Peakonomics: kiss your boarding pass goodbye

This article was first published in Unconventional Oil & Gas Monitor, 4 June 2012.

Peak oil used to be a debate in which geologists and economists threw bricks at each other but seldom talked. Like so many cross-disciplinary problems, each side believed only their own calling offered any useful insights. Geologists insisted that oil decline is immutable because the geology is in control; economists countered that a rising oil price changes everything and there could never be a shortage. But now, finally, a team of economists from the International Monetary Fund (IMF) has incorporated peak oil analysis into an econometric model, and the results are not pretty.

For a decade the peakists have had the best of the evidence: global oil production has stagnated in spite of an oil price that marched relentlessly upwards. These trends were especially pronounced during 2005-2008, when production essentially flat-lined while the price spiked to its all-time high of $147 per barrel. It was this that led the IMF group to wonder if there wasn’t something in this peak oil business after all, and their paper draws some conclusions that look as dark as a peak oiler’s worst nightmare.

The IMF Working Paper*, published last month, which the group stresses does not represent official IMF thinking, sets out to test the idea that the ten-year rise in the oil price can be explained by geological constraints. The group found that allowing for oil depletion in their model led to “dramatically improved” forecasts of both production and price. It also suggests real oil prices will almost double in the next decade. “By far the most important reason is that geology doesn’t want to release that much more oil unless you really press the price button”, said Michael Kumhof, one of the authors, in an interview with NewsBase.

The IMF group investigated an approach known as Hubbert linearization – after the father of the peak oil school of thought – further developed more recently by Kenneth Deffeyes, a professor of geology at Princeton. Hubbert linearization works on a deceptively simple equation that codifies the notion that oil becomes harder to produce, the less there remains to be produced. Ingeniously, it also allows you to predict the peak of global oil production without relying on an inevitably uncertain prior estimate of the total quantity of oil that will ever be produced. To find out how, I recommend Deffeyes’ lucid book, Beyond Oil: the view from Hubbert’s peak.

Deffeyes used this method to predict global production would peak in 2005, but the IMF group argues this was premature: the forecast fitted the data well until 2003, but then real-world production started to outstrip the prediction. They go on to suggest this was because oil prices surged to levels far higher than anything seen during the previous 20 years. However, they also note there was a sharp inflection in 2005, since when production has been stuck on a plateau, with output growth far below the historical average of 1.5%-2%. Prices mattered, the authors conclude, but not enough to allow production growth to regain its previous trajectory.

When the IMF team combined Hubbert linearization with a more conventional approach that allows oil production to respond to GDP and the oil price, the results were striking. By back-testing the model against the historical data, they found its oil production forecasts were more accurate than those of both peak oilers, traditionally too pessimistic, and authorities such as the US Energy Information Administration, which have generally been far too optimistic.

Their price forecasts were also far more accurate than traditional economic models that take no account of oil depletion, forecasting a strong upward trend that closely fits what has actually happened since 2003. Temporary shocks to supply, demand and GDP caused major deviations above and below the trend, for example in 2008 and 2009, but the underlying trend quickly reasserted itself. “This confirms that the problem of oil becoming harder and harder to produce in sufficient quantities was an important factor that would have significantly increased oil prices regardless of shocks”, the authors conclude. It’s a point reinforced by Kumhof in interview: “When you look at the oil price [over the past decade], the trend is almost entirely explained by the geological view”.

Although incorporating the oil price was important to improving its accuracy, the model also demonstrates how limited an impact rising oil prices have on oil production. The paper finds that output responds to rising oil prices far more in the short term, when producers can dip into existing spare capacity, and worryingly little over the long term, suggesting oil companies are struggling to find and develop new resources. It is this that drives the model’s forecast of $180 per barrel by 2020. “Our prediction of small further increases in world oil production comes at the expense of a near doubling, permanently, of real oil prices over the coming decade”, write the authors.

Nor is the advent of the so-called shale oil revolution a ‘game-changer’, as its supporters claim. While the paper makes no mention of specific technologies, the model does capture the impact of shale oil, says Kumhof, and the message is far from bullish. Shale is a sign of the desperate measures required to maintain any kind of supply growth, because conventional oil is declining so fast. “We have to do these really expensive and really environmentally messy things just in order to stand still or grow a little bit”, he says, “it doesn’t mean the picture is all rosy”.

There’s perhaps not much to surprise a peak oiler here, nor perhaps many in the oil industry, or some in the City; Barclays Capital has forecast $185 per barrel in 2020 for some time now. But the paper is significant in that it is the first mainstream economics model to demonstrate the overwhelming importance of oil depletion in driving oil prices, and effectively to debunk the shibboleth that rising prices will solve peak oil, as many economists have long insisted.

My main reservation is that the model manages to combine a near doubling of the oil price in the next decade with a world economy that continues to grow at 4% per year. How can that be right, when oil price spikes have preceded almost every recession since WWII, and when the spike to $147 was so clearly implicated in the Great Recession of 2008-09? The idea that a doubling of oil prices could be ‘permanent’ also flies in the face of recent history; the Great Recession took oil from $147 to $35 in less than six months as demand slumped. A scenario of repeated oil price spikes alternating with deep recessions seems more likely than one that combines permanently high oil prices and sustained economic growth.

To be fair, the IMF team acknowledges the uncertainties here. If the impact of rising oil prices on oil production is minimal, the impact on global demand could be far greater, and, they suggest darkly, may be ‘non-linear’: a 10% rise starting at $200 could be far worse for the economy than a 10% rise at $100. “There is likely to be a critical range of oil prices where the GDP effects of any further increases become much larger than at lower levels, if only because they start to threaten the viability of entire industries such as airlines and long-distance tourism”, says the paper. Kumhof stresses there is not yet enough data to prove the point – because we have never experienced oil prices as high as those predicted – but fears it is correct. “When oil prices become too high”, he says, “something has got to give”. This will be the focus of the IMF team’s future research.

If their hunch is right, at worst it could mean we soon reach what you might call ‘economic peak oil’, when oil prices repeatedly spike so high they kill all prospect of sustained economic growth. At best we could soon be kissing our boarding pass goodbye. Welcome to the last oil shock.

*IMF Working Paper WP/12/109 The Future of Oil: Geology versus Technology, Jaromir Benes et al, May 2012