XXXI. Israel

a) Brief Summary of the Facts

6A law firm brought an action against two people for unpaid legal fees, the court of first instance (Magistrate Court) found for the plaintiff but the court of appeals (District Court) reversed this decision and a leave for appeal to the Supreme Court was denied. After several years, the firm brought another action for the unpaid fees, which was denied in accordance with the doctrine of res judicata and the statute of limitations. The firm then brought a tort action against the State, claiming that its contractual actions were denied due to the legal system’s negligent operation. The court of first instance denied this claim and the court of appeals affirmed. The firm applied for leave to appeal to the Supreme Court.

b) Judgment of the Court

7Justice Amit held that judicial immunity is substantive, not procedural. Judges cannot be liable for torts committed in their judicial capacity, and this immunity cannot be circumvented in any way. Thus, the State cannot be vicariously liable for torts attributed to judges. Similarly, judicial immunity cannot be circumvented by depicting a series of judicial actions as the operation of the legal system as such. The plaintiff’s attempt to sue the State for its own negligence in pursuing its claims must fail.

c) Commentary

8Judicial immunity is defined in sec 8 of the Civil Wrongs Ordinance (CWO):^[5]

‘(n)o action will be brought against any person constituting, or being a member of, any court or tribunal or against any person lawfully performing the duties of any such person, or against any other person performing judicial functions, including an arbitrator, in respect of any civil wrong committed by him in his judicial capacity’.

9A distinction between substantive and procedural immunities exists in Israeli law. A substantive immunity totally negates the tort and any ensuing right. A procedural immunity prevents legal action based on the tort but does not undermine its very existence, so the victim may recover by using other tools, such as bringing a counterclaim when relevant, setting-off, or suing the perpetrator’s employer under the doctrine of vicarious liability. Similarly, while procedural immunity prevents an action by the victim, it does not prevent an action for contribution by a joint tortfeasor.

10Judicial immunity is considered substantive.^[6] Therefore, victims of judicial torts can neither bring a civil action against the relevant judge nor recover for any loss incurred by the tort in any other way. Previously, the Supreme Court held that the State cannot be held vicariously liable for torts committed by judges in their judicial capacity. In Weinberg the Court correctly added that the immunity cannot be circumvented by bringing an action against the legal system as such for what is in fact a series of judicial actions.

11Judicial immunity is distinct from the general public-employees’ immunity. Section 7A(a) of the CWO provides that: ‘(a) claim will not be brought against a public employee in respect of an action performed in the fulfillment of a government office as a public employee’. As the language of sec 7A(a) implies, public-employees’ immunity is procedural. The victim cannot bring an action against the employee, but the tort is not negated, and the State can be held vicariously liable (as sec 7A(b) explicitly states). Although judges are public employees, they enjoy a much broader (substantive) immunity under sec 8, which also prevents any action against the State for judicial misconduct.

a) Brief Summary of the Facts

12The plaintiffs (petitioners in the leave to appeal) brought an action for discrimination under the Prohibition of Discrimination in Products, Services and Entry into Places of Entertainment and Public Places Act (hereinafter the PDA).^[7] The defendant (respondent) was a restaurant which offered complementary cocktails to ‘exclusively female tables’ during the FIFA World Cup and refused to offer similar treats to male patrons. This policy was also published on the restaurant’s Facebook page. The plaintiffs argued that men were discriminated against by the practice and by the advertisement.

13The court of first instance found the defendant liable under sec 3 of the PDA, which prohibits discriminatory practices at public venues but not under sec 4 which prohibits discriminatory advertising because the advertisement did not cause any of them actual harm (they learned about it only after the incident at the restaurant). The court of appeals upheld this ruling and the plaintiffs applied for leave to appeal to the Supreme Court, arguing that the amount awarded for the discriminatory practice was too small and that they were also entitled to ‘damages without proof of harm’ for the discriminatory advertisement under sec 5(b) of the PDA.

b) Judgment of the Court

14The Supreme Court denied the leave to appeal, concluding that the case did not meet the conditions for a second appeal because it did not give rise to a general legal question that transcends the facts of the particular case.

15According to sec 5(a) of the PDA, violations of secs 3 and 4 are civil wrongs subject to the general principles of the CWO. Section 3 of the CWO provides that people injured or harmed by torts committed in Israel are entitled to remedies from the wrongdoers. From this section, Justice Stein concluded that only a person harmed by a tort is entitled to legal remedies. Conversely, a person who was exposed to a wrongful conduct but did not suffer harm has no remedy. Thus, even if publishing the discriminatory advertisement is a tort, only those who incurred harm can recover damages. Section 5(b) enables only those who were harmed but cannot prove the extent of the harm (as opposed to those who did not incur any harm) to recover ‘damages without proof of harm’.

16In addition, Justice Stein doubted that the advertisement was discriminatory. In his view, offering special treats to women during the FIFA World Cup was a legitimate business practice based on experience and statistics rather than gender stereotypes. The defendant assumed that male patronage would decrease during the games and attempted to balance this by temporarily increasing female patronage.

c) Commentary

17Section 3(a) of the PDA stipulates that providers of products and public services, as well as operators of public places shall not discriminate in providing the product or service, permitting entrance to the public place or providing service in such place, on the basis of race, religion or religious group, nationality, land of origin, sex, sexual orientation, views, political affiliation, personal status or parenthood. Section 5(a) provides that violation of sec 3 is an actionable tort, subject to the general provisions of the CWO. Section 5(b) adds that in a claim brought for such violation, the court can award the victim up to NIS 50,000 without proof of harm.

