Part I
“Pam and Matt Keith spent Memorial Day weekend on a houseboat on Lake Oroville in Northern California. But because of high gasoline prices, the Keiths never even untied the boat from its mooring slightly offshore. When they ventured away from the shore, they supplied their own power — in kayaks.”
So began The New York Times take on the start of the summer driving season in an age of $3 gas: “Holiday Travelers Hit the Road, but Scrimped a Bit.”
The Times’ page-one piece was guaranteed to bring smiles to both economists and despisers of motorized recreation. As a member of both camps, I ate it up. I loved that the Keiths were kayaking instead of houseboating around the lake, and that another California couple, Celia and Michael Shane, had shelved their annual jet-skiing trip in Lake Mead National Recreation Area in Nevada. “To save the $70 per tank it now costs to fill up their minivan,” the Times reported, “the Shanes were barbecuing instead.” For guys like me who can let a single roaring jet ski ruin an entire beach day, fewer decibels mean more happiness. And after a year’s drumbeat of articles insisting that higher gas prices hadn’t dented Americans’ “love affair with their cars,” it was heartening to see the paper of record start acknowledging the No. 1 tenet of economics — higher prices mean lower demand.
The world’s thirst for petroleum breeds war, props up dictators, and imperils the climate. Known oil deposits are shrinking by the day. So no question in economics is more pressing than whether, and by how much, changes in the price of gas reduce the demand for it. I’ve been examining this question since May 2004, when the price first edged past two bucks. Every month I faithfully enter the latest price and consumption data into a spreadsheet.
This has to be done just right. For one thing, because gas use follows seasonal patterns, monthly data must be compared over intervals of 12 months (or 24, etc.). For another, changes in price must be adjusted for general inflation. Most important is netting out the upswing in gasoline use that ordinarily accompanies expanding economic activity when the price of gas is stable. Only after taking these steps can one isolate the effect of higher pump prices on gasoline demand.
What my spreadsheet shows is that higher gas prices are dampening gas use. Not hugely, but enough to command attention. True, demand isn’t down in absolute terms, but it’s practically flat, which means it’s lagging the higher levels it would otherwise be reaching due to economic growth. For all of 2005, with an economy 3.5% bigger than the previous year’s, gasoline usage was up only 0.2%. Similarly, purchases of gasoline in April 2006 (the most recent month available) were just 0.1% greater than in April 2005, even though the economy was almost 4% larger.
What emerges from these numbers is a “price elasticity” of around 20% (economists write this as 0.2). This means that declines in gasoline usage due to rises in price are averaging 20% as much as those same percentage changes in price. For example, a 10% rise in the price of gas engenders a 2% drop in demand. This 20% ratio has held remarkably constant over the past 24 months, never straying beyond a range of 11%-26% (provided I use three-month “moving averages” to smooth out monthly variations in inventories).
It turns out, then, that Americans have been responding to the rising price of gas by using less. As a smattering of articles have reported, many of us are cutting out a few car trips here and there, shortening a few others, occasionally carpooling, now and then leaving the SUV in the driveway and taking the sedan, perhaps trying a lighter touch on the gas pedal, even walking or biking occasionally. The overall impact, while modest, is visible in my spreadsheet nonetheless. And it’s been proportionate: small price rises evoke small changes, big hikes bring bigger ones.
But, you ask, if the price-elasticity of gasoline is 20%, how come the recent doubling of gas prices — in percentage terms, a 100% price jump — hasn’t dropped demand by 20%?
There are two reasons: first, gas prices haven’t doubled. In fact, they’re not even up by half. Notwithstanding the mass near-hysteria over prices at the pump, the average 2005 price, $2.34, was just 43% higher than the average 2003 price, $1.64. Adjusted for inflation, the two-year real increase was 34%, or roughly one-third — a long way from doubling. Even $2.80 gas this April, the highest monthly price ever save for the post-Katrina spike last fall, was only 16% higher in real terms than the April 2005 price and just 39% higher than April 2004. By these terms, compared to two years ago demand should be down just 7-8% (I got that by multiplying the 20% elasticity by the roughly 35-40% real price increases).
The second point is that the basically flat demand compared to two years ago indeed constitutes a drop of 7-8%, relative to where demand would have been due to the economy’s being 7-8% bigger.
Part II
Now another question arises: so what? At the end of the day, what’s the practical difference between the actual price-elasticity of 20% and the popular conception of zero elasticity if the effect of higher gas prices is only going to be offset by economic growth?
I offer two answers. Combined, they just might hold a solution to our era’s twin overriding crises: the oil-dependence crisis and the climate crisis.
The first answer is that as we extend our time horizon, gasoline’s price-elasticity, or price sensitivity to break free of the jargon, gets larger — a lot larger. Going out several years or more, individuals have greater scope to take actions that economize on gasoline. They can junk the gas-guzzler, or at least not replace it with another one when the old one gives out. They might calculate the dollar tradeoffs between density (high rents but less need to drive) and sprawl (the reverse) and pick up stakes for a less car-dependent area. They may gravitate toward job opportunities closer to home. And they can make more durable commitments to behavioral changes that reduce the need to drive, like forming a carpool or buying a roadworthy bicycle or selling the far-away vacation home.
The consensus of economists who have studied gasoline use is that the “long-term” price elasticity — the effect on demand eight or ten years hence — is between 50% and 70%, or roughly triple the 20% “short-term” elasticity I’m seeing in my spreadsheet. That is, over the long haul, rises in the price of gas are likely to dampen demand several times as much as the modest changes we’ve seen in the past year or two.
