Reasons to be cheerful (Part 1): Peak gasoline
There are plenty of reasons to be gloomy about the prospects of stabilising the global climate. The failure at Copenhagen (partly, but far from wholly, redressed in the subsequent meeting at Cancun) means that a binding international agreement, let alone an effective international trading scheme, is a long way off. The political right, at least in English-speaking countries, has deepened its commitment to anti-science delusionism. And (regardless of views on its merits) the prospect of a significant contribution from nuclear power has pretty much disappeared, at least for the next decade or so, following Fukushima and the failure of the US ‘nuclear renaissance’.
But there’s also some striking good news. Most important is the arrival of ‘peak gasoline’ in the US. US gasoline consumption peaked in 2006 and was about 8 per cent below the peak in 2010. Consumption per person has fallen more than 10 per cent.
There are a couple of ways to look at this. One is in the standard economics terms of supply and demand. Given that oil production reached a plateau some time ago, and that demand from China and other developing countries is growing fast, equilibrium can only be reached if prices rise enough to limit the growth in Chinese consumption and generating an offsetting reduction in consumption elsewhere (I’m assuming little or no supply response, which seems consistent with the evidence).
We have of course seen oil prices rise substantially. The effect on demand depends on the percentage change in fuel prices and on the elasticity (a measure of responsiveness) of demand. Because the US has very low taxes on gasoline and other fuels a given change in oil prices produces a much larger percentage change in fuel price than in other developed countries. The US gasoline price rose, in real terms, by around 40 per cent between 2000 and 2010, and have risen by another 30 per cent or so since then, for a total increase of about 70 per cent. So, the US is the place to look for big price effects.
The big question is the elasticity of demand, that is the percentage change in demand arising from a 1 per cent change in price. In the short run, this elasticity is quite low, reflecting the fact that fuel is a small part of the running costs of a car. The short run elasticity (measured over periods less than a year) is relatively easy to estimate and is about -0.25, that is, a 1 per cent price increase will reduce demand by 0.25 per cent, and a 40 per cent increase will reduce demand by 10 per cent. That’s roughly in line with the observed outcome. However, given that factors such as income growth tend to raise demand, the observed reduction is a bit more than would have been expected with constant prices.
The long run elasticity is much higher, since in the long run people can change their driving habits, reduce their stock of cars, and choose more fuel-efficient cars The meta analysis cited here suggests values around -0.6, suggesting that the price increases we’ve already seen should reduce demand in the long term by around 40 per cent, relative to the trend with constant real prices.
In my view, even the long-run estimates are too low. A sustained upward trend in prices will induce the development of energy-saving innovations (the reverse is true – when energy is cheap and getting cheaper, people invent new ways to use more of it). I suspect that the full long-run elasticity, including induced innovation, is near 1, meaning that if current real prices are sustained, consumption could fall as much as 70 per cent below the level that would be expected if prices had remained at the 2000 level.
The alternative is the ‘bottom-up’ approach of looking at changes in driving habits, the car fleet and so on, then working back to total demand.
* The number of vehicle miles driven has peaked. This is partly a response to higher prices, but I suspect there may also be an element of saturation (Americans already spent far more time behind the wheel than people in any other country) and the emergence of substitution opportunities through IT and telecommunications, such as Internet shopping replacing trips to the mall (the construction of new malls has just about ceased).
* Cars are becoming more fuel efficient. That’s partly a market response, reflected in the demise of the Hummer, and partly the result of regulation. The US is tightening its fuel economy standards, and has finally blocked the loophole under which SUVs like the Hummer were treated as “light trucks” and counted separately from cars. The 2009 regulations require a 40 per cent improvement in the average efficiency of new cars, relative to the existing fleet, by 2016. That will take a decade or so to feed through (but efficiency standards for post-2016 cars can be increased further in that time).
* Car sales have been below previous peaks for some years. That’s partly a response to the Global Financial Crisis and the subsequent recession, but there’s evidence that the US car fleet is past its peak size.
In all of this, the GFC has had an effect, which is mostly temporary. So, we should expect some recovery in demand as the general economy recovers, but the peak gasoline phenomenon is real.
Finally, what does peak US gasoline imply about Peak Oil, which I’ll interpret as the point at which the current plateau in oil production turns into a clear, though gradual decline?
* First, we won’t really notice it happening (except as it’s manifested as a further increase in oil prices). Rather, we’ll have to look back at the stats to identify when the decline began
* Second, the adjustment will be a combination of many different processes (less travel altogether, less of that by car, more fuel-efficient cars) rather than one big shift
* Third, given that oil accounts for something like a third of all CO2 emissions, the sooner Peak Oil arrives, the better