Projecting future oil demand

Joyce Dargay and Dermot Gately have a new paper out that projects 134 mb/d global oil demand in 2030, which is far higher than the standard projections (IEA, EIA, OPEC), all of which are in the neighborhood of 105 mb/d. It got a good bit of attention about a month ago (see here, here, and here, among others). I have a few questions about its conclusions.

Here’s the main thrust of the paper’s argument:

Most of the demand reductions since 1973-74 were due to fuel switching away from fuel oil, especially in the OECD; in addition, the collapse of the Former Soviet Union (FSU) reduced their oil consumption substantially. Demand for transport and other oil was much less price-responsive, and has grown almost as rapidly as income, especially outside the OECD and FSU. World oil demand has shifted toward products and regions that are faster growing and less price-responsive…. we project rest-of-world growth that is consistent with historical patterns, in contrast to the dramatic slowdowns which [others] project.

It’s an interesting paper that’s well worth reading. I have a few questions, though:

1. The demand and price trajectories don’t seem mutually consistent. The paper assumes the future price trajectory given in the 2009 EIA International Energy Outlook (IEO). That trajectory, of course, is consistent with demand meeting supply at the level the IEO projects – about 107 mb/d in 2030, not 134 mb/d. If demand climbs much faster than in the 2009 IEO (because of the factors that Dargay and Gately cite), prices will presumably rise; that, in turn, should curb demand growth. Of course, that wouldn’t fully offset the increased demand. But it should result in a number somewhere between the IEO and this paper.

2. The paper seems to largely ignore changes in subsidy policies. If demand grows the way that Dargay and Gately project, prices will go way up, making existing subsidies extremely difficult to sustain, particularly in non-exporting countries. That would make demand much more price responsive than in the model that the paper presents. There’s already one example of this: the paper finds zero price elasticity for China, which reflects historical government price-setting, and projects that into the indefinite future, even though the Chinese government has moved in the last couple years to largely bring product prices into line with world markets.

3. I’d like to see the projections for fuel oil separated out, particularly for non-OECD countries. As Dargay and Gately note, the OECD is mostly out of using oil for these purposes. But non-OECD oil consumption is currently dominated by this segment – and it is far from obvious to me that historical patterns will continue to dominate there (as I infer is the case in the numbers underlying the Dargay and Gately paper), rather than, say, non-OECD countries following the OECD path and switching to other fuels. This is a case where some more fine-grained analysis seems necessary before past performance can be translated into future projections.