International Law: Self-Defense

More on Obama v. Bush on Preemption: Response to Lederman

By Jack Goldsmith
Tuesday, April 12, 2016, 11:48 AM

Marty Lederman says in response to my posts that the big difference between the Bush and Obama preemption doctrines was that the Bush Administration “argued that international law permits the United States to engage in a ‘first use’ strike, in a nonconsenting state, against a state or nonstate actor that has not already engaged in an armed attack against the United States, before any threat of attack is ‘fully formed’ — indeed, even where the probability of any such future attack is ‘relatively low.’” 

In fleshing out this claim, Marty goes very deep in the weeds, presents the Obama position and practices in their most favorable possible light, adopts an uncharitable view of the preemption principle the Bush Team publicly defended, links together positions that the Bush administration never linked, and relies extensively on a statement that was not in the public record until 2009.  The need for this sort of analysis is a pretty good initial indication that there is not much difference between Bush and Obama on preemption.  But I must acknowledge that we can never know for sure precisely how close the two preemption principles actually are since each administration’s principle is supported by many different and not always consistent or complete statements, and each administration contemplated the principle for different contexts.  With that important caveat, I continue to believe that the Obama administration has adopted the same preemption principle as the Bush administration, and applied it more extensively than did Bush. 

Before offering my arguments, I should be clear why I think the Obama continuity on preemption is important.  It is not a “gotcha” point.  Nor is it a charge of hypocrisy.  I don’t believe that the Obama Team has professed to hold beliefs that it refused to adhere to in practice.  The point, rather, is that the vociferous resistance to the Bush preemption principle was always informed less by its content than by the larger context (Iraq) and the manner (unilateral and chest-thumping) in which some members of the Bush Team talked about how to deal with terrorist threats generally.  The Obama Team’s embrace of the same basic preemption principle, its invocation of the principle in actual use-of-force situations, its less scary and more diplomatic presentation of the principle, and its success in getting others in the international community on board for the principle (and related ones) in the context of the dispersed Islamist terrorist threat, together constitute a signal achievement in the progressive development of international law.

First Use

In none of the passages Marty cites did the Bush administration proclaim a “first-use” notion of self-defense.  That is Marty’s term.  However, anticipatory self-defense is by definition a defensive strike by a nation prior to an attack by another nation or or group that threatens attack.  Otherwise the strike is just part of armed conflict or at most self-defense, not anticipatory self-defense.  So in seeking to decouple a “first-strike” from anticipatory self-defense, Marty is already muddying the water.  

The Bush administration embraced a flexible notion of “imminence” in its anticipatory self-defense calculus, and its preemption principle certainly did not rule out the possibility of what Marty calls first strike.  Still, all of the plausible scenarios where the Bush administration might have applied that principle—in Iraq, and against al Qaeda in Afghanistan and in other places—involved situations where the target had previously used or threatened force against the United States and had violated jus ad bellum rules.  Moreover, it is wrong (and deeply misleading) for Marty to link the Bush preemption doctrine to President Bush’s September 2001 pledge to find, stop, and defeat “every terrorist group of global reach.”  President Bush did not claim that he would use preemption to attain this goal, which might of course be reached through many strategies.  Finally, as Marty acknowledges, the Bush administration did not claim self-defense as the justification for invading Iraq (though it argued that it could have done so). 

While the Bush administration never to my knowledge exercised the “first-strike” anticipatory self-defense that Marty decries (cf. Olympic Games, below), the Obama administration almost certainly did.  Consider the 2014 strikes on the Khorasan Group in Syria.  The administration was all over the map in describing the nature and imminence of the threat from the Khorasan group.  But recall that Admiral Kirby justified the strikes on the grounds of “prevent[ing] the planning from going any further, and certainly to prevent them getting into an execution phase, which we don’t believe they were in yet.”  Of those same strikes FBI Director Comey said, “I can't sit here and tell you whether [Khorasan’s] plan is tomorrow or three weeks or three months from now.  Given our visibility, we know they're serious people, bent on destruction. And so we have to act as if it's coming tomorrow.”  The important point for present purposes is that the Obama administration used force in a nation without its consent to attack a group that had never attacked the United States based on a stretchy conception of imminence—the very sinful act that Marty accuses the Bush administration of supporting in theory. 

Marty tries to avoid this awkward conclusion by asserting that the Khorasan Group consists of “elements of al-Qai’da,” a group that has attacked the United States.  One response is that this just shows how malleable and unconstraining the term “no first-use” is (cf. “imminence” and “associated forces”).  Another response is to wonder whether this argument commits Marty to the controversial view that the United States can use force in self-defense under the Charter in any country where al Qaeda associates may be found regardless of the actual threat they pose, as long as the "unwilling or unable" criterion is also satisfied.  (This seems to have been the administration’s view when John Brennan’ claimed that “we have the authority to take action against al-Qa’ida and its associated forces without doing a separate self-defense analysis each time.”)  But the most fundamental response is that the U.N. Charter guarantees Syria’s sovereignty against the use of U.S. force.  The fact that a group in Syria contains “elements” of a group that the United States is in armed conflict with elsewhere does not, I think, relieve the United States from performing an independent and comprehensive jus ad bellum analysis when it attacks the related group in a new nation without the nation’s consent.  (See here for more on this point.)  The Obama administration had to perform a new self-defense analysis before it used force in Syria, which meant that it had to determine (among other things) that the Khorasan group in Syria, which had never attacked the United States, itself posed a threat.

