1 Sec. 2(4) of the Income Tax Ordinance (New Version), 1 L.S.I. [N.V.] 145 (hereinafter referred to as the Ordinance), taxes income from interest and linkage differentials. In Zimmerman v. Assessing Officer (1973) 27(i) P.D. 802, the Court held that “interest” referred to in sec. 2(4) covers not only an addition to the capital sum of a loan which a borrower pays the lender under a voluntary loan, but also interest awarded under the Adjudication of Interest and Linkage Law, 1961 (15 L.S.I. 214) which was added to a sum won in a lottery, and which did not have the nature of income. Subsequently, in Benano v. Assessing Officer (1979) 33(ii) P.D. 296, it was held that the rule in Zimmerman decided in a matter relating to pecuniary damage – applies also to interest adjudicated in a matter of physical injury. Therefore, if the award in respect of loss of earnings and expenses up to the date of judgment is the nominal amount of loss of salary and expenses together with the maximum legal interest or together with linkage differentials and interest (not linked) from the date the damage was caused to the date of judgment – the major part of the interest in the first case, and the entire interest in the second case, are in the nature of salary or refund of expenses and not interest for deprivation of the use of the money. Despite this, they will be taxed as interest, and not as compensation for loss of earnings or for expenses. Refund of expenses incurred by the injured person clearly ought not to be liable to tax. Salvation may be found in sec. 2(4) of the Income Tax (Exemption from Tax on Linkage Differentials) Regulations, 1984 (K.T. (1984) no. 9620, p. 1322), made under the power inferred on the Minister of Finance with the approval of the Knesset Finance Committee by sec. 16B of the Ordinance, which exempts from tax “linkage differentials which an individual received as an addition to a sum of compensation, to an amount paid out under a claim or to a sum in respect of waiver of a claim, under any law, provided that the sum to which the linkage differentials are added is exempt from tax or does not constitute income in his hands”.

In Israel the argument may be advanced that receipt of compensation from the wrongdoer or indemnity from the insurer is a “sale” (“waiver” or “exchange”) of an “asset” (a right of action in tort against the wrongdoer or in contract against the insurer which is a “chose in action”) as these concepts are defined in sec. 88 of the Ordinance, so that capital profits tax applies. If this assertion is accepted, any compensation or indemnity (even if not liable to ordinary income tax and not amounting to a sale of a physical asset attracting capital profits tax) will incur liability to capital profits tax.

2 See discussion at pp. 40–41, infra.

3 Kahane, and Yoran, , “Compensation for Loss of Income and its Taxation: A Policy Analysis” (1979) 32 Nat. Tax J. 117, at 118Google Scholar.

4 Ibid.

5 Surrey, , “Tax Incentives as a Device for Implementing Government Policy: A Comparison with Direct Government Expenditures” (1970) 83 Harv. L.R. 705, at 706CrossRefGoogle Scholar.

6 A number of drawbacks to a tax expenditure as against direct budgeting: A tax expenditure (unless it is in the form of a credit) is regressive. It is not subject to annual parliamentary approval, unless parliamentary approval is required not only for the budget itself but also for the tax expenditure budget, as in Germany and the USA. The attempt to enact in Israel a statute under which every annual budget law should have appended to it an estimate of amounts that will not be collected owing to exemptions, reliefs or allowances from tax, has recently failed. See the Foundations of Budget (Amendment No. 4) Bill (H.H. (1987) no. 1816, p. 157). In that draft law it was also proposed that no enactment be passed nor any regulation promulgated which have the effect of increasing tax benefits or diminishing State revenue, unless a source to finance them is found, and that a draft law or a law involving a further reduction of taxes or of State revenue in any other way should include a provision stating the source from which they are to be financed.

There exists an optical illusion that a tax expenditure costs less than direct budgeting so that it is easier to pass tax expenditure laws than laws providing for direct support. Where a part of the subvention is given through direct budgeting and part in the form of a tax exemption, there is no overall picture of the amount of support; it is the tax authority that is responsible for the support, not the administrative authority responsible for and expert in the particular field of activity. See Yoran, , “Tax Incentives or Grants – Which is Preferable?” (1973) 24 Ro'eh HaHeshbon 6Google Scholar.

7 Regarding the exemption in sec. 104 (a)(2) in the United States, it has been stated that it “is rooted in emotional and traditional, rather than logical, factors”: Harnett, , “Torts and Taxes” (1952) 27 N.Y.U.L.R. 614, at 626Google Scholar.

