We have already noted the key point for the understanding of the proportionality requirement within the framework of private defence; this is an essential but flexi- ble requirement. This requirement is well accepted in Anglo-American law as expressed in the literature, in court rulings,713in the existing legislation and—in a less direct way—in draft laws.714In the decisive majority of the existing laws over- seas that were examined, the requirement is formulated in a flexible manner, using
711It is emphasised, that said requirement of proportionality does not imply that the citizen is obliged to relinquish his interest to the aggressor in all those cases where the attacked interest is of lit- tle value. In the great majority of the cases it will be sufficient to use moderate defensive force in order to protect the attacked interest. Thus, eg, when the aggressor wishes to take property with little value from the attacked person (such as a simple pencil), it should usually be sufficient to provide a warning, and at the most to use moderate defensive force (such as a push or a light blow) in order to protect the asset. Accordingly, the justification for the use of tremendous defensive force will be negated by the requirement of necessity. Only in rare cases in which moderate means would not suffice, and therefore a large amount of force would be necessitated, would the question of proportionality arise. Thus, in my view, obliging the citizen to waive his right in such rare cases is not as outrageous as it would appear at first glance.
An additional trend in German law was to establish that private defence would not apply where an excessive deviation from proportionality was concerned. Thus, eg, it is recognised according to this perception that fatal shots aimed at a child that is about to steal an apple are not to be justified. Even those who support this approach do not negate lethal force to protect property, but would, for exam- ple, justify the killing of a person who ran away with a jewelry box that he stole—see Bernsmann, n 265 above, at 177–78 (Bernsmann also has reservations regarding the justification of taking human life in order to protect property, even if property of great value is concerned).
712Compare the position of Eser, n 26 above, at 633 (claiming that what is important is the very recognition of the principle of proportionality, even if the simple and honest method of a direct and explicit requirement is not chosen).
713See, eg, Robinson (1984), n 37 above, at 5; Kadish, n 34 above, at 887; Gordon, n 1 above, at 715;
Williams (1983), n 1 above, at 506ff; Ashworth, n 183 above, at 297.
714See, eg, s 41 of The Crimes Bill of New Zealand (Wellington, 1989).
principally drafting that includes an element of reasonability,715or formulations from a negative perspective (ie, negating the defence only if the defensive force is out of all proportion to the prevented evil).716
As to the method of applying the test of proportionality, I am of the opinion that, in distinction from the drafting that is sometimes used of a comparison between the harms,717it is preferable to relate to the defensive force itself and not to its results, which may be completely random.718The central factors that should be considered are therefore the expected results (or those that can be foreseen719) of private defence and of its avoidance. Thus, for example, if the defender hit the aggressor with a punch of his fist without knowing that the aggressor had a ‘thin skull’, there is not much point—either for the purpose of guiding behaviour or for considerations of fairness and justice—to take into consideration the surprising fatal result, but the punch of the fist itself or its anticipated (or foreseeable) results (which would not usually include death) should be considered. From the opposite direction, lethal shots in the direction of the aggressor’s head in order to protect property should be prohibited as contravening the principle of proportionality, even if the actor actually missed the aggressor’s body.
With regard to the evaluation of the anticipated results for the aggressor and for the person attacked, the main characteristic factors that should be taken into account are the relative strengths of the sides; a multiplicity of aggressors; an attack with an especially violent character but without a weapon; previous violent behav- iour of the aggressor; the type of force used by each side and its strength; etc (this is, of course, not an exhaustive list).720Williams, for example, ventured an opin- ion that it is so common for brutal aggressors in England to kick the heads of their opponents after they fall to the ground, that it is definitely reasonable to take such a serious possibility into account in the evaluation of the danger.721
715 See, eg, s 34(6) of the Argentinian Penal Code (1921); s 21 of the Korean Penal Code (1953);
s 33 of the Swiss Penal Code (1937); ss 271–73 of the Australian (Queensland) Penal Code; s 8 of the Spanish Penal Code (1944; 1963).
