Earlier this week, after the National Review endorsed military action in Syria, the editors let Jim Manzi respond and explain why an intervention would be misguided. One of his reasons is that “the risks of open-ended entanglement are severe”:
How many people would we kill, and how much public money would we spend? Why do we believe that the rebels would form a government that would not be worse for us? How would Iran attempt to counter such an intervention, since they have an extremely strong interest in the outcome? And so on. The litany of costs and dangers ought to be familiar to anybody after Iraq and Afghanistan. Would you voluntarily take on one-tenth the cost in deaths and money of either of those wars to replace Assad with whatever is likely to follow him? Wandering into that kind of a commitment based on what has been presented to the American people so far would be extremely rash.
This intuitive argument has the form, “The uncertainty involved in doing X is so great that we should avoid doing X.” I agree, but (in the interests of writing about Syria as little as possible) I’m struck by the fact that the same logic would refute Manzi’s arguments against taking action to prevent global warming. Manzi is probably best known for his writing on climate change, and he is one of the few people (maybe the only person) with a conservative position on the issue who is worth reading.
His thesis on global warming is a rejection of a similar argument to the one he advocates here. Policymakers can’t just examine X, he argues. They must also consider the ramifications of Y. On addressing climate change, X is inaction, while Y would be a national or global scheme for reducing emissions. Both have adverse consequences, Manzi argues, and we shouldn’t necessarily choose the one that involves less uncertainty. Instead, we should do our best to estimate and, if possible, quantify the risks involved in both actions and choose which action we can reasonably expect to have a better result. He believes the costs we can expect to incur from reducing emissions would be greater than the benefits we can expect from avoiding climate change. “We must weigh risk against risk,” he writes.
Nick Kristof is asking the same question in the context of the Syrian crisis, making the case for an intervention:
To me, the central question isn’t, “What are the risks of cruise missile strikes on Syria?” I grant that those risks are considerable, from errant missiles to Hezbollah retaliation. It’s this: “Are the risks greater if we launch missiles, or if we continue to sit on our hands?”
Kristof argues that we shouldn’t let the uncertainty involved in an attack on Syria prevent us from considering all of our options and the risks involved in each. How did he and Manzi end up on the opposite sides of this question? Kristof approaches the problem with exactly what Manzi has tried to portray as intellectual discipline in relation to global warming.
On the question of Syria, Manzi is clearly correct to trust his instincts, and Kristof has simply thought himself into a corner. In fact, the crisis in Syria neatly shows what is wrong with Manzi’s argument on global warming.
Manzi argues that in making decisions about global warming, we should rely on the best probability distributions available, which we have to trust capture everything the scientific community knows about the problem, and that we have to ignore risks that are “inherently unquantifiable.” Yet it would be obviously impossible to apply a probability distribution to a military intervention in Syria. To begin with, a distribution has only one dimension, and it cannot provide useful information about how to compare outcomes for different groups of people. Should we be more concerned about U.S. sailors than say, Lebanese and Israeli civilians? Using probability distributions to talk about climate change, as Manzi does, is similarly reductive. He compares various policies along the axis of gross domestic product, but the economic burdens of climate change will not be shared equally from north to south, Even if preventing global warming would be costly in the aggregate compared to allowing it (and I don’t concede that point), then one could still argue that people near the equator should not be forced to bear the costs of global warming while people elsewhere reap whatever the benefits may be.
More important, there are limits to what can be quantified in questions of policy. In the case of Syria, these limitations are obvious. No one can attach a meaningful probability to the risk of an Iranian counterattack in the event that the United States launches missiles at Syrian President Bashar al-Assad’s troops. These limitations might be less obvious when it comes to a changing climate, but they are no less real. Environmental economists can try to model the effects of global warming, but those models necessarily involve any number of assumptions about, for example, future growth, technological change, or demographic trends. The models can’t account for the probability that their assumptions are flawed. They also can’t reflect mechanisms by which a changing climate could affect the economy that haven’t yet been identified. Who predicted, for example, that low water levels on the Mississippi would affect shipping?
It isn’t that the economic models are poorly made. It’s just that they’re essentially an academic exercise and relying on them to make policy would be foolish. When the future is inherently unquantifiable, national leaders have to make decisions about which unquantifiable risks are acceptable and which are not. On both climate change and on the Syrian crisis, policymakers are burdening all of us with unacceptable risks.