8 Frolik, , “Personal Injury Compensation as a Tax Preference” (1985) 37 Maine L.R. 1Google Scholar.

9 See development of this analysis, infra at pp. 58–61.

10 In the case of Perma Sharp v. Assessing Officer (1974) 6 P.D.E. 361, it was provided under a compromise arrangement that a certain sum should be paid to the General Manager of a company, whose employment had terminated, “in settlement of his claim … for compensation for damage caused to his name and his reputation as well as compensation for distress caused to him …” The District Court rejected the argument of tort damages, and expressed misgivings both as to the very existence of a wellfounded claim for payment of tort damages and as to the level of damage caused. In the Supreme Court, sitting as a Civil Appeals Court, Asher J. held that the plea as to considerable damage alleged through harm to the name and reputation of the plaintiff does not create the impression of being a serious one, and that “hesitations on the part of the assessing officer, and, following thereon, also on the part of the learned judge, as to the existence of a genuine claim in tort, … were sufficiently well-founded to dismiss the case for a claim in tort”; ibid., at 364. The law is as held in Stein v. Assessing Officer (1973) 27(i) P.D. 182, at 186: “that the income must be examined according to its real nature, and that it is not within the power of the parties to the transaction, and not even within the power of an arbitrator, to alter reality by conferring on the payment a certain form and calling it a name, which do not conform to its nature”. See also Rosenzweig v. Rosenzweig Bakery (1962) 16 P.D. 2548, where an individual sued in tort a company that was entirely controlled by himself, and of which he was the sole director, for breach of duty of proper fencing of machinery.

11 Kahane and Yoran, supra n. 3, at 119.

12 Ibid.

13 In Israel: the provision in sec. 9(7) of the Ordinance which exempts from tax “any capital sum received as consolidated compensation for death or injury”. In the United States: the provision of sec. 104(2) of the Internal Revenue Code which exempts from tax lump-sum or periodical compensation for personal injury or sickness. In England, since judicial interpretation has determined that no tax is payable, it will be necessary to have a specific statutory provision imposing liability for tax. In the past, proposals have been put forward for making damages liable to tax. For the U.S., see Frolik, supra n. 8, and Yorio, , “The Taxation of Damages: Tax and Non-Tax Policy Considerations” (1977) 62 Cornell L.R. 701, at 731–735Google Scholar.

Lord Keith, the minority judge in British Transport Commission v. Gourley [1956] AC 185, at 218, was of the opinion that if the injured person ought to have his compensation reduced owing to tax liability, this has to be effected by legislation providing for the tax to be paid to the Treasury rather than into the plaintiff's pocket. The same view is expressed in McGregor, , Damages (London, Sweet & Maxwell, 14th ed., 1980) 419Google Scholar. This was also one of the minority opinions in the committee appointed to examine the implications of the Gourley decision. See Report of the Law Reform Committee, Seventh Report, Effect of Tax Liability on Damages, August 1958, Cmnd. 501Google Scholar. Bale thought that this solution was an equitable one but he remarked: “Is it feasible to expect the [Canadian] Federal Parliament to adopt such a solution? I think that the answer is probably no”: Bale, “British Transport Commission v. Gourley Reconsidered” (1966) 44 Can. B.R. 66, at 101.

Dworkin considers this to be the right solution, although not a perfect one, since the tortfeasor would have to bear the estimated tax liability of the injured person in respect of the compensation, and thus would be likely to pay excess compensation: Dworkin, , “Damages and Tax – A Comparative Survey” (1967) B.T.R. 315, at 329Google Scholar. In my opinion, Dworkin is mistaken on this matter. The tortfeasor does not pay the tax on the compensation for which the injured party is liable; rather he pays a gross amount of compensation, so that the question of a wrong estimate of the tax on the compensation does not arise at all. Moreover, the tax liability on the compensation will be determined according to the current estimate of income, so that the compensation minus the tax together with the tax liability on the compensation is equal to gross compensation. See Kahane and Yoran, supra n. 3, at 117. In the case of Khalif v. State of Israel (1981) 35(ii) P.D. 242, at 251, Barak J. refers to our comment on this matter and remarks that “this matter goes beyond the realm of assessment of damage in tort, and is a legislative consideration which the law-maker ought to take into account”. He does not express any opinion on the question of whether the tax laws should be amended. Later, in his article “Damages for Personal Injury: The Law and its Reform” (1983) 9 Iyunei Mishpat 243, at 256, n. 86Google Scholar, he points out that the reasonable solution is that of gross compensation liable to tax, in accordance with our proposal. For a similar conclusion, see recently, Shavell, S., Economic Analysis of Accident Law (Harvard U.P., 1987) 143–144CrossRefGoogle Scholar. See also n. 36 infra.

14 It would seem that the injured party will be entitled to deduct his legal expenses, including lawyers' fees, from the compensation, since they have become expenses incurred in producing the income. As to the “loading” required in order to compensate the injured person by increasing the compensation in consideration of the fact that he is not refunded his legal expenses, see Kahane and Yoran, supra n. 3, at 125, n. 10.

15 29 L.S.I. 311, as amended by the Road Accident Victims (Compensation) (Amendment) Law, 1977 (32 L.S.I. 61).

16 34 L.S.I. 92.

17 Sec. 4 (a)(2) of the Road Accident Law provides: “where the said compensation is exempt from income tax, the losses of the victim shall, for the purposes of that compensation, be calculated in accordance with his income after deduction of income tax due thereon at the time the compensation was fixed; provided that the reduction of the income through the deduction of tax as aforesaid shall not exceed 25 per cent of the income in accordance with which the compensation is calculated.”; Sec. 5(a) of the Defective Products (Liability) Law, 1980, is identical in content. It follows that even if the deduction provisions are not repealed, the deduction will not be effected, since the condition as to exemption of the compensation from tax will not have been fulfilled. However, the rules for deduction ought to be repealed if the tax exemption is abolished.

