LCA 8925/04

1. Solel Boneh Building and Infrastructure Ltd

2. Aryeh Insurance Company Ltd

v.

1. Estate of the late Ahmed Abed Alhamid deceased

2. Abed Alhamid Mudib

3. Hatam Mohammed Halef

4. Engineer Dov Yahalom

5. Noga Insurance Company Ltd

6. Farid Attallah

The Supreme Court sitting as the Court of Civil Appeals

[27 February 2006]

Before President A. Barak, Vice-President Emeritus M. Cheshin and
Justices D. Beinisch, E. Rivlin, A. Grunis, M. Naor, Y. Adiel

Appeal by leave of the judgment of the Haifa District Court (Justice B. Bar-Ziv) on 16 August 2004 in LCA 1494/04.

Facts: Ahmed Alhamid died in a work accident. His estate and dependents (the respondents) filed a claim against the appellants for compensation. During the proceedings, the respondents reached a settlement with the appellants, according to which the appellants would pay a sum of NIS 100,000 to the respondents. This settlement was given the force of a court judgment on 22 February 2004. Three weeks later, the Supreme Court gave its judgment in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter ([2004] IsrLR 101). In that judgment the Supreme Court held that if a person is injured as a result of a tort and his life expectancy is shortened (the ‘lost years’), he is entitled to compensation for the loss of earning capacity in those years. The estate is also entitled to compensation for this head of damage, if the life expectancy of the injured person is shortened and he dies during the tortious act or soon after it. This decision overruled Estate of Sharon Gavriel v. Gavriel, which had been given twenty years earlier, and in which it was held that compensation would not be awarded for the ‘lost years.’

Following the decision in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter, the respondents applied to the trial court to cancel the settlement and to amend their statement of claim. Their application was granted. The appellants’ appeal to the District Court was denied. The appellants applied for leave to appeal to the Supreme Court, and leave to appeal was granted.

The questions before the court were whether the ruling in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter should apply retrospectively to events that occurred before that ruling, and if so, whether the respondents were entitled to cancel the settlement because of the subsequent change in the law.

Held: (President Barak) As a rule, case law has both retrospective and prospective effect. There is no reason why the ruling in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter should not apply retrospectively.

(President Barak) The question whether the respondents may cancel the settlement because of the (retrospective) change in the law should be resolved with reference to the doctrine of mistake in the law of contracts. The respondents’ mistake, however, was only a mistake in the ‘profitability of the transaction.’ Such a mistake is not a ground for cancelling an agreement, and therefore the settlement could not be cancelled.

(Vice-President Cheshin) As a rule, case law has only prospective effect. Retrospective application of case law is the exception to the rule. The plaintiff has the burden of persuading the court that considerations of justice require the relevant case law to have retrospective application. In the present case, considerations of justice supported the retrospective application of Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter.

(Vice-President Cheshin) The respondents did not make any mistake in real time. The question whether the law would change was not one of the risks that the parties took into account when they made the settlement. Consequently there was no basis in the doctrine of mistake for cancelling the settlement.

Appeal allowed.

Legislation cited:

Basic Law: Freedom of Occupation.

Basic Law: Human Dignity and Liberty.

Contracts (General Part) Law, 5737-1973, ss. 14(b), 14(d).

Interpretation Law, 5741-1981, ss. 1, 22.

Unjust Enrichment Law, 5739-1979, s. 2.

Israeli Supreme Court cases cited:

[1] CA 140/00 Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [2004] IsrSC 58(4) 486; [2004] IsrLR 101.

[2] CA 295/81 Estate of Sharon Gavriel v. Gavriel [1982] IsrSC 36(4) 533.

[3] CFH 4011/04 Jerusalem Municipality v. Estate of Ettinger [2005] IsrSC 59(4) 8.

[4] HCJ 716/86 Moriah Spas Hotel, Dead Sea v. Tamar Neveh Zohar District Council [1987] IsrSC 41(2) 389.

[5] LCrimA 1127/93 State of Israel v. Klein [1994] IsrSC 48(3) 485.

[6] CA 6585/95 M.G.A.R. Computerized Collection Centre Ltd v. Nesher Municipality [1996] IsrSC 50(4) 206.

[7] HCJ 3648/97 Stamka v. Minister of Interior [1999] IsrSC 53(2) 728.

[8] RT 8390/01 Axelrod v. State of Israel (not yet reported).

[9] HCJ 221/86 Kanfi v. National Labour Court [1987] IsrSC 41(1) 469.

[10] CA 2000/97 Lindorn v. Karnit Road Accident Victims Compensation Fund [2001] IsrSC 55(1) 12.

[11] HCJ 680/88 Schnitzer v. Chief Military Censor [1988] IsrSC 42(4) 617; IsrSJ 9 77.

[12] HCJ 2722/92 Alamarin v. IDF Commander in Gaza Strip [1992] IsrSC 46(3) 693; [1992-4] IsrLR 1.

[13] CA 2622/01 Director of Land Appreciation Tax v. Levanon [2003] IsrSC 57(5) 309.

[14] HCJ 1113/99 Adalah Legal Centre for Arab Minority Rights in Israel v. Minister of Religious Affairs [2000] IsrSC 54(2) 164.

[15] CA 376/46 Rosenbaum v. Rosenbaum [1948] IsrSC 2 235.

[16] HCJ 9098/01 Ganis v. Ministry of Building and Housing [2005] IsrSC 59(4) 241; [2004] IsrLR 505.

[17] HCJ 19/56 Brandwin v. Governor of Ramla Prison [1956] IsrSC 10 617.

[18] LCA 2413/99 Gispan v. Chief Military Prosecutor [2000] IsrSC 54(4) 673.

[19] CA 180/99 Director of Purchase Tax v. Tempo Beer Industries Ltd [2003] IsrSC 57(3) 625.

[20] CA 3602/97 Income Tax Commission v. Shahar [2002] IsrSC 56(2) 297.

[21] CA 5/84 Yehezkel v. Eliyahu Insurance Co. Ltd [1991] IsrSC 45(3) 374.

[22] LCA 1287/92 Buskila v. Tzemah [1992] IsrSC 46(5) 159.

[23] AAA 1966/02 Majar Local Council v. Ibrahim [2003] IsrSC 57(3) 505.

[24] CA 110/86 Gevaram v. Heirs of the late Shalom Manjam [1988] IsrSC 42(2) 193.

[25] HCJ 6055/95 Tzemah v. Minister of Defence [1999] IsrSC 53(5) 241; [1998-9] IsrLR 635.

[26] HCJ 1715/97 Israel Investment Managers Association v. Minister of Finance [1997] IsrSC 51(4) 367.

[27] CA 2495/95 Ben-Lulu v. Atrash [1997] IsrSC 51(1) 577.

[28] CA 3203/91 Azoulay v. Azoulay (unreported).

[29] CA 4272/91 Barbie v. Barbie [1994] IsrSC 48(4) 689.

[30] CA 2444/90 Aroasty v. Kashi [1994] IsrSC 48(2) 513.

[31] CrimA 4912/91 Talmai v. State of Israel [1994] IsrSC 48(1) 581.

[32] CA 6821/93 United Mizrahi Bank Ltd v. Migdal Cooperative Village [1995] IsrSC 49(4) 221.

[33] HCJ 5843/97 Bar-Gur v. Minister of Defence [1998] IsrSC 52(2) 462.

[34] HCJ 6126/94 Szenes v. Broadcasting Authority [1999] IsrSC 53(3) 817; [1998-9] IsrLR 339.

[35] HCJ 4804/94 Station Film Ltd v. Film and Play Review Board [1996] IsrSC 50(5) 661; [1997] IsrLR 23.

[36] HCJ 606/93 Advancement Promotions and Publishing (1981) Ltd v. Broadcasting Authority [1994] IsrSC 48(2) 1.

[37] CFH 7325/95 Yediot Aharonot Ltd v. Kraus [1998] IsrSC 52(3) 1.

[38] LCA 6339/97 Roker v. Salomon [2001] IsrSC 55(1) 199.

[39] HCJ 57/67 Gross v. Income Tax Commissioner [1967] IsrSC 21(1) 558.

[40] HCJ 4157/98 Tzevet, Association of Retired IDF Servicemen v. Minister of Finance [2004] IsrSC 58(2) 769.

[41] CA 8972/00 Schlesinger v. Phoenix Insurance Company Ltd [2003] IsrSC 57(4) 817.

[42] CA 1761/02 Antiquities Authority v. Station Enterprises Ltd (not yet reported).

American cases cited:

[43] Kuhn v. Fairmont Coal Co., 215 U.S. 349 (1910).

[44] Great Northern Railway Co. v. Sunburst Oil & Refining Co., 287 U.S. 358 (1932).

[45] Linkletter v. Walker, 381 U.S. 618 (1965).

[46] Chevron Oil Co. v. Huson, 404 U.S. 97 (1971)

[47] United States v. Johnson, 457 U.S. 537 (1982).

[48] Griffith v. Kentucky, 479 U.S. 314 (1987).

[49] James B. Beam Distilling Co. v. Georgia, 501 U.S. 529 (1991).

[50] Harper v. Virginia Dept. of Taxation, 509 U.S. 86 (1993).

English cases cited:

[51] National Westminster Bank plc v. Spectrum Plus Ltd [2005] UKHL 41; [2005] 4 All ER 209.

European Court of Human Rights cases cited:

[52] Marckx v. Belgium (1979) 2 E.H.R.R. 330.

European Court of Justice cases cited:

[53] Defrenne v. Sabena [1976] E.C.R. 455.

[54] Deutsche Telekom A.G. v. Vick, Conze and Schroder [2000] I.R.L.R. 353.

Indian cases cited:

[55] Golak Nath v. State of Punjab [1967] 2 S.C.R. 762.

[56] India Cement Ltd v. State of Tamil Nadu [1990] 1 S.C.C. 12.

[57] Orissa Cement Ltd v. State of Orissa [1991] Supp. (1) S.C.C. 430.

Jewish law sources cited:

[58] Babylonian Talmud, Rosh HaShana 25b.

[59] Babylonian Talmud, Bava Batra 21a.

For the appellants — J. Asulin.

For respondents 1-3 — G. Tannous, R. Tannous.

For the fourth respondent — T. Tenzer.

For the fifth respondent — Z. Rapaport.

For the sixth respondent — D. Attallah.

JUDGMENT

President A. Barak

The Supreme Court decided that a person who is injured as a result of a tort and whose life expectancy is shortened is entitled to compensation for the loss of earning capacity in the years by which his working life expectancy was shortened. His estate is also entitled to compensation for this head of damage, if the life expectancy of the injured person is shortened and he dies during the tortious act or soon after it. This is the ‘lost years’ rule. It was decided in CA 140/00 Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1]. In that case the Supreme Court departed from a case law ruling that had been decided twenty years earlier in CA 295/81 Estate of Sharon Gavriel v. Gavriel [2]. When the judgment was given in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] there was a large number of claims concerning compensation for loss of earning capacity pending in various courts. What effect does the new ruling have on those cases? That is the general question that arises before us. The specific question is what effect does Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] have on a settlement concerning the quantum of damages between an injured person and several tortfeasors that was given the validity of a partial judgment, while the action continued with regard to the relationship between the tortfeasors inter se.

The facts and the proceedings

1. The deceased Ahmed Alhamid Mudib Abu Sahon was killed in a work accident. An action was filed with regard to his death by his estate and his dependents against the employer, the owner of the site where he worked and the insurers. In the course of the proceedings, the parties, at the recommendation of the court, reached a settlement. According to this, the plaintiffs would be paid a sum of NIS 100,000. The trial would continue with regard to division of the liability between the parties. On 22 February 2004, this settlement — which was called in the court’s decision a ‘procedural arrangement’ — was given the force of a court decision.

2. On 15 March 2004, judgment was given in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1]. The application for a further hearing was denied (CFH 4011/04 Jerusalem Municipality v. Estate of Ettinger [3]). In consequence, on 5 April 2004 the plaintiffs filed an application to cancel the procedural settlement and to amend the statement of claim. The defendants opposed this. The Magistrates Court (Justice I. Ganon) granted the application. He held that his decision (of 22 February 2004) amounted to a ‘procedural arrangement,’ and was not a ‘partial judgment.’ It was not proper or just to prevent the plaintiffs from cancelling the settlement. The defendants appealed to the District Court. The appeal was denied. It was held (per Justice B. Bar-Ziv) that the decision of the Magistrates Court amounted to a partial judgment. For reasons of justice — and according to case law — it was possible to repudiate this partial judgment. The defendants applied to this court for leave to appeal. We granted the defendants’ application and gave leave to appeal. In view of the importance of the questions that arise before us the panel was expanded.

The questions that require a decision

3. The appeal before us raises two main questions. First, does Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] apply prospectively only (from now onwards) or does it also have retrospective effect (changing the position in the past)? If Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] has no retrospective effect, it does not apply to the accident in this case, and therefore there is no argument that allows the agreement between the parties to be repudiated. But if Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] does have retrospective effect, the second question arises: this concerns the effect that Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] has on the agreement between the parties. Let us turn to consider the first question.

