Two Conceptions of Proportionality

George Fletcher and Jens David Ohlin have co-authored a new book arguing that the international law regulating the use of force by states should be informed by domestic criminal law regulating the use of force by individuals.  Their central argument is that the U.N. Charter should be read to incorporate the French criminal law doctrine of “legitimate defense”, which encompasses both self-defense and defense of others, in part because such a reading permits the defense of other states as well as national groups within other states without Security Council authorization.  They also argue that international law should follow the MPC and allow the use of force if “immediately necessary on the present occasion” to eliminate a threat that is inevitable but not imminent.  These positions are debatable but I want to focus on one area regarding which I think the authors misapply domestic criminal law concepts to the international context. 

Fletcher and Ohlin observe that in domestic criminal law self-defense and necessity involve two very different conceptions of proportionality.  The permissive conception of proportionality which derives from the law of self-defense allows one to inflict a greater harm on an attacker to avoid a lesser harm to a victim so long as the two harms do not differ too much in relative seriousness (for instance, we may kill to prevent a rape).  The restrictive conception of proportionality which derives from the law of necessity requires an impartial balancing of the wrong to an innocent person that one proposes to commit against the harm to others that one would thereby prevent.  Fletcher and Ohlin argue that the more permissive conception of proportionality should govern the killing of civilians (p.100) as necessary side-effects of direct attacks on legitimate military targets.  But this seems badly wrong since the permissive conception was designed to govern the use of defensive force against attackers while the more restrictive conception was designed to govern the infringement of the rights of the innocent. 

The authors write that “the action should be regarded as privileged because of the legitimacy of the dominant purpose” but it is hard to know what this could mean.  A soldier’s actions are privileged in the sense of being immune from prosecution if and only if they are legal, so this conception of privilege can’t be used to establish the legality of an action.  If the authors mean that the attack on the military target is privileged in the sense of being presumptively legal then this may be true, but only as true as saying that the use of defensive force against a wrongful aggressor is presumptively legal.  By contrast, damage to civilians is presumptively illegal and in need of further justification just as infringements of the domestic law rights of the innocent are presumptively illegal and in need of further justification.  The strange parallel the authors wish to draw, between criminal aggressors and innocent civilians, simply does not materialize. 

The authors also suggest that their position fits the Rome Statute’s description of “clearly excessive” damage to civilian persons and objects but this seems wrong as well.  “Excessive” and “disproportionate” are synonyms and “clearly” is an epistemic concept not a moral concept like, say, “grossly.”  The Rome Statute can’t possibly mean that it’s okay to kill an excessive number of civilians just as long as the excessiveness isn’t “clear”.  It seems pretty obvious to me that this is a conduct rule prohibiting excessive damage to civilians mashed together with a decision rule directing courts to convict only where the excessiveness is “clear”.  The reason is that if the excessiveness of the damage is clear then the order to cause that damage is “manifestly illegal” and soldiers have a duty to refuse.  If the legality of the order is arguable then soldiers should be protected if they obey.  The wording of the Statute is conceptually confused (non-clearly excessive damage is not permitted but merely excused), but that’s no reason to advocate treating innocent civilians as if they were wrongful aggressors. 

Incidentally, the authors may have missed an opportunity to make a genuinely interesting and morally defensible point, which is that the permissive conception of proportionality applicable to criminal aggressors should be applied to enemy soldiers, giving them far more protection than they currently enjoy.  My own view is that the killing of enemy soldiers is impermissible unless doing so is a necessary means to achieving a military objective whose value is proportionate to their deaths.  This contrasts with civilians, who can only be killed as a necessary side-effect of achieving such an objective.  I used to think that there was only one difference between soldiers and civilians, namely the appropriate causal relationship between their deaths and military objectives (means versus side-effects).  I now see that there is an additional difference, namely that the killing of soldiers should be evaluated using a more permissive proportionality standard and the killing of civilians should be evaluated using a more restrictive standard.