Chapter 7
International Air Carrier Liability and Social Perceptions of Risk: Third Parties on the Ground

I. A Brief Overview

7.1 In 1933, the Convention for the Unification of Certain Rules Relating to Damage Caused to Third Parties on the Surface (1933RC) was signed in Rome in order to uniformly regulate liability for damage caused by aircraft to third parties on the surface, and to provide for a compulsory system of third party liability insurance. However, the limited number of countries that ratified the 1933RC led to its replacement in 1952 by the Convention on Damage Caused by Foreign Aircraft to Third Parties on the Surface (1952RC) that was also signed in Rome. Furthermore, the Protocol to Amend the Convention on Damage Caused by Foreign Aircraft to Third Parties on the Surface Signed at Rome on 7 October 1952 (1978 Protocol) was signed in Montreal in 1978.

7.2 The aim of this chapter is to demonstrate the influence of intentional, man-made risks on the legal standards governing the international liability of air carriers for damages to third parties on the ground. Risk allocation achieved in these Conventions was essentially designed to satisfy social perceptions as expressed in domestic legislation to the undermining of the technical realities of aviation. In contrast to the WCS and the MC, the Rome Convention “System” (RCS)1 was not attempting to balance the interests between the parties involved, but predominantly was trying to protect third parties on the ground in an internationally uniform way. This legislative lack of cooperation between subjective and objective elements of aviation risk is arguably the main reason behind the limited acceptance of the RCS.

7.3 Furthermore, the two Conventions – Convention on Compensation for Damage Caused by Aircraft to Third Parties (General Risks Convention) and Convention on Compensation for Damage to Third Parties, Resulting from Acts of Unlawful Interference Involving Aircraft (Terrorism Convention) – which were recently drafted as a legislative response to the events of 9/11 are arguably missing a similar point. In a risk society, the main aim of an instrument of private international air law should be the restoration of the protective cocoon that is shattered by the impact of intentional, man-made risks. This requires the consideration of broad behavioural and technological factors that are not always compatible with theories of channelling liability to air carriers.

II. The Rome Convention “System”

1933 Rome Convention: introduction

7.4 In the case of liability of air carriers to passengers, the adoption of a presumed liability regime “was the product of an historical bargain”:2 to reconcile the need to protect the financial interests of carriers in the 1930s and to shelter passengers from the operational and technical complexities of air travel.

7.5 When it comes to the liability of air carriers for damage to third parties on the ground, the bargain was different. During a time when a paternalistic social environment was taking over the laissez faire approach of the industrial era, the deliberations of the 1933RC followed a different path to the travaux préparatoires of the WC, and engulfed near-absolute liability as the basis of the liability system. This time, the starting point was not the industry, but the need to protect the innocent bystanders on the ground.

The liability scheme of the 1933 Rome Convention

Domestic perceptions of ultra-hazardous activity influence international drafters

7.6 Josef Koval provides a synopsis of the reasoning behind the imposition of near-absolute liability on the operator of the aircraft:

  • “(a) inequality of the situation between the person causing the injury and the person receiving it;
  • (b) the impossibility of the person injured to show proof of the fault of the aviator;
  • (c) the use of a machine which creates a new risk for the life and material welfare of the public.”3

7.7 Similar arguments were put forward during the drafting of the WC. They were ignored because they run contrary to the aim of drafting a convention that would balance the interests of air carriers and passengers. However, in the case of the 1933RC, this suggestion was accepted, albeit reluctantly at first.4 This difference of approach can be explained if one looks into domestic legislations at the time the 1933RC was drafted: the vast majority of states were imposing a system of near-absolute and unlimited liability on the aircraft owner and/or possessor.5

7.8 The considerations behind this domestic law choice were similar to the suggestions of the drafters of the 1933RC: “… aviation presents a unique situation from the one-sidedness of risk, the one-sidedness of the ability to minimize the risk, the singularly favourable position of the aircraft operator to distribute the loss with the least social hardship”.6 It becomes apparent that as early as the 1930s, primitive theories of enterprise liability, together with the widely held social belief that aviation was an ultra-hazardous activity, influenced the drafting of domestic laws, as well as the 1933RC. In both cases the subjective reality of aviation risks took precedence over the technical reality that statistically questioned the classification of aviation as an inherently hazardous activity and called for the application of fault as the determining liability standard.

7.9 Admittedly, it was still difficult in the early 1930s to persuade the public that aviation was as safe as carriage by train. Statistical analysis could provide some comfort regarding the unusually hazardous nature of carriage by air. Yet it could not attenuate the fears of the public as to the extensive effects of a fatal aircraft accident, both in terms of its passengers and damage on the surface. It is true that the judiciaries in the UK and US were providing considerable liability immunity to air carriers during the early years of aviation. Yet this immunity was directed to the liabilities arising from the carriage of passengers on the basis that the intentional use of an unproven, new means of transportation justifies a wide-ranging assumption of risks by passengers.

Near-absolute liability as a cost-benefit test

7.10 In the case of liability for personal injury to persons on the ground, air carriers could not invoke risk-sharing as a reason for obtaining preferential treatment: “aviation creates considerable risks even for persons who take absolutely no part in the exploitation of air traffic, and … such persons must enjoy full protection”.7 As such, both domestic legislations and the 1933RC embraced near-absolute liability and utilised it as a general basis for the allocation of risks.8 At the same time, the defences available to carriers were limited to the fault of the injured person,9 and domestic legislations imposed unlimited liability. In contrast, the 1933RC provided for breakable limits of liability in cases of “gross negligence or wilful misconduct on the part of the operator and its agents”. The limits were based upon the weight of the aircraft.10

7.11 This way, persons on the ground were relieved of the requirement to prove fault and stood more chances of recovery than passengers, since carriers were not given the opportunity to demonstrate that they took “all necessary measures” to avoid the injury or death. Obviously, the default of the claimant as a defence was not providing any meaningful protection to air carriers, and even more noticeably, the force majeure defence was missing from the Convention. In an era when the causes of aviation accidents remained more often than not unexplained, the force majeure defence could have played the role of the “all necessary measures” defence in the WCS: it would require carriers to prove that the injury or death was caused by an event utterly unforeseeable that could not be managed by any means whatsoever.