A. The temporal application of Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter

Time and law

4. Does the ruling in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] apply to tortious acts that took place before it was decided? Does it have retrospective effect? The answer to this question lies in the status of time in the law. Indeed, every legal norm applies not only in space but also in time. Against this background, we should consider a wide variety of problems in which time, at the heart of the law, is a common factor. One group of problems concerns laws that apply when the law changes at a certain point in time. These are the problems of intertemporal law (droit transitoire). Within this framework, the question of the retrospective, effective or prospective application of the new law plays a central role (see A. Rodger, ‘A Time For Every Thing Under The Law: Some Reflections On Retrospectivity,’ 121 L. Q. R. 57 (2005); R.H.S. Tur, ‘Time and Law,’ 22 Oxford J. L. Stud. 463 (2002); see also A. Barak, Legal Interpretation (vol. 2, 1993), at p. 609). This is the case with regard to the temporal application of new legislation; it is also the case with regard to the temporal application of new case law — whether this overrules previous case law or whether it determines a new case law ruling. In all of these, the question of the temporal application of the new norm arises. We shall focus on the solution to this question in a case where a new judicial ruling gives a new interpretation to a statute by overruling a previous interpretation. What is the temporal application of the new case law ruling? Does it apply both from this moment onward (prospectively) and also to earlier events (retrospectively)? Or does it perhaps apply only from this moment onward (purely prospectively)? If the latter, what is the law with regard to the case in which the new law is decided: does the new law apply to it (a kind of general prospectivity and a specific retrospectivity)? And does it apply also to all the other cases that are being litigated before the courts? This is not a new question in Israel. There is academic discussion of it in Israel (see G. Tedeschi, ‘Case Law for the Future,’ Essays in Law 25 (1978); E. Kaplan, ‘Prospective Application of Supreme Court Precedents,’ 9 Hebrew Univ. L. Rev. (Mishpatim) 221 (1979); A. Barak, Judicial Discretion (1987), at p. 417; E. Kaplan, ‘Future Application of Supreme Court Precedents,’ Avner Hai Shaki Book 125 (2005)). It arose in the past in several judgments, and several obiter statements have been made on this subject (see HCJ 716/86 Moriah Spas Hotel, Dead Sea v. Tamar Neveh Zohar District Council [4], at p. 392; LCrimA 1127/93 State of Israel v. Klein [5], at p. 504; CA 6585/95 M.G.A.R. Computerized Collection Centre Ltd v. Nesher Municipality [6], at p. 220; HCJ 3648/97 Stamka v. Minister of Interior [7]; RT 8390/01 Axelrod v. State of Israel [8]).

The premise: retrospective and prospective application

5. The fundamental premise is that a new judicial ruling acts both retrospectively and prospectively (see HCJ 221/86 Kanfi v. National Labour Court [9], at p. 480). Justice Holmes rightly said that ‘Judicial decisions have had retrospective operation for near a thousand years’ (in Kuhn v. Fairmont Coal Co. [43], at p. 372). This is the position with regard to the development of the law within the framework of the common law, and it is also the position where case law interprets a legislative provision (a constitution, statute, regulation), or fills a lacuna in it (for the distinction between these, see A. Barak, ‘The Different Kinds of Judicial Creation: Interpretation, Filling a Lacuna and Development of the Law,’ 39 HaPraklit 267 (1990); A. Barak, Selected Articles (H.H. Cohn and I. Zamir eds., vol. 1, 2000), at p. 755). There are three arguments that support this approach (see Barak, Judicial Discretion, at p. 421): a jurisprudential argument, a constitutional argument and a practical argument.

The jurisprudential argument

6. The jurisprudential argument is the following: since the court decides the law — whether within the framework of the common law or within the framework of interpreting legislation or filling a lacuna therein — it declares the law. It does not create it. When the court departs from a previous judgment, it is deciding that the erroneous judgment never was the law. The overruling judgment does not create new law. It declares what the law always was. This is the declarative theory of law. It was developed by Blackstone. His well known statement was that:

‘... if it be found that the former decision is manifestly absurd or unjust, it is declared, not that such a sentence was bad law, but that it was not law’ (1 Blackstone, Commentaries 71 (1769)).

The declarative theory of law leads to the conclusion that a judgment that overrules a previous judgment acts retrospectively. If the overruled judgment was never law, and the law was never as declared in the overruled judgment, this means that the judgment that overruled it acts temporally in a retroactive manner. An additional jurisprudential argument is this: when a change in case law is merely prospective and it does not act in favour of the parties in the trial (pure prospectivity), the new case law is an obiter dictum, and it is not binding at all.

The constitutional argument

7. The constitutional argument that supports the retrospective application of new case law is this: a central element in any democratic constitution is the separation of powers. According to this, the legislative branch enacts statutes, and the judicial branch decides disputes. In enacting a statute, the legislative branch is competent to determine its temporal application. This determination will usually be prospective, for constitutional and other considerations. If the judgment can also determine a prospective application of the case law ruling, it will be indistinguishable from legislation. This was well expressed by Lord Devlin, when he said that if new case law has only prospective application, then it —

‘... crosses the Rubicon that divides the judicial and the legislative powers. It turns judges into undisguised legislators’ (P. Devlin, ‘Judges and Lawmakers,’ 39 M. L. R. 1 (1976), at p. 11).

Preserving the proper separation between the legislative and judicial functions leads to a recognition that the application of legislation is only prospective, but the application of case law is otherwise. A merely prospective change in case law makes the judge into a legislator (M.D.A. Freeman, ‘Standards of Adjudication, Judicial Law, Making and Prospective Overruling,’ 26 Curr. L. P. 166 (1973), at p. 204). In addition to this constitutional consideration of the separation of powers, there is an additional constitutional consideration. As we shall see, various prospective approaches distinguish between the litigant who asked the court to overrule the previous case law, to whom the new case law ruling will apply retroactively, and other litigants, whose cases are being considered before the courts and have not yet been decided, to whom the new case law ruling will not apply. This creates a forbidden discrimination that violates the principle of equality. In America there is an additional constitutional consideration that operates against a merely prospective overruling of the previous case law, and this concerns the constitutional requirement that the courts may only decide ‘cases’ and ‘controversies.’ When the new case law is given only a purely prospective force, that new case law ruling does not decide the dispute before the court; it constitutes an advisory opinion with regard to that case, and it is therefore prohibited.

The practical argument

8. In addition to the jurisprudential and the constitutional arguments, it is possible to find support for the retrospective application of new case law in several practical arguments. First, it is argued that the ability to give only prospective validity to a new case law ruling that overrules its predecessor releases the judge from the constraints that limit his discretion as to whether to depart from a previous case law ruling or not. According to this approach, the retrospectivity of the case law ruling acts as a barrier against too great a departure from the previous law. When this barrier is removed, there is a fear that the proper framework may be undermined, and that there will be too many departures from previous case law rulings (see J. Stone, Social Dimensions of Law and Justice (1966), at p. 663; P. Mishkin, ‘Foreword: The High Court, The Great Writ, And The Due Process of Time and Law,’ 79 Harv. L. Rev. 56 (1965), at p. 70). Second, there are several systems of merely prospective changes in case law (see Barak, Judicial Discretion, at p. 420, and G. Calabresi, A Common Law for the Age of Statutes (1982), at p. 280). Choosing between these systems is complex. The litigants will usually not know which system the court will choose. As a result, the whole judicial process is undermined. Third, if we choose from among the different systems the one that advocates a purely prospective overruling of previous case law — according to which the new case law does not apply even to the litigant who was successful in his argument that the previous case law should be changed — this will reduce the motivation of litigants to argue that the case law should be changed, since in any case they will not benefit from the change. This is a negative consequence that will lead to stagnation in the development of case law (see R. Dworkin, Law’s Empire (1986), at p. 156). Fourth, often a mere prospective application of the new judicial ruling undermines public expectations of the judiciary. This leads to a loss of public confidence in the judiciary, which should be protected at all costs (see A. Barak, A Judge in a Democracy (2004), at p. 49).

Criticism of the jurisprudential argument

9. The jurisprudential argument is not convincing. Admittedly, often a judgment only declares the law and does not create it. Similarly, sometimes a previous judgment is absolutely wrong, and it should be overruled retroactively. All of this is correct sometimes, but not always. Sometime the new judgment does create new law, which is appropriate for its time and place. The previous law — which the new judgment overruled — was not absolutely wrong. It may be that it was correct and proper in its time, but now the time has come to change it. In these circumstances, there is no jurisprudential reason not to give the new case law ruling only a prospective application. Take a law that was interpreted in the past in a certain way, and now the court departs from that interpretation and adopts a new interpretation. This overruling is not always based on an original error in the first judgment. It is based on the current needs and values of society. Indeed, interpretation of statutes is dynamic (see A. Barak, Purposive Interpretation in Law (2003), at pp. 200, 412; see also R. Eskridge, Dynamic Statutory Interpretation (1994)). ‘Yet their words remain law’ (see F.A.R. Bennion, Statutory Interpretation: A Code (third edition, 1997), at p. 687). I discussed this in one case, where I said:

‘The statute integrates into the new reality. Thus an old statute speaks to modern man... Interpretation is an ever-changing process. Modern content should be given to the old language. Thus the disparity between the statute and life is reduced. Against this background it is correct to say, as Radbruch did, that the interpreter may understand the statute better than the creator of the statute, and that the statute is always wiser than its creator. This leads to the interpretive approach that is accepted in England, whereby statute should be given an updating interpretation... Indeed, the statute is a living creature. Its interpretation should be dynamic. It should be understood in a way that is consistent with and advantageous to modern reality’ (CA 2000/97 Lindorn v. Karnit Road Accident Victims Compensation Fund [10], at p. 32. See also HCJ 680/88 Schnitzer v. Chief Military Censor [11], at p. 629 {90}; HCJ 2722/92 Alamarin v. IDF Commander in Gaza Strip [12], at p. 705 {16-17}; CA 2622/01 Director of Land Appreciation Tax v. Levanon [13]).

The same is true of the interpretation of constitutions and Basic Laws. These are living documents. A modern meaning should be given to the values enshrined in them. A similar approach applies to the development of common law. Since its inception, it has undergone wide-ranging changes that have created new case law principles that are suited to the needs of the time and place. This was discussed by Lord Nicholls, who said:

‘... judges themselves have a legitimate law-making function. It is a function they have long exercised. In common law countries much of the basic law is still the common law. The common law is judge-made law. For centuries judges have been charged with the responsibility of keeping this law abreast of current social conditions and expectations. That is still the position. Continuing but limited development of the common law in this fashion is an integral part of the constitutional function of the judiciary. Had the judges not discharged this responsibility the common law would be the same now as it was in the reign of King Henry II’ (National Westminster Bank plc v. Spectrum Plus Ltd [51], at para. 32).

In situations where the change in the common law is intended to bridge a gap between the law and life, the old precedent is overruled not because it was originally wrong, but because it is unsuited to the new reality. The declarative theory does not give any proper answer to this situation. Naturally, it is always possible to say that changes sprout forth from the fertile soil of the common law, and that the judge brings out the potential latent in it from theory into practice. Even if this is the case, it involves judicial creation. Just as a new statute, which brings out from theory into practice what is latent in the constitution, constitutes a new creation, so too does a new judicial ruling that springs forth from the soil of the law constitute a new creation. Indeed, the declarative theory is incapable of explaining the entirety of judicial activity. It has passed its time. It is based on a fiction that should not be recognized (see Axelrod v. State of Israel [8], at para. 10). It is to be hoped that though we may have buried it, it will not rule us from its grave (in the words of Maitland on the forms of action: see F.W. Maitland, The Forms of Action of Common Law (1941), at p. 2).

Criticism of the constitutional argument

10. The constitutional argument against the merely prospective application of a judicial ruling is also not convincing. The reason for this is that in the course of deciding a dispute, the court is obliged to determine the law according to which the dispute will be decided. Sometimes this decision is merely a declaration of what already exists. Sometimes this decision creates a new law, whether within the framework of the common law or by means of interpretation or filling a lacuna in legislation. Creating this law constitutes ‘judicial legislation’ (see A. Barak, ‘Judicial Legislation,’ 13 Hebrew Univ. L. Rev. (Mishpatim) 25 (1983); Barak, Selected Articles, at p. 821). This is not ‘legislation’ in the institutional sense. That is solely within the jurisdiction of the legislature. This is ‘legislation’ in the functional sense, since it creates a norm that did not exist in the past. This functional legislation does not violate the principle of the separation of powers. Recognizing it does not blur the boundary between legislation (in its institutional sense) and the administration of justice. Giving only a prospective force to a new case law ruling is an expression of judicial creation. It involves no crossing of the Rubicon that divides legislation from the administration of justice (see the judgment of Justice Cardozo in Great Northern Railway Co. v. Sunburst Oil & Refining Co. [44], at p. 366). With regard to the argument that prospective application of a new case law ruling violates equality, this is true only if we do not adopt the system of the purely prospective change. With the purely prospective system, there is no violation of equality. And as for the other systems, even though they involve a violation of equality, we need to consider whether this violation is a proper one. Equality is not an absolute right. It can be violated for proper purposes by means of proportionate measures (see HCJ 1113/99 Adalah Legal Centre for Arab Minority Rights in Israel v. Minister of Religious Affairs [14]).

The decisive consideration — the practical consideration

11. I have therefore reached the conclusion that the jurisprudential and constitutional arguments are incapable of preventing the court from departing from its previous path in giving the new case law ruling retrospective application. Prospective application, in its various forms, is consistent with the jurisprudential and constitutional status of judicial activity (see P.J. Fitzgerald, Salmond on Jurisprudence (twelfth edition, 1966), at p. 127; K. Diplock, The Courts as Legislators (1965), at p. 17). Indeed, the decision as to whether the court should deprive its new case law ruling, in appropriate circumstances, of retrospective application will not be decided by jurisprudential or constitutional considerations. It will be decided by the proper balance between practical considerations. This was well expressed by Justice Cardozo, when he said that the question of the retrospectivity of a case law ruling —

‘… will be governed, not by metaphysical conceptions of the nature of judge-made law, nor by the fetich of some implacable tenet, such as that of the division of governmental powers, but by considerations of convenience, of utility, and of the deepest sentiments of justice’ (B.N. Cardozo, The Nature of The Judicial Process (1921), at p. 148).

