With gas at $3.50 a gallon in April, the U.S. mainstream media is replete with stories of drivers abandoning SUVs, hopping on mass transit, and otherwise cutting back on gasoline. Yet a year or two ago, when pump prices were approaching and even passing the $3.00 "barrier," the media mantra was that demand for gasoline was so inelastic that high prices were barely making a dent in usage.
Which story is correct? I lean toward the more "elastic" view, and here I'd like to share some of the data that inform my belief.
I've been tracking official monthly data on U.S. gasoline consumption for the past five years and compiling the numbers in this spreadsheet. You'll find that it parses the data in several different ways: year-on-year monthly comparisons (e.g., March 2008 vs. March 2007), three-month moving averages that smooth out most of the random variations in reporting, and full-year comparisons that allow a bird's-eye view.
Here's what I see in the data:
- Gasoline demand is trending downward, though only slightly. In the 49 year-on-year comparisons in the spreadsheet, monthly gasoline use dipped below the year-earlier level only eight times, but these include each of the last five months (see Moving Avgs worksheet tab).
- Gasoline's short-run price-elasticity is rising. After a low of -0.04 in 2004, the short-run price-elasticity increased to -0.08 in 2005, -0.12 in 2006, and -0.16 in 2007. (I assume an "income-elasticity" of two-thirds in calculating price-elasticity; see Full Years worksheet.)
- A big reason that gasoline use kept rising until recently was the growing economy. Demand is heavily affected by economic activity. The minimum year-on-year GDP growth for any month in all four years was plus 1.7 percent (see Moving Avgs worksheet).
- Another reason gasoline demand was slow to drop is that the price signal, while significant, was less than advertised. Adjusted for general inflation, the average 2007 pump price was only 54 percent higher than the 2003 price. Amid all the talk of a doubling or even tripling in gas prices, it's sobering to learn that you have to go all the way back to 1998 to find the last year that the real price was just half the 2007 price.
- The biggest market barrier of all may have been gasoline price volatility. The spreadsheet spans 63 months, allowing 62 month-to-month comparisons. In 29 of these, the price went down (see 1-yr comparison worksheet). That's right: the average gasoline price decreased from the prior month an astounding 47 percent of the time (see graph). Pump prices have been so volatile that consumers didn't know whether the price three months later would be up or down. The result? American families and automakers alike found it hard to justify long-term investments in more-efficient cars. And allied policies like de-subsidizing sprawl didn't get taken seriously.
- Nevertheless, gas prices have now risen five years in a row and are virtually certain this year to chalk up a sixth. There hasn't been a comparable period of sustained increases since the late 1970s.
The big takeaway for carbon taxes is that the short-run price-elasticity of gasoline demand is rising (point No. 2). (The long-run price-elasticity is probably around minus 0.4, as we discuss here.)
While a rising elasticity contradicts the standard economic model in which price-sensitivities don't change much over time, point No. 5 provides a reasonable explanation: Gasoline prices (and energy prices in general) had fluctuated so wildly for decades, and a sense of entitlement to cheap gasoline had become so ingrained in American society, that it took a long time for households and businesses to internalize the rise in pump prices -- to regard it as real.
Perhaps now, however, a line has been crossed. Maybe the trigger was the price of crude breaching $100 a barrel, or the unfolding credit crisis signaling a fundamental change in the U.S. economy. Or it may simply have been the accumulating weight of price increases noted in point No. 6. Whatever the reason(s), Americans finally seem to be getting the message that higher gas prices are here to stay.
That's good news for the climate, national security, and green jobs ... but it's bitter medicine for hard-pressed families as well as business and jobs that aren't oil-intensive but are being pulled under by gasoline-caused belt-tightening.
Now imagine if the price rises had been delivered not by a rapacious market but via socially mandated ramped-up increases in the gasoline tax (as some commentators have proposed since the 1970s, including, with renewed urgency, after 9/11 [PDF]).
Americans would have had time to adapt, along with real choices such as truly fuel-efficient cars and smaller houses in more-compact developments. And the extra revenues from the higher-priced gasoline would have belonged to all of us rather than just the owners of oil reserves. Those revenues could have been returned to households and businesses via tax-shifts or dividends, not skimmed off for private enrichment.
The analogy to a revenue-neutral carbon tax couldn't be more clear.
Top 25 reasons to give a damn about climate change
Oil: enough energy to melt glaciers!
Cash for Clunkers brought us ... more clunkers!
Comments
View as Flat
Colin Wright Posted 3:46 pm
13 May 2008
Permalink
amazingdrx Posted 5:17 pm
13 May 2008
Peak oil? No problem if cars transition to renewable electric plugin hybrid. Oil demand will dip. To meet up with supply.