7.12 This is a major hurdle to overcome today, yet at the beginning of the twentieth century the limited technical knowledge would have made it easier for air carriers to establish a complete lack of fault on their part. This way, the risk of an injury or death caused from unforeseeable circumstances would have been transferred to the persons on the ground. As such, carriers would have been in a position to control the formation of the notion of foreseeabilty through their technical progress.

7.13 The absence of any such defences in the 1933RC created a near-absolute liability regime that in effect barred carriers from transferring any risks to the people on the ground, and forced them to internalise any costs arising from the industry’s development. This liability scheme contrasts sharply with the scheme of the WC, which was created a few years earlier to shape, rather than follow, the contemporary social perceptions of aviation risk. The 1933RC adopted a more pragmatic approach to legal risk management in that it allocated risks based upon the demands of society at the beginning of the century when technical achievements of aviation were admired, but most importantly the public was concerned about its catastrophic potential.

7.14 By imposing near-absolute liability on air carriers, its drafters gave voice to lay persons’ estimations of this high-risk activity. For them, permitting carriers to transfer risks to parties that stood to gain nothing from this activity was unacceptable. Aviation had not proven its social value yet, and as such any protection afforded to the industry was running contrary to the policy considerations expressed in domestic legislations.

7.15 This approach, arguably, satisfied the paternalistic social environment of the first half of the twentieth century. At the same time, it attempted to make more apparent that the fault-based liability system of the WC was a liability scheme justified by the special relationship between passengers and carriers. However, it did not match contemporary social attitudes towards aviation risk. The drafters explicitly rejected the presumption of the carrier’s fault as the basis of the liability system, although its practical effects would have been similar.

7.16 In both schemes, the carrier bears the burden to provide evidence that he was not at fault, which would have been sufficient protection to the victim on the ground. However, the near-absolute liability system was preferred on the basis that it would provide a strong paternalistic tone to the 1933RC and also enable the assessment of the social usefulness of flying. If this assessment was positive, the expected formation of a relationship of trust between carriers and the public could enable the lowering of the liability standards from near-absolute to strict or even fault based.

7.17 This way, a revised version of the “wait and see” approach of the WC was adopted: “wait and see” whether carriers will be able to survive the zero accident policy that the 1933RC was imposing. No immunities or transfer of liability risks were provided, and carriers were required to prove that they could manage the risks that flying imposed to third parties on the ground, without resorting to any “legal subsidisation”.

7.18 One might argue that the imposition of liability limits diluted the strong paternalistic overtones of the 1933RC, and provided air carriers with a level of protection equivalent to the WC. Admittedly, the liability limits were provided as a quid pro quo for the near-absolute regime, yet their usefulness for the carriers is seriously questioned. The absence of any meaningful defence up to the limits would have put such a strain on the ill-defined terms “gross negligence” or “wilful misconduct” that breaking the limits would not have been as difficult as the drafters planned.

7.19 Bearing in mind the extent of damages that can accrue on the ground as the result of an aircraft crash, it becomes apparent that safeguarding the liability limits of the 1933RC was more significant for carriers than preserving the limits of the WC. The provision for breakable liability limits was not a genuine attempt to balance the interests between carriers and third parties on the ground. Instead, it was a way of making the 1933RC more appealing to carriers without actually hindering the claimants’ prospects of recovery. The injection of truly individualistic elements into the 1933RC would require the imposition of unbreakable, albeit higher, limits of liability that could offset “the excessively ‘fatal’ character and [the] lack [of any] stimulating factor for prevention” that the near-absolute liability scheme had.11

7.20 Furthermore, the near-absolute liability scheme prevented the 1933RC from having any risk-informative role. On the one hand, the WC, with its presumption of liability, enabled the formation of behavioural standards that were unique to international aviation and also facilitated courts by shedding light on air carriers’ operations. On the other hand, the 1933RC focused upon the outcome of flying, ignoring the operational realities of carriage by air at the beginning of the twentieth century. Apart from imposing an onerous liability system on air carriers, the near-absolute liability system did not encourage courts to look into the efforts of air carriers to improve their safety management systems. Ironically enough, the liability system of the 1933RC meant that courts would interfere less in the operation of air carriers, giving them more flexibility when it comes to the management of their risks to third parties on the ground.

Near-absolute liability as an impediment to creating a trust relationship between air carriers and third parties on the ground

7.21 This “hands-off” approach meant that technical and lay perceptions of aviation risks would remain apart. In an era when aviation risks were reasonably manageable by technical means, yet socially unacceptable, the creation of a long-term trust relationship between carriers and third parties on the ground was of the utmost importance.

7.22 However, the formation of this relationship could not be achieved through an instrument of international law that adopted a passive approach to the management of risks, and did nothing but replicate the contemporary social environment. Instead, it required a legal regime that would try to shape the exact norms that it is supposed to follow – in a similar fashion to the drafting of the WC.

7.23 The similarity in approach does not necessarily mean that the end result needed to be the same. It signifies that a Convention that attempts to regulate a pioneering activity on an international level is required to treat the risks arising from this activity in a manner that does not hinder its future development. It then becomes inevitable that the drafters are required to find a balance between retaining and transferring the risk of the activity that will take into consideration the prospects of the activity and will not just be a summation of domestic legislations.

7.24 It is understandable that the drafters wanted parties on the ground to be treated better than passengers. Yet the near-absolute liability system did not permit any kind of balancing between the various risk treatment techniques that would give to the 1933RC the required flexibility to shape social perceptions and follow the foreseeable technological development of the industry.

7.25 On the contrary, this liability system requested that air carriers retain risks and then find ways to eliminate them. Bearing in mind the level of technological progress and the financial fragility of air carriers at the time, it becomes evident that the elimination of risks flying posed to third parties on the ground, or their reduction to the extent that they were posing insignificant threats, was not a realistic aim. It might have been what the public expected from air carriers, or what the regulators expected from aviation technology, yet the role of air carriers in protecting the non-flying public was undermined, if not completely ignored.