We have mentioned several practical considerations that support the retrospective application of the new judicial case law ruling. What are the practical considerations that support the other approach, that it is possible to deny the retrospective application of a new case law ruling? Which considerations have the upper hand? Let us now turn to consider these questions.

12. Rejecting retrospective application and recognizing only prospective application (in one of its forms) is supported by several practical considerations: first, the need to reject the retrospective application of a new case law ruling arises usually when the court examines the previous case law rule and comes to the conclusion that it ought to overrule it. Notwithstanding, the court is concerned about the damage that overruling it will cause those persons and bodies who have relied on the previous case law rule, and who have regulated their relationships on the basis of this reliance. In such a situation, the court faces the following dilemma: either it must leave an undesirable case law rule as it stands because of the reliance interest, or it must change case law and determine a new and better case law rule in its place, even though this harms the reliance interest). The approach that a change in case law should not be retrospective and should act only prospectively extricates the judge from the dilemma in which he finds himself. It allows him to make a change to an erroneous case law rule and to establish a new case law rule in its place, without harming the reliance interest. Thus security and stability are maintained in addition to adapting the law to social change. We have before us a kind of ‘wonder remedy’ that allows both stability and progress (see Barak, Judicial Discretion, at p. 421, and R.J. Traynor, ‘Quo Vadis Prospective Overruling: A Question of Judicial Responsibility,’ 28 Hastings L. J. 533 (1977), at p. 542).

13. Second, the truth is that several systems of prospective application are recognized (see para. 8, supra). This multiplicity does not lead to complexity or confusion. Within a short time it can be determined in what conditions one prospective system will be adopted and when the court will adopt another system. The ‘supply’ of prospective systems is not large, and it is possible without difficulty to choose the appropriate law in this regard.

14. Third, a merely prospective application of a new case law ruling is consistent with the sense of justice. It allows a new and just ruling to be made, without harming the reliance interest. It averts the need to made a decision — such as the one that President Zamora made with regard to the question of precedents — that ‘between truth and stability — truth prevails’ (CA 376/46 Rosenbaum v. Rosenbaum [15], at p. 254). It makes it possible to achieve both ‘truth’ and ‘stability.’ Thereby it increases confidence in the judicial system. This confidence will be harmed if a proper change does not take place because of the reliance interest, or if the change does take place and harms the reliance interest.

15. The practical considerations lead to conflicting conclusions. How can we decide between or balance the conflicting considerations? It should be stated immediately that every legal system has decisions and balancing points of its own. This is a product of the strength of the jurisprudential and constitutional considerations in that legal system. The decision is also affected by the way in which the society understands the judicial role, and its willingness to examine realistic arrangements and practical balances. All of these vary from one legal system to another. They also vary over time within the framework of the same legal system. A good example of this can be found in American law. There the courts of the various states first recognized the prospective overruling of case law as long ago as the nineteenth century (see T.S. Currier, ‘Time and Change in Judge-Made Law: Prospective Overruling,’ 51 Va. L. Rev. 201 (1965)). If found recognition in the Federal courts in the 1960s and the beginning of the 1970s in the judgments in Linkletter v. Walker [45]; Chevron Oil Co. v. Huson [46]). Since the 1980s there has been a significant retreat in this sphere. Today the case law of the United States Supreme Court rejects a merely prospective application of new case law rulings (see United States v. Johnson [47]; Griffith v. Kentucky [48]; James B. Beam Distilling Co. v. Georgia [49]; R.H. Fallon and D.J. Meltzer, ‘New Law, Non-Retroactivity, and Constitutional Remedies,’ 104 Harv. L. Rev. 1731 (1991); Harper v. Virginia Dept. of Taxation [50]; J.E. Fisch, ‘Retroactivity and Legal Change: An Equilibrium Approach,’ 110 Harv. L. Rev. 1055 (1997); B.S. Shannon, ‘The Retroactive and Prospective Application of Judicial Decisions,’ 26 Harv. J. L. & Pub. Pol’y 811 (2003); M. Katz, ‘Plainly Not “Error”: Adjudicative Retroactivity on Direct Review,’ 25 Cardozo L. Rev. 1979 (2004)). The original American approach — the one that recognized the possibility of changing case law prospectively — is accepted in India (see Golak Nath v. State of Punjab [55]; India Cement Ltd v. State of Tamil Nadu [56]; Orissa Cement Ltd v. State of Orissa [57]). The original American approach has also been applied by the European Court of Justice in Luxembourg (see Defrenne v. Sabena [53]; Deutsche Telekom A.G. v. Vick, Conze and Schroder [54]), and the European Court of Human Rights in Strasbourg (see Marckx v. Belgium [52], at p. 353).

16. English law wavered for a long time over the question of the prospective application of new case law (for an analysis of the various positions, see the opinion of Lord Nicholls in National Westminster Bank plc v. Spectrum Plus Ltd [51]). In that case, which was decided only a few months ago, it was held, by a majority, that the question whether to adopt only a prospective overruling of previous case law was within the discretion of the court (ibid. [51], at para. 39). It was held that there might be circumstances in which the court would adopt this approach. Lord Nicholls wrote:

‘... there could be circumstances in this country where prospective overruling would be necessary to serve the underlying objective of the courts of this country: to administer justice fairly and in accordance with the law. There could be cases where a decision on an issue of law, whether common law or statute law, was unavoidable but the decision would have such gravely unfair and disruptive consequences for past transactions or happenings that this House would be compelled to depart from the normal principles relating to the retrospective and prospective effect of court decisions’ (ibid. [51], at para. 40).

In that case it was decided to give the new case law ruling retrospective application, since the conditions for prospective application only were not fulfilled.

17. What is the law in Israel? The fundamental premise is that a new judicial ruling applies both retrospectively and prospectively. Notwithstanding, I am of the opinion that there is nothing in principle that prevents us from recognizing the power of the Supreme Court to give its precedents merely prospective force. The declaratory theory of law has not acquired great strength in Israel; there is no constitutional obstacle that prevents recognizing this possibility. The legal community in Israel would not regard this as judicial activity that is inconsistent with the character of the judicial system. The possibility of adopting this approach was raised in several judgments (see para. 4, supra) and it seems to me that Israeli law is ready to absorb it. Therefore the question is not whether we should recognize this possibility in principle. The answer to this is yes. The question is on what conditions and in what circumstances should we adopt this approach. I shall now turn to examine this question.

Protection of the reliance interest

18. What supports the need to resort solely to a prospective overruling of old case law by a new judicial decision is the reliance interest of individuals and (private and government) bodies who have managed their affairs on the basis of the old judicial ruling. ‘The interest of reliance is like a golden thread that runs through Israeli law’ (HCJ 9098/01 Ganis v. Ministry of Building and Housing [16], at para. 19). Indeed, the reliance interest is one of the most protected interests in the law. This is the position in the sphere of administrative law (see D. Barak-Erez, ‘The Protection of Reliance in Administrative Law,’ 27 Hebrew Univ. L. Rev. (Mishpatim) 17 (1996)). The same is true of private law (see D. Friedman and N. Cohen, Contracts, at p. 151; G. Shalev, The Law of Contracts — General Part: Towards a Codification of Civil Law (2005), at p. 247; see also L.L. Fuller and W.R. Perdue, ‘The Reliance Interest in Contract Damages,’ 46 Yale L. J. 52 and 373 (1936-1937)). The rule of binding precedent is also based, in part, on the protection of the reliance interest (see Barak, Judicial Discretion, at p. 441). The outlook concerning a solely prospective application of a case law ruling that changes the previous law is also derived from the need to protect the reliance interest. Indeed, a retrospective change of the existing law may seriously harm someone who relied on it, to such an extent that it may prevent the change in the law. It follows that the examination of this issue should focus mainly on the reliance interest (see Stamka v. Minister of Interior [7], at p. 746; see also P.J. Stephens, ‘The New Retroactivity Doctrine: Equality, Reliance and Stare Decisis,’ 48 Syracuse L. Rev. 1515 (1998)). Therefore, if the issue is new and has never been decided in the past, it cannot be said that there is a reliance interest that is worthy of protection. The same is true if the old case law ruling did not in practice create any real reliance, or if the reliance was unreasonable, or if it should not be given any significant weight in view of the issue under discussion and the nature of that reliance. In all of these cases, and in others, we should not give much weight to the reliance factor, and there is a basis for applying the new case law ruling retrospectively (see W.V. Schaefer, ‘The Control of “Sunbursts”: Techniques of Prospective Overruling”, 42 N. Y. U. L. Rev. 631, (1967), at p. 638). Examples of this situation can be found in the following situations: the previous case law was not a decision of the Supreme Court; the previous case law was unclear, and it has been interpreted in different ways; the previous case law was accompanied by opposition and proposed changes; in several obiter statements judges have expressed reservations concerning the previous case law; the previous case law was not known to the parties; the parties relied on the old law but each took the risks that it might be changed (see: Note, ‘Prospective Overruling and Retroactive Application in the Federal Courts,’ 71 Yale L. J. 907 (1962)). In these situations and in many others, anyone who relies on the previous case law takes a risk and it is therefore possible to give the new case law retrospective validity. Indeed, in many cases the change in case law is not a surprise. It does not come — in the language of Lord Devlin — ‘out of a blue sky’ (Devlin, ‘Judges and Lawmakers,’ supra, at p. 10). Justice Cardozo rightly said that:

‘The picture of a bewildered litigant lured into a course of action by the false light of decision, only to meet ruin when the light is extinguished and the decision overruled, is for the most part a figment of excited brains’ (B.N. Cardozo, The Growth of The Law (1924), at p. 122).

See also Schaefer, ‘Precedent and Policy,’ 34 U. Chi. L. Rev. 3 (1966), at p. 15.

Indeed cases of reliance that justify giving only prospective force to a new case law ruling are, in the natural course of events, not many (see Traynor, ‘Quo Vadis Prospective Overruling: A Question of Judicial Responsibility,’ supra, at p. 542).

19. The existence of a reliance interest and a violation thereof are essential conditions for a merely prospective application of a new case law ruling. But they are not sufficient conditions. The court should go on to examine whether it may be possible to overcome the reliance problem without adopting a solely prospective application of the new case law. Indeed, the reliance interest is given broad protection by the law. In those cases where general laws protect the reliance interest, there is no basis for giving the interest any additional protection in the form of prospective application. An example of this is the doctrine of the de facto civil servant. According to this doctrine, when a civil servant has acted in a situation where he believed in good faith that he was acting by virtue of legislation that the court declared to be unconstitutional or unlawful, the acts that he carried out during the period of the illegality should be regarded as valid (see HCJ 19/56 Brandwin v. Governor of Ramla Prison [17], at p. 630). In this situation, validity will be given to the reliance interest by means of the doctrine of the de facto civil servant so that it is not necessary to declare the legislation invalid prospectively. We have before us an example of a broader approach, which concerns relative voidance (see LCA 2413/99 Gispan v. Chief Military Prosecutor [18], at p. 684; D. Barak-Erez, ‘Relative Voidance in Administrative Law: On the Price of Rights,’ Itzchak Zamir Book: On Law, Government and Society 283 (Y. Dotan and A. Bendor eds., 2005)). This doctrine distinguishes between a violation of the law and the relief for the violation. Within the framework of the relief, it is possible to take the principle of reliance into account.

20. Another example can be found in a case where tax was paid by virtue of legislation that was set aside because it was contrary to a Basic Law or to a statute. A restitution of the taxes that were collected naturally harms the reliance interest of the government body that collected the tax. Protection for this interest can be found in the argument that the government body is entitled to rely on the general protection given by the laws of unjust enrichment with regard to unfair restitution (s. 2 of the Unjust Enrichment Law, 5739-1979). To the extent that this protection is available to the government body, this is capable of solving the reliance problem, without it being necessary to determine that the decision concerning the unconstitutionality or the illegality of the tax does not act retrospectively. Indeed, the application of this protection to the restitution of tax payments varies from one legal system to another. In our legal system, no ruling has yet been made in this regard. It has been left undecided on several occasions and in this appeal we shall also not adopt any position on this issue (see CA 180/99 Director of Purchase Tax v. Tempo Beer Industries Ltd [19], at p. 644; CA 3602/97 Income Tax Commission v. Shahar [20], at p. 337).

21. In these examples and in many others, there is no basis for resorting to a solely prospective overruling of previous case law in order to protect the reliance interest, since other legal doctrines are capable of giving sufficient protection to this interest. Naturally, we should examine in each case whether the protection of the reliance interest, which these other doctrines provide, is comparable with the protection that the reliance interest would have been given by virtue of a solely prospective overruling of the previous case law. Sometimes the two are not interchangeable: sometimes the cost of resorting to general doctrines is so great — whether from the viewpoint of the parties concerned or from the viewpoint of the courts — that it is better to give the new case law solely prospective validity.

