The wrong symmetry

Morning Edition had an interview on Friday with Marcia McNutt, described as "one of the country's most influential scientists. A former opponent of the Keystone XL pipeline, she's now come around to support it. Her arguments were interesting and had a certain amount of sense in them, but she wrapped up with a couple of statements that strained reason, and one of them would get an "F" on a logic exam.

The argument against the pipeline is that it's being built to bring the Athabascan tar sands to market, and that we need those tar sands to stay in the ground if we're going to have any chance at all of preventing run-away climate change. McNutt observed that the tar sands are moving to market anyway, just by truck and by rail, and those means of transportation use a lot more energy than would a pipeline.

People have also raised concerns about the safety of the pipeline, and several recent spills and/or explosions have increased those fears. But truck and rail transport are also prone to accidents, as has also been recently demonstrated. McNutt argued that environmentalists would do better to cooperate, relinquishing opposition to the pipeline in return for (among other things) advanced safety features to make the pipeline really safe.

Those seem to me like serious arguments that are worth weighing in a discussion of Keystone, but other parts of her argument undercut her credibility. She observed that transport by pipeline is much cheaper than alternatives, and suggested that in this "austere" time there might be a way to use those savings as a source of funding for alternative energy sources.

Well, sure, in principle. From an aggregate perspective if the economy is saving money through cheaper transport of gas, it should have more means available for other things, like investments in green energy. But the oil companies were planning to pocket those savings, or maybe split them with consumers, so—again from an aggregate perspective—redirecting some of those savings toward green energy would be tantamount to a tax on tar sands extraction.

Look at the political situation today.
The fossil-fuel industry has successfully fought off any serious action on climate since the Clinton administration. Among the things that they've resisted have been carbon taxes or cap-and-trade systems, the two least "interventionist" ways of doing even a little bit about climate-changing emissions. That history doesn't give much confidence that the deal McNutt proposes would actually fly.

Besides, the framing is misguided. By "austere" she presumably means there isn't much money available for worthy projects like green energy, but that's a political choice, not an act of the gods. If you look at the underutilization of labor and of manufacturing capacity, we're physically capable of doing lots more than we're doing. And the government can borrow long-term at rates that are near historic lows (or it could raise taxes to the "disastrous" levels of the 1990s), so the lack of money is a choice, not an unalterable constraint.

And from a physical perspective, the framing is flat-out backwards. McNutt wants to use tar-sands extraction as a way of paying for green energy. But look at the physical situation.

What does it take to build pipelines and infrastructure for the extraction and processing of tar sands? It takes metal, energy, and labor. What does it take to build green-energy infrastructure? It takes metal, energy, and labor. If we're worried about scarcity, and we recognize that the tar sands are a temporary solution with the long-term answer lying in green energy, what sense does it make to devote metal, energy, and labor to a dead-end in order to have the means to apply those inputs to the long-run solution?

She's essentially saying we have to waste a lot of useful inputs in order to be able to put some other useful inputs to an actually desirable end. Perhaps she's right, but if so, our whole political-economic system has become nothing but a waste engine.

The coup de grace, however, was a little earlier in the interview. The interviewer cited the argument of pipeline opponents that if we build the pipeline, then we're guaranteed to extract everything that's feasible to extract from the tar sands. McNutt's reply was that she hadn't seen any evidence that not building the pipeline would guarantee that the stuff wouldn't be extracted. And that's where the big logic fail comes in.

Opponents are saying that A guarantees B. If A, then definitely B. They're arguing that the pipeline is a sufficient condition for complete extraction. The alternative would be the claim that the pipeline is a necessary condition for complete extraction: B happens only if A.

If you negate the sufficient condition (by not building the pipeline), you get, If not A then maybe not B. If we stop the pipeline, then maybe we won't extract all the tar sands.

If you negate the necessary condition, you get, If not A, then definitely not B. If there's no pipeline, then we're guaranteed not to extract all the tar sands.

The interviewer asks about what is in essence an argument about a sufficient condition, and McNutt's rebuttal is that the negation of the necessary condition isn't true. It sounds clever, but it's actually a pretty slippery debating tactic, and it suggests she doesn't have a strong case.

The problem for her argument is that the opponents' point is pretty hard to refute. Yes, we are currently hauling the tar sands out by truck and by train, more than was predicted. But that is an expensive operation, and very little infrastructure has been built just for the purpose of carrying it out. Reining in tar-sands extraction is still hard, but potentially doable. Once we build a big, expensive piece of infrastructure, the only purpose of which is to move the tar sands to market, the political-economic balance shifts pretty decisively against slowing down.

Stopping the pipeline doesn't guarantee that any of the tar sands will stay in the ground. It merely increases the probability of that outcome. So it's pretty important, and McNutt somehow misses that.