Has Anybody Figured Out the Rough Price Elasticity of Demand For Gas in the US?

[countjulian]

I'm assuming its fairly inelastic, and probably very low below 1. With all of the price changes that go on and all of the statistics collected by the government, I think it should be rather easy to figure out. Has anybody done the maths?

(just for those who know nothing about economics but have the relevant statistics on hand, the formula is

[% change in quantity demanded]/[% change in price]

In order to get an accurate # though, you would have to average in a number of price shifts, which is why I assume somebody important has done this)

[The Central Scrutinizer]

I don't have the relevant data, but I assume it's approaching 0.

[funinspace]

I doubt that it is a linear construct, so the above equation would not come close to fitting the reality of mass human behavior. I think there are many factors involved, including things like the amount of disposable income available, the perception of the state of the economy, the perception of how long the price will remain at a certain value, et.al.

I did see an article the other day that suggested that the average number of miles Americans drove has actually gone down a little this year. I think it was in a newspaper, so I don't have a link.

[laughing dog]

The price elasticity of demand is measured over a certain time frame: the shorter the time frame, the more inelastic, since consumers have less time to adjust the purchasing patterns.

[AdamWho]

This is an old article but it may be start

http://www.mackinac.org/article.aspx?ID=1247

[countjulian]

The price elasticity of demand is measured over a certain time frame: the shorter the time frame, the more inelastic, since consumers have less time to adjust the purchasing patterns.

It's also different in different price ranges (e.g., from $3-$5 a gallon it is fairly inelastic, whilst once you ratchet the price up to somewhere around $25 a gallon I think you would see a lot more elasticity).


Thanks for the link, Adamwho.


Gasoline, short-run


0.2

Gasoline, long-run


0.7

Long run it is higher than I expected. And his sources also seem rather valid


Source: Economics: Private and Public Choice, James D. Gwartney and Richard L. Stroup, eighth edition 1997, seventh edition 1995; primary sources: Hendrick S. Houthakker and Lester D. Taylor, Consumer Demand in the United States, 1929-1970 (Cambridge: Harvard University Press, 1966,1970); Douglas R. Bohi, Analyzing Demand Behavior (Baltimore: Johns Hopkins University Press, 1981); Hsaing-tai Cheng and Oral Capps, Jr., "Demand for Fish" American Journal of Agricultural Economics, August 1988; and U.S. Department of Agriculture.

[AdamWho]

This Economist Blog is pretty good and talks about these issues often. He is sighted by the major bloggers and is considered strong in his field

http://www.angrybear.blogspot.com/

Specifically, there is a post about this topic
http://angrybear.blogspot.com/2006/04/higher-gasoline-prices-shift-along.html

[drewjmore]

A participant in another thread provided this source of gasoline data:
http://tonto.eia.doe.gov/oog/info/twip/twip_gasoline.html#stocks

Price is at the top of the page, demand is at the bottom.
Excel data is available for download there if you look carefully click "more data" and then "Download Series History".

[drewjmore]

http://i117.photobucket.com/albums/o42/drewjmore/GasolinePriceElasticity.jpg

Since I was interested to see this myself, here you go.

Perhaps I went about the calculations too naively:
The data provided is weekly, so I calculated the weekly % change in both price & product demanded.
Then I divided %change in demand by %change in price, per the OP.
(In weeks when price change was 0%, I set the elasticity to 0, rather than divide by 0.)
Then I fit the data back into a graph I created last week showing the 'relationship' among supply, demand and price, setting 'zero' elasticity to 4000 on the graph, also scaling that value up by 100.

Is this calculation supposed to be illuminating? I don't see any meaningful correlations.
Do I need to average over more time than the one-week period I used?
Do real economists use something more caculus-based to represent rates of change? If so, how do they model real-world data in their functions?

-djm [took micro & macro in the 90's, never saw the practical application of what was learned]

[laughing dog]

Is this calculation supposed to be illuminating? I don't see any meaningful correlations. This would be illuminating if the only influence on the change in QD was the price of that period. Otherwise, the calculated elasticity includes the effects of other influences.


Do real economists use something more caculus-based to represent rates of change? If so, how do they model real-world data in their functions?
Typically, economists estimate a demand function QD = F(P,other variables). Then take the derivative of the function with respect to price, and then plug in the relevant values.