7.26 This way, the 1933RC achieved a just allocation of risks between innocent bystanders and carriers. Yet it was unable to contribute to the creation of a trust relationship between carriers and the general public, since it was keeping the objective and subjective reality of aviation well apart. This discrepancy was becoming more obvious by the conciliatory approach that the WC adopted. Although liability of carriers to passengers is a legal field distinct from liability to third parties on the ground, it is inevitable that these two legal instruments that were drafted around the same time were competing over the appropriate legal standard to govern aircraft operations. Having conflicting legal standards regulating two sub-species of the same activity in an international level was a luxury at the time: the financial and operational state of the industry could not afford reactive management of technical risks when it came to passengers and a proactive approach when it came to third parties on the ground.

7.27 In an era when aircraft crashes into the ground due to technical failures were the main sources of liability to passengers, any protection afforded to the carriers by the WC was automatically taken away by the 1933RC. This way, the decision of whether to implement the 1933RC into domestic legislation was exceeding its exact provisions and became a choice between: (i) a legal system that left the existing balance between technical risks and their social perceptions intact; and (ii) a legal system that was designed to control social perceptions until the improvement of the technical reality enhanced the social acceptability of aviation risks.

7.28 The international community at the beginning of the twentieth century opted for the second option, arguably sensing that the benefits arising from flying would eventually outweigh the costs, but also under the influence of aviation’s early romanticism. At this point, international carriage by air was in need of a unified protective approach that would have been seriously damaged by the 1933RC: the 1933RC could have unleashed strong arguments in favour of applying a standard of near-absolute liability in the carriage of passengers – a proposal that was made during the drafting of the WC but was rejected on the basis that it was rather harsh towards carriers.

Social perceptions v. technical perceptions of risk: 1933 Rome Convention v. Warsaw Convention 1929

7.29 The debate between the 1933RC and the WC demonstrates that the everlasting conflict in aviation between technical demands and social needs was transferred into the legal field from an early point in the history of aviation. Within four years from the drafting of the WC, an instrument of international law changed priorities and focused upon social needs. The 1933RC might have been an intentional attempt to counteract the industry-protective effects of the WC. Yet the refusal to implement it domestically was a strong signal that the international community was prepared to grant legislative immunity to air carriers despite the paternalistic social environment.

7.30 At the same time, it demonstrated the delicacy of the balance achieved in the WC. The WC managed, through legal techniques such as the reversal of the burden of proof and the limitation of liability, to protect the interests of carriers enough to allow them safety improvements without jeopardising their financial stability. However, it did not prevent voices arguing for stricter liability standards from finding a forum in the 1933RC. Even the limitation of liability “was reluctantly adopted” in order to make the 1933RC appealing to “countries most active in international air carriage [that otherwise] would probably reject any convention dealing with such liability”.12

7.31 This further shows how difficult it is to contain social perceptions of risk and create a legal system that defies the prevailing levels of risk acceptability. Although the 1933RC was not accepted to the extent that its drafters expected, it provided a form of expression to the contemporary popular feelings of concern for the consequences of this unknown new activity. Any further developments had to wait until the end of WWII, which had an impact on both the technical growth of aviation and the way it was perceived by the public.

7.32 As we know, the post-WWII amendments of the WC failed to adapt fully to this new socio-technological aviation reality. This failure has arguably contributed to the dismantling of the international regime governing the liability of air carriers to passengers. The next section of this chapter will attempt to analyse how the 1952RC – the successor to the 1933RC – regulated the liability of air carriers to third parties on the ground in the post-WWII world, focusing on the management achieved therein of the technical realities of aviation and the social perception of aviation risks.

1952 Rome Convention: introduction

7.33 The aim of the 1952RC is no different than the WCS: “ensure adequate compensation for persons who suffer damage caused on the surface by foreign aircraft, while limiting in a reasonable manner the extent of liabilities incurred for such damage in order not to hinder the development of international civil air transport”.13 Again, the mandate of the Convention’s drafters was to balance the interests of third parties and air carriers. This express reference to the need of achieving a balance further suggested that the 1952RC was not intended to be a mere amendment to the 1933RC.14

7.34 Instead, it was intended to go a step further and create a liability system that would accommodate the characteristics of the post-industrial society that was under formation in the aftermath of WWII: on the one hand, safety improvements that made carriage by air critically safer and financially stable; and on the other hand, the prevalence of paternalism which demanded greater protection for the public and instilled doubts about the social responsibility of air carriers.

7.35 In that way, the drafters of the 1952RC were in a position similar to the position of the drafters of the HP, yet at the same time their role had substantial differences. The similarity is that both were assigned the task of reallocating the risks between carriers and the public to reflect the changes in the post-WWII world. However, in the case of the 1952RC, the basis of the deliberations was a paternalistic legal instrument that required an insertion of individualistic elements to make it more appealing to carrier-fare nations. In sharp contrast to the drafters of the HP, who were required to translate the changes in social perceptions of aviation risks to a legal instrument, the drafters of the 1952RC were more concerned with aviation technology and how its development could affect the near-absolute liability regime of the 1933RC.