18Turning to the general provisions of the CWO, which apply to claims under the PDA, sec 3 provides that any person injured or harmed by a tort defined in the CWO and committed in Israel is entitled to a remedy in accordance with the CWO. According to the prevalent view, this section lays down a private international law rule rather than a substantive tort law rule.^[8] It stipulates that Israeli law applies to torts committed in Israel and does not specify the preconditions for liability under Israeli law. Justice Stein’s notion that this section imposes a substantive restriction on liability is therefore dubious.

19More importantly, sec 3 applies to injury or harm. Injury is defined in sec 2 of the CWO as an unlawful infringement of a legal right, even without harm (traditionally known as injuria sine damnum). This wording emphasises a crucial point: there is no general harm requirement in tort law. Claiming, based on sec 3, that civil wrongs cannot be actionable without harm is analytically false.^[9] A thorough analysis of the common law of torts reinforces this conclusion. A clear distinction exists between harm-based and non-harm-based wrongs. Some of the most important torts, in particular negligence and breach of statutory duty, consist of an element of harm, and are not actionable in the absence of harm. Yet other torts, such as battery, assault, false imprisonment, passing off, and defamation (or at least their Israeli versions) do not require harm. Recently, I argued that the harm requirement should also be abolished in the case of negligence.^[10] Admittedly, this proposal does not reflect existing law, but the underlying analysis demonstrates that harm is not an inevitable element of a tort.

20Lastly, sec 3 speaks of entitlement to remedies, not to compensatory damages. Because Justice Stein interpreted sec 3 as requiring harm, his inevitable conclusion was that, in the absence of harm, there can be no remedy. This is incorrect. The use of the term ‘remedy’ is consistent with the inclusive reference to injury or harm (rather than just harm). If there is neither harm nor injury, there is no remedy. If there is no harm, there can be no compensatory damages. But the absence of harm does not in itself preclude all remedies: non-compensatory remedies, such as injunctions or punitive damages, may still be available. Strangely, after concluding that there are no remedies without harm,^[11] Justice Stein admitted that non-monetary remedies may be available without harm.^[12] This conceptual inconsistency is not explained and remains baffling.

21As regards the discriminatory nature of the advertisement, I agree that the special treatment of women was motivated by business considerations. But a business motivation does not in itself make a business practice non-discriminatory. Promising only women special treats is discriminatory under the PDA’s definition, regardless of the motivation. The defendant could adopt non-discriminatory practices serving the same goals, for example by offering ‘exclusively male tables’ similar treats. Assuming that women are not deterred by the presence of men at the restaurant, this could increase patronage even more.

22At any rate, the court’s reliance on statistical assumptions in assessing the business practice is problematic for two reasons. First, statistical data often reflects or replicates existing stereotypes and biases that the PDA aimed to eradicate. Relying on such data, even if accurate, might perpetuate the underlying stereotypes and biases and hinder progress. The ‘statistical assumption’ that men are much more interested in competitive sports, particularly the World Cup, might perpetuate gender-based stereotypes and exclude women. Second, ‘statistical assumptions’ are often subjective (potentially biased) intuitions rather than hard statistical data. For example, in 2018, women constituted 33 %–40 % of the viewers of FIFA World Cup games in Israel.^[13] The assumption that women were not interested in the games was simply untrue. Accordingly, women’s organisations demanded proper representation in relevant broadcasts, condemning the ‘anachronistic gender stereotypes’ reflected in all-male media coverage.

a) Brief Summary of the Facts

23A pedestrian was injured by an electric bicycle while crossing the road at a crosswalk and brought an action under the Road Accidents Victims’ Compensation Act (RAVCA) against the bike rider and against the Road Accident Victims’ Compensation Fund (Karnit). Under this scheme, a ‘motor vehicle’ user is liable for personal injuries caused in a road accident in which that vehicle is involved, this liability is covered by mandatory insurance, and, if the user is uninsured, Karnit takes the insurer’s place.

24Section 1 of RAVCA defines a motor vehicle as a vehicle propelled by mechanical power over land, whose main purpose is land transportation. The Supreme Court held in 2020 that an electric bicycle is not a ‘motor vehicle’ for the purposes of RAVCA, so an accident in which an electric bicycle injures a pedestrian does not give rise to liability under this Act.^[14]

25Justice Amit opined that the language of sec 1 does not provide a conclusive answer to the question of electric bicycles because they are propelled by a varying combination of mechanical power and human pedalling. Legislative history does not assist in the statutory interpretation either, because electric bicycles were not envisaged by the legislature. Section 1 must therefore be interpreted in light of the underlying goals of RAVCA and public policy, and these considerations lead to the conclusion that electric bicycles should not be regarded as motor vehicles. First, the fact that the use of electric bicycles does not require insurance undermines any attempt to impose liability for injuries caused by their use under RAVCA. Second, holding that electric bicycles are motor vehicles would secure compensation under RAVCA to pedestrians hit by these bicycles, but leave riders hit by cars without any remedy. In contrast, holding that electric bicycles are not motor vehicles secures compensation under RAVCA to riders hit by cars, but not to pedestrians hit by cyclists (who must resort to the general principles of tort law). Because accidents in which cyclists are injured by cars are much more common and acute than those in which pedestrians are injured by electric bicycles, securing compensation to electric bike riders is more important.