18 Harris, , “Compensation for Loss of Income and its Taxation: Comment” (1981) 34 Nat. Tax J. 135, at 135–136Google Scholar. This is a response to Kahane and Yoran, supra n. 3.

19 Ibid., at 135.

20 In Israel, with regard to the insurance required by sec. 3(a)(2) of the Motor Vehicles Insurance (Third-Party Risks) Ordinance (New Version), 1970, (2 L.S.I. [N.V.] 74) as distinct from liability insurance under sec. 3(a)(l) of that Ordinance.

21 See infra n. 40. For a discussion of various criteria for allocating the premium of road accident victims insurance, see Kahane, Y., An Examination of the Financial Implications of Treatment of Road Accident Victims in Israel by the National Insurance Institute (research study submitted to the National Insurance Institute, 1977)Google Scholar.

22 See Hadari, , “Damages and Income Tax: The Rule of Gourley in the Light of Later Cases and in No Fault Liability Statutes” (1978) 6 Iyunei Mishpat 373, at 393Google Scholar: “It would be right to limit compensation for loss of earnings or of earning capacity to the amount of three times the general average wage, and to provide that income tax should be deducted from the payment. In both cases, the limitation would be on the amount of payment serving as a basis for calculating the compensation. From the point of view of the Treasury, this would amount to its participation in the no-fault compensation plan. At the same time that the scope of the liability of insurers would be extended towards victims who would not obtain compensation through a traditional claim in tort, i.e., proof of fault, the scope of the liability of insurers for notional tax on payment would be reduced. The State on its part would leave the exemption on capital compensation in force, thereby participating in the cost of the law”.

23 The aims of increasing the cover and reducing insurance payments are mutually contradictory, and in Posner's view are also not consistent with the aim of preventing accidents. See Posner, R.A., Economic Analysis of Law (Boston, Little, Brown, 3rd ed., 1986) 189Google Scholar.

24 See infra n. 36 as to the remedial aim of the law of compensation, infra n. 43 as to the deterrent theory of Posner and infra n. 49 as to the aim of efficiency, according to the Calabresi's formula. There are those who maintain that from a purely economic point of view resources should be transferred to the victim only on the basis of his contribution to the economy after the tortious act. Consequently, there is no room for full compensation. Burnovski, and Liber, , in their article “Compensation for Road Accident Victims – Some Economic Aspects” (1978) 6 Iyunei Mishpat 132, at 134–135Google Scholar, take this as their starting point and reach the conclusion that the preferable approach is an intermediate one – of preserving the economic status as it was prior to the accident, up to a ceiling, and establishing minimum compensation that will enable a victim whose actual and expected income is low to subsist in a dignified way. In view of this criterion, they criticize the ceiling provided in sec. 4 of the Road Accident Law as being too high. Their article was written before the Law was amended and the deduction of 25% for tax exemption was enacted.

25 As from 1983 periodical payments for personal injuries are exempt from tax in the U.S. See supra n. 13. This was the position under French law, as well as under long-standing practice in Hungary. See McGregor, , “Personal Injury and Death”, in International Encyclopaedia of Comparative Law: Torts, Part 2, 9–136, at 52Google Scholar. In Germany, until 1964 the law provided for net periodical compensation subject to tax; ibid. These are exceptions to the consensus as to gross periodical compensation subject to tax.

26 As in Germany, Switzerland and Sweden. See ibid., at 52–53.

27 This is the law in Sweden, where the period of averaging is up to 10 years. See Ibid., at 53, n. 456. In Germany a part of the lump-sum compensation is exempt from tax; ibid., at n. 457.

28 See Kahane and Yoran, supra n. 3, at 122-123. A benefit in respect of a short period is granted only if averaging is permitted over a period which is greater than the period of the damage.

29 We put it as follows, ibid., at 123: “The damages awarded to the victim represent a substitute for the flow of income lost as a result of the injury. The principle of equity would be preserved by taxing the damages in a way that would create the same tax liability as before the accident. This can be achieved by creating tax flows identical to those which would have been generated by the lost earning flows. Alternatively, the same tax liability is preserved when the present value of the tax flows is collected. These amounts can be estimated by the use of basic actuarial formulae:

Assume that a victim is injured at age x, and that his loss of income at year t after the injury is S_x+t. If his relevant tax rate at that time is T_x+t, then the amount he should pay as taxes is T_x+tS_x+t. The loss of income is conditional on the probability of survival at age x + t. Such probabilities are stated in mortality tables and denoted by L_x+t/L_x, where L_x is the number of people living at their x^th birthday.

The damages, D, which the victim is entitled to receive, equals the expected present value of the pre-tax loss.

where v is the discount factor . The duration of the loss of income is denoted by n. (This would be, typically, the period from age x up to retirement; e.g., if retirement at age 65 is assumed, n = 65 - x.) The damage, D, can easily be calculated using actuarial tables.

