Confused Thinking on Keystone XL

The endgame appears nigh for the seemingly endless battle over the Keystone XL pipeline, which would carry about 800,000 barrels of oil sands product from Alberta to the Gulf Coast. I still stand by what I wrote in a 2009 study – the potential climate damages and energy security advantages of oil sands development are both widely overblown – but given the amount of misinformation and confusion circulating, I thought I’d weigh in on one important detail in the debate.

Opponents of the pipeline have been making two big claims: further development of the oil sands would massively increase global greenhouse gas emissions, yet such development would have almost zero moderating effect on oil prices. But here’s the thing: it’s pretty much impossible for both of these claims to be simultaneously true.

To make the case that Keystone XL would greatly increase greenhouse gas emissions, one has to start by arguing that blocking the pipeline would substantially curtail oil sands production. That’s a debateable point, but let’s take it as true. To come to the proposed conclusion, though, one has to also argue that the foregone production won’t be replaced by another source – say, OPEC crude. Otherwise, the emissions impact is largely a wash.

But there’s a catch: the models that project essentially zero impact of oil sands production on global crude oil prices do so because they assume that changes in oil sands production would be largely offset by countervailing changes in production volumes elsewhere. Remove that assumption, and the impact of changes in oil sands production on world crude prices would be much more substantial.

Some will counter that I’m being narrowminded in assuming that lower oil sands production would be substituted for by higher oil production elsewhere. Why couldn’t it be substituted for by renewable fuels or efficiency? Alas, the accounting doesn’t work that way. Let’s say we massively increase the use of alternative fuels and efficiency. We’ll still face the tradeoffs I’ve just outlined: cuts in oil sands production will still either be washed out by production increases elsewhere, leading to minimal climate and price impacts, or they won’t, leading to non-trivial consequences on both fronts.

So analysts have a choice: either they can argue that Keystone XL would have a substantial impact on both greenhouse gas emissions and oil prices, or they can conclude that it will have neither. Both of these are respectable positions. (I lean strongly toward the latter.) What is not right is to mix and match assumptions so that one predicts an outcome that is not consistent with itself.