26Justice Mintz reached the same conclusion on different grounds (holding that the motor only assists the pedalling rider and therefore does not change the nature of the bicycle as a non-motorised vehicle), whereas Justice Barak-Erez concluded in dissent that electric bicycles are motor vehicles. The victim filed a motion for further hearing, arguing that the Supreme Court had set an important, difficult (as the controversy among the judges implies), and novel precedent in its prior ruling.

b) Judgment of the Court

27Section 30 of the Courts Act (Consolidated Version)^[15] provides that further hearing can be allowed if the Supreme Court sets a rule that conflicts with prior precedent or is sufficiently important, difficult, or novel to justify further hearing. These conditions are not met in the case at bar.

28Justice Hendel held that the Supreme Court’s 2020 decision was not inconsistent with prior precedent. The classification of electric bicycles has not been previously decided, and prior decisions concerning the interpretation of the term ‘motor vehicle’ have been discussed and respected. The marginal development of prior case law, though clearly new, does not in itself justify further hearing. Moreover, the controversy among the judges (with respect to the outcome and its justification) does not indicate an exceptional novelty. Finally, the 2020 majority’s conclusion was the more conservative (hence less ‘important, difficult, or novel’) option as it did not have a real impact on reality: liability insurance was and still is unavailable for electric bicycles, so pedestrians could not benefit from imposing liability on bicycle users under RAVCA anyway; automobile users have always been and still are liable when injuring cyclists, and their liability is covered by mandatory insurance.

c) Commentary

29The Court’s application of sec 30 of the Courts Act concerning further hearing seems generally compelling because the original 2020 decision is neither in conflict with prior precedent nor sufficiently important, difficult, or novel. However, the argument that the adopted interpretation does not change reality because of the lack of insurance coverage for electric bicycles puts the cart before the horse. It may well be that the lack of insurance reflects courts’ reluctance to classify electric bicycles as motor vehicles subject to the RAVCA liability regime. If lack of insurance is the result of no liability under RAVCA, it cannot serve as an excuse for denying such liability. Had the court classified electric bicycles as motor vehicles, insurance would have been mandatory and offered either in the private insurance market or through special arrangements (as in the case of motorcycles^[16]).

30More importantly, the original decision hereby upheld remains problematic. That decision, discussed in the 2020 report,^[17] relied on the need to avoid the worst of two evils. Classifying electric bicycles as motor vehicles would leave cyclists without redress when injured by automobiles, because motor vehicle users can only recover from their own insurers under the road accidents victims’ compensation scheme and electric bicycles are not currently insured. Not classifying electric bicycles as motor vehicles would leave pedestrians without redress when injured by electric bicycles because their injuries would not be covered by the road accidents victims’ compensation scheme and suing under the general principles of tort law is practically futile (in the absence of mandatory liability insurance). Justice Amit, with whom Justice Mintz concurred, concluded that securing compensation to cyclists hit by automobiles was more important because the automobile-bicycle scenario is more frequent and more acute than the bicycle-pedestrian scenario.

31Still, the court (in its 2020 decision) did not truly provide empirical evidence in support of the ‘worst of two evils’ argument. Justice Amit contended that only 10 % of the victims of electric-bicycle accidents are pedestrians.^[18] Alas, nowhere in his decision did he cite information about the relative frequency of the automobile-bicycle scenario. The fact that only 10 % of the victims of accidents involving electric bicycles are pedestrians does not mean that the remaining 90 % are automobile-bicycle cases, because cyclists can get injured in self-accidents. Data from other countries suggest that most injuries to electric bicycle users result from self-accidents.^[19] Moreover, victims of electric bicycle accidents are not always the respective cyclists or pedestrians. They may include passengers, users of traditional bicycles, etc. Not classifying electric bicycles as motor vehicles leaves many cyclists injured in self-accidents,^[20] passengers, and others, not only pedestrians, with no redress.

32Even if the automobile-bicycle scenario is more frequent and acute than the bicycle-pedestrian scenario, this fact in itself cannot justify an interpretation that addresses the former and neglects the latter. The relative frequency and impact of each type of accident depend on many dynamic variables, including the regulatory framework and the existing infrastructure. For instance, if electric bicycles can or must be used on shared-use paths or if cyclists must take precautions for their own protection (such as wearing helmets), the relative frequency of pedestrian injuries will increase and the relative frequency of automobile-bicycle accidents will decrease. In contrast, if electric bicycles can be used on roads only, cyclists will be exposed to a much greater danger whereas pedestrians will be relatively safe. Legislative interpretation cannot rely so heavily on facts which are the result of an existing regulatory scheme or an existing infrastructure, given their dynamic and malleable nature.