In the special case where the same income flows are expected every year (S_x+t = S), the damages can be calculated by simply multiplying the income, S, by a constant which is found in actuarial tables and represents the present value of a life annuity of $1 per annum; i.e.,

The present value of taxes to be collected would be:

Assuming a uniform tax rate over the period and a constant earnings flow, R becomes .

Thus the theoretical problem is quite simple. In order to avoid overtaxation or undertaxation, the damages awarded should be taxed in a way which would result in a present value, R. There is an infinite number of possible tax vectors which add up to the present value R, however; the most reasonable method is to deduct the sum R as a lump sum from the lump sum awarded. Alternatively, the compensation may be spread over a period of years and be taxed at rate T. Both ways are identical as long as the taxes are calculated according to the same assumptions that are used in determining the damages, i.e., the same mortality table, the same interest rate, the same period. etc.” Compare, Bishop, and Kay, , “Taxation and Damages: The Rule in Gourley's Case” (1987) 103 Q.L.R. 211, at 225–227, 232–233Google Scholar.

30 In the past, the Supreme Court dismissed the possibility put forward by the District Court of a further application to the Court by the widow of a victim after judgment was given, and held that “it has no basis in the Law, and furthermore it is not consistent, and may even come into conflict, with the limited period of limitation laid down… [in] the Civil Wrongs Ordinance, 1944”; Panz v. Feldman (1955) 9 P.D. 1711, at 1713-1714. However, the Supreme Court has recently acknowledged the competence to award periodical damages in exceptional cases; Na'im v. Barda (1982) 36 (iii) P.D. 762.

31 For an economic analysis of awards of periodical compensation as against lump-sum compensation, see Rea, , “Lump-Sum Versus Periodic Damage Awards” (1981) 10 J. Leg. Studies 131CrossRefGoogle Scholar. For a comparison of various systems of compensation, see also More, D., “Periodical Payments for Road Accident Victims” (1978) 6 Iyunei Mishpat 645Google Scholar; and Fleming, , “Damages: Capital or Rent” (1969) 19 U. Toronto L.J. 295CrossRefGoogle Scholar.

32 The argument against reduction in compensation owing to benefit from a third party is that such reduction will provide an inadequate deterrent. See discussion at pp., 49-50, supra. In addition, the injured party has paid for the benefit, for example, where personal insurance against accidents is concerned, by payment of the premium. The premium reflects the discounted estimated value of the risk (together with the insurance company's cost and profit). If the insurance policy were to exclude events for which compensation was paid or if it had a subrogation clause, the premium would be lower. And what about a benefit conferred gratis? Posner, supra n. 23, at 186, holds that the injured person nearly always pays indirectly for benefits, somewhat along the lines of “no free lunches”. Barak, supra n. 13, at 268, has put forward a fundamental approach as to the law that ought to apply to the tripartite relationship between benefactor - tortfeasor - victim: “The benefit should not be allowed to reduce the amount of compensation which the tortfeasor will pay; nor should the benefit increase the amount of compensation which the injured person will receive. The benefactor ought to get his benefit back, either by receiving it from the tortfeasor (at the same time deducting the sum that the tortfeasor will pay to the injured party), or by receiving it from the injured party (after full payment of compensation by the tortfeasor to the injured party). This basic method would not apply where the benefit is intended as a gift to the tortfeasor or to the injured party, or where the tortfeasor or the injured party had previously purchased the benefit since they expected the damage to occur, or in special and exceptional situations”. Since I do not see any justification for making a gift to the tortfeasor or to the injured party in the form of a tax benefit, my proposal is to abolish it. See pp. 40-43, supra.

33 See discussion at pp. 40-43, supra.

34 See supra n. 1.

35 Benano v. State of Israel, supra n. 1.

36 In Israel the accepted view is that the aim of the law of tort is remedial, so that the aim of the law relating to compensation is to give the injured party such a sum of money, which as far as is feasible, will put him in the same situation as he would have been in if he had not suffered harm by the tort. The expression “restoration of the situation to what it was previously”, used to describe the remedial aim, is not a felicitous one. We are speaking of compensation, not restoration, and indeed where physical injury is concerned, restoration of the situation to what it was is in no way possible. Nevertheless, this expression has become a “code word” for the compensatory aim, and I shall therefore continue to use it in its accepted sense. Although compensation is intended to fulfil a remedial aim, “such a remedial purpose is not necessarily the sole purpose… additional purposes may exist, which in appropriate cases would allow for granting compensation even where no remedial purpose is to be fulfilled. Nevertheless… these additional purposes, even if existing and recognized, do not detract from the compensation payable under the remedial system, although they may allow for payment of compensation where no remedial aim can be achieved”: Barak, supra n. 13, at 248. It seems to me that implementation of such rules in this context should result in grossed-up compensation wherever compensation is awarded liable to tax in respect of a loss which is exempt – otherwise the remedial purpose will not be attained. Where compensation is awarded free of tax in respect of a loss which is liable, net compensation would indeed realize the remedial purpose, but in such a case there is room for also taking into account the economic aim of efficiency of the law, and adjudicating gross compensation.