The liability system of the 1952 Rome Convention

US social and technological perceptions of aviation risk fail to influence the international liability choices

7.36 A proposal to change the near-absolute liability standard in favour of a presumed fault standard came in the Conference on Private International Air Law that led to the drafting of the 1952RC from the country at the cutting edge of aviation technology, namely the USA:

7.37 This suggestion did not come as a surprise, and can be explained both from legal and socio-technological points of view. During the 1920s, 24 states in the USA, influenced by the now obsolete Uniform Aeronautics Act, had statutes imposing near-absolute liability on the part of the owner or lessee of an aircraft for damage caused to property or persons on the ground.16 In 1943, the Act was withdrawn by the Commission on Uniform State Laws, and the number of states imposing negligence liability steadily increased.17 US jurisprudence was quick to endorse this legislative change of direction: as Judge Peters demonstrated in Boyd v. White,18 both courts and legislature in the 1950s were of the same mind on the risk-characterisation of flying.19

7.38 The changes in both judicial and legislative attitudes did not come alone, but were the results of the much improved safety standards of US air carriers. In Chapters 3, 4 and 5, it was noted that the technological rationality of the post-WWII era might have reduced accident rates, but increased passenger-related liability exposures of carriers, since the accidents incurred were more easily attributed to the fault of air carriers. Yet when it came to liability for damages on the ground, technological rationality played a significant role in the rationalisation of US domestic legislation:

“Over 16,000,000 take-offs and landings were made in the USA during 1951 without even one fatality to persons on the surface … There were only six accidents [which involved fatalities to persons on the ground in the United States from March 1946 to 7 March 1953] and there were only four more such accidents in which persons on the ground were non-fatally wounded.”20

7.39 This significant improvement in accident rates automatically made it difficult for legislatures to justify the imposition of a very high standard of care on air carriers. It was becoming increasingly obvious that the initial disbelief in the ability of the industry to protect the property and the lives of people on the ground was unfounded. Flying had matured to the level where it was comparable to other modes of transportation. Carriage by air was becoming commonplace, and the technical reality of aviation was giving considerably less cause for social concern.

7.40 It is true that the paternalistic influences of post-industrial society were arguing for greater protection to the public, flying or on the ground, an aim that the near-absolute liability regime was fulfilling. Yet the exceptionally low number of deaths and injuries on the ground as a result of aircraft crashes was living proof that near-absolute liability had outlived its usefulness. Aviation in the US managed to reach the near zero-accident aim regarding injuries and property damage on the ground that both domestic and international laws were requiring. In return it was requesting a fault liability standard that would facilitate its adjustment to the post-war social reality.

7.41 Admittedly, third parties on the ground should have been “protected against unreasonable risk of harm; ‘unreasonableness’, however, [was] to be interpreted in the light of the period”.21 In that respect, it was reasonable to interpret it in accordance with methods of quantitative risk analysis. This does not mean that broader social considerations were excluded or intentionally ignored. Instead, it signified that the technical reality of aviation, arguably for the first time, convinced, rather than forced, social perceptions of risks, as expressed by the judiciary and domestic legislatures, to follow its choices of liability standards.

7.42 This way, an enosis of the two worlds was achieved in the US, and the middle ground between the underlying questions of “how safe is safe enough” and “how fair is safe enough” was found. This balance, though, was the result of an active dialogue between air carriers, legislature, courts, and the general public which lasted over half a century, and not the outcome of an international conference that was convened with a predetermined aim as the WC or the 1952RC did.

7.43 It is not surprising, then, that the US advocated adopting a fault liability standard on an international level. The success of the WC, which opted for the same liability standard, was a contributing factor to this proposal. Yet the most important feature was the belief that the liability system, which strengthened the relationship of trust in the domestic field, could work as well in the international stage.

7.44 However, the majority of states that participated in the travaux préparatoires of the 1952RC did not share the views of the US and argued for the near-absolute system of the 1933RC.22

Near-absolute liability again: misinterpretation of contemporary social and technological aviation perceptions

7.45 The resistance to the US proposals demonstrated a strong disbelief in the safety standards of aviation, which exceeded by far the social perceptions of aviation risks at the time. Post-industrial society was gradually requesting both courts and the legislature to provide a combination of proactive and reactive risk management techniques that would increase the protection offered to the public. Yet the growth of the industry was still a consideration, especially in Europe, where civil aviation was severely damaged by WWII. As such, a near-absolute liability standard, which called for a precautionary risk management approach, did not accord with the technical advances and the changes in social perceptions in the post-WWII world, and also with the mandate to provide a balanced international convention.

7.46 At the beginning of the twentieth century, domestic and international laws influenced by the hazardous characteristics of flying imposed what Michael Milde characterised as “fatal” liability systems referring to both the 1933 and the 1952RC.23 In the 1950s, though, aircraft were not “looked upon with fear and dread as a dangerous instrumentality” on a worldwide basis and “no one in the world that I have heard of, in 1951 or formerly, suffered either from inability to get proper damages from air carriers or from their lack of financial responsibility”.24 This demonstrates that from a technical point of view, near-absolute liability was not justified.

7.47 Admittedly, most domestic laws still provided for near-absolute liability for deaths, injuries, and property damage caused on the ground by aircraft. Yet for the most part these laws were not revised to reflect the changes in aviation, arguably because the low rate of accidents on the ground took them off the spotlight.25 As such, the active dialogue among the various stakeholders that was achieved in the US as a result of the extensive use of aircraft for transportation, and the beneficial effect of WWII on the state of US civil aviation, did not take place in other parts of the world.26 George Orr, a strong advocate of fault liability, provided a description of the legislative situation outside the US:

“The wave of socialism that has engulfed … Europe and the United Kingdom in the past few decades, result[ed] in much unsound class legislation in favour of the masses at the expense of any constructive enterprise which has earned assets that can be looted. Absolute liability – the making of one party the insurer of another regardless of merit or justice – is reactionary in the extreme.”27

7.48 George Orr’s disapproval of what he describes as “absolute liability” contains elements of hyperbole. Yet it reinforces the assumption that the complex interplay between the technical and social perceptions of aviation risks was delayed in Europe. This delay, though, was instrumental in the failure of the 1952RC to provide a balanced liability system that would be accepted worldwide. It would be no exaggeration to argue that the discussions over the appropriate liability system of the 1952RC were based upon an outdated interpretation of socio-technological aviation reality.