In addition to the Khorasan group/Syria example, the Obama administration invoked a self-defense rationale (and not just collective self-defense) for the use of force in Syria against ISIL, another group “related” (though in a different way) to al Qaeda.  In this context Marty acknowledges that the use of force in Syria against ISIL is “the most forward-leaning and potentially controversial of those that Western nations have made in connection with the jus ad bellum and ISIL, because they at least imply that the U.S. and U.K. could use force in Syria against ISIL even before ISIL has operationalized or planned any armed attacks on those two nations.”  Marty tries to take the sting out of this point by arguing that it was “largely academic” due to the collective defense rationale.  If only he were equally charitable to the Bush administration, whose “academic” positions that weren't relied on in practice he views in quite the opposite light.   

And then there is Olympic Games, the operation of cyberattacks against the Iran nuclear weapons facility.  The deployment of the Stuxnet worm to damage 1000 or so Iranian nuclear centrifuges was almost certainly a “use of force” under the Charter.  Since Iran had not yet attacked Israel or the United States, or indeed even developed a nuclear weapons capability, it appears that the only conceivable legal justification under the Charter for the cyberattacks was an exercise of “first-use” anticipatory self-defense with a temporally very generous conception of imminence (including speculative likelihood of use) that Marty says is the core Bush administration evil.  The Obama administration thus appears either to have done in Iran exactly what Marty criticizes, or it acted unlawfully under the Charter.  (Olympic Games began under Bush and accelerated under Obama; we do not know if the Bush-era cyberattacks rose to the level of a use of force, but I presume that the Bush administration, like the Obama administration, concluded either that the cyberattacks were unlawful under the Charter or constituted anticipatory self-defense.)

Fully-formed threats and low-probability attacks

I don’t know why Marty emphasizes as distinctive to Bush the idea that anticipatory self-defense is sometimes justified “before any threat of attack is ‘fully formed.’”  That rationale was a justification for attacks on ISIL in Syria, the principle that Egan defended, what Kirby and Comey described in the passages above, and the probable justification for Olympic Games.

I think the main point that Marty seeks to make in his piece is that the Bush administration argued that it could exercise anticipatory self-defense when there is a “relatively low” probability of attack by the other nation or group within it.   Marty derives this central point of supposed difference between Bush and Obama from a 2002 OLC opinionan opinion that was not published until 2009, and that the Obama administration did not (in contrast to other Bush-era opinions) reverse.  Here is the full passage:

Third, as we have discussed earlier, the degree of harm that could result from Iraq’s use of WMD could well be catastrophic. The combination of the vast potential destructive capacity of WMD and the modest means required for their delivery make them more of a threat than the military forces of many countries. Chemical weapons and biological agents are easy to hide, and small quantities can have a devastating effect on the civilian population.  Perhaps even more frightening is the tinder-box that would result were Iraq to transfer WMD to terrorists.

We observe, therefore, that even if the probability that Iraq itself would attack the United States with WMD, or would transfer such weapons to terrorists for their use against the United States, were relatively low, the exceptionally high degree of harm that would result, combined with a limited window of opportunity and the likelihood that if we do not use force, the threat will increase, could lead the President to conclude that military action is necessary to defend the United States.

OLC is here fleshing out the “degree of harm” component in determining whether a threatened attack is adequately imminent to justify anticipatory self-defense.  OLC says that an exceptionally high degree of harm and a limited window of attack will justify a “relatively low” degree of probability of attack that “could” (i.e. might) justify anticipatory self-defense.  OLC does not specify what “relatively low” means.  It does not say that the term means an objectively low probability.  Rather, it says only that it is relatively low compared to the certainty one would require for less destructive weapons.

This 2002 analysis comes in just the type of broadly reasoned OLC opinion that used to keep me up at night, and that I have no interest in defending generally.  But this passage is not nearly the big distinguishing factor that Marty makes it out to be (and not just because the opinion was not published until 2009).

First, nothing in Egan’s speech or any other Obama explanation about the scope of anticipatory self-defense rules out this rationale.  To the contrary, Egan’s speech makes plain that the anticipatory self-defense balancing analysis includes “the likely scale of the attack and the injury, loss, or damage likely to result therefrom in the absence of mitigating action”—i.e. the very factor that the 2002 OLC opinion is fleshing out (along with the other factors that Egan mentions).  I am not saying that the Obama Team embraces the notion of anticipatory self-defense in response to “relatively low” probability but hugely destructive attacks.  But such a position is consistent with Egan’s balancing test, which contemplates that the degree of harm is relevant to preemption. 

Second, Marty’s apparent aim in invoking language from a 2002 opinion that the public did not see until 2009 is to link the Bush administration’s preemption principle to something akin to the 1% doctrine.  I have no doubt that some in the early Bush administration were motivated by this doctrine, though I do not recall anyone linking that doctrine to an interpretation of preemption.   Instead, the Bush administration’s contemporary public justifications for the preemption principle (see, e.g., here and here), which Marty ignores, articulated the same basic principles as Brian Egan.