Affordability of Gasoline

At Commons Blog, guest author Indur Goklany makes a salient point about the affordability of gasoline:

The price of gasoline is on everybody's mind these days, but affordability is just as important. To help put recent gasoline price hikes into perspective, some analysts and media outlets even provide us with information on the real price of gasoline, that is, price adjusted for inflation of the dollar. But even that provides, at best, an incomplete picture. A broader perspective is obtained by looking at trends in price relative to the personal income of the average American, as shown in the following figure.

He then posts a graph showing the history of nominal gas prices (higher than they’ve ever been), inflation-adjusted gas prices (almost as high as they were at their peak in the early '80s), and gas prices as a fraction of income (higher than in the '90s, but still a good bit lower than in the early '80s).

In short, real income has grown faster than the real price of gasoline. Once we take into account the growth of income, gas is substantially more affordable now than during the last big spike in gas prices.

However, Jonathan Adler linked Goklany’s post on the Volokh Conspiracy, and one commenter questioned Goklany’s use of average household income.

One problem with this measure is that the income disparity between rich and poor has increased dramatically since 1978. In my opinion, a more useful comparison would be ratio between price and the income excluding the top quartile.

I’ll go one better, by looking at the price-to-income ratio for each quintile. The graph below shows the fraction of each quintile’s average household income needed to buy 1000 gallons of gas for every year since 1973. (The choice of 1000 gallons was arbitrary, but the same results would hold regardless of how many gallons I chose. [See update below.])


(Sources: U.S. Census Bureau and Energy Information Administration. I only had household income by quintiles through 2004; for 2005 and 2006, I assumed each quintile’s income would grow by the same percentage growth it had averaged over the previous four years. Gas prices for 2006 are based on the first quarter only, and thus don’t include the high prices of the last two months.)

So what can we see? Even looking at the poorest fifth of the population, the fraction of income required to buy gasoline is still lower than it was in the early '80s. Not surprisingly, the fraction has risen a great deal over the last few years, but it still has not surpassed its historical peak. The same holds true for every other income quintile, but the effect is more muted, since higher income means any given price difference will correspond to a smaller fraction of income. (If gas prices stay at their current price of about $2.90/gallon, however, then we could pass that early-80s high-water mark this year.)

How is it possible that the same qualitative pattern holds for both rich and poor, given the rising disparity in income? It’s true that the income gap has increased since the 1970s. But that’s not the same as saying the poor have gotten poorer, because in fact real incomes have risen for every income group over this period. The gap has grown because the rich have gotten a lot richer, while the poor have gotten a little richer.

UPDATE 1: Although I noted that my choice of 1000 gallons per year was arbitrary, I fear that some readers may have gotten the wrong impression. I've noticed some people concluding from the graph that poor people are now spending 22% of their annual income on gasoline! In fact, according to this page, the average American car consumed 550 gallons in 2003. I don't know how that average differs by income class, and obviously some households have multiple cars. My guess is that lower-income households have fewer cars and drive fewer miles on average. But in any case, the point is that 1000 is arbitrary. I could have picked any other number -- say, 100 or 500 -- and the pattern would still have been the same, though the actual numbers on the vertical axis would have differed. (Thanks to reader Patrick Livingood for the link.)

UPDATE 2: And yes, my commenters are correct to note that gas mileage has improved over time, and taking that into account would make gas look more affordable in recent years. On the other hand, people might be commuting longer distances now -- or shorter ones, for all I know. Both mileage and commuting times would be felt as changes in the number of gallons/year, whereas the graph above is based on a constant number of gallons/year.