Yesterday's post of gas prices around the world--and John's post today* asking "how expensive does gas have to get, in the US, for there to be a dip in consumption"--got me thinking about possible economic strategies for reducing gas consumption. I'm a firm believer that prices are the only incentive strong enough to motivate people to drive less. Moral suasion, guilt and celebrity pleas will only go so far. But how far do prices have to go before the public takes note?
I'm going to answer the question env-econ style--that is superficially, sloppily and with numerous leaps of faith.
First, I took the 2007 gas price numbers from The Oil Drum graph, and combined them with two other pieces of information. a) per capita gas consumption figures for each country in 2003--because this were the first set of numbers that popped up in Google and b) per capital income figures for each country from 2005. I know all the dates don't match up, but I don't think there's been much relative movement in prices or consumption across countries over the last 5 years--and if I'm wrong I'm sure someone will tell me.
WARNING: Statistical content in the next paragraph. If you're not interested in how I got ht results, skip the next paragraph and just read the results.
Given my new data set, I ran simple linear regressions of the natural logarithm of per capita gas consumption on the natural logarithm of gas prices and the natural logarithm of per capita income. The resulting parameter estimates are quick and dirty estimates of the price and income elasticities of demand. For reasons that will become apparent later, I ran two separate regressions. One for the countries with above median per capita income and one for countries with below median per capita income. Here's what I found.
For the 22 countries with per capita income in the bottom half of the sample, I find that the price elasticity of demand is -.17 and the income elasticity of demand is virtually zero (.0003). This means that a 1% increase in the price of gas will result in a .17% decrease in gas consumption for low-income countries, while a 1% increase in income will have almost no effect on gas consumption. One possible explanation is that individuals in low income countries don't spend a lot of their income on gas, because they can't afford the amenities that need gas. The per capita annual gas consumption in these countries is about 65 gallons per person per year and only four--Saudi Arabia, Slovenia, Greece and New Zealand--average over 100 gallons per person per year. It could easily be argued that at least a couple of these belong in the higher income group.
By contrast, the 22 higher income countries average 165 gallons per person per year with 16 of the 22 averaging over 100 gallons per person per year. But, are higher income countries more or less sensitive to gas price changes than lower income countries. My quick and dirty estimates show that the elasticity of demand for these higher income countries is roughly -.75 which means that every 1% increase in gas prices reduces gas consumption by .75%--almost 4.5 times the change for the lower income countries. An encouraging result in that high income countries tend to at least be somewhat sensitive to price changes and that means prices may work as incentives for reducing gas consumption.
But, if we look at the estimated income elasticity of demand for these countries, a troubling result emerges. Using the thrown together data set, I estimate the income elasticity of demand to be .72 in these higher income countries. Which means for every 1% increase in income, gas consumption increases by .72%--almost the exact same amount as the consumption reduction caused by a 1% increase in gas prices.
What does all of this mean? Hell if I know--just kidding. It means that any gas consumption reduction is going to have to come from high income countries--the top 22 countries consume 4 times the gas of the bottom 22. This sounds reasonable, because these may very well be the countries with the lowest marginal benefit from the last gallon consumed--so they will lose the least by reducing consumption.
But, these are also the countries with the fastest growing incomes. To reduce gas consumption--at least according to this simple analysis--gas prices will have to rise faster in percentage terms than incomes in these countries. For every 1% increase in income , there would have to be a greater than 1% increase in gas prices to reduce gas consumption. In the U.S., real GDP growth has averaged about 3% per year for the last 30 years. To reduce gas consumption, real gas prices would have needed to increase by a factor of 2.4 over that time (1.03^16).
So why has gas consumption risen? Low gas prices combined with rising incomes. Not too profound, but we often forget the income effect when talking about gas prices.
*I finished this post before I saw John's post. Luckily they tie together nicely--I think.