22. Finally, sometimes there will be a basis for giving retrospective validity to new case law even if this harms the reliance interest. It is well known that this interest does not have absolute force. It should be balanced against the values and the principles that conflict with it. Sometimes the court may think that the considerations that support a change of the law are of greater weight than the considerations that support the old law, and the damage that is caused to the reliance interest by the actual change (see CA 5/84 Yehezkel v. Eliyahu Insurance Co. Ltd [21], at p. 384; LCA 1287/92 Buskila v. Tzemah [22], at p. 172; AAA 1966/02 Majar Local Council v. Ibrahim [23]). Indeed, the determination of the question whether to give a new case law ruling solely prospective validity should take into account all of the considerations relevant to the case; the judge should balance these, by giving weight to the conflicting considerations, in the circumstances of the case before him. In all of this, the fundamental premise is the retrospective and prospective validity of the new case law.

The reliance interest and the law of torts

23. What weight should be given to the reliance interest when case law is changed in the field of the law of torts? In order to answer this question, we should examine each issue on its merits. We should examine to what extent the parties relied on the old case law, and to what extent this reliance is harmed by changing that law. The accepted view in legal literature is that, as a rule, reliance in the field of the law of torts is minimal. This was discussed by Justice Traynor, who said:

‘... neither the tortfeasor nor the victim nurses any reasonable expectations about injury that has yet to occur. When everyone’s daily life is prone to risk, it is hardly realistic to suppose that people are assiduously studying current rules of liability so that they may set out to hit or be hit advantageously’ (R.J. Traynor, ‘The Limits of Judicial Creativity,’ 29 Hastings L. J. 1025 (1978), at p. 1036; see also Traynor, ‘Quo Vadis Prospective Overruling: A Question of Judicial Responsibility,’ supra, at p. 545).

Notwithstanding, even in the field of the law of torts, there is a basis for taking the interest reliance into account. This is especially the case with regard to imposing new obligations that were not recognized in the past. It was precisely in the field of the law of torts that the courts in America first recognized the possibility of a merely prospective overruling of previous case law. They did this in the past in those cases in which the old law did not recognize liability in torts (such as the case law ruling that held that hospitals are immune from liability in tort), whereas the new law recognized liability (by cancelling the immunity). The courts decided that the new case law would only have prospective application, since the hospitals had not insured themselves in reliance on the old law (see Currier, ‘Time and Change in Judge-Made Law: Prospective Overruling,’ supra). Naturally, these considerations do not apply where there is insurance. As a rule, significant weight should not be given to an argument that the scope of the old case law ruling determined the amount of the insurance premiums (R. Keeton, Venturing to Do Justice (1969) 42). There are many different considerations according to which insurance premiums are determined, and the extent of liability under case law is only one of them. In any case, the power of insurance companies to ‘spread the loss’ among all of its insureds reduces their reliance interest. There may, of course, be exceptional cases in which the amount of the compensation has a decisive effect on the insurance, but this is not usually the case.

Should the rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter be merely prospective?

24. Against the background of all the considerations that we have discussed, I am of the opinion that there is no real reason why we should not give Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] retrospective force. It will therefore apply both retrospectively and prospectively. It will apply to every tortious act that occurred before it and after it. This is the fundamental premise and there is no reason to depart from it in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1]. With regard to this case, the main reason underlying my approach is that a retrospective application of Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] will not harm the reliance interest to any great degree. Potential tortfeasors and injured parties did not rely on Estate of Sharon Gavriel v. Gavriel [2], which was overruled by Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1], in determining how they would conduct themselves. Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] did not impose new obligations; it only affected the quantum of damages, and here too its effect is not significant. Moreover, the liability of the tortfeasor is usually covered by insurance. Even if the insurance company relied in some way or another on Estate of Sharon Gavriel v. Gavriel [2] in determining the premium, it is capable of absorbing the additional payments for which it will be liable under Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1]. In his discussion of the weight of the reliance interest in the law of torts where there is insurance, Keeton says that the need to protect the reliance interest in this situation is small, since the harm to the reliance interest of a specific insurer or a specific insured is less serious. Keeton also says that as a rule it is difficult to determine the effect of a legal doctrine on the amount of the premium (see Keeton, Venturing to Do Justice, at p. 42). In any case, no figures were brought before us to show that this approach does not apply with regard to the ‘lost years.’ The burden in this regard lies with the party that argues for a merely prospective application of the new case law.

25. Moreover, Estate of Sharon Gavriel v. Gavriel [2] has passed its time. In England, Canada, Australia and the United States the approach that was expressed in Estate of Sharon Gavriel v. Gavriel [2] has not been accepted. In several countries express statutory provisions have been enacted in this regard (see Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1], at pp. 528 {143} et seq.). It has been criticized in case law (see the opinion of Justice H. Ariel in CA 110/86 Gevaram v. Heirs of the late Shalom Manjam [24], at p. 199). Criticism was also levelled at it in Israeli professional literature (see D. Katzir, Compensation for Personal Injury (fifth edition, 2003), at p. 381; A. Porat, ‘The Law of Torts,’ Israel Law Year Book 1991, 221 (A. Rosen-Zvi, 1991), at p. 250). In the draft civil codex, the Civil Law (2004), it was proposed that it should be abandoned (see section 544). Justice Rivlin in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] pointed to ‘a change in the legal climate,’ which led to the need to change Estate of Sharon Gavriel v. Gavriel [2] (see Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1], at p. 559 {177}). A change in this climate naturally affects the actual reliance, its strength and its reasonableness. Against this background, it would appear that the weight of the reliance interest of insurers on Estate of Sharon Gavriel v. Gavriel [2] does not justify giving only prospective force to Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1]. In any case, we do not have any reason to assume that the financial burden that Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] is likely to impose on insurance companies falls outside the scope of the professional risks for which insurance companies should be liable.

26. In so far as the retrospective operation of Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] harms the reliance interest, we should turn to the general laws that protect this interest, in order to find a remedy in them. Therefore we should allow parties in the trial court — who filed their written pleadings before the judgment in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] — to amend them as a result of that decision. We should also allow arguments in this matter to be raised in an appeal, as long as the judgment has not become final.

27. Before we conclude this topic, we would like to point out that our approach with regard to a merely prospective change is unrelated to and does not affect those cases in which it is held that a law is unconstitutional and it is also held that the unconstitutionality will come into effect at a future date (see HCJ 6055/95 Tzemah v. Minister of Defence [25], at p. 284 {687}; HCJ 1715/97 Israel Investment Managers Association v. Minister of Finance [26]; see also Y. Mersel, ‘Suspending a Declaration of Voidance,’ 9 Mishpat uMimshal (2006) 39). In all of these cases, the declaration that the provision of statute was void acted retrospectively. All that was decided was that the declaration concerning the retrospective voidance should be suspended temporarily. We therefore adopted an approach that applied the new case law rule retrospectively, by attaching a ‘time fuse’ that postpones the time when the declaration comes into effect. The considerations underlying this approach are fundamentally different from the considerations that we have discussed in our opinion.

B. The effect of Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter on the agreement between the parties

28. The estate and the dependents made an agreement with the tortfeasors that a certain amount of compensation would be paid to end the dispute between them. This agreement was given the force of a court decision. The agreement was made and was given judicial force before judgment was given in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1]. Now that case has come and changed the law of compensation retrospectively. According to the new law, it is possible that the estate and the dependents are entitled to additional compensation. Against this background, the question arises as to whether the estate and the dependents are entitled to repudiate the agreement, in view of the change in case law that was caused by Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1]. This is the second question before us in this appeal. The answer to this question should be found in the law of mistake in contracts. Admittedly, the agreement between the parties was enshrined in a judicial decision, but the law is that for the purpose of the rescission of such an agreement on the ground that it was tainted by a mistake, we should refer to the law of contracts (see CA 2495/95 Ben-Lulu v. Atrash [27]; CA 3203/91 Azoulay v. Azoulay [28]; CA 4272/91 Barbie v. Barbie [29], at p. 699). Thus the question is whether the estate and the dependents have a claim that they were mislead into thinking that the law in their case had been determined in Estate of Sharon Gavriel v. Gavriel [2] and therefore ‘the court may... cancel the contract, if it thinks that it is just to do so’ (s. 14(b) of the Contracts (General Part) Law, 5737-1973).

29. In our opinion, the answer is no. The mistake of the estate and the dependents related to ‘the profitability of the transaction’ and such a mistake does not empower the court to cancel the contract (s. 14(d) of the Contracts (General Part) Law). Indeed, each of the parties to the agreement took upon himself the risk that in view of ‘the change in the legal climate,’ there might be a change in the law of compensation in so far as the lost years are concerned. In such circumstances, there is no ‘operative’ mistake (see Friedman, ‘Contractual Risk and Mistake and Misrepresentation with regard to Profitability,’ 14 Tel-Aviv University Law Review (Iyyunei Mishpat) 459 (1989), at pp. 466-471). This was discussed by Justice Tz.E. Tal:

‘In settlements the parties take upon themselves the risk concerning the legal position. No party can be certain that he will win the case, and even if he wins, perhaps it will not be worth his time and costs, so he therefore makes a settlement. Moreover, the party making a settlement takes upon himself the risk that the law will change retroactively, and on the basis of this knowledge he settles’ (CA 2444/90 Aroasty v. Kashi [30], at p. 527).

In the same spirit, D. Friedman and N. Cohen said:

‘It is assumed that the parties to the settlement take upon themselves the risk of new case law, including that this may change the law retroactively’ (Friedman and Cohen, Contracts (volume 2, 1997), at p. 729).

Indeed, not only the estate and the dependents, but also the tortfeasors (and the insurance company that stands behind them) cannot repudiate the contract that they made because of a retroactive change in the law of compensation. Each of the parties took upon himself the risk that the new law may increase or reduce the compensation.

The appeal is allowed. The decision of the Magistrates Court (of 29 April 2004) and the judgment of the District Court are set aside.

Justice Y. Adiel

I agree with the opinion of President A. Barak.

Vice-President Emeritus M. Cheshin

I have read the opinion of my colleague President Barak, and it seems to me like a miniature that contains everything with unbelievable detail. My colleague presents the complex subject of ‘prospectivity-retrospectivity’ clearly and straightforwardly, and this presentation makes it easier for me to present a slightly different version from my colleague’s version, even though I agree with his final conclusion.

Opening remarks

2. My colleague says (in para. 5 of his opinion; see also para. 17) that in Israeli law ‘The fundamental premise is that a new judicial ruling acts both retrospectively and prospectively,’ and after examining and clarifying the issues on their merits, he comes to the conclusion that the new case law ruling that was determined in this matter ‘will... apply both retrospectively and prospectively. It will apply to every tortious act that occurred before it and after it. This is the fundamental premise and there is no basis for departing from it...’ (para. 24 of the opinion). With regard to these statements I would like to make two comments before I discuss the heart of the matter.

3. First, even though my colleague speaks of a new case law rule that acts ‘both retrospectively and prospectively,’ the real interpretation is that according to his understanding the new case law acts retrospectively. After all, no one disputes that according to all approaches a new case law ruling operates prospectively. My colleague wishes therefore to establish a presumption — albeit a rebuttable presumption — that a new norm which is determined in case law and overrules a norm that preceded it is valid retroactively; that it applies almost automatically to acts that were done in the period when the old norm that was overruled prevailed. In this matter I disagree with my colleague, since in my opinion a new norm that is determined in case law will apply to events that take place after it, whereas its applicability to events in the past will be the exception. The application of a new norm will therefore be prospective, and someone who wishes to apply it retrospectively — to past cases — will have the burden of proving that it is right and proper to apply that new norm to acts that were done when the previous norm prevailed and in reliance on its existence. Moreover, as I shall explain later on in my remarks below, my opinion is that determining a sweeping retrospective rule is inconsistent with the varied character of the law, and we know that when we are dealing with the retrospective application of a norm, civil law cannot be compared to criminal law, the law of contracts cannot be compared to the law of torts, and even one area of the law of torts cannot be compared to another area of it. Each area of the law should be examined separately, and the determination of a uniform rule will not be successful.

Second — and this is the main point — when President Barak says that ‘the fundamental premise’ is that a new case law ruling acts retrospectively and prospectively, my colleague is relying on the remarks of Vice-President Miriam Ben-Porat in HCJ 221/86 Kanfi v. National Labour Court [9], at p. 480). But to the best of my knowledge, case law has not until now considered in depth the question that is now before us — the question of the ‘prospectivity-retrospectivity’ of the case law rulings of the Supreme Court — and since the panel hearing this case is an expanded panel, I think that it is not right to determine a ‘fundamental premise’ for the voyage of inquiry and interpretation on which we are embarking. If this is true as a rule, it is certainly true in view of the fact that the common law has been influenced considerably, at its roots, by Blackstone’s theory that the courts do not create law but merely reveal it. As we know, there were reasons for this theory, and these reasons are not accepted by us today. See also the penetrating remarks of Prof. Tedeschi in his article ‘Case Law for the Future,’ which was published in his book Essays in Law (1978), at p. 25 (the article is also mentioned in the remarks of Vice-President Ben-Porat, in Kanfi v. National Labour Court [9]).

My opinion is therefore that we are starting without any premise, and what we write will form a first impression.

4. At the outset I should say that, subject to what we will write below, I agree with my colleague that jurisprudential arguments, a priori legal doctrines and constitutional arguments do not have the power to decide the matter. My colleague considered these arguments and I agree with his opinion in its entirety (see also the remarks of Benjamin Cardozo cited in para. 15 below). But the question that we are called upon to decide is not which considerations will not decide the matter but which considerations will decide the matter. My colleague is of the opinion that practical considerations should prevail, and I will address these. Notwithstanding, I will add, as we shall see below, that the concept of practical considerations assumes, self-evidently, basic principles of law, which are themselves also based, inter alia, on practical considerations.