7.49 Admittedly, the new social trends were evident mainly in the US, making it more difficult for the rest of the world to follow. Even so, some states acknowledged that the proposal of the US for a system of presumed liability requires careful consideration in the future.28

Near-absolute liability again: enterprise liability

7.50 Harold Poulton, the Australian Delegate gave a different dimension to the liability choice of the Convention’s drafters:

“… the Convention was not based upon the principle that aircraft were to be governed by the same legal principles as dangerous animals. The theory of the Convention was rather to provide a means of distributing the losses arising from certain foreseeable risks inherent in aviation, and that the most scientific method of achieving this result was to attach the liability to pay compensation for the loss on the operator who could take out insurance which would in the last analysis be paid by the passengers, government subsidies, and to a certain extent, by the operator himself.”29

7.51 One can find traces of enterprise liability being applied to an international level in his opinion, and also can witness a rather unsophisticated attempt to allocate risks among air carriers and third parties on the ground.30 His main argument was that the reluctance to characterise carriage by air as an abnormally dangerous activity does not prove the inaptness of near-absolute liability. Ample justification for imposing it may be found in the fact that accidents on the ground are “a part of the costs of operation of airline air traffic. It is in reality just as much an actual part of the overhead costs of air traffic as is the wear and tear on the machinery and equipment used in such air traffic”.31

7.52 Implicit in this argument is that people on the ground do not derive any benefit from carriage by air, do not participate in the activity, and as such, they should not be allocated any of the risks of flying. This way, the Australian delegate expressed in an international forum why many states remained sceptical about the use of the fault liability standard – presumed or not – as a way to domestically regulate the liability of air carriers to third parties on the ground. Their scepticism did not stem anymore from the hazardous nature of flying, but from the inevitable effect of the post-industrial society upon social and political priorities:

7.53 The suggestion that liability for damage on the ground should be channelled to air carriers is not without merit, at least at a domestic level. When one party controls the instrumentality of harm and the other is essentially passive, the case for one-sided allocation of risks is strong even from a compensatory point of view.33 Yet the actual choice is to be made by following the active dialogue between the various stakeholders in each country. In the US, where the theory of enterprise liability was born and developed and this active dialogue took place immediately after WWII, domestic aviation was excluded from its application. Instead, an approach that favoured carriers – the fault liability standard – was adopted.

7.54 In an international environment, though, the mandate to drafters was very clear, and called for a balanced allocation of risks. Due to the short-sightedness of the 1933RC, the position of the 1952RC was close to the WC, since it was required to build social risk perceptions and also play a risk-education role. At the same time domestic tort laws were in the process of adapting to the new socio-technological environment either: (i) by lowering the liability standards to a fault principle because the benefits of flying were believed to exceed its costs, and thus society was ready to be burdened with a greater share of the risks brought about by the growth of the activity; or (ii) by retaining the near-absolute liability standards on the basis that the initiator of the activity should take responsibility for the consequences of its actions, irrespective of the social benefits offered by this activity.

Disregarding the need to balance social and technical perceptions of risks in the 1952RC

7.55 The aim of the 1952RC was not to take sides and create a liability system that would favour one option over the other. Instead, it was supposed to acknowledge the progress of the industry and its value to the international community, but at the same time further society’s interests in compensation. This way, the one-sided approach of the 1933RC would be repealed, but enough room would be left for the trends of post-industrial society to develop. However, the drafters of the 1952RC decided to make aesthetic changes to the 1933RC instead of overhauling its liability system, resembling the attitudes of the HP’s drafters when amending the WC. They attributed the limited ratifications of the 1933RC to the interruption caused by WWII, and in essence worked towards its refinement rather than “to achieve a result that will be in greater harmony with our times and with the ever-changing requirements of air transport”.34

7.56 In that respect, the system of near-absolute and limited liability based on the weight of the aircraft was retained. The liability attached to the operator of the aircraft, which was defined as “the person who was making use of the aircraft at the time the damage was caused”.35 However, in case the “control of the navigation of the aircraft was retained by the person from whom the right to make use of the aircraft was derived, whether directly or indirectly, that person is considered the operator”.36 The registered owner of the aircraft was presumed to be the operator, unless he proved during the trial that someone else was the operator and joined him in the proceedings.37 This way, the injured third party on the ground would always have an entity against whom to file a claim.

7.57 In wet leases the owner would be the operator for the purposes of the 1952RC, since he provided the crew. In dry leases, though, where the owner provided only the aircraft, the lessee would be the operator, provided that he had “the exclusive right to use the aircraft [at the time the damage was caused] for a period of more than fourteen days, dating from the moment when the right to use commenced”.38 In any other case, the person from whom such right was derived would be jointly and severally liable with the operator.39

7.58 It is clear from the above provisions that the liability system of the 1952RC was created to provide the best available legal protection to the public and to guarantee recovery from the air carriers in case of an accident. Surprisingly, though, the 1952RC did not provide for mandatory insurance of any aircraft registered in the territory of a Contracting State, but left the decision of whether aircraft should carry insurance to the Contracting States. This comes in dire contrast to the 1933RC, which provided for a system of compulsory insurance, and linked it to the limitation of liability. It also demonstrates that the regulation of the risks inflicted by flying to third parties on the ground achieved in the 1952RC lacked the innovative features that were required for a rapidly developing industry: “it was … noted that several States had not yet introduced compulsory insurance with regard to aircraft belonging to their own citizens. Therefore, it was proposed that the insurance should not be compulsory but only permissive”.40

7.59 At this point, the 1952RC contradicts itself. The provision for a system of optional insurance acknowledged that underwriting aviation risks was a nascent activity in the 1950s, and in many parts of the world air carriers were operating without insurance coverage, at least as it is understood today. The implied admission that aviation insurance has limited availability on a worldwide basis, inevitably questions the decision of the drafters to channel liability to carriers on the basis that they had a far superior ability to transfer the risks by insuring them. It is true that a number of delegations noted that the liability limits would enable air carriers and their insurers to calculate ex ante their liability exposures and also “to avoid the hazard of a catastrophic loss which might wipe out assets accumulated over many years”.41 Yet this assumption is partially accurate, because the breakability of the liability limits inevitably weakens its founding basis, i.e. that insurers would have been able to estimate accurately the liability exposures of air carriers and as such charge a premium that corresponded to the risk transferred.