33The language of sec 1 is clear. A motor vehicle is a vehicle which is propelled by mechanical power and whose primary objective is land transportation. Electric bicycles are propelled by electric motors which constitute mechanical power, at times in addition to pedalling and at times exclusively. Consequently, they are motor vehicles for the purposes of RAVCA. The 2020 majority’s interpretation is counter-textual in failing to classify a vehicle operated by a motor (sometimes exclusively) as a motor vehicle. Although the Supreme Court generally favours a hybrid (subjective-objective) dynamic interpretation theory rather than textualism or originalism, it is understood that counter-textual interpretation should be highly exceptional and sufficiently justified. The current context does not present the exceptional circumstances which may justify deviation from the statute’s plain language, as Justice Barak-Erez explained in her 2020 dissent.^[21]

34In 2012 and again in 2013, the legislature attempted to amend some of the definitions in sec 1 of RAVCA. The Road Accidents Victims’ Compensation Bills,^[22] discussed in previous reports,^[23] intended to replace the rigid list of excluded vehicles with a reference to a more expansive and more flexible list in a new Appendix. The proposed Appendix included a wheelchair, an invalid carriage, an escalator, a mobility scooter, a segway, a motorised scooter, and a motorised bicycle. The Bills were supposed to empower the Minister of Justice to modify the list with the Minister of Finance’s consent, after consulting with the Minister of Transport.^[24] Neither passed the second and third readings, so the exclusion of motorised bicycles failed. This failure is important for two reasons. First, the fact that the legislature considered adding motorised bicycles to the list of excluded vehicles implies that, without such an amendment, motorised bicycles may be deemed ‘motor vehicles’. Second, in holding that electric bicycles are not motor vehicles, the Supreme Court did something that the legislature ultimately decided not to do, subtly transgressing the separation of powers.

a) Brief Summary of the Facts

35The petitioner-plaintiff’s fiancée was killed in a road accident. The two cohabited and were supposed to get married two and a half months after the event. The plaintiff brought an action for his alleged economic losses under the wrongful death provisions of the CWO. The court of first instance allowed the claim, and the court of appeals reversed, holding that the deceased did not support the plaintiff at the time of her death. The petitioner applied for leave to appeal to the Supreme Court.

b) Judgment of the Court

36A cause of action for wrongful death under secs 78–80 of the CWO arises only if two conditions are met. First, the plaintiff must establish that he or she is the deceased’s spouse, parent, or child (sec 78). According to Lindorn v Karnit,^[25] the term ‘spouse’ encompasses a married spouse and a common-law spouse. The tests for recognising common-law partnerships were previously in dispute. Today, a common-law partnership arises when two people intend, explicitly or implicitly, to subject themselves to the civil-economic implications of marriage and divorce without formally getting married. The parties’ subjective intent is determined by relevant evidence, including but not limited to cohabitation. Nowadays, people often choose to cohabit without an intent to form a marriage-like bond. Thus, for example, when two people are in the beginning of their relationship and have just started living together, possibly for a trial period, courts should be hesitant to classify them as common-law spouses. The law in a liberal State should not burden individuals with far-reaching obligations they did not intend to assume.

37Second, the plaintiff must demonstrate that he or she was an actual or potential dependant, namely that he or she was supported or expected to be supported in the future by the deceased (sec 80). Consequently, loss of a deceased spouse’s support can be established based on expected future income even if the deceased was not working at the time of death, because he or she was invested in rearing the children, on parental leave, studying, temporarily unemployed, etc.

38The court of first instance did not determine whether the plaintiff and the deceased were common-law partners in accordance with the relevant tests, and the court of appeals failed to consider loss of future support. Thus, the case should be remanded to the court of first instance for determining whether the two conditions for recovery under secs 78–80 were met.

c) Commentary

39Israeli law, in line with the common law tradition, does not impose liability for negligently caused relational economic loss (that is, economic loss consequent on physical injury to the person or property of another). Even though the Supreme Court has never laid down a general rule on this matter, it has rejected all claims for negligently caused relational economic loss in specific contexts.^[26] A few statutory exceptions exist, of which the most important is the right to recover for economic losses arising from the wrongful death of a close relative.

40In the Western legal culture, there are three methods for deciding which dependants are entitled to compensation following the wrongful death of another. The French model allows recovery by every person who incurred certain loss following another person’s wrongful death, namely every person who can establish factual dependence. Courts have gradually recognised additional categories of people who can establish certain loss, and currently allow actions by fiancé(e)s, unmarried spouses, and even extra-marital partners.^[27] The German model allows recovery only by people to whom the deceased was legally obliged to provide maintenance or render services.^[28] The Anglo-American model allows recovery only by relatives specifically enumerated in statutory lists. The list originally consisted of the deceased’s spouse, children, and parents, but has been extended in various jurisdictions.