37 McGregor, supra n. 13, at 324-325, advances in this spirit the possibility that if capital gains tax on compensation applies in respect of loss which itself would not have been liable to that tax, then the wrongdoer must increase the compensation, in accordance with the rule that “a wrongdoer must take his victim as he finds him”. See also Bishop and Kay, supra n. 29, at 213.

38 Kahane and Yoran, supra n. 3, at 119.

39 On the other hand, if we are speaking of receipt which a self-employed person loses, constituting private expenditure on the part of the payer (such as expenses involved in consultation with a lawyer on private matters), the expenditure would not be an allowable deduction in the hands of the payer. See secs. 17 and 32(1) and (2) of the Ordinance.

40 In Friedmann v. Assessing Officer (1962) 29 P.M. 234, it was held that transportation of workers is part of the process of producing income, so that the accident had also occurred during that process. In Assessing Officer v. Silberstein (1962) 16 P.D. 1175, the Court held that an outlay on damages paid by a lessor to a lessee owing to a defect in the premises “did not go beyond the scope of expenses relating to domestic management which he is likely to incur as one of the natural, even if not regular, features of this source of income”. On the other hand, it was held in England in the well-known case of Strong & Co. Ltd. v. Woodfield (1906) 5 T.C. 215, that damages paid by an innkeeper to a customer as compensation for damage due to a chimney falling on the customer were not paid out in the process of producing income.

41 See n. 39, supra.

42 One should add that owing to the damage national income is also reduced (unless the wrongdoer is a foreigner or the damage is fully insured abroad, and the damage was compensated in full). Consequently, income tax receipts will also be reduced owing to the injury.

43 Atiyah, P.S., Accidents, Compensation and the Law (London, Weidenfeld & Nicolson, 3rd ed., 1980) 547–548Google Scholar, refers to the question of just apportionment of damage caused by road accidents, and states: “Society could impose the loss principally on those who indulge in the activities which cause accidents, e.g., those who take part in motoring, or it could decide to spread the burden equally over the shoulders of the nation – or at least as equally as the tax burden is ever spread, which means in fact making the wealthier pay a large share and the less wealthy a smaller share. Or society could compromise between the two and finance the whole system rather as it does with the social security system, with a mixture of national insurance and taxation. No one of these three methods of distributing losses can be said to be more ‘equitable’ than another, except on the basis of a fundamental political judgment about the amount of inequality which is tolerable in our society…” The remarks refer to “transfer” payments (i.e., direct support), but they are also applicable to support through tax expenditure.

Elsewhere we have proposed that in the context of tort liability distributive justice should be determined by the following criteria: a) The wrongdoer should pay the full amount of damage; b) the injured party should receive full compensation; c) there should be no shifting, i.e., third parties should not be affected; Kahane and Yoran, supra n. 3, at 118.

44 As to the question of who within the triangular relationship of tortfeasor–victim–beneficiary is entitled to enjoy a compensatory benefit, Ogus argues that with regard to particularly hazardous activity, the full burden should fall on the tortfeasor, and the beneficiary is entitled to restitution from either the victim or the tortfeasor: Ogus, A.I., The Law of Damages (London, Butterworths, 1973) 230Google Scholar.

45 U.S. v. Carroll Towing Co., 159 F. 2d 169 (2d Cir., 1947). “Hand's Formula elegantly combines considerations of deterrence and efficiency by converting the economic possibility of preventing accidents into social justification for transferring the damage”; Veljanovski, C.G., The New Law and Economics (Oxford, 1982) 74Google Scholar.

46 R.A. Posner, supra n. 23, at 147-151. Where the wrongdoer is required to pay less than the damage, the situation is not efficient according to the definition of efficiency given by Kaldor and Hicks, since by taking precautions an advantage would be gained (by the injured party) that would be greater than the cost (to the tortfeasor); Kaldor, , “Welfare Propositions of Economics and Interpersonal Comparison of Utility” (1939) 49 Econ. J. 549CrossRefGoogle Scholar; Hicks, , “The Valuation of Social Income” (1940) 7 Economica 105CrossRefGoogle Scholar. The result is a reduction in the “value of life” (or the value of rescuing the life of one statistical person), measured by cumulative readiness to pay for a marginal reduction in the risk that will rescue the life of one person; Veljanovski, supra n. 45, at 72.