The relevant question

5. The question that I intend to answer is this: the Supreme Court makes a case law ruling — as in our case — on an issue that is mainly found in the field of case law, i.e., an area that has not been regulated expressly in statute. Years later the court once again considers the same issue, and after considering the issue, it overrules the original case law ruling. What is the law that applies to events that occurred between the first case law ruling and the second one, which come before the courts for a decision after the second ruling? Does the law that prevailed at the time of the event — i.e., the first case law ruling — apply, or does the law provided in the new ruling apply? We should note, and we will discuss this further in our remarks below, that we have presented the question that is troubling us on a (relatively) low level of abstraction. We are not speaking in general of a case law ruling made by the court — as to whether it merely acts prospectively or whether it also acts retrospectively — but of a ‘common law’ ruling only. Indeed, we could lower the level of abstraction and restrict our remarks to the law of torts, but for reasons that will become clear in our remarks below, we prefer to consider the question in the way that we have presented it.

Past events and retrospective norms

6. Events in the past are different from those in the present (which immediately becomes the past) and those in the future. The past is frozen in time and cannot be changed. That is how things are in the physical world and this is how things are in the normative world. And if someone asks — what, then, is a retrospective norm? — we shall answer as follows: a retrospective norm is a norm that, once it comes into effect, means that we no longer judge the past in accordance with the norms that prevailed when the events occurred but in accordance with that norm. We discussed the past, norms that apply to the past and questions that concern these in CrimA 4912/91 Talmai v. State of Israel [31], at pp. 619-620, and this is what we said:

‘... We are unable to change the past (to the regret of some and to the relief of others). Acts that were done, were done; omissions that were committed, were committed; events that occurred, occurred; vows that were made, were made; vows that were broken, were broken. All of these are as if they froze on the spot and became stone, and what has been done cannot be undone. We are incapable of doing anything other than describing and recording things that have happened — or that have not happened — but we are unable to change them. The freedom of choice and selection remains only for the future, but as to the past the choice has already been made, and the choice and selection — as choice and selection — are no more.

This is the case in the physical world and it is also the case in the world of norms, in the world that we have created and that is the product of the human spirit. Norms that existed in the past — including principles and rules of law — cannot be changed retrospectively: what was, was, and what was not...

What then is a retrospective norm, and what is a law that acts retroactively? Do these not have the power to change the past, at least in the world of norms? … Our answer to the question is no. This is what we say: the meaning of a norm that is supposed to apply retroactively is this, that from the day on which the norm begins, and thereafter, we shall no longer judge cases from the past in accordance with the norms that originally applied to them but as that norm directs us… All norms are prospective, by their very definition; they look to the future. But some of them also look to the past with regard to their application in the future to acts or omissions in the past…’

Thus, as a premise for our deliberations, the past is like Lot’s wife, whom we cannot return to life. But this is not the case in the normative sphere: if we only wish it, we can change in the future our attitude to what happened in the past. But if this is what we want, we will need to explain why and wherefore we wish to ignore what actually happened in the past and the norms that applied at the time of the event, and to apply to the past, from now on, different norms from those that prevailed at the time of the event.

7. Every act, every omission, every transaction and everything else that has legal significance is done, or not done, within the framework of a certain legal system at a given time and place. The moment that those things come into the world certain rights and duties are formed and created. Those things are born into a certain legal system, the legal system that surrounds them, and it also gives them a certain character, a certain ‘status.’ That system of rights and duties is born, one might say, with a certain genetic-legal character. Physically that system cannot be changed. Normatively, in the future, it can be changed, and this is within our power. In CA 6821/93 United Mizrahi Bank Ltd v. Migdal Cooperative Village [32], I mentioned the statement that Parliament in Westminster can do everything apart from turn a woman into a man and a man into a woman, and I commented on that statement (ibid. [32], at p. 527):

‘This statement is, of course, imprecise. If the author’s intention is that Parliament does not have the power to make a man a woman and a woman a man — taking the words literally — then the remarks are certainly correct. But then they have no significance whatsoever, since in the same way Parliament does not have the power to move a pencil from one side of the table to the other. Parliament — as such — does not concern itself at all with physical actions, and it does not have the power to make physical changes in the world about us directly. Parliament only concerns itself with norms and normative activity, and it is in this field that it has power and authority. If the intention of the author is therefore that Parliament is “unable” — from a normative point of view — to make a woman a man and a man a woman, it is obvious that the statement is incorrect. In the wonderful world of norms — a world that cannot be perceived by the five senses but rules our lives — the Knesset “can make” a man a woman and a woman a man. It is a separate question whether those persons to whom the norms are supposed to apply will abide by them. That question, it need not be said, falls outside our jurisdiction.’

(See also HCJ 5843/97 Bar-Gur v. Minister of Defence [33], at p. 473). I went on to say in Ganis v. Ministry of Building and Housing [16], at para. 38 of my opinion:

‘And so, in the creation of norms in the world of norms, Parliament is all-powerful. Parliament does not have — nor did it ever have — a surgeon’s scalpel that can draw blood. But it had, has and always will have a normative surgeon’s scalpel.’

Indeed —

‘From the viewpoint of the legislation, in and of itself — or we might say, from a merely normative viewpoint — there is no difficulty in this. Just one stroke of the pen, and a statute that is enacted today carries itself into the past at the whim of the legislator. Such is the act of legislation’ (ibid. [16], at para. 29 of my opinion).

And as I went on to say (ibid.):

‘Therefore a kind of dichotomy arises: reality does not allow us to change events in the past, but from a normative point of view we find legislation that seeks to take control of events in the past that were originally governed by a different law.’

8. But as long as no change is made to a norm, the norm that prevails at the time when the event occurred is the norm according to which the event will be judged, and it will determine which rights and obligations were created and what is the reciprocal relationship between them. This is true of the law and it is also true of social customs, social ethics, human behaviour and interpersonal relationships between human beings. Every act and all conduct is judged according to its time and place. An ‘enlightened’ emperor is enlightened relative to his time and the times that preceded him. It cannot be otherwise. It would not be right to judge — favourably or unfavourably — persons in the past and acts in the past with the tools that we have today. ‘Jephtah in his generation is like Samuel in his generation’ (Babylonian Talmud, Rosh HaShana 25b [58]).

9. Everyone agrees without exception that the legal status of acts and omissions that have legal significance should be determined, first and foremost, in accordance with the law that prevailed at the time the act or the omission was committed. This consensus is also the source of the doctrine of acquired rights. In the words of s. 22 of the Interpretation Law, 5741-1981:

‘Qualifications to the power of cancellation	22. The cancellation of a law is not capable of — (1) reviving something that was not valid at the time when the cancellation came into effect;
	(2) affecting an earlier act of the law that is cancelled or something that was done thereunder;
	(3) affecting a right or an obligation under the cancelled law and a sanction for an offence against it.

Here, then, is the principle of prospectivity: a new law is valid from the date of its commencement and thereafter, and it follows from this that when a later law repeals an earlier law, the repeal does not affect rights and obligations (in the broad sense of these concepts) that came into existence by virtue of the previous law. This is, of course, ‘if there is nothing in the matter under discussion or in its context that is inconsistent’ with this provision (s. 1 of the Interpretation Law), i.e., this rule will not apply where the new law itself says that it applies retroactively (on the self-evident assumption that the retroactive application was done lawfully).

This fundamental assumption, that the operation of a statute is prima facie prospective, gives expression to our intuitive feeling and the sense of fairness that is innate in us that this is how it should be. As we said in Talmai v. State of Israel [31], at p. 621:

‘The doctrine concerning “acquired rights” is an effective tool in the law, and usually it gives legal expression to the intuitive feeling of the expert jurist and the sense of fairness innate in us; moreover, all of these are consistent with public order and public security. A contract that is made and that is binding under the law that was in force when it was made will bind the parties to it even if the law, and with it the preliminary conditions for the making of a contract, is subsequently changed (subject to public policy); a tort that was done does not cease to be a tort merely because after the act that particular tort was repealed, and vice versa: an act that did not amount to a tort when it was done will not become a tort merely because after the event the legislature decided that such an act would constitute a ground for a tort; and so on and so forth.’

This is the situation with regard to rights and obligations that have arisen from within Israeli law. It is also the case with regard to rights and obligations that have arisen from within a legal system outside Israel, where that legal system is a legal system to which the rules of private international law that prevail in Israel refer. This is the case with regard to the existence, or the non-existence, of a right or an obligation, and this is the case with regard to the scope of a right or an obligation. This is the theory of acquired rights, even though this theory is capable of making us dizzy by its circular nature (Talmai v. State of Israel [31], at p. 622). We should also add this: a recognition of rights and liabilities under the law that prevailed at the time of the act or omission is not only required by common sense and logic, but this determination, and this alone, is capable of introducing security into practical life. This is how people acquire rights and this is how obligations are imposed on them. This is how people acquire immunity, etc..

10. This, therefore, is the first rule of legislation — the rule of prospectivity. A statute has, in principle, prospective application; its purpose is to create rights and liabilities for the future. And even though it is possible to give a statute retroactive application (subject to the general restrictions of the law), someone who argues this has been done has the burden of proving it.

All of this concerns legislation. What is the position with regard to case law?

A norm determined in case law

11. A long time has passed since we abandoned Blackstone’s theory that the courts merely ‘reveal’ law and do not ‘create’ law. We do not need to look far, for this can be seen in our case: in the earlier case of Estate of Sharon Gavriel v. Gavriel [2] the court created law, and this is also what happened in the later case of Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1]. This is true of every judgment of the Supreme Court, especially of judgments that knowingly and intentionally determine case law rules, whether they are rules that have a larger effect or rules that have a smaller effect. Case law in a judgment, whether it is an important case law ruling or not, whether it is mainly declaratory or it is mainly constitutive, is case law that is created on the date of giving the judgment. On that date the right of the litigant is created de facto. On that day the case law rule is made. In judgments and decisions made by the court, it creates law (‘judicial legislation’) and rights, and in principle there is no reason why I should distinguish between a system in which a statute repeals a statute and a system in which case law overrules case law. Just as when a new statute repeals an old statute the new law does not — prima facie — affect rights and obligations that were created by the old statute, so too when case law overrules case law the new case law does not — prima facie — affect rights and obligations that were created by the old case law. Subject to what we shall say below, there is no difference — prima facie — between statute and case law, whether it is from the viewpoint of practicalities, the viewpoint of justice, the viewpoint of social ethics or any other viewpoint. A norm is cancelled by a later norm, and the same logic that applies in the one case should also apply in the other case.

12. As in all the literature that has been written on the subject of retrospectivity, my colleague the president also raises the element of reliance (which is a close relative of the doctrine of acquired rights) as an element that support the principle of prospectivity. I agree with his remarks, provided that we realize that we are not speaking of a specific or an individual reliance, in a particular case, but of an element of ‘constructive’ reliance. We are speaking of a phenomenon of reliance that derives from an examination of the conduct of human beings; a reliance that originates in experience that is acquired in practical life; a reliance that is based upon the ordinary person and the ordinary case. This extralegal element underwent a process of crystallization, its essence was formulated into a legal rule, and from the time it was formulated the rule prevails and we no longer need to ask the question whether, in one specific case or another, the element of reliance actually occurred. In other words, the element of ‘constructive’ reliance led to the creation of a rule in the law, and it can be said — and this is what we do say — that today we have an institutional rule according to which the application of norms in the law — whether in statute or in case law — is prospective. The meaning of this is that a new statute or a new case law ruling does not purport to affect rights and liabilities that arise and were created by the law or case law that prevailed and existed before the new statute or case law; this, of course, is subject to the exception that we may decide and determine that in the circumstances of a certain case or a certain type of case, the application of a certain norm will be retroactive, and for what period of time it will apply retroactively.

13. To summarize this far, we can say the following: rights and obligations exist by virtue of the legal system; where the court decides a certain rule, rights and obligations, within the scope of that rule, exist by virtue of that rule. The case law that was decided is the law of the state — there is no other law of the state — and everyone is supposed to act accordingly. And if at a later date the court considers the ruling and overrules it, the new case law is valid from the date on which it was decided. The new case law ruling does not, prima facie, affect rights and obligations that were created by virtue of the old law. And if someone argues that the new case law acts retrospectively and that it can change or cancel rights or liabilities that were created and exist by virtue of the old case law, the person making that argument has the burden of justifying and explaining why and how the new case law is capable of cancelling rights and liabilities that were created and exist in accordance with the old law. See also and cf. A. Barak, Judicial Discretion (1987), at para. 283, the excellent article of Dr. E. Kaplan, ‘Prospective Application of Supreme Court Precedents,’ 9 Hebrew Univ. L. Rev. (Mishpatim) 221 (1979), and her revisiting of this article: ‘Future Application of Supreme Court Precedents,’ Avner Hai Shaki Book, 4 Mozenei Mishpat (Netanya Law Review) 125 (2005). I think that Dr Kaplan and I both travel along the same route, each of us in his or her own way.

14. Hitherto I have discussed one half of the picture. The other half, which is relevant to the current case, concerns the question of the application of a norm retrospectively. In so far as we are speaking of statute, the question of its retrospective application will be decided and determined by interpreting the statute and on the basis of the assumption that the statute satisfies the constitutional tests (in our legal system — satisfies the tests of the Basic Law: Human Dignity and Liberty and the Basic Law: Freedom of Occupation). Thus, if the statute is intended by the legislature to be retrospective, i.e., to change in the future rights and obligations that were created by the law that prevailed before the statute, and this designation arises from its interpretation in accordance with the accepted rules of interpretation, then such will be the case. The question that we are asking concerns the status of new case law that overrules previous case law. Can the new case law have retrospective effect? In other words, can case law in the future retroactively change rights that were acquired and obligations that were imposed under the previous law?