Near-absolute liability and breakable liability limits: a litigation minefield

7.60 There was considerable debate at the Conference as to the degree of fault that was required to break the liability limits. Italy and a number of other European countries were in favour of applying the “wilful misconduct” threshold of the WC that has already been tested in practice and was putting third parties on the ground on a par with passengers.42

7.61 It becomes apparent then that the drafting of the 1952RC was a continuous comparison with the WCS. The majority of the drafters of the 1952RC did not want third parties on the ground to be afforded less protection than passengers, and their choices were frequently guided solely by this consideration. It is not surprising then that the “wilful misconduct” threshold was suggested upon its perceived success in the WC and without considering its implications for the liability system of the 1952RC.43

7.62 The delegate of IUAI pointed out the difficulties that domestic courts faced with establishing a uniform interpretation of the term “wilful misconduct” in the WC and hinted at its serious repercussions for the carriers in the event of accidents on the ground:

“If aircraft operators were to know where they stood and what liabilities they ought to insure, it was essential to find some wording which would not enable a jury to award unlimited damages on a presumption of intention drawn by them from the facts … [In an] action brought under the Warsaw Convention … the operator had been held liable without limits because one of his mechanics had committed an error in the adjustment of a sensitive altimeter.”44

7.63 This way the IUAI expressed the valid proposition that the availability of compensation which is based on the availability of insurance coverage, is better served by a system of near-absolute liability which has unbreakable or at least difficult to break limits. By referring to the attitude of the juries, the delegate of IUAI implicitly highlighted the importance of building a relation of trust between the public and the air carriers which eventually would form a protective cocoon for the industry. However, the creation of such a relation is seriously hindered by a Convention that endorses adjudication instead of the quick and amicable settlement of claims. Especially in the case of injuries to innocent bystanders, the importance of expeditious recovery becomes even more important both because of the magnitude of the losses and the absence of any participation in the injury-causing activity by the victims.

7.64 The desire of air carriers was for the 1952RC to contain as few subjective elements as possible, so that any judicial influences that had the potential to expand the liability system were kept outside the Convention.45 Still, they were admitting that in the 1950s “the great majority of damage caused on the surface was lower than the limits found in the draft Convention”.46 However, it was undisputable that the level of damages could only increase as the cost of living worldwide was gradually rising, putting thus more pressure upon the limits. Most importantly, air carriers wanted to safeguard their interests, as they could sense that the halcyon days of the early twentieth century were fading away as a result of the slow, albeit inevitable, transition to a risk society.

No provision made for intentional risk events: modified corrective justice approach still the preferred choice

7.65 It was becoming increasingly obvious to air carriers that intentional, man-made accidents threatened to change beyond recognition both the technical and the social perceptions of aviation risks. Aircraft were not used as weapons of destruction yet, and even in the cases where they were used for purposes other than transportation, they were rarely causing damages on the ground that could have been covered by the RCS. Their “abnormal” uses were limited to hijackings, which were either initiated by “political refugees seeking a safe heaven” or motivated by greed and resulting in “holding passengers hostages for ransom”.47 Nevertheless, they signified the beginning of a new era for aviation that: (i) increased the maximum probable losses on the ground as a result of flying; and (ii) required to factor behavioural, social, and political features in both technical and legal risk management decisions.

7.66 As such, the post-WWII environment called for the drafters of the 1952RC to look beyond both the 1933RC and the WC and form an innovative liability system. The need to do so was more critical than the updating of the WC, because in the case of liability to passengers, the maximum probable loss irrespective of liability limits was a combination of the number of passengers carried in an aircraft and the four fora. As such, the losses were predictable and the question was how to allocate the risks in a legal instrument so that a balance between the technical and social perceptions of aviation risks was achieved. Despite the failure of the HP to provide a liability system that would simplify the recovery process and accommodate the prevailing social perceptions, the existence of the aviation industry was not jeopardised. What happened eventually was a multiplicity of amendments and a judicially-driven expansion of the liability exposures of carriers in the second half of the twentieth century.

7.67 In direct contrast, the element of predictability was seriously weakened by the intentional use of aircraft as risk-creators. Paul Dempsey explained the reasons:

7.68 With the wisdom of hindsight, we now know how far politically or religiously motivated militants can go when it comes to the use of aircraft, and we also are well aware of the social, technical, financial, and legal consequences of such a use. It would be an impossibility to have required from the drafters of the 1952RC the ability to regulate the liability of air carriers for ground damage so that it could cope with the challenges of the twenty-first century. Yet the period from the end of WWII to the 1980s was already providing strong signals that an international liability system would need to achieve a greater socialisation of risks in order to control the side-effects of the use of aircraft by terrorists and other political activists. A first step towards this socialisation could have been achieved by disassociating blameworthiness from liability and excluding considerations of motive from the liability system. The near-absolute liability system of the 1952RC provided this disassociation, yet the element of motive was directly linked to the breakability of the limits.

7.69 The drafters of the 1952RC opted for breakable limits, demonstrating thus a disregard for the management of intentional, man-made risks. For them, the 1952RC was a Convention that had the task of regulating the effects of safety-related accidents on the ground. Terrorism-related or other similar intentional, man-made accidents were not visible enough at the time to be taken seriously into consideration. To achieve a compromise between the “wilful misconduct” standard and unbreakable limits, it was agreed that the claimant ought to prove that the damage was caused “by a deliberate act or omission of the operator, his servants or agents [acting in the course of their employment and within the scope of their authority], done with intent to cause damage”.49

7.70 The delegate of the UK, sensing the forthcoming socio-political changes, attempted to influence the judicial interpretation of this provision by highlighting the need to prove both deliberate act and intention: “The word ‘deliberate’ had been left in Article 12 to make it more difficult to establish unlimited liability against an air operator, because it was thought that a deliberate act done with intent to cause damage was something worse than an act done with intent to cause damage”.50 His intervention highlights the concerns that even this comparatively high threshold would not be able to control the secondary effects of an accident on the ground, and especially the amplified individual and social desires to seek financial retribution against the carrier involved, irrespective of its degree of culpability.