Who cares about all the complexities of oil markets and oil wars?
Start a subsidy diversion now. 66 cents per electric equivalent to a gallon of gas, sounds like an antidote to gasoline fueled economic disaster.
http://amazngdrx.blogharbor.com/blog
Permalink
Sean Casten Posted 7:12 am
14 May 2008
I agree that volatility confuses the elasticity calculation, but in my experience it behaves contrary to the way you suggest. ACEEE has done some nifty analysis (that I wish I could lay my hands on) showing that volatility is a much better predictor of when people will invest in energy efficiency than absolute price. Which makes intuitive sense, in a frog-in-a-pot way. The frog doesn't move with slow, steady temperature increases, but hops when it swings. Similarly, energy users learn to live with expensive energy, but have - at least in my experience - been much more prone to spend money on efficiency when the price is bouncing. This would seem contrary to your analysis.
Gasoline as a transportation fuel always seems to have an inherent inelasticity simply by virtue of the fact that most of the cost of a car is capital recovery. (Or, to put it another way, a car spends most of it's lifetime parked.) Compare that to a boiler or power plant that runs 24 hours a day, and it becomes rather apparent that while owners of the latter have a very strong incentive to invest in capital that will lower energy costs, the relative incentive for vehicles is much smaller. (Hell, even at $4 gas, an average, 15,000 mile/year driver with a 20 mpg car "only" spends $3000/yr on gasoline. Doubling their fuel economy would therefore save them $1500 a year - which makes it hard for that driver to justify much in the way of capital or lifestyle changes beyond a couple thousand bucks.) This seems to me to explain much of why fuel oil shows so much greater elasticity and - at least until we get to European level fuel prices - will hold gasoline elasticity down.
Finally, a theory that I can't prove. I don't think it is entirely coincidental that the massive increase in commodity volatility (not just oil, but also gas, copper, and corn) has coincided with a general lack of decent investment opporutnities elsewhere. Equity investments stink. Mortgage markets have imploded. Even US treasuries are lousy. But all that money in the financial system needs to flow somewhere - and in a lot of cases, we're seeing commodity pricing that really can't be explained on the basis of historic fundamentals. Natural gas and fuel oil have completely decoupled from one another. Natural gas prices don't make any sense relative to reserve/demand ratios. Copper prices are absurd. This leads me to the nagging feeling that - concerns about peak oil and carbon notwithstanding - the current high commodity prices are a bit of a mirage, likely to come down as other investment opportunities arise. Not because I can prove it, but because when prices are increasing and historic fundamentals no longer make sense, it's usually a pretty good warning sign that a speculate bubble is getting set to pop.
Your thoughts?
Permalink
Charles Komanoff Posted 8:10 am
14 May 2008
I'd love to see that ACEEE analysis too. It would rattle my world if you're right about volatility driving EE investment more than do increasing energy prices per se (I've said it a little differently than you, please note.) Indeed, the standard line about OPEC is that they always open up the pumps to drop the price just as Americans and other oil consumers start seriously investing in alternatives. Anyway, until I see counter-evidence, I'm sticking with the idea that steadily rising prices historically have been the strongest price signals for EE investment.
Your exposition helps explain why even the long-run price-elasticity for gasoline is probably no greater than (negative) 0.4, whereas for fuel oil it's 0.6 or 0.7. FYI, in my analyses I generally crank gasoline costs into a generalized cost-to-travel function that includes other variable but not fixed costs. You're certainly right that few can afford or otherwise prematurely retire a serviceable vehicle solely to save on fuel costs. OTOH, rising fuel costs will prompt less driving, consumer selection of more fuel-efficient cars, and mfg'er design and provision of same. Which is why gasoline's price-elasticity differs from zero.
That's an interesting hypothesis about commodity prices. As it happens, I share your skepticism that prices for metals; construction staples like conduit, pipe and rebar; and machinery will keep rising or even stay at present "high" levels.
Amazingdrx -- When you say We can "price carbon without a tax [by removing] the subsidies from fossil fuel," you're either envisioning a far lower carbon tax than we at the Carbon Tax Center feel is necessary, or neglecting to do some math. Please allow me: In very round numbers, the U.S. economy uses 80 quads a year of fossil fuels and these are fiscally subsidized (e.g., tax dodges) at no more than $30 Billion a year. The subsidy thus equates to around 40 cents per million Btu, then, which, for gasoline, say, is just a nickel a gallon, which in turn corresponds to a carbon charge of just $5 per ton of CO2. Peanuts, no?
Charles
http://www.komanoff.net
Permalink