15. It is plain and simple that in the absence of any statute that tells us otherwise — and there is no statute in this regard — the question of the retrospective application of case law is also a question of case law. How then should we decide the matter? At this crossroads, we shall find it difficult to draw an analogy from ‘statute repeals statute’ to ‘case law overrules case law.’ The reason for this is that there are many different considerations that lead the legislature to enact or to grant retroactive application to a statute and these include considerations that by their very nature are foreign to the way in which a court works. In the words of R.J. Traynor, in his article ‘Quo Vadis Prospective Overruling: A Question of Judicial Responsibility,’ 28 Hastings L. J. 533 (1977), at pp. 537-538:

‘… In the legislative process there is neither beginning nor end. It is an endless free-wheeling experiment, without institutional restraints, that may have rational origins and procedures and goals or that may lack them…’

The legislature therefore has a broad horizon of considerations. The court is different. Its considerations are restricted to the world of the law, and they are mainly considerations of justice, reasonableness and utility.

In his book, The Nature of the Judicial Process (Yale University Press, 1921), Benjamin N. Cardozo addressed the question whether we should distinguish — for the purpose of prospectivity-retrospectivity — between case law that changes previous case law concerning the validity of statute (from a constitutional point of view), case law that changes previous case concerning the interpretation and scope of a statute and case law that changes previous case law concerning the interpretation or scope of a common law rule. He said in this regard (ibid., at pp. 148-149):

‘… Where the line of division will some day be located, I will make no attempt to say. I feel assured, however, that its location, wherever it shall be, will be governed, not by metaphysical conceptions of the nature of judge-made law, nor by the fetich of some implacable tenet, such as that of the division of Governmental powers, but by considerations of convenience, of utility, and of the deepest sentiments of justice.’

It follows that since the considerations of a legislator in applying a statute retrospectively are different from the considerations of a court in applying case law retrospectively, we will find it difficult to draw an analogy from statute to case law.

16. The main difficulty that stands in our way in applying case law retrospectively lies in those rights that have been acquired and those expectations that have arisen as a result of the previous case law — rights and expectations in the broad sense of these concepts — which the new case law wishes to cancel or restrict. At the time of the event, the law of the state was the original case law, and now we are seeking not to apply to the event that law of the state, but rather case law that was determined later and that overruled the previous case law. We should not take this injury lightly, since it is capable of completely changing legal relationships to the point of causing an injustice. As Lord Diplock said in a lecture in 1965 on the subject of ‘The Courts as Legislators’:

‘… judge-made law… is in theory retrospective. A precedent which reverses or modifies a previous precedent is applicable to all such cases which are tried subsequently even though they arise out of acts done before the new precedent was laid down. This is unjust, and because it is unjust it is itself a factor which makes the courts more hesitant than they would otherwise be to correct previous errors or to adapt an established rule of conduct to changed conditions. And yet the rule that a new precedent applies to acts done before it was laid down is not an essential feature of the judicial process. It is a consequence of a legal fiction that the courts merely expound the law as it has always been. The time has come, I suggest, to reflect whether we should discard this fiction’ (cited in Traynor, ‘Quo Vadis Prospective Overruling: A Question of Judicial Responsibility,’ supra, at p. 535, note 7).

See also the remarks of Lord Nicholls in National Westminster Bank plc v. Spectrum Plus Ltd [51].

17. The conclusion that inevitably follows is therefore that where case law ruling A prevails and subsequently case law ruling B overrules it, we need a substantial reason of great weight in order to agree to apply case law ruling B retrospectively to an event that occurred when case law ruling A was in force, i.e., to an event that occurred after case law ruling A and before case law ruling B. When we consider that the main purpose of the court is to do justice, we will realize automatically that the substantial reason of great weight that tells us to apply the new case law retrospectively also needs to be a reason that is based entirely on considerations of justice. Because if that reason is not entirely based on considerations of justice, it will not have the strength to overcome the premise that an event that occurred when a certain case law ruling was in force ought to have its legal character determined by that case law. This is what ought to happen, and this is how we ought to act. This justice that we should seek may be an individual, specific justice, between a plaintiff and a defendant, and it may be a justice that applies to a whole branch of law. The greater the requirements of justice, the greater the retrospectivity. But we must find justice, which is the force motivating the decision to apply the case law retrospectively.

18. We said at the beginning of our remarks (in para. 4) that we would only consider in this opinion of ours the type of case that is before us, i.e., a case law ruling that overrules a case law ruling in the field of case law (the Israeli version of common law). This is what we said, and for good reason. The reason for this is that this field of ‘case law overrules case law’ extends to various branches of the law and to very different types of cases. It is so wide in the areas that it applies that it would not be right and proper to speak of an all-embracing formula that is supposed to extend to all the different kinds of case. If we find an all-embracing formula of this kind, its wording will be so general and so diluted that we will be unable to make use of it as a tool for examining and considering cases. Indeed, the less the wording is fine-tuned, the greater the erosion of the mechanisms of scrutiny and wisdom.

19. We can find an analogy to our case in the subject of the freedom of expression. Freedom of expression, as we have said elsewhere (HCJ 6126/94 Szenes v. Broadcasting Authority [34], at p. 854 {384}), is not monolithic. It protects different kinds of interests, some of which are interests of great weight and some interests of little weight, and it would not be right and proper for us to give equal protection and equal treatment to all of these interests:

‘… the freedom of expression (like the freedom of creation) is not monolithic; it is a kind of federation, a federation of rights and interests. There are historical chronicles and there are speeches; there are commentaries and there is fiction and poetry; there is political comment and there is commercial advertising, there are marches and there are demonstrations, there are plays and there are films. Each of these methods of expression reflects a certain interest, and not all the interests are the same. Thus, for example, commercial advertising will not receive — and should not receive — the same protection as historical chronicles. It follows that instead of speaking of the freedom of expression in general, we ought to select carefully from the whole gamut of freedom of speech the aspect that is relevant. We are not talking of mere labels — fiction or historical chronicles, a poem or a demonstration. We should investigate to the very roots and do our very best in order to establish the nature of the interest that seeks protection — the essence and the substantial content of the right presented before the court.’

See also HCJ 4804/94 Station Film Ltd v. Film and Play Review Board [35], at p. 689 {57}; HCJ 606/93 Advancement Promotions and Publishing (1981) Ltd v. Broadcasting Authority [36], at p. 25; CFH 7325/95 Yediot Aharonot Ltd v. Kraus [37], at p. 78. Indeed, were we to regard the interest of freedom of expression as a monolithic interest, then we would be mixing together a large number of different ingredients. For everyone will agree — even the most ardent supporters of the freedom of expression — that political comment cannot be compared to commercial advertising.

20. Our case is like the case of freedom of expression, because the issue of ‘case law overrules case law’ is not monolithic and is not made of one material only. Thus, for example, the relevant considerations in a criminal proceeding are different from the relevant considerations in a civil proceeding, and any child will realize this. It is not at all surprising, therefore, that in the article of Richard H. Fallon and Daniel J. Meltzer, ‘New Law, Non-Retroactivity, and Constitutional Remedies,’ 104 Harv. L. Rev. 1731 (1991), the authors speak separately of the question of retroactivity in criminal cases, of limited immunity in constitutional torts cases and in proceedings to impose taxes. Moreover, they discuss separately the various approaches of the Supreme Court on the question of retroactivity in criminal cases in the time of Chief Justice Warren and in the time of Chief Justice Rehnquist (for the doctrine that prevailed in the time of Chief Justice Warren, see also: M. Cheshin, ‘Further on the Reassessment by the Income Tax Commissioner,’ Tax Quarterly, 1968, at p. 3). Moreover, civil law is also not monolithic. Thus, for example, the law of contracts is different from the law of torts, and the factor of reliance — as a factor that runs through the length and breadth of the law — is of prime importance in the law of contracts, whereas it is of secondary importance in the law of torts. Possibly we may even distinguish — in the field of torts — between torts involving property damage and those involving personal injury. The same is true with regard to the question whether in the sphere of ‘case law overrules case law’ the same rule should apply in a common law matter, i.e., in a matter than is not expressly regulated in statute; in a matter of declaring a certain statute to be unconstitutional and therefore void; and in a matter that concerns the interpretation and scope of a statute. All of these cases, and others too — so Benjamin Cardozo taught us (see para. 15, supra) — will be governed by the same basic considerations: justice, utility, etc., but when a specific case is brought before us in the field of ‘case law overrules case law,’ we have the burden of investigating the nature of the matter thoroughly. And we shall decide the law only after examining the specific force of the competing interests.

21. Moreover, as we have said above, case law that has been made becomes the law of the state, and therefore it is supposed to guide people in their actions. Even if it is later held that a case law decision in the past was made in error, that case law was still the law of the state until it was overruled. The Rabbis of the Talmud have already taught us that ‘an error, once made, has effect’ (Babylonian Talmud, Bava Batra 21a [59]). The same is true of a case law ruling that is made (case law A) and that the court later decides to overrule (case law B). This is what has happened in our case; the decision in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] has overruled the decision in Estate of Sharon Gavriel v. Gavriel [2]. The question that arises concerns the interim period, namely what is the law concerning those events that took place in the interim period, between case law A and case law B, which come before the court for consideration after case law B? (Actually, the question also arises with regard to events which occurred before case law A and which come before the court for consideration after case law B). The premise for our case is that in both the first case law ruling (in our case: Estate of Sharon Gavriel v. Gavriel [2]) and in the second case law ruling (Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1]) the court created law — law that applies to the parties and that has normative application for everyone.

22. In so far as the new case law ruling is supposed to apply to events that occurred after it was given, we shall encounter no difficulty. But this is not the case when we seek to apply the new case law retrospectively to the interim period between the two case law rulings. An analysis of this scenario of ‘case law overrules case law’ against the background of the recognition and consensus that both the first case law ruling and the second case law ruling created law necessarily leads us to the conclusion that when the court considers whether the second case law ruling — a ruling that creates law — should apply retrospectively, it should take into account, among the considerations that oppose this, those considerations that conflict with the inclination of applying the new law retroactively. These opposing considerations may lead us to a conclusion that the new law should not be applied retroactively — whether in general, in a class of cases or in one specific case or another — and they may also go on to create qualifications and defences, whether these are qualifications and defences with normative effect that are required by the new rights, or they are qualifications and defences that are required by the general law. After all, during the interim period the first case law prevailed, and we shall find it difficult to accept that after a period of months or years — sometimes quite a long period — the first case law will be struck down, retroactively, without any attention being given to what happened in the interim period. The most important factor for our case here is, of course, the reliance element.

23. I should add in this context, with all due caution, that it is possible to argue that reliance in the context of our case here does not only concern a situation in which someone relies on the existing law (the first case law ruling) and changes his position; reliance also concerns the normal course of events and the reasonable expectation that notice will be given in advance of a change in the law, and that the individual as a rule benefits from certainty with regard to the law. See also Prof. A. Barak, ‘Judicial Legislation,’ 13 Hebrew Univ. L. Rev. (Mishpatim) 25 (1983), at p. 73. Even if a person does not change his position specifically by relying on the existing law, retrospective application of a law is capable of confusing and undermining the confidence of the individual in the system of government. We have known for a long time that a statute requires prior notice, and a change in the proceedings for enacting a statute should also require prior notice. See United Mizrahi Bank Ltd v. Migdal Cooperative Village [32], at pp. 533-534. As we have already said in our remarks above, the main factor in the struggle between stability and change is the principle of justice and fairness — mainly distributive justice — together with the factors of proper practice and utility.

When is retrospectivity appropriate?

24. It is plain and simple that in the absence of a statute that gives us directions pointing one way or another — and there is no statute in the Israeli legal system on this subject — the question of the retrospective application of a particular case law ruling is a question that should be determined by the relevant legal system, whether normatively or on an individual basis. The question is one of determining a rule and establishing the exceptions to it: what will be the rule and what will be the exceptions? Common law, for example, never questioned the retrospective application of new case law — this was the rule that it determined — especially as a result of the doctrine outlined by Blackstone, according to which the courts do not create law but only reveal it. At the same time, the courts in England were aware of the injustice that may be caused by applying a new case law ruling retrospectively, and they sought to remedy this defect by determining a balancing formula that takes into account any exception to the rule. See, for example, para. 40 of the opinion of Lord Nicholls in National Westminster Bank plc v. Spectrum Plus Ltd [51] (which is cited in para. 16 of the opinion of my colleague, the president).

25. My colleague President Barak refers to comparative law, and in his opinion he reviews a broad selection of case law in various countries. In National Westminster Bank plc v. Spectrum Plus Ltd [51] the justices of the House of Lords also referred extensively to comparative law and the various case law rulings that have been made in various countries, and it is possible to say that over the years these have been of all types and kinds. The selection is a wide one, and anyone who wishes to rely on comparative law may choose what he wants. See, for example, P.J. Stephens, ‘The New Retroactivity Doctrine: Equality, Reliance and Stare Decisis,’ 48 Syracuse L. Rev. 1515 (1998). There is much confusion, especially in the United States, and the literature on the subject is extensive and burdensome. It is difficult to avoid the impression that the course of case law in the United States — case law that changes direction from time to time — is affected mainly by the fundamental outlooks of the justices of the United States Supreme Court with regard to the role of the court in the system of government. See also and cf. J.E. Fisch, ‘Retroactivity and Legal Change: An Equilibrium Approach,’ 110 Harv. L. Rev. 1055 (1997). As for us, true to our approach we will say that where ‘case law overrules case law,’ the second case law ruling will apply prospectively, but the court, like the legislature, may apply it retrospectively while taking into account the distribution of justice between those who benefit and those who lose out under the later case law.

