elasticity.html), you'll see that only the elasticity ratio is linear-- try playing with slider a bit. The effect is generally that there is some inflection point at which price and demand changes are reflected "one-for-one", but if you go significantly higher or lower than that, you won't change demand very much.

The point is that we are still on the inelastic side of the curve for gas-- where it's so cheap that everyone can afford it even if you jack it up quite a bit. The discussion on this thread is around the idea that we may be approaching the point where price changes start resulting in significant demand shifts-- where the ratio is around 1:1.

There is theoretically another price point at which pretty much only the wealthy can afford to buy gas, and price changes above that wouldn't result in significant demand decreases. That is when oil is a luxury, not a mainstream fuel source-- it would be used for explosives and technical/medical supplies at that point, I think.

So if you're thinking of the demand (y) as a function of price (x), then it looks like a wave with trough on the left and crest on the right and a inflection point at the middle, such as you suggested. If you think of it as a ratio (the economist's way of looking at it), it's a straight line like the primer I referred to above. Make sense?

Any economists want to correct my description of price elasticity curves?

Posted by: Jason | Sep 21, 2005 11:26:21 AM

I agree, these have all been great comments thus far. You've hit all the points I just happen to be reiterating to my class yesterday.

Price elasticity of demand is influenced by:
-Availablility of Substitutes
-Proportion of Income spent
-Time period considered (ie. time to adjust)
-Perceived permanency of price change
-Degree of necessity (versus luxury)

You can see that the recent gas price increases are a great example of a product whose price elasticity of demand may tend to be low given all of these criteria.

Posted by: Jeremy Brown | Sep 21, 2005 12:03:56 PM

Back in the 1980's, when I was a student at the University of Oregon in Eugene, I had a Norwegian housemate. He was interested in purchasing a nice car to show-off to his visiting French girlfriend.

As a budding environmentalist back then, I asked him to please consider buying a car, in his price range, that would be the most fuel-efficient. He scoffed, and said that because America's gas prices were SO CHEAP compared to Scandanavia and Europe, there was absolutely NO INCENTIVE for him to buy a fuel-efficient car in America.

That was the first time I had heard about America's gas prices being "dirt cheap" compared to countries across the Atlantic. 20 years later, it seems this (environmentally unflattering) comparison still stands...

Posted by: Michelle Parker | Sep 21, 2005 1:04:22 PM

As Hurricane Rita zeroes in on Galveston/Houston, we may soon found out if Wayne's question regarding non-linear price elasticity has merit. If some of the refineries in the area are taken out, we may soon see $5 gas with long lines for good measure.

Posted by: Ron | Sep 21, 2005 2:06:09 PM

As Hurricane Rita zeroes in on Galveston/Houston, we may soon found out if Wayne's question regarding non-linear price elasticity has merit. If some of the refineries in the area are taken out, we may soon see $5 gas with long lines for good measure.

Posted by: Ron | Sep 21, 2005 2:20:32 PM

Corner the market on a price inelastic product - the surefire way to exorbitant wealth.

If we had an ethical government which actually worked for the common good, they would institute a cap-and-trade system for personal and commercial transportation consumption of petroleum products. Just cap the national consumption level and divide that by the population 16 and over. Pay a market price for gasoline up to the per capita cap, then pay the trading market price for anything over that cap. It would reward people who consume less than their "fair share" and would also eliminate need for things like CAFE standards etc. Exemptions for the transport industry could be made.

Then, just keep lowering the cap and have an adjustable tax that keeps the market price of gasoline within a predictable band, in case the wholesale cost of gasoline declines as demand falls off.

Could apply the same logic to natural gas - another fuel whose price is out of control.

Posted by: Joseph Willemssen | Sep 22, 2005 12:03:42 AM