41The Israeli wrongful death provisions follow the English model. According to secs 78–80 of the CWO, in the case of death, the spouse, parent, and child (‘the dependants’) of the deceased may recover damages for their consequent economic losses from the person responsible for that death. Section 2 defines the term ‘parent’ as including grandparents and stepparents, and the term ‘child’ as including grandchildren, stepchildren, adopted children, ‘illegitimate’ children, and even foetuses, but does not define the term ‘spouse’. The cause of action is dependent on the deceased’s cause of action in two respects. First, the dependants can bring an action only if the deceased would, if death had not ensued, have been entitled at the time of his death to damages for bodily injury caused by the defendant’s conduct. Second, if the deceased contributed to the occurrence of the fatal injury, the amount of damages awarded to the dependants must be reduced accordingly.^[29]

42In Israel, fiancé(e)s are not included in the statutory list of dependants. Therefore, the plaintiff was at pains to establish that he was the deceased’s spouse. The Supreme Court held more than two decades ago, in the seminal case of Lindorn,^[30] that the term ‘spouse’ encompasses not only married but also common-law spouses (including same-sex common-law spouses as the plaintiff in Lindorn). The importance of the current case lies in clarifying the conditions for recovery by common-law partners under secs 78–80. The Court emphasised that the plaintiff must establish both the necessary legal relationship and factual dependence. As regards the legal relationship, an intent to form a marriage-like bond (endorsing the civil-economic implications of marriage and divorce) is required. The public view or perception of the specific relationship is irrelevant. Furthermore, while cohabitation is indicative of such intent, it might be insufficient because in modern times people often cohabit without an intent to form a legal relationship. As regards factual dependence, the Court explained that either actual or potential (expected) dependence will suffice. This is crucial where the deceased spouse was not contributing to the household income at the time of death but expected to do so at a later stage.

a) Brief Summary of the Facts

43A patient suffering from bipolar disorder committed suicide by hanging herself from an outdoor gutter and jumping off a picnic table, which was placed underneath that gutter at a State psychiatric hospital. Her estate and dependants brought an action against the State, arguing that the hospital’s staff were negligent and in breach of statutory duties. The District Court of Haifa found neither negligence nor breach of statutory duties and denied the claims. The plaintiffs appealed to the Supreme Court.

b) Judgment of the Court

44All judges agreed that there was no negligence whatsoever in the deceased’s medical treatment and focused on the hospital’s alleged failure to take the necessary security and supervision measures to prevent her suicide. The court examined (1) whether the plaintiffs positively established the hospital’s negligence and (2) whether the burden of proof of negligence could be shifted to the hospital under the doctrine of res ipsa loquitur.

45Starting with the first question, the Court unanimously reiterated that the reasonableness of medical conduct is determined by comparing the patient’s expected harm to the cost of precautions required to reduce or eliminate the risk (the economic position) and that medical practice is an important but inconclusive indicator of reasonableness. Moreover, Justices Sohlberg and Stein agreed that two different kinds of negligence can be attributed to medical institutions and their staff. In cases of negligence in medical treatment (diagnosis, development of treatment plans, and implementation), medical practice will normally determine the standard of care, and only substantial deviation therefrom will be deemed negligent. In cases of negligence in administration and logistics (particularly allocation of budgets, equipment, and other resources), the court must set the standard of care based on its own analysis, and the expert’s testimony will only serve as one piece of evidence to be considered among all others. Justice Hendel rejected this distinction, holding that professional practice is always useful but never binding.

46The distinction is crucial in psychiatry. Subject to exceptional cases, courts should hold psychiatrists negligent in treating patients only if they substantially deviate from common professional practice. Otherwise, psychiatrists might be over-deterred and psychiatric treatment might become unavailable for those in need. In contrast, courts may use a traditional cost-benefit analysis to consider whether a psychiatric institution took reasonable administrative and logistical steps. Here, the risks of judicial error and over-deterrence are much smaller. The higher the probability that a mentally ill patient would attempt to commit suicide, the greater and costlier the administrative and logistical measures that the hospital is required to take to protect that patient.

47On the administrative-logistical level, Justice Stein added a secondary distinction between locked wards, open units, and non-hospitalised care. In locked wards, the duty to prevent suicide is almost absolute; the great risk that patients pose to themselves, and the special conditions of hospitalisation in such wards, entail very strict supervision. In open units, the risk of suicide is lower, and the conditions of hospitalisation are generally more lenient (making preventive measures more complex and expensive), so a lower level of supervision may be reasonable. In cases of non-hospitalised care, providing proper diagnosis, medication, and guidance would normally suffice. These general observations are always subject to the circumstances of each case. Justice Sohlberg accepted this distinction in principle but opined that it was unnecessary to consider the standard of care in types of cases that were not discussed here.

48At this point the judges diverged. Justice Sohlberg (with whom Justice Hendel concurred) explained that imposing liability for omission entails proof that a specific omission was unreasonable. Without knowing which omission was negligent, it is also impossible to discuss and decide questions of causation. Additionally, the cost of precautions that could have reduced the risk of suicide, such as additional supervision or removal of picnic tables, trees, or gutters from the hospital yard, would be prohibitive considering their effects on the welfare and recovery of patients in the open unit.

49Justice Stein concluded in dissent that while the medical treatment was non-negligent, there was negligence on the administrative-logistical level. The probability of a suicide attempt at the hospital was high (greater than 50 %) given the patient’s bipolar disorder and recent suicide attempt. Consequently, the hospital was required to take steps to prevent further suicide attempts. Simple supervision through patrolling the hospital yard or installing video cameras may suffice, but the hospital did nothing on top of its ordinary psychiatric institution routines.