The second reason price-elasticity matters is that prices of gasoline and other fuels will probably climb in the future. Or, to be candid, fuel prices need to climb far and fast if America is to ever get off the oil spike and the world as a whole is to avoid disastrous climate change.
For all the promising antidotes to oil dependence, from ethanol and hybrid cars to rearranging living patterns so people and goods don’t have to move as much, there’s a growing awareness that the only surefire way to advance on all fronts is to create an irresistible and universal market pull by pricing gasoline at a very high level — perhaps in the $10 a gallon range. And now that the climate crisis is overtaking oil dependence as the ultimate energy nightmare, people are starting to face the fact that only vastly higher prices for all fossil fuels can reduce CO2 emissions across the board, through conservation, not just of gasoline but of all petroleum products as well as natural gas and coal.
Yes, I’m talking about a carbon tax — the only mechanism powerful and direct enough for the daunting task of phasing out fossil fuels. Conventional “market forces” would be too volatile in the short term and too weak in the long term to provide the needed incentives, and would ravage the poor and middle class besides. And sole reliance on market forces — say, rising crude oil prices due to shrinking supplies — would simply open the door to massive development of synthetic fuels such as tar sands, liquefied coal and oil shale, bringing environmental and climate consequences many times worse than the oil they would replace.
In contrast, carbon taxes could be predictable, sending unmistakable price signals to individuals, businesses, and government policymakers. Carbon taxes need not fall disproportionately on the poor — quite the reverse, if the tax revenues are redistributed in equal per-capita rebates or are dedicated to tax-shifting away from regressive taxes (social security checkoffs, for example). And if set at levels that truly “internalized” the damage costs of using fossil fuels, carbon taxes could create incentive enough to usher in full-scale deployment of every alternative, from wind farms and photovoltaics to non-sprawl cities and less driving and flying.
How high must we tax fossil fuels to accomplish, say, a halving of their use? As a thought-experiment, let’s extrapolate from gasoline and postulate that the long-term price elasticity of all U.S. fossil fuel use is 60%, the middle of the consensus range for gasoline over the long haul. It’s tempting to say that we only need an 80-85% increase in price (since an 83% price rise multiplied by 60% elasticity should give us the desired 50% drop in usage). But that would ignore the Law of Diminishing Returns, by which increasing increments to the price are required to keep sending usage lower, once we’ve picked the low-hanging fruit. And there’s also the need to offset the “natural” increases in fuel use due to future growth in the economy.
Reflecting these factors requires a bit of higher algebra — exponents and logarithms — so you’ll have to trust me on the numbers. Twenty years out, if we want to have cut fossil fuel use by half while our economy has kept growing at 3% a year, fuel prices will need to have increased — get ready — eight- or nine-fold from last year’s levels. Gasoline will sell in 2026 for $16 or $18 a gallon, home heating oil for close to that, and a kilowatt-hour of coal-fired electricity for 20 cents. You’ll probably pay a quarter for a plastic bag at the checkout line. And air travelers will be charged by the pound.
Sound awful? Only if we obsess over the empty part of the glass. Don’t forget, a heavy carbon tax means enormous revenues, enough to eliminate not just workers’ social security payments but, most likely, all federal income taxes on everyone’s first $100,000 of income, and state sales taxes to boot. Recycling the tax windfall through rebates or tax shifts will ensure that in the aggregate the nation’s hundred million households have as much money as now. Although energy will cost a lot more, families that now use less energy than average (that’s almost all poor families and many middle-class ones) will end up with more spending money than now.
Moreover, while carbon-taxing fuels to raise their prices nearly ten-fold in twenty years may sound outlandish, ramping up the tax in equal percentage installments could keep the increases manageable. Through the magic of compound growth, the necessary annual price increases would be only 10-12%. That’s a smaller real increase than has already occurred in each of the past three years — and not a penny of those increases, remember, has been rebated to consumers.
Most important, the carbon tax could revitalize our society. Picture the changes ten years out, just halfway to our 2026 target year: Wind farms have replaced coal-fired electricity as the mainstay of U.S. electricity supply. Road-hogging SUVs are out, trim sedans are back in. Energy-thrifty compact fluorescent lamps have made power-hungry incandescent and halogen bulbs collectors’ items. Costlier air and road travel has created the “market pull” for a 300-mph intercity rail network linking major American cities. Urban trips by bicycle are up 10-fold, to 10% of travel, making cycling safe and stress-free (through the “safety-in-numbers” phenomenon). Suburbs stop sprawling. Cities thrive. And, of course, oil dependence, like crack cocaine, is yesteryear’s problem, while the climate crisis is halfway toward solution.
Carbon taxes, carefully constructed, can do all this by providing the motivation to turn literally billions of individual and social decisions toward lower energy demand and carbon-free supply. That’s why it’s important for the media to get the story right on drivers’ adaptations so far to higher gasoline prices — the fate of our world may literally depend on the human propensity to rein in consumption when the price goes up.
What about those two California couples whose curtailed holidays were reported in the Times? We can imagine that in 2026 Pam and Matt Keith’s houseboat lives on in the family scrapbook. The Keiths now take the train and ride their bikes to their favorite kayaking spots (rentals are always available). And Celia and Michael Shane have turned their jet ski into a backyard fountain. The neighborhood kids frolic in it at the Shanes’ annual Memorial Day barbecue.