The term ‘reasonability’ is also used in Art 34p of the Israeli Penal Code 1977.
716 See, eg, s 24(1) of the Swedish Penal Code (1962, 1972); s 5 of the Greenland Penal Code (1954) and s 48 of the Norwegian Penal Code (1902).
717 Eg, the original formulation of s 22 of the Israeli Penal Code 1977 was ‘provided that . . . the harm inflicted by the act was not disproportionate to the harm avoided’.
718 For similar opinions see Perkins and Boyce, n 85 above, at 1188; Gordon, n 1 above, at 761;
Ashworth, n 183 above, at 296.
719 I will refer to the distinction between the expected results and the results that can be foreseen below—see the text accompanying n 769.
720 See on this matter: La Fave and Scott, n 43 above, at 653; Card, Cross and Jones, n 180 above, at 628.
721 See Williams (1983), n 1 above, at 506.
Such an evaluation must take into account a very important factor that I have noted previously—the probability of the occurrence of the danger.722The need to weigh the probability of the occurrence of the danger (and the occurrence of the evil), of course, add to the great complexity of performing the test of proportion- ality. This complexity constitutes an additional reason (with a limited but existing weight) for favouring a flexible test over the requirement for perfect balance.
Another interesting suggestion that was raised is to take into account in the evaluation of proportionality the possibility that the victim might gain full com- pensation for his injury if he avoids the use of force.723I have referred previously to this possibility in the context of the discussion of the necessity requirement.724 Another case that deserves mention is that of multiple aggressors. In my view, in such a case, the existence of the requirement of proportionality should be examined in relation to the injury to each aggressor separately. There should be no joining together of all the aggressors for this purpose. It would be possible, for example, to justify a defensive action during which the attacked person kills a number of aggressors if this is necessary to save his life. This opinion can be based directly on the rationale for private defence: although it is possible that the injury to the autonomy of the person attacked is not increased by the multiplicity of the aggressors, there is no doubt that each of them has his own ‘aggressor’s guilt’ and that each of them independently infringes the social-legal order.725
722See the discussion of the severity of the damage in section 3.5.2 above and also the text accom- panying n 636 above. Gorr performed an interesting attempted analysis that compares the various pos- sible risks in light of the level of danger and the probability of its occurrence. He constructed a table of nine possible combinations of danger level (that might be trivial, moderate or serious) and the proba- bility of occurrence (that could be almost certain; significant but not certain; or insignificant). Gorr tries to establish certain determinations with regard to these nine possible combinations in respect to the permission for lethal defensive force. The attempt, in itself, is interesting. However, the reality seems to me to be far more complex. Eg, there is also a need for solutions to the exercise of defensive force that is not lethal. It is also doubtful whether the division of the level of the danger and the prob- ability of its occurrence into (only) three ranks for each provides the minimally necessary accuracy. I am of the opinion that there is no other resort but to leave the principal decision with regard to the requirement of proportionality in the hands of the court in light of the general test. This general test can and should be supplemented with specific instruction, especially with regard to lethal defensive force, but not as exclusive instruction standing alone without a general test. See Gorr, n 191 above, at 259ff.
723See Gorr, n 191 above, at 261ff. Gorr restricts his proposal to lethal defensive force and to cases in which the victim knows that he will gain full compensation for his injury, restrictions for which he does not provide a basis and for which I see no reason.
724See the text above accompanying n 635.
725For a similar approach see Robinson (1984), n 37 above, vol 2 at 70 and Montague, n 181 above, at 32.
At the end of this discussion on the content of the proportionality requirement, it is necessary to note a very important issue in the rule of private defence that is usually seen as a derivation of the require- ment for necessity; this is the duty to retreat. As I shall demonstrate in the following chapter, which is devoted to this important issue, my opinion is—in contrast to what is usually accepted—that the duty to retreat is not derived from the necessity requirement alone, but also (or—more accurately—princi- pally) from the proportionality requirement.