50Turning to the second question, the doctrine of res ipsa loquitur (embodied in sec 41 of the CWO) applies if three conditions are met: (1) the plaintiff had no knowledge or means of knowledge of the actual circumstances which caused the occurrence which led to the damage; (2) the damage was caused by some property of which the defendant had full control; (3) it appears to the court that the happening of the occurrence causing the damage is more consistent with the defendant having failed to exercise reasonable care than with his having exercised such care.

51Justice Sohlberg concluded that the first condition is not met. Res ipsa loquitur shifts the burden of proof when the circumstances leading to the injury are unknown; it does not generate a normative presumption of negligence when the circumstances are known. The events leading to the deceased’s suicide in this case are known and the measures taken by the hospital in accordance with its protocols are also known. In the absence of factual ambiguity, the doctrine does not apply. In addition, the third condition is not met. A patient’s suicide during hospitalisation in a psychiatric institution does not indicate negligence because, according to the scientific literature, mentally ill patients frequently attempt to commit suicide during hospitalisation regardless of any steps taken by the hospital.

52Justice Stein opined in his dissent that the res ipsa loquitur doctrine applies. The first condition is met because the deceased could not know what the hospital did or did not do to prevent her suicide. The second condition is met whenever a person is injured during treatment at a hospital, because the hospital controls all facilities, equipment, medication, and other objects from which the risk of injury could arise (here the table, the gutter, the sheet used for hanging). The death of a patient who was known to be suicidal at a psychiatric hospital also satisfies the third condition. The burden of proof thus shifts to the hospital, which did not establish its lack of negligence.

c) Commentary

53The case discusses the substantive tests for reasonable conduct and the possible shift of the burden of proof under the res ipsa loquitur doctrine. The dominant definition of reasonableness in Israeli law has been economically oriented, integrating the probability of harm, the extent of harm, the cost of precaution, and sometimes also the utility of the risk-creating activity.^[31] Even though several judges tried to diverge from or qualify this definition,^[32] the current case reinforces the traditional stance.

54The economic definition can be traced back to a series of American cases decided by Judge Hand in the 1940 s. Hand famously related three variables in an algebraic inequality: If the probability of harm is labelled P, the severity of harm L, and the burden of precautions needed to eliminate the risk of harm B, ‘liability depends upon whether B is less than L multiplied by P: (that is), whether B < PL’.^[33] Put differently, failure to take cost-justified precautions is negligent.^[34] Imposing liability on negligent injurers forces potential injurers to take into account, or internalise, the externalities of inefficient conduct, thereby preventing such conduct.^[35] According to economic wisdom, this deterrence of unreasonable risk is the primary objective of tort liability.^[36] Traditionally, the cost of precautions for the defendant was compared to the victim’s expected loss. The novelty of the current case seems to be Justice Sohlberg’s inclusion of the benefits for third parties (other patients at the hospital and their families) of not taking certain measures in the ‘cost of precautions’ compared to the victim’s expected loss. This development seems consistent with economic theory.

55In discussing the role of professional practice in determining the standard of care, Justice Stein put forward a new distinction, based on his decade-old scholarly article.^[37] According to Stein, when medical treatment is considered, professional practice generally determines the standard of care, whereas when administration and logistics are considered, evidence on common practice is considered with all other available evidence in the court’s cost-benefit analysis of the defendant’s conduct. Justice Sohlberg endorsed this distinction, but Justice Hendel rejected it, holding that professional practice is always useful but never binding. I tend to agree with Justices Stein and Sohlberg that evidence on professional practice must be assigned much greater weight in the case of medical treatment, where the court does not have sufficient expertise to challenge professional decisions and actions, and the risks of judicial error and over-deterrence are real. This does not mean that compliance with medical practice can never be deemed negligent, only that finding negligence in such circumstances would be exceptional.

56Another contribution of the case to the development of tort law is Justice Stein’s distinction (also accepted by Justice Sohlberg) between locked wards, open units, and non-hospitalised care for the purpose of determining the extent of administrative-logistical measures required from psychiatric institutions. A more intensive mode of treatment usually indicates a greater underlying risk to the patient on the one hand and the availability of additional precautions on the other hand, giving rise to an expectation of a higher level of supervision and security.

57With respect to res ipsa loquitur, several comments are in place. To begin with, according to Fogel v The Municipality of Tiberias,^[38] the ‘lack of knowledge’ condition is met only if the facts are obscure. Once the plaintiff presents a detailed account of the circumstances leading to the injury, he or she can no longer establish lack of knowledge or means of knowledge of these circumstances: there is no obscurity which justifies the application of sec 41. This rule applies even if the court finds the plaintiff’s account of the events inaccurate or untrue, as in Fogel itself. In summary, sec 41 can apply only if the facts are obscure. It cannot apply here, as the majority concluded, because the facts were known to all parties. Note further that the plaintiff must establish ‘lack of knowledge’ at the time of the trial, not at the time of the occurrence which led to the injury.^[39] This relatively recent interpretation generally favours defendants, because things that were unknown at the time of the occurrence often transpire at a later stage. It has no real impact on the current case because the circumstances leading to the injury were known from an early stage.