26. The first question is: what is the law concerning the plaintiff who won in the later case (case law B) and brought about the overruling of the original case law? In our opinion, the question concerning the distribution of justice between a plaintiff and a defendant should be asked also in the case of this plaintiff, just as it will be asked in the case of plaintiffs who will come after him and seek to benefit from the new case law. We should, however, add that we shall have difficulty in finding a case where that plaintiff will not be found worthy to benefit from the fruits of the new case law. First it should be said — and others have already said this — that if the plaintiff in the later case does not benefit from the application of the new case law to his case, when it is the case law that he himself brought about, we shall not find plaintiffs who bring about a change in case law that ought to be changed, and the public will be the loser. Second, justice demands that the new case law will apply to the person who brought about the change, and that he will benefit from his labours. But we should emphasize that where the interest of the defendant and the public interest outweigh the interest of the plaintiff, the new case law will not apply even to the plaintiff who brought about its creation. The effect of the case law ruling will be merely prospective — i.e., the case law will apply only to events that occur after it is given — and in the language of American jurists it will be said that the case law ruling is ‘purely prospective.’ It need not be said that if the defendant has a specific defence against the right that the court is establishing for the first time, he is entitled to raise that defence in exactly the same way that defendants in the future will be entitled to raise specific defence arguments against the new right that was established in the second judgment. All of the aforesaid concerns the plaintiff who brought about the change in case law. But what is the position with regard to others whose case is identical or very similar to the case of the person who brought about the change?

27. My answer to this question is that the ‘legal burden’ lies with this other plaintiff to persuade the court that the new case law ruling should be applied retrospectively, namely that it is right to change arrangements that were in force and systems of rights-obligations that were established before the new case law, and especially that it is just — in distributing justice between a plaintiff and a defendant — to apply the new case law retrospectively. Indeed, unlike my colleague, who is of the opinion that the new case law applies retrospectively unless the court decides otherwise — I am of the opinion that case law should not apply retrospectively unless the court decides otherwise. The burden of persuading the court that a case law ruling should be applied retrospectively rests with the plaintiff.

28. In this context I would like to ask a question of the supporters of retrospective application. Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] recognized — or perhaps we should say, created — a head of damage that previously had been denied by case law, and thereby it changed the substantive law of torts with regard to personal injury. We should emphasize that we are not speaking of a change of procedure but a change in substantive law. See Dicey and Morris, The Conflict of Laws (thirteenth edition, L. Collins ed., 2000), vol. 1, rule 17, at para. 7-034 et seq.. This gives rise to the question: let us suppose that our case did not concern the creation of a head of damage but the creation of a tort, i.e., the creation of a new cause of action whose existence the court rejected in previous case law. Would we decide to apply this case law retrospectively?

What is the law in Israel?

29. Since we have, until now, spoken at length, we can now speak briefly. My colleague the president is of the opinion that restrospectivity is the rule and non-retrospectivity is the exception. In his words (at para. 17 of his opinion):

‘The fundamental premise is that a new judicial ruling applies both retrospectively and prospectively. Notwithstanding, I am of the opinion that there is nothing in principle that prevents us from recognizing the power of the Supreme Court to give its precedents merely prospective force.’

My opinion is otherwise. Unlike my colleague, who assumes — as a premise — that new case law acts retrospectively unless there is a statement to the contrary, my assumption is that new case law acts prospectively unless there is a statement to the contrary. In other words, in my opinion prospective application is the rule, whereas retrospective application is the exception. Since this is the case, the premise is that new case law has prospective application, and therefore the onus lies with the person arguing that it should have retrospective application. This is how we should address the issue.

From general principles to the specific case

30. Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1], as case law that creates law and as case law that overrules case law, applies to events that will occur after it. Everyone agrees upon this. The relevant question is whether this case law should be applied retroactively, to events that occurred when Estate of Sharon Gavriel v. Gavriel [2] was valid. It would be true to say that the question is not an easy one for us. Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] reversed the ‘lost years’ rule as held in Estate of Sharon Gavriel v. Gavriel [2], and by doing so it created an earthquake in this specific field of the law of torts. A ‘minor revolution’ took place, in the language of President Yitzhak Kahan in Estate of Sharon Gavriel v. Gavriel [2] (at p. 570). Should Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] be applied retrospectively?

The question whether Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] should be recognized as having retrospective application or only prospective application will be determined by the conflict between the competing considerations, and in this regard I was especially impressed by the moral warmth that can be seen in the judgment of the court in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] — a judgment that was written by our colleague Justice Rivlin, with the agreement of President Barak, Vice-President Or and Justices Mazza and Dorner — and from the intensity with which Justice Rivlin spoke of the right of the injured party to receive justice, namely compensation for the lost years. Let us cite several passages from the opinion of Justice Rivlin in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1]:

‘The compensation for the “lost years” is… capable of remedying the unequal state of affairs that was created as a result of the tortious act. It offers a solution to the injustice that is inherent in the denial of the right of compensation to someone who is not able to realize his earning capacity, because of a reduction of his life expectancy brought about by a tort, while at the same time compensation is awarded to someone whose inability to realize his earning capacity derives from his being injured by a tort. This results in it being cheaper to kill than to wound. It should be noted that we are not concerned with punishing the tortfeasor but with balancing the scales and refraining from an unjust reduction of the compensation merely because of the fact that in addition to the harm to the injured person’s earning capacity the tortfeasor also caused him a reduction of his life expectancy’ (ibid. [1], at para. 15).

‘Money cannot replace a damaged limb, the suffering involved in loss of a place of work, and it can certainly not replace years of life that have been lost. However, this alone cannot undermine the power of the courts to award compensation, in so far as this is necessary in order to bring the injured person as close as possible to the position he would have been in, had the damage not occurred… The compensation will not prevent the suffering, but it can make the suffering bearable’ (ibid. [1], at para. 18).

‘Indeed, if compensation for the “lost years” is not awarded, the result obtained from the provisions of s. 78 of the Ordinance, in cases where the deceased does not have, when he died, a claim for compensation, is, from the viewpoint of the dependants, harsh and unjust. Take the case of a person who had a working life expectancy of twenty years, and because of a tortious act his life expectancy is reduced to only two years. The vast majority of the potential earning years, which will not be realized because of the act of the tortfeasor, will not be given any expression in the award of compensation, and the dependants, even if they inherit what he was awarded in his claim, will be left with an empty shell, unless the injured person chose — and to put such a choice before him is inconsistent with criteria of justice and logic — not to file a claim for his damage’ (ibid. [1], at para. 29).

‘… the award of compensation for the loss of earning in the “lost years” corrects — admittedly not in the full sense of the word but in important senses — the major imbalance in the external balance that was caused by the wrongful act of the tortfeasor. The injured person has been deprived, by the wrongful act, of the ability to earn income and to make use of it for his needs and for those of his family. Awarding compensation addresses the need to take this into account, and ensures that the lack of balance caused by the tort will not remain unaddressed especially in cases where the result of the tortious act is particularly serious…

… the award of compensation for the “lost years” prevents the arbitrary results according to which compensation is not awarded for the loss of earnings to an injured person whose life expectancy is shortened, while compensation on this head of damage is awarded to an injured person in a permanent vegetative state, or to the estate for pain and suffering and reduction of life expectancy, all of which without any really adequate justification for the distinction… Perhaps most importantly of all, the awarding of the compensation for the “lost years” (to the living injured person) ensures that a situation will not arise in which, although the dependants have been deprived by the tortious act of the support of the injured person — support that they would have received had it not been for that act — this damage will remain unremedied’ (ibid. [1], at para. 70).

If justice is on the side of the plaintiff — or perhaps we should say, on the side of the injured person or his dependents — then justice appeared in its full glory in the opinion of Justice Rivlin in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1]. This justice, warm human justice, has great weight — maybe even decisive weight — when determining the question of retrospectivity. I have also taken into account the fact that our case concerns differences of opinion between an insurance company and a worker who was killed in the course of his employment, and the plaintiffs are the dependents of the deceased and his estate. In the distribution of justice between these two parties, who are not of equal force, and in view of the ability of the insurance company to spread the damage, the scales tip in favour of the injured person and those dependent upon him. There was a time — a long time ago — when counsel for a plaintiff was not allowed to mention — especially before a jury — that the defendant was insured and that the damages would be paid by the insurance company and not by the defendant personally. That time has passed, and we are in the present.

31. In summary, I agree with the conclusion of my colleague the president that the rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] should be applied retrospectively, subject only to specific arguments — including arguments of reliance and other arguments — that defendants may raise in proceedings against them.

The agreement between the parties and the rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter

32. On the basis of the assumption that Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] applies to their case — and this is indeed what we are deciding — the defendants raise a defence argument that relies on an agreement that was made between them and the plaintiffs and that was given the force of a court decision. The rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] was made on 15 March 2004, but a short time before that the parties made a settlement according to which the defendants would pay the plaintiffs a sum of NIS 100,000 in settlement of the claim. On 22 February 2004 — approximately three weeks before the rule was made in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] — this agreement was given the force of a court decision. The question is therefore whether this agreement, which was given the force of a court decision, stands in the way of the plaintiffs and denies them their (retrospective) right. My colleague, President Barak, is of the opinion that the agreement is a barrier to the plaintiffs’ claim, and I agree with his conclusion. But my method is different from his method.

33. In my colleague’s opinion, the question should be decided in accordance with the provisions of s. 14(b) of the Contracts (General Part) Law, 5733-1973, which provide and tell us the following:

‘Mistake

14. (a) …

(b) If someone entered into a contract as a result of a mistake, and it may be assumed that had it not been for the mistake he would not have entered into the contract and the other party did not know or should not have known this, the court may, upon an application of the party that made the mistake, cancel the contract, if it thinks that it would be just to do so; if it does this, the court may hold the party that made the mistake liable for compensation for the damage that was caused to the other party as a result of making the contract.’

This provision of statute concerns a ‘mistake,’ and the relevant question is whether the plaintiffs did indeed fall victim to an operative ‘mistake’ when they signed the settlement. Personally, I find it difficult to see any ‘mistake’ to which the plaintiffs allegedly fell victim.

34. There is no doubt that when it was made the agreement was not tainted by any mistake. But the case law rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1], as my colleague says, should be regarded as case law that changed the law of compensation retroactively — in our case, at least to the date of making the agreement — and if this is so, the plaintiffs should be regarded as having fallen victim to a mistake when they made the agreement: the plaintiffs thought that the rule in Estate of Sharon Gavriel v. Gavriel [2] applied to them, whereas it was in fact the rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] that applied to the case (as we found out shortly afterwards). I do not accept this line of reasoning. The mistake of which s. 14(b) of the Contracts (General Part) Law speaks is a mistake that is contemporaneous with the time of making the agreement. In other words, the concept of ‘mistake’ in a contract, by its very nature, applies on the date of making the contract. We do not find any mistake of this kind. And if it is argued that the retroactive application of the rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] made an agreement that was originally untainted by any mistake (‘in real time,’ as the saying goes) into an agreement that is supposedly tainted by a mistake, it seems to me that we are stretching the concept of ‘mistake’ to the point of bursting, such that its whole content will be spilled. See LCA 6339/97 Roker v. Salomon [38], at p. 253. In a paraphrase of remarks that were written in Roker v. Salomon [38], at p. 254, we can say the following: the concept of mistake is built on the essence of ‘mistake,’ and around that essence there are events and cases that are attracted to its centre of gravity. The essential meaning is what will determine the scope of the concept. The D.N.A. is what will decide it. Introducing an objective element into this concept of mistake will completely undermine the arrangements (see and cf. D. Friedman and N. Cohen, Contracts (vol. 2), at p. 727, para. 14.57).

35. With regard to the present case, we shall say this: there is no doubt that the plaintiffs did not make a ‘mistake’ when the agreement was made — in the fundamental and accepted meaning of the concept of ‘mistake’ — and I have not found any justification for imputing any mistake to them after the rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] came into existence, by distorting the language. This is the case even if we adopt the retrospective perspective. But according to our approach, there was certainly no mistake in the agreement, since the retroactive force of the rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] was given to it by the court, and not automatically by the general law. And if there was no mistake in the agreement, there is therefore no basis for applying the provisions of s. 14(b) of the Contracts (General Part) Law.

36. I think that the remarks of Justice Tzvi Tal in Aroasty v. Kashi [30] are in agreement with our remarks. At pages 522 et seq. of the judgment, Justice Tal addresses the question of the retrospective application of case law that interprets a law and the issue of a ‘mistake of law’ as a defect in a contract (according to the provisions of s. 14(d) of the Contracts (General Part) Law)), and in his summary of the matter he determines (at p. 524) that ‘a later interpretation of a “statute,” which changes its meaning from what a party to a contract originally thought, cannot be considered a “mistake of law”.’ Later on (at p. 525) Justice Tal goes on to say that —

‘It is difficult to entertain the idea that it is possible to open a matter that has been concluded, such as a contract that has already been performed, with a claim that one of the parties made a mistake of law, as a result of new case law, maybe years later, that changed the previous law.’

