Fernando supports the US attack on Syria, if tentatively. I’m (even) less sure. Like Fernando, I think that the moral questions involved are complex. Unfortunately, many commentators fall back on slogans.

I agree, of course, that Assad is an awful tyrant. The world would be better off without him. No quarrel there. But that’s not enough to justify military strikes. What’s needed, in addition, is for the harm that such a strike imposes to be justifiable in their own right.

In our forthcoming book Debating Humanitarian Intervention, I distinguish between two kinds of justifications. First, harms may be justified because people are liable to them – most notably, because they’ve forfeited their rights against those harms. Second, there are what we might call lesser-evil justifications. These involve harms to people who are not liable to them, but whose rights can be justifiably overridden in order to achieve something morally more important.

When we are dealing with this second kind of justification, I argue, attacks must meet a so-called success condition. This condition, roughly, requires that attacks have a good enough chance of bringing about a just goal (just in light of the expected harms, that is). Only then can we justify imposing harms on people who have rights against those harms being imposed.

Reports about the actual harms imposed by these strikes vary wildly. Fernando linked to a report suggesting no collateral damage. The L.A. Times reports that Syrian officials claim 15 people were killed, nine of whom lived in surrounding villages. It is hard to know which is closer to the truth at this point. Even if we assume – which may not be true – that the airport personnel was liable to being attacked, should these reports of civilian casualties be true, it is hard to see the harms as proportional, and thus justifiable. I think Fernando would agree.

Here’s something about which Fernando and I disagree. In my view, insofar as the morality of the strikes is concerned, the most important thing isn’t how they actually worked out. More precisely, if we want to know whether the attack was morally justified, whether they were the right thing to do, we should not look at its actual results. Instead, we should look at its ex ante prospects – what the chances of success or failure were at the time the attack was undertaken.

This, too, is difficult to ascertain. Each case must be judged on its own merits, and the people actually making this decision (fortunately) have more and better information at their disposal than we do. However, generally speaking, the prospects of interventions being successful – and thus satisfying the success condition – are very bad.

The empirical literature on intervention is not particularly uplifting. Interventions fail much more often than they succeed, and they frequently end up making things worse. In what’s the best empirical study of military intervention I’ve seen, political scientist Patrick Regan studied 175 military interventions that occurred in the period 1944-1994. Regan finds that interventions in general succeed to reduce violence and loss of life in only about 30 percent of the cases.

Of course, not all interventions are alike. Under the most favorable circumstances, interventions manage to reduce violence and loss of life in about 50 percent of the cases. These cases are interventions that are undertaken unilaterally by major powers in intense conflicts on behalf of the existing government. While the Syrian conflict is certainly intense, and the strike was taken unilaterally, it did oppose the existing government. Such interventions typically succeed much less frequently in reducing violence. Indeed, they often lengthen and worsen a conflict. (Especially when the intervention attracts a counter-intervention on behalf of the other side.)

There are good reasons why we should expect results like these. Foreign military and political leaders – despite knowing more than you and me – typically know very little about the situation on the ground. Interventions must often cope with internally conflicting goals. And there are strong political pressures on the leaders to select strategies that are not well suited to the conflict.

Let me focus on this last point here. In Debating Humanitarian Intervention, I write:

Since democratic leaders are accountable to their voters, the political system in which they operate is designed to skew their decisions towards the interest of their people. Generally, of course, this is a good idea, as protects people against the use of political power for the private interests of the leaders. But the system does not just steer decisions away from the private interests of politicians, it steers them away from the interests of anyone to whom the decision-makers are not accountable. This includes, of course, the citizens of countries that are the target of military interventions.

Unfortunately, the interests of the people (and interest-groups) of intervening countries are generally not well aligned with the interests of the people whom intervention is supposed to help. The voters in intervening countries have a strong interest in limiting potential casualties among their own armed forces, in limiting the financial cost of intervention, and in avoiding a lengthy military presence in the target country. Each of these moves interventions away from what’s required for a good chance of success.

This, of course, explains why airstrikes are the preferred method of intervention these days. We call them “surgical”, but they are mostly clean in the sense of making it easier for us to ignore how dirty our hands might be. One more quote:

In politics, the lives of foreigners have almost no value. And so the same dynamics plague political decisions about how to intervene. Interveners choose their strategies primarily with an eye on maintaining domestic support, not on optimizing the chances of success. Since the main goal is to minimize casualties on one’s own side, interventions often depend heavily on air power. But while this kind of engagement minimizes domestic political costs, it is rarely strategically optimal. Committing ground troops is often a better strategy, yet too tough to sell to voters back home.

If it’s true, of course, as Fernando said, that “this means commanders almost never are permitted to act.” He seems not willing to live with that conclusion. I think it’s the only morally acceptable position to take, given the dangers we impose on innocent people whenever we do intervene.

One last thing. It is controversial to say, as I have, that the most important thing, in terms of justification, are the ex ante prospects of the attack, not the actual consequences. Many of you will want to resist this. Many of you will say that the ex ante prospects determine whether Trump was culpable for attacking, but the actual outcomes determine whether the attack was the right thing or the wrong thing to do. I think this is an important mistake, for reasons to which I will return in a later post.