58Next, according to previous case law cited by the Supreme Court, the second precondition (full control of the property that caused the injury) is almost automatically met if the harm was caused during the plaintiff’s hospitalisation, because the doctors have control over the totality of the facilities, equipment, medications, and any other objects used at the hospital, and any injury to a patient can be attributed to the use of these things.^[40] This liberal interpretation of the full control requirement was not required in the case at bar, since the death was directly caused by specific items which were under the hospital’s control (the sheet, the gutter, and the table).

59Lastly, the probabilistic (third) condition is met if the occurrence of injury is more consistent with a finding of negligence than with a finding of non-negligence. The occurrence of injury does not in itself satisfy this condition, and statistical data can and should be used to assess whether the particular type of incident usually occurs with or without negligence. The fact that many of the mentally ill patients in psychiatric institutions attempt to commit suicide (and often succeed) regardless of any measures taken by the hospital staff appears to demonstrate that a suicide attempt is not more consistent with negligence than with lack thereof. Again, the majority’s view on this matter is more compelling.

a) Injuries Caused by Military Action

60According to sec 5 of the Civil Wrongs (Liability of the State) Act,^[41] the State is not liable in tort for an act performed through a ‘wartime action’ of the Israel Defense Forces (IDF). The term wartime action is defined in sec 1 as including any action of combatting terror, hostile actions, or insurrection, and any action intended to prevent terror, hostile actions, or insurrection and involving danger to life or limb.

61In Estate of Eleish v State of Israel,^[42] the plaintiff’s daughters and niece were killed when a tank fired two shells at their house during Operation Cast Lead in 2009. He sued the State of Israel for the alleged misconduct of the Israel Defense Forces. The court concluded that the incident was part of a military operation against terrorists in Gaza, and that the shells were fired at the house because hostile observers were believed to operate in it. In denying liability based on the wartime action defense, the Supreme Court made three important observations. First, the fact that no one attacked the military force at the time of the event underlying the claim does not bar the application of sec 5. One cannot break a military activity down into discrete events and artificially isolate a single component, regardless of the overall context and factual continuum, for the purposes of the wartime action defense. The court must assess the activity as a whole. Second, the defense under sec 5 is substantive and absolute. There are no exceptions based on the degree of the defendant’s fault, so even if the military force was negligent, grossly negligent, or reckless, no liability can arise. Third, the wartime action defense bars claims not only under domestic private law but also under international law.

62In State of Israel v Estate of Shawamra,^[43] a teenager was killed by an IDF unit when approaching the Israeli West Bank Barrier and failing to respond to warnings. His estate brought an action against the State. The court of first instance rejected the claim, holding that the military conduct was intended to prevent attempted sabotage of the barrier, and therefore constituted a wartime action. The court of appeals reversed, ordering the lower court to examine whether the military conduct was negligent and reconsider its classification as a wartime action. The Supreme Court concluded that the court of appeals erred. According to sec 5(b) of the Liability of the State Act (added in 2012), if the State invokes the wartime action defense, the court must decide this matter immediately, and deny the claim upon finding that the defense applies. The applicability of the defense must be discussed and decided before any other legal question. Thus, the court of appeals should have discussed the defense before requiring the lower court to evaluate the military’s alleged negligence.

b) Road Accidents Victims’ Compensation

63As explained in Part A above, personal injuries caused by road accidents are governed by the Road Accidents Victims’ Compensation Act (RAVCA). The interpretation of the Act’s provisions has burdened the courts despite the legislature’s explicit intent to create a simple road accident victim compensation scheme. Izhak Englard, author of one of the main treatises on the subject, observed that thousands of cases, of which hundreds had reached the Supreme Court, discuss the interpretation of the Act’s provisions.^[44] The Israeli report cannot cover all relevant cases, but it usually offers commentary on the most important ones and mentions some of the others in the ‘personal injury’ section.

64Recall that the compensation scheme has three pillars: (1) motor vehicle users’ strict liability for personal injuries caused in road accidents; (2) mandatory insurance covering users’ liability and personal injuries; (3) a supplementary fund (Karnit), covering cases in which liability insurance is unavailable. Motor vehicle users are liable for personal injuries caused in road accidents in which their vehicles are involved and are also covered by the mandatory insurance for personal injuries they sustained in road accidents. The term ‘road accident’ has a complex statutory definition in sec 1 of RAVCA. Under the main provision, a road accident is an event in which bodily injury is caused as a result of the use of a motor vehicle for the purpose of transportation. The concepts used in this definition (‘motor vehicle’, ‘use of a motor vehicle’, and ‘bodily injury’) are also defined in sec 1. Courts often need to determine whether a particular event constitutes a road accident.

65For example, in Shlomo Insurance Company v John Doe,^[45] the plaintiff slipped and fell off the back of his truck when trying to close the fuel cap after filling up the tank. He sued the truck’s insurer. The Supreme Court held that the plaintiff’s acts cannot be subsumed under any of the statutory instances of ‘use of a motor vehicle’, particularly ‘driving’, ‘entering or exiting’ a vehicle, and ‘roadside treatment or repair’. Filling up a fuel tank enables driving but is not a natural and integral part of the driving or entering, as it is performed after the vehicle parks and the driver fully exits it and before the driver starts to re-enter. It is not a roadside treatment either because it is not necessitated by a sudden event (as required by prior case law). Therefore, an event in which such conduct causes personal injury is not a ‘road accident’ for the purposes of the compensation scheme.