The theory of deterrence embodied in the law of liability for negligence in tort has been subject to criticism. Atiyah (who prefers to speak of prevention of accidents rather than of deterrence) remarks that anyone not deterred by criminal liability and the risk of injury to himself from becoming involved in an accident will not be deterred by tortious liability. He adds that accidents are generally not the result of the taking of a conscious risk, but of lack of observation and perception, erroneous discretion, lack of basic qualifications and other factors which do not react to deterrence. Liability insurance neutralizes or considerably weakens the deterrence factor. The effect of insurance on the deterrence factor depends on the premiums system. The total premiums of all insured reflect the estimated capitalized value of the total risk plus cost and profit of the insurance companies. Increase in the risk will increase the total premiums. However, at the level of the insured individual, the effect may be negligible, marginal or insignificant. If there is no self-participation and increase in insurance payments owing to claims, the effect is negligible. To the extent that the structure of premiums is a function of reduction of the risk by the insured, the effect will increase: P.S. Atiyah, supra n. 43, at 558-576. See also Englard, , “The System Builders: A Critical Appraisal of Modern American Tort Theory” (1980) 9 J. Leg. Studies 27CrossRefGoogle Scholar, and in particular his criticism at p. 38 of treatment of the matter of benefits from third parties under Calabresi's system. See infra n. 49.

47 See discussion at pp. 40–43, supra.

48 The Supreme Court of the United States did not find in the legislative history of the provision exempting personal injury from tax - sec. 104(a)(2) of the Internal Revenue Code – a legislative intent that the exemption should favour the injured party and not the wrongdoer: Norfolk & Western Railway v. Liepell, 444 U.S. 490, at 496, n. 10; rehg. denied, 445 U.S. 972 (1980). On the other hand, it was held in England that since the legislator overturned the ruling of the House of Lords in Riches v. Westminster Bank (1947) 28 T.C. 159, and exempted from tax interest forming part of an award of compensation for death or personal injury, by sec. 375A which in 1971 was added to the Income and Corporation Taxes Act 1970, it would be a reasonable conclusion that he intended to benefit the plaintiff and not the defendant: Mason v. Harmon [1972] R.T.R. 1.

49 We are dealing with the question of the amount of compensation within the scope of a given law as to liability. Thus, the controversial question of how one liability law or another provides an incentive for reducing accidents, which is the main purpose of the law relating to accidents, does not concern us here. See on this, inter alia, Shaved, , “Strict Liability versus Negligence” (1980) 9 J. Leg. Studies 1Google Scholar; Polinsky, , “Strict Liability versus Negligence in a Market Setting” (1980) 70 Am. Econ. R. 363Google Scholar; Rizzo, , “Law Amid Flux–The Economics of Negligence and Strict Liability in Tort” (1980) 9 J. Leg. Studies 291CrossRefGoogle Scholar.

According to Calabresi, economic efficiency in the law of torts means reducing to a minimum three items of cost: the cost of accidents, the cost of accident prevention and the cost of administering a system of accident law. The purpose of efficiency of the law of accidents is to reduce the cumulative sum of accident costs and costs of their prevention. This can be achieved by imposing the damage on the “cheapest cost avoider”: Calabresi, G., The Costs of Accidents: A Legal Analysis (Yale, 1970)Google Scholar.

Net compensation, by reducing the deterrence, would increase the cost of accidents. It does not have the effect of reducing the social cost incurred as a result of uncompensated damage, unless the operation of damage allocation mechanisms is improved due to the reduction of compensation. The cost of operating the system of net compensation is higher than that of gross compensation, owing to tax calculations and litigation on the question of whether the compensation is exempt from tax. Thus it appears that even according to Calabresi's approach, when the various considerations are balanced, gross compensation promotes the efficient allocation of resources for prevention of accidents more than net compensation does.

50 Pigou, A.C., The Economics of Welfare (London, 4th ed., 1950) 172, 183–184Google Scholar.

51 Pigou's solution for closing the gap was imposing tax. Other possible solutions include merging the activity creating the externality with the activity influenced thereby, or laying down appropriate rules of liability. See Mishan, E.J., Introduction to Normative Economics (New York, 1981) 407–415Google Scholar. Each solution has its merits and defects: Coase, , “The Problem of Social Cost” (1960) 3 J. L. and Econ. 1CrossRefGoogle Scholar.

In our matter, resort to the rule of liability requiring gross compensation is the only way to avoid the negative externality on the tortfeasor. Gross compensation does indeed grant the injured party excess compensation whereas net compensation implements the principle of restoring the victim's position to what it was previously. However, gross compensation prevents the formation of technological externality which would disrupt the process of reallocation necessary for attaining the optimal position. See Prest, and Tuney, , “Cost Benefit Analysis: A Survey” (1965) 75 Econ. J. 683, at 688CrossRefGoogle Scholar; Arrow, , “Uncertainty and the Welfare Economics of Medical Care” (1963) 53 Am. Econ. R. 941, at 945Google Scholar.

For a comprehensive discussion of the economic approach to law in the context of “internalizing” externalities in tort, see the doctoral thesis of my colleague Gilead, I., The Economic Approach to the Law of Torts and the Defective Products (Liability) Law, 1980 (Jerusalem, The Hebrew University, 1983, in Hebrew) 9–10Google Scholar. The terminology and references in this note are based on his thesis. He is of the opinion that the/ “Coase Theorem”, which asserts that in a perfect market, in the absence of transaction costs, efficient allocation can be attained even without liability rules, does not operate for the laws on product liability. The reason for this is that even if there were a perfect market with free competition for products, there would not be in such a market direct negotiations between producers and consumers, and consequently there would be transaction costs. As a result, only if the law would provide liability rules which place the full social cost on the producer would an efficient allocation of resources be attained. Gilead, , “‘Coase Theorem’, Competitive Market and Products Liability Law” (1985) 20 Is.L.R. 39Google Scholar.