7.71 Michael Milde proposes that the equivalent threshold of the HP – recklessness and with knowledge that damage would probably occur – should have been applied to the 1952RC for the sake of the better protection of third parties:

7.72 This view endorses the belief that the 1952RC should adhere to the modified corrective justice approach of the WCS and resist a move towards the socialisation of the risks on the basis that the deterrent effects of law will be compromised. For supporters of this argument the aviation industry was not mature enough to have the level of safety self-regulation that would give adequate protection to third parties without the threat of quasi-punitive damages. The concerns about security and safety defects demonstrate scepticism about the competence of the aviation industry to protect third parties on the ground, and consequently their own financial and operational viability.

7.73 These reservations about the organisation of air carriers and the limits of their technological state were understandable in the 1950s. Yet they should be read into the perspective of the entire Convention and not in isolation. The near-absolute liability system was predominantly designed to counteract the effects of these deficiencies: to strengthen the reparatory function of the Convention by imposing a precautionary risk management regime. Its role was not so much to prevent accidents rather than to ease the compensation process in the aftermath of an accident. A limited preventative function was achieved indirectly, since the best way to avoid a large payout, which at the time would have come from the funds of the air carrier, was to improve the safety and security of its operations. In turn, the liability limits were there to balance the interests of carriers and third parties on the ground. It is true that the more difficult it was to break the limits, the more compromised the reparatory and the preventative functions of the 1952RC would be. Yet this was the price to be paid for having a liability system that in principle does not prioritise one interest over the other.

7.74 This provision was never applied, so it is difficult to test whether the prediction of the UK delegate was accurate or if the criticism of Michael Milde was justified. Judging from the judicial treatment of the corresponding provisions in the WC and the HP, it is most likely that the amplified social perceptions in the aftermath of accidents on the ground would exercise severe pressures upon the judiciary to break the limits. In safety-related accidents, arguably, the dual requirement of intention and deliberate act might have been a safeguard, yet in intentional, man-made accidents, Article 12 would have an impossible task to protect the industry.

Vicarious liability and breakable liability limits in an era of global operations

7.75 In cases where the carrier’s personnel have somehow contributed to the accident by facilitating or actively causing it, the elements of intentional and deliberate act would have been more easily satisfied. By way of protection, the 1952RC provides that the operator would not bear unlimited liability if he could prove that the employees that caused the damage were acting contrary to his instruction or outside the scope of their normal duties.52 In this defence – that was proposed by the industry – air carriers attempted to link their exoneration from liability with their ability to control the decisions of their employees.53

7.76 This was a reasonable standard in an era when the size of operations was permitting line managers to exercise direct supervision over employees at the end of the production line. Until the end of the 1970s, all they had to do was to make sure that minimum safety and security standards were met. Furthermore, accident investigations were still focusing upon the analysis of individual errors, demonstrating thus a systematic tendency to exonerate the management of air carriers of any wrongdoings. As such, it was relatively simple for the carrier to demonstrate that its instructions were in compliance with regulatory requirements and to prove “whether or not the act had been done with his express authority”.54

7.77 With technological improvements came the worldwide expansion of operations, and inevitably operational complexity. In turn, the increasing levels of flying required the adoption of proactive safety management systems in order to further reduce accident rates. In addition to ensuring compliance with a solid framework of legislation and regulatory requirements, line managers were increasingly required to define best practices or desired standards, and implement comprehensive safety-management systems which reflected them. At the same time, accident investigations began to focus on latent causes, identifying the carriers’ deficiencies in organisation, quality management, and safety culture as the main causes of individual errors. As a result, regulators were requiring greater operational management control, yet the complexity of operations was often an impediment for managers to be more involved in day-to-day operations and check the effective application of standards.

7.78 Bearing in mind the tendency of courts to bring into the WCS elements of domestic law by linking liability to the carriers’ duty of control, contrary to the intentions of its drafters, it would not have been surprising if courts adopted a similar approach in the case of the 1952RC. As such, it is seriously questioned whether the defence intended to protect the carriers from intentional, malicious acts of their employees would have actually done so. Most probably courts would have devised ways to link the carriers’ unlimited liability to their potential for control over their employees, instead of their actual instructions to their employees or the employees’ normal duties. Automatically, then, the system of near-absolute liability of the 1952RC becomes even more demanding, since the carrier is called to safeguard the liability limits by adopting a precautionary risk management approach to the employment and supervision of its employees and contractors. In that respect, even if Michael Milde’s criticism on Article 12 of the 1952RC is prima facie accepted, it is estimated that its practical application would have given to third parties the indirect protection that Milde is requesting, both in terms of safety-related and terrorism-related risks.

Defences: reinforcing the channelling of liability to air carriers

7.79 One might argue that the 1952RC was not providing for near-absolute liability, but for a system of strict liability, in a similar fashion to the MA and the MC. If Article 1(1) is considered in isolation, this argument might be accepted. It is also true that the Convention included a number of defences that in theory relaxed the near-absolute character of the Convention. Yet in practice none of them could limit the liability exposures of the carrier. The carrier could avoid liability: (i) if it proved that the damage was caused solely through the negligence or other wrongful act or omission of the injured person;55 and (ii) if the damage was the direct consequence of armed conflict or civil disturbance, or if the liable person has been deprived of the use of the aircraft by act of public authority.56

7.80 Especially regarding (i), people on the ground who suffered damage (in direct contrast to passengers), more often than not would not have any means to influence the operation of aircraft. It becomes apparent, then, that the defence of contributory negligence could not achieve any meaningful reallocation of risks, but it was included in order to give a false sense of balance in an otherwise prejudiced Convention.