66In 2021, the Supreme Court also discussed Karnit’s right to reimbursement from uninsured drivers. In Karnit v Jane Doe,^[46] a minor was injured while crossing the road when an uninsured vehicle hit and threw her at her mother’s (also uninsured) vehicle, which was waiting after drop-off at a no-parking area. The minor sued the uninsured driver (hereinafter D1) and Karnit. The latter sued D1 and the mother, arguing that both were involved in the accident and, as uninsured drivers, had to reimburse it for any payment to the victim. The case illustrates two important principles in the road accident compensation scheme.

67First, according to sec 3(b) of RAVCA, if a person was injured outside a vehicle in a road accident involving several vehicles, the drivers are jointly and severally liable for the injury and bear equal shares of the burden. A vehicle is deemed involved in the accident if, at the time of the accident, there was contact between that vehicle and either another vehicle or the victim (sec 3(b)). Both vehicles in this case were involved in the accident because both were in contact with the victim and therefore both drivers are liable. Because neither was insured, Karnit covered the loss.

68Second, according to sec 9(a)(2) of RAVCA, one who has paid compensation due under the Act has a right of recourse against a person who is liable for the same injury under the Act but has no insurance coverage. The outcome was that the victim’s uninsured mother (along with D1) had to reimburse Karnit for the payments to her own daughter. In addition, the Supreme Court reiterated the general principle in the law of damages, whereby the loss of earning capacity of a minor is assessed based on the national average wage.

c) Infringement of Autonomy

69In 2021, the Supreme Court continued developing and applying the unique doctrine of damages for infringement of autonomy. This head of damages originated in the case of Daaka v Carmel Hospital.^[47] In that case, the plaintiff sued the defendants for injuries caused by a surgery. The defendants were found negligent in not obtaining the plaintiff’s informed consent (although they were not negligent in performing the surgery or in the subsequent medical treatment). The court concluded that there was no causal link between that negligence and the bodily injury because if the plaintiff had received all relevant information, she would have given her consent. Put differently, the plaintiff could not establish that, but for the negligence, the injury would have been prevented. The court held, however, that every person has a fundamental right to autonomy and self-determination, which encompasses the person’s right not to endure any bodily intrusion without his or her consent and that infringement of that right is a compensable harm. The court thus recognised a new head of damages.^[48] While this head of damages originated and is frequently used in personal injury cases, especially where the causal link between the negligence and the personal injury is uncertain, it is not limited to such cases.

70In John Doe v HaEmek Medical Centre,^[49] the plaintiff was admitted to the defendant’s hospital following a workplace accident and the surgery at the hospital resulted in permanent disability. The court found that the surgery and the pre- and post-surgery procedures were carried out without negligence. However, the plaintiff did not receive the required explanations before giving his consent, so his autonomy was infringed. The Supreme Court held, first, that damages for infringement of autonomy are awarded for the deprivation of an informed choice, regardless of the plaintiff’s expected decision with full information. Thus, the plaintiff can recover even if he would have decided to undergo the surgery had he been given all relevant information. Second, although damages are awarded for the subjective emotional reaction to the deprivation of informed choice, such a reaction is assumed whenever autonomy is infringed. Third, any benefits that the victim is expected to receive from the National Insurance Institute (NII) following a particular injury must be deducted from tort damages awarded to the victim for the same injury. Because any injury incurred at the hospital could be causally attributed to the initial workplace accident, NII benefits for the accident had to be deducted from any tort damages payable by the hospital, including those for infringement of autonomy. Expected NII benefits considerably exceeded the extent of damages for infringement of autonomy, so the hospital had to pay nothing.

71In Jane Doe v Clalit Health Services,^[50] the Supreme Court was asked to decide whether damages can be awarded for infringement of autonomy of a foetus. The defendant’s doctors informed the plaintiff’s mother that her foetus suffered from the Turner syndrome and advised termination of the pregnancy. The termination procedure failed, and the plaintiff was born with disabilities related to her premature birth. Subsequently, it transpired that the doctors did not provide accurate information to the mother: the plaintiff did not truly suffer from the Turner syndrome but from a milder condition. The court of first instance concluded that the parents would have decided to terminate the pregnancy even if they had received accurate information. Thus, the disabilities caused by the premature birth could not be attributed to the doctors’ negligence. However, the mother’s autonomy was infringed by the doctors’ failure to provide accurate information. The question on appeal was whether the foetus’s autonomy was also infringed.

72The Supreme Court answered in the negative: a foetus is not a legal person and cannot have any legal rights. Thus, it does not have a right to autonomy that can be infringed. More importantly, a foetus (as opposed to the parents) does not and cannot make informed choices concerning its future and therefore cannot be deprived of the power to make such choices. Furthermore, a foetus cannot sustain the subjective emotional reaction that underlies a claim for infringement of autonomy. Finally, the infringement of the parents’ autonomy cannot be extended to the foetus based on their status as natural guardians. Even if the parents were the natural guardians of the foetus (which is questionable), guardians can only protect the rights and interests of their wards; they cannot confer their own rights and interests on their wards.