37. I should also point out that had the element of mistake existed in the settlement in our case — and in my opinion, as aforesaid, the agreement was not tainted by any mistake — I would have been disposed to consider seriously the plaintiffs’ application — an application based on the provisions of s. 14(b) of the Contracts (General Part) Law — to cancel the settlement ‘for reasons of justice.’ Indeed, were we to agree that the agreement was tainted by a mistake, then the question would have arisen as to whether the plaintiffs’ mistake was an operative mistake, i.e., a mistake that makes a contract defective, or whether it was only a mistake in the ‘profitability of the transaction’ (in accordance with s. 14(d) of the Contracts (General Part) Law, and in such a case there would be no grounds for cancellation of the contract. My colleague the president is of the opinion that the mistake in our case was a mistake in the ‘profitability of the transaction,’ and I find this conclusion problematic.

38. The parties before us made a settlement between themselves, and we agree of course — how could we do otherwise? — that a settlement tells us that the two parties to the settlement made reciprocal concessions with regard to the chance of being entitled to more than what the settlement gave them, something that is commonly known as ‘risk management.’ See D. Friedman, ‘Contractual Risk and Mistake and Misrepresentation with regard to Profitability,’ 14 Tel-Aviv University Law Review (Iyyunei Mishpat) 459 (1989), at p. 469; Friedman and Cohen, Contracts, supra, at p. 736; HCJ 57/67 Gross v. Income Tax Commissioner [39], at pp. 559-560 (per Justice Silberg). The question is simply what chance did the parties to the settlement give up and what risk did they take upon themselves? This question also contains the answer to the question: what is a ‘mistake in profitability?’ We accept the definition of Friedman and Cohen that ‘a mistake in profitability… is a mistake with regard to a risk that the party took upon himself, whether expressly or according to the correct interpretation of the contract or in view of the understanding that we have of contracts of this kind’ (Friedman, ‘Contractual Risk and Mistake and Misrepresentation with regard to Profitability,’ supra, at p. 466; Friedman and Cohen, Contracts, supra, at p. 727). With regard to this risk that a party ‘took upon himself,’ Friedman goes on to tell us (ibid.) that:

‘… We do not necessarily mean that the party took this risk upon himself willingly and knowingly. Sometimes this is indeed the case, but in other cases the law attributes to him the taking of the risk under discussion. In other words, in view of the approach that we have to the nature of contracts and in view of our understanding of the ordinary risks involved therein, we assume (unless it is determined otherwise) that a party took on a certain risk even if he was not actually aware of this.’

The question in a settlement is, therefore, which fact, legal rule or possible development in the future was in dispute between the parties (Friedman and Cohen, Contracts, at p. 736). A settlement is an act of risk management, but ‘the question is always what was the risk that was minimized and what were the assumptions underlying that settlement’ (Friedman, ‘Contractual Risk and Mistake and Misrepresentation with regard to Profitability,’ supra, at p. 470). Thus, a mistake is a mistake in the ‘profitability of the transaction’ — it is not an operative mistake — if it falls within the scope of the risk that each party took upon itself. By contrast, a mistake that falls outside the scope of the risk will not be a mistake in the ‘profitability of the transaction.’ In the language of Friedman (ibid., at p. 466):

‘… There is no basis for the claim of mistake with regard to a risk that a party took upon himself within the framework of a contract (it is possible, if one wishes, to call this mistake a “mistake of profitability”), but there is a basis for a claim of mistake with regard to a matter that was not included within the scope of the risk that the party took upon himself.’

And as Friedman and Cohen say on the subject of settlements (Contracts, at p. 737):

‘… Where a settlement is based upon a fundamental mistake, on a point that was not in dispute and with regard to which the parties did not compromise, the settlement may, like any other agreement, be rescinded provided that the conditions for this are satisfied. Even if the settlement was given the force of a judgment, the settlement and the judgment that is based on it may be cancelled because of a mistake… A settlement is admittedly a case of risk management, but the question is always what is the risk that was minimized and what were the basic assumptions that served as a basis for that settlement.’

39. In our case we can say that when the settlement was made — in ‘real time’ — the question of the ‘lost years’ was far removed from the areas of risk that the parties took upon themselves or from the hopes that the parties entertained. Indeed, like in any settlement of a pecuniary nature, the parties settled with regard to the amount of the compensation that the employer should pay the worker, but the question of the ‘lost years,’ as a question in itself, was very remote from their thinking. We should assume — this is what any reasonable understanding tells us — that when calculating the compensation the parties were mindful of the rule in Estate of Sharon Gavriel v. Gavriel [2], and they made this case law ruling the basis for the negotiations between them.

But now the rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] has appeared. The rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] is not merely ‘another rule’ in the law of compensation. This is not an ordinary rule, a rule of the kind that we encounter every day. This is a rule that brought into the world a new head of compensation. It is a creation ex nihilo, or perhaps we should say, a creation ex negativo. It is like a case law ruling that creates a new cause of action in torts that never existed when the settlement was made (and what is more — a cause of action whose existence was expressly rejected in case law). This is expressed in the remarks of Justice Rivlin in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1], when he said (in para. 70 of his opinion) that following the rule in Estate of Sharon Gavriel v. Gavriel [2], as the years passed, we expected that legislation would change the case law ruling, but we waited in vain and the legislature failed to step forward and enact legislation. Indeed, in our opinion Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] did cause a minor revolution in the field of the law of compensation in torts. Had Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] been the prevailing case law at the time the settlement was made, then if they had assumed that Estate of Sharon Gavriel v. Gavriel [2] was actually the prevailing case law the parties to the settlement would have made an operative mistake. Their mistake would not merely have been a mistake in ‘the profitability of the transaction,’ since the question of the ‘lost years’ would have been beyond the scope of the concessions that were made and the reciprocal risks that the parties took upon themselves. The plaintiffs did not take into account in their considerations the possibility that the decision in Estate of Sharon Gavriel v. Gavriel [2] would be overruled. In the settlement they did not ‘waive’ the possibility of a change in the law, and it is not just that we should attribute to them a waiver of a chance that the law would be changed.

40. As we have said in our remarks above, Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] is, in our opinion, like a case law ruling that created a new cause of action, and, what is more, a cause of action whose existence was rejected in the past. Just as in the latter case — had it occurred — the question of the existence of that cause of action would have fallen, so it may be assumed in the ordinary case, outside the scope of the risks and chances, so too in our case. Thus, the principle of justice that led us to apply the decision in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] retrospectively is the very same principle of justice that tells us to recognize a ‘mistake’ — in so far as there was one — as an operative mistake, as a mistake that is not merely ‘a mistake of profitability,’ as a mistake that undermines a transaction and cancels a settlement. But in our opinion, as aforesaid, there was no mistake in the settlement. The doctrine of ‘mistake’ is too limited to include a set of facts that did not include a real mistake in ‘real time,’ and for this reason I agree, albeit with some regret, with the conclusion of my colleague the president that the settlement brought an end to the claim of the plaintiffs, who are the respondents before us.

Summary

41. I agree that the rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] ought, in principle, to apply retrospectively, but it is subject to defence arguments that a defendant may raise in the specific circumstances of his case. As I said at the beginning of my opinion, I agree with my colleague the president that the appeal should be allowed and the judgment of the trial court should be set aside.

Justice E. Rivlin

I have read the illuminating opinions of both my colleague President A. Barak and my colleague Vice-President M. Cheshin. I agree with the result that my colleagues have reached that the appeal should be allowed and the judgment of the trial court should be set aside. With regard to the fundamental disagreements between my colleagues, which concern the method of reaching the result, my opinion is in accordance with the opinion of the president, for his reasons.

Justice A. Grunis

I agree with the opinion of my colleague, President A. Barak.

Justice D. Beinisch

My colleagues, President Barak and Vice-President Cheshin, have spoken extensively on the question of the temporal application of new case law and in their opinions the question of prospectivity and retrospectivity has been presented in all its multi-faceted complexity. I agree with their opinion concerning the result, according to which the appeal should be allowed and the judgment of the trial court should be set aside. With regard to the fundamental dispute I will add only a few words with regard to my position.

There is no dispute between my colleagues that the application of new case law can be retrospective. The vice-president is of course correct when he says that the real and only question is the question of the retrospective application of new case law, since there is always prospective application. My colleagues also agree that the application of legislation of the Knesset, which is usually prospective unless it contains a statement to the contrary, cannot be compared to the application of case law, which is the ‘common law’ in our legal system. The disagreement revolves around the question of what is the fundamental premise: what is the rule and what is the exception?

In this disagreement, I agree with the opinion of President Barak that ‘The fundamental premise is that a new judicial ruling acts both retrospectively and prospectively.’ The different approaches were presented by my colleagues, as aforesaid, comprehensively and fluently and to attempt to add to them would be merely to detract. In his opinion, the president defined the ‘practical consideration’ as the decisive consideration for his conclusion and he gave his reasons for this. Indeed, in my opinion too the practical consideration outweighs the other considerations and touches upon the heart of the judicial role and the essence of judicial creation.

Let us mention once again the consensus in our legal system that the Supreme Court does not lightly depart from its precedents. New case law is created against a background of new circumstances, and as a rule these are not commonplace in our judicial work. New case law is made when the court is persuaded that the previous case law was erroneous or when its time has passed because circumstances have changed. The need for new case law arises when the law needs to be brought in line with reality, whether this is social reality, practical reality or legal reality. Only then is case law likely to change and thereby develop the law.

A change in case law requires a balancing of the existing position and the extent to which it corresponds with reality against the extent of the harm to legal stability and its consequences. When the judge reaches the point of decision and comes to the conclusion that the legal reality should be changed, from that point onward he will have great difficulty in making a decision that only has prospective application. In the course of applying the law on a daily basis, it will be a very complex if not impossibly difficult task to continue to make judicial decisions that are based on the case law ruling that has been overruled, or to contend with the need to examine the validity of the new case law ruling on a case by case basis. This difficulty is resolved when the rule is that new case law will apply retrospectively.

This conclusion does not ignore all the situations and difficulties that may arise. It does not ignore the existence of circumstances in which decisive weight should be given to the need to respect rights and obligations that were crystallized in the past and to refrain from a serious injury to protected interests. The aforesaid conclusion does not require us to ignore differences between different branches of public and private law that may justify special treatment, as Vice-President Cheshin has said in his opinion. The approach that recognizes the retrospectivity of new case law as a rule determines a fundamental position but it does not compel us to ignore exceptional circumstances in which new case law should not be applied retrospectively because of the extent of the injury to acquired rights or a protected reliance interest. The decision when to restrict new case law and to give it prospective application only, or suspended retrospectivity, is a decision that depends upon the circumstances and the context. The proper balancing point in each specific case will usually be decided from the viewpoint of and in accordance with the new case law, and restrictions will also determined on the basis of the new case law. Cases in which we are required to limit the application of the case law and to make it merely prospective will be examined by means of legal doctrines that run the length and the breadth of the legal system and through all of its branches, and this was discussed by President Barak when he presented a non-exhaustive list of possible solutions in difficult cases.

By way of generalization it is therefore possible to say that when the court has crossed the ‘stability barrier’ presented by existing case law and sees a need to make a new case law ruling that is appropriate to the time and the social and normative reality that prevails when it is made, there is a need for consistency in deciding cases in accordance with case law as it stands at the time of giving judgment, while adapting it to the specific solutions that are provided in exceptional cases, in order to prevent damage and harm that are disproportionate according to the fundamental principles of the legal system.

I therefore agree with the opinion of President Barak.

Justice M. Naor

1. In my opinion, in the circumstances of the case before us there is a settlement between the plaintiffs and the defendants, and this was given the force of a judgment on 22 February 2004 (although this judgment was called a ‘decision’). The proceeding between the plaintiffs and the defendants therefore ended before the rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] was made. This is sufficient in order to determine that there was no basis for allowing the plaintiffs to repudiate their consent to the settlement in these circumstances (see Ben-Lulu v. Atrash [27]; HCJ 4157/98 Tzevet, Association of Retired IDF Servicemen v. Minister of Finance [40], at pp. 790-791; CA 8972/00 Schlesinger v. Phoenix Insurance Company Ltd [41], at p. 843). Therefore, because of the principle of finality, the question of the retroactive application of new case law does not arise at all in this case, just as it does not arise with regard to other cases that already ended in a settlement or a final judgment before the decision in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] was made.

2. The question of the retroactive application of judgments that change case law in general, and the judgment in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] in particular, is an important question. The disagreement of opinion between my colleague President Barak and my colleague Vice-President Cheshin is ultimately a question of what is the rule and what is the exception. Neither my colleague the president nor my colleague the vice-president recommend making an absolute rule that allows no exceptions (cf. the judgment given very recently with an expanded panel in CA 1761/02 Antiquities Authority v. Station Enterprises Ltd [42]). Since in my opinion a discussion of this issue is not required for the decision in this case, I shall limit myself to addressing the question of the retrospective or prospective application only with regard to the decision in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1].

3. In this matter, I am of the opinion, like all my colleagues, that the rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] should be applied retroactively to cases that are pending in the judicial system. The decision in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] did not come into the world from nowhere and its spirit hovered over legal proceedings for a long time before it was made. Many parties sought to amend statements of claim and to argue that they are entitled to compensation for the lost years before the decision in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] was made. Many cases, in all the courts, waited for the litigation in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] to end, and it is not right that the Ettinger estate should be the only one that benefits from the change in case law. Moreover, as my colleague Vice-President Cheshin said, the rule in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] is of great force, and it was very just that my colleague Justice Rivlin called in his opinion in Estate of Ettinger v. Company for the Reconstruction and Development of the Jewish Quarter [1] for compensation to be awarded for the ‘lost years.’

4. I therefore agree that the appeal should be allowed, the judgments of the Magistrates Court and the District Court should be set aside, and the judgment (called a ‘decision’) of 22 February 2004 should be reinstated.

Appeal allowed.

29 Shevat 5766.

27 February 2006.