52 See sec. 9(5) of the Ordinance.

53 The courts in Israel have held that one should ignore not only the tax implication of the tax exemption on compensation, but also the implication of tax exemption, as a result of the injury, on the income after the injury. See Gafnan v. Zoro (1958-9) 20 P.M. 212 and Englard v. Amram 14 P.D. Summaries 11, 170.

54 (1957) 11 P.D. 225Google Scholar.

55 Adeto v. Zackzer (1972) 76 P.M. 70, at 80.

56 In Kohavi v. Becker (supra n. 54, at 237) Witkon J. left intact the rule of Grossman v. Roth (1952) 6 P.D. 1242, remarking that “the income tax factor is liable to complicate tort cases by introducing into the calculation of the damage an element which cannot be anything but a matter of guesswork”. Moreover, he bases his determination on the double taxation implicit in the rule of gross compensation – either owing to the tax burden being transferred in the future from tax on income to tax on consumption, or owing to taxation of the yield on the lump-sum compensation.

57 The rule of gross compensation prevails in Canada – see The Queen v. Jennings (1966) 57 D.L.R. 2nd 644 (Can. S.C.) – as well as in most of the States of the U.S. See Frumer, L. and Friedman, M.I., Personal Injury, Vol. 3, cumulative supplement (Albany, N.Y., Matthew Bender, 1985) 3.04 [4][a][vii]Google Scholar, and authorities quoted there. However, there are States where gross compensation is only awarded in respect of high incomes. See Harper, F.V., James, F. Jr., Gray, O.S., The Law of Torts, Vol. 4 (Boston, Little,Google Scholar

58 The general formula for retaining monthly interest I_m producing at compound interest an I_a annual interest is:

or

59 This simplistic calculation ignores the effect of the progressiveness on the taxation of the return on the lump-sum compensation. It is presented here in order to demonstrate the difference between gross compensation discounted at a gross rate and net compensation discounted at a net rate. A more exact calculation will be presented infra.

60 Norfolk & Western Railway Co. v. Liepelt 444 U.S. 490 (1980). The ruling that a net rate should be used for discounting was in the nature of obiter dictum.

61 Brady, , Brookshire, and Cobb, , “Calculating the Effects of Income Taxes on Lost Earnings” (Sept. 1982) Trial 65Google Scholar. In this article a table is set out presenting the present value of all future losses of income, assuming a 10% rate of annual pay increase and 12% annual interest. The discount tax factor is then:

where i is the rate of interest; t_k is the tax on the interest in year k; n is the number of years into the future, from 1 until the last year; π is the outcome of [1 + i (1–t_k)] from k = 1 to k = n.

At the amount of disposable income required at the commencement of each year so as to give the plaintiff during the year a sum equivalent to his loss the authors arrive by means of an iterative process. The authors remark that “…in general, as inflation increases, and hence the assumed wage growth rate and interest (discount) rate increase, the value of the “tax adjusted” loss estimate rises relative to the “no tax” loss estimate. At some point, as assumed wage growth and interest rates rise, the “tax adjusted” estimate exceeds the “no tax” estimate, all other factors remaining the same. Further, as the wage level of the injured/deceased is higher, and as the number of years over which wages will be lost increases, the likelihood increases that the “adjusted” estimate will exceed the “no tax” estimate.” Ibid., at 67-68. See also Franz, , “Should Income Taxes be Included when Calculating Lost Earnings?” (Oct. 1982) Trial 53, at 57Google Scholar.

Bell, Bodenhorn and Taub have argued that it is not possible by accepted techniques of capitalization to determine under a system of progressive tax a lump-sum compensation that will yield exact compensation, since the amount of compensation is a function of the tax rate, and vice versa. They therefore reach the conclusion that without knowing the annual tax liabilities which would have applied to the lost income, it is not possible to accurately determine the compensation, neither according to the method of determining present value nor according to that of allowing for the same opportunities for consumption as prior to the accident. The reason for it not being possible to reconstruct opportunities for consumption is that the relative cost of transit from one year to the next has changed; before the injury the transit would have been at a certain marginal rate of tax, whereas after the accident, owing to the compensation, the marginal tax rate will have changed. They conclude that it is possible to determine the compensation that will allow at least the same consumption every year. Such compensation would impel the recipient to alter the distribution of his consumption over the years in such a way that his position would be improved as compared to what it was before the injury; Bell, , Bodenhorn, & Taub, , “Taxes and Compensation for Lost Earnings” (1983) 12 J. Leg. Studies 181CrossRefGoogle Scholar.

Burke and Rosen disagree with the above conclusion. They contend that by a process of continual experiment it is possible to find an exact iterative solution to the determination of lump-sum compensation at an amount equivalent to loss of income after all the tax implications have been taken into account: “The model introduced here employs an iterative solution. It determines the exact amount needed to provide consumption choices as well as the taxes to be paid on the future earnings of any award. It is expressed in [the] equation… where PV = the desired value; = the estimated present value; , the error; and K. = the percentage of the error made up in the next iteration. The model is in equilibrium when and E = O…. First, a preliminary estimate of the present value is made. Second, taxes on the loss of labor earnings are calculated and deducted in each period… Then the taxes on the earnings of the award are also deducted in each period. Third, an annuity fund is calculated such that the error can be determined. The error will be either the sums of money left in the fund or the deficit in the fund at the end of the work-life expectancy. Fourth, this error in the fund is subtracted if there is a surplus of funds or added if there is a deficit. The amount subtracted or added is equal to this deficiency or surplus times the percentage K. Finally the sum of the prior period's present value plus or minus the adjustment factor becomes the next period's estimated present value. This iteration continues until the error is eliminated. Thus present value is precisely determined without any foreknowledge of either the award or the tax rates which may apply”. Burke, and Rosen, , “Taxes and Compensation for Lost Earnings: A Comment” (1983) 12 J. Leg. Studies 195, at 198–199CrossRefGoogle Scholar.

The meaning of this determination is not that it is possible to determine present value without knowing what future tax rates will be. The meaning is only that, assuming that certain tax rates prevail, it is possible to overcome the difficulty arising from the mutual effect of the amount of compensation on the marginal tax rate, and vice versa, by means of an iterative process. The final outcome is determination of present value by the system demonstrated by Brady, Brookshire and Cobb.

62 Linkage differentials on linked debentures (whether index-linked or linked to the rate of exchange) issued by the State, the banks or commercial companies, are exempt from tax, whereas the interest is liable to tax at a lower rate (at present 35%). The reduced rate of tax on the interest, is granted by the Minister of Finance, with the approval of the Knesset Finance Committee, under sec. 5 of the Encouragement of Saving (Guarantee of Loans, Income Tax Reduction) Law, 1956 (10 L.S.I. 49), as amended 11 L.S.I. 168. As a result of the reduced tax on the interest, linkage differentials on the debentures are exempt from tax under sec. 9(13) of the Ordinance, which exempts from tax “linkage differentials on preferred loans which are not income under sec. 2(1)”. This result derives from the definition of “preferred loans” in sec. 1 of the Ordinance as “loans or deposits under a savings scheme, the interest on which is wholly or in part exempt from tax under any law, unless otherwise provided in that law, as well as a certificate of participation therein”.

63 See infra pp. 76–96. These tables were prepared by my friend, the economist Mr. Michael Gutman, and I am deeply grateful to him for the considerable effort he put into this.

64 In Khalif v. State of Israel, supra n. 13, at 248-249, Barak J. stated: “In order to overcome this factor [concerning taxation of receipts], the income tax charged on the dependants has to be grossed-up and added to the amount of compensation to which they are entitled. This method was indeed adopted by the House of Lords in England in Taylor v. O'Connor. But if so, what good has been achieved? Of course, it is possible that the tax rate, deducted under the Gourley ruling from the basis for calculation of the compensation, is not identical with the tax rate by which the amount of compensation should be increased under the rule in Taylor. But can such perfection be achieved in estimating damage?”

65 The purpose of the law of torts in Israel is remedial. See supra n. 36.

66 See discussion at p. 42, supra.

67 The proper rate for discounting in Israel is the rate of return prevailing at the time of judgment on State-linked debentures, less the expected rate of increase of real wages and the rate of risk that the return on debentures will be lower in the future (since it is not possible to acquire debentures payable monthly during the period of loss of income, so that debentures have to be realized on the market and new ones acquired). The courts in Israel hold the rate of discounting to be 3%. That rate is based on the assumption of a return of 5.5% per annum on linked debentures, and on the assumption that real wages will increase by 2.5% per annum; Zada v. Behar (1976) 30 (i) P.D. 169, and Sa'ada v. Hamdi (1979) 33(i) P.D. 589. This rate of discounting is excessive, ignoring the fluctuations in the rate of return on debentures and the risk with which victims are consequently confronted.

68 This is the position in many States in the U.S.A. where the system of gross compensation prevails (see supra n. 57), in view of the tax exemption for compensation (whether lump-sum or periodical) per personal injury under sec. 104(a)(2) of the Internal Revenue Code.

69 This is the law in various jurisdictions in the U.S.A. and Canada (see supra n. 57), as well as in Israel (see pp. 63-64, 71 supra), except in respect of compensation under the Road Accident Law and the Defective Products Law which provide for reduced compensation of up to 25% owing to the exemption from tax on compensation awarded in place of income subject to tax. See supra n. 17.

70 Englard, , Road Accident Victim Compensation (Jerusalem, Yahalom, 1978, in Hebrew) 184Google Scholar.

71 Harris, supra n. 18, at 136.

72 See Yoran, , Tax Aspects in Tort Compensation (Jerusalem, The Hebrew University, 1987, in Hebrew) 188–189Google Scholar.

73 See supra n. 62.