7.81 Furthermore, defence (ii) is incomplete, since it does not make any reference to force majeure. As a matter of fact, the inclusion of force majeure was explicitly rejected by the drafters, because out of the 41 states requested, just a minority had included it in its domestic legal framework.57 Once again, the drafters of the 1952RC adopted a domestic-centric rationale when designing its liability system, demonstrating thus a very cautious, backwards-looking approach to the allocation of aviation risks. IATA pointed out the irrationality of this choice:

“Traditionally the law of carriers under all systems has linked defences of ‘act of God’ with defences of the act of ‘a public enemy’ … It is difficult to understand why under the same legal philosophy the aircraft operator should not be permitted to prove that the aircraft fell as the result of force majeure. The householder insures himself against loss to his property from lightning. But if the loss results from lightning striking an aircraft pursuing its normal course through the airspace, and the airplane as a result falls and hits a house, the operator … takes the place of the insurer.”58

7.82 Even more interestingly, the drafters of the 1952RC explicitly rejected the proposal to exclude the carriers’ liability for damage resulting from the wrongful act of a third party, done with intent to cause damage and without the operator being able to prevent it.59 A similar provision was included in the 1933RC, and its inclusion in the 1952RC had the potential to alleviate the effects of both near-absolute liability and the breakable limits.60 This is so because it was excluding the liability of air carriers for what would be the most significant future threat against the stability of the aviation industry and the insurance market, i.e. terrorists taking control of aircraft and intentionally causing damage on the ground to promote their objectives.

7.83 Obviously, this provision was open to interpretation by domestic courts, especially the part which provides that the carrier can take advantage of the defence only when it can demonstrate that it took “all necessary measures” to prevent the damage. Yet it would have set a precedent that intentional, man-made risks should be treated differently from safety-related ones, and not channelled indiscriminately to air carriers.

7.84 Instead, the 1952RC provided that:

“[i]f a person makes use of an aircraft without the consent of the person entitled to its navigational control, the latter, unless he proves that he has exercised due care to prevent such use, shall be jointly and severally liable with the unlawful user for damage giving a right to compensation under Article 1, each of them being bound under the provisions and within the limits of liability of this Convention.”61

7.85 This provision was not designed with the above scenario in mind, and it is doubtful whether a carrier could have used it as a defence in the aftermath of an accident on the ground caused by the unlawful use of an aircraft.

7.86 Even if it could use it, though, the end result would have been very different compared to the aforementioned proposal: the carrier is required to retain the risk of unlawful interference with an aircraft and to bear the subsequent losses on the ground, despite exercising due care. Joint and several liability is yet another euphemism for near-absolute liability, since it would be impossible for a carrier to recover any damages from an individual terrorist or a terrorist organisation. As such, the carrier who bears a minor part of the responsibility unfairly shoulders the burden of damages.

7.87 The analysis of the defences provided in the 1952RC, or their lack thereof, confirms what this work argues all along, i.e. that the 1952RC created a legal system of broad accountability that rendered air carriers liable for virtually all damages on the ground caused by their aircraft. Furthermore, it demonstrated that the 1952RC contained an unworkable liability system. The main reason behind this deficiency was the deliberate attempt of the drafters to express in this Convention an outdated version of the social fears and scepticisms regarding the effects of aviation technology on third parties on the ground. By doing so they did not consider the problems that the interaction between technological progress and international political disputes will create for the industry and third parties.

7.88 The end result was a Convention that could have been successful in the pre-WWII environment, yet it failed to attract attention in the post-WWII world that was urging for a more balanced approach in the management of aviation risks: on the one hand, demanding better treatment for passengers and the non-flying public, but on the other hand, keen to take advantage of the full potential of flying commercial jets.

The Guatemala City Protocol as a prototype for the regulation of air carriers’ liability to third parties on the ground

7.89 This is not to argue that the 1952RC was a complete failure. Although it was ratified by very few states,62 it contained all the ingredients that, if refined, could provide a balanced and forward-looking Convention that would be minimally manipulated by judicial interpretations. My purpose is not to redraft the 1952RC, yet a few fundamental changes, that can assist the analysis of the two Conventions that were recently drafted to replace the RCS, need to be mentioned.

7.90 First, a liability system based upon the presumption of fault as suggested by the US would have improved the position of carriers and, as it demonstrated herein, it would not have weakened the position of third parties. Yet fault as the basis of a liability system was still perceived as a pro-defendant legal notion, and as such the US proposal might not have been as successful as it was hoped for in the early 1950s. To overcome this impediment, a system of strict liability could have been designed. Article 1(1) would have remained as it is, but three new defences would have been introduced: (i) force majeure; (ii) damage resulting from the wrongful act of a third party provided that the carrier has taken “all necessary measures”; and (iii) assumption of risks by third parties (referring specifically to damage caused to persons or property in airports).

7.91 This way, a more balanced allocation of risks could have been achieved. The carrier would still be liable in the vast majority of cases where damage has been caused on the ground. At the same time, though, the carrier would have the opportunity to escape or reduce its liability in situations that are beyond its control or where a third parties’ position resembles that of passengers. The reparatory function of the liability system will remain intact, and more incentives will be given to the carrier to safeguard its operations from intruders.

7.92 Second, the drafters could have provided for unbreakable, yet considerably increased, liability limits. One might argue that high liability limits could have placed an excessive burden on international civil aviation due to the increase in the cost of third-party insurance. Yet the predictability and certainty of exposures guaranteed by the unbreakable limits would have eventually reduced the cost of insurance and provided stability in operations.

7.93 Admittedly, it would have been an innovative and daring proposition at the time to have unbreakable limits of liability, irrespective of their levels. It was required to drop the proposed defences in order to have it approved, creating thus a pure no-fault liability scheme that the third party was just required to prove the level of damages in order to get compensation.

7.94 This quid pro quo was worth it, because an objective liability system would have been created that would simplify the compensation process, disassociate notions of social blameworthiness from liability, and resist domestic legal intrusions. In order to cover the compensation to third parties in extremely catastrophic accidents, a compensation fund could have been created over and above the limits. This way, any fears that the unbreakable liability limits would not be sufficient to pay for the damages incurred would have been satisfied.

7.95 These proposed principles were the basis of the GCP. Although the GCP was never put into force, its structure was more suitable to regulate the liability of air carriers to third parties rather than to passengers. This is so because the unbreakable limits and the compensation fund could reduce the inherent unpredictability of the relation between carrier and third parties, and as such reduce the asymmetry of information between them as to the level of risk or the effort undertaken to manage it. This liability scheme would have operated as a buffer against intentional, man-made accidents, achieving at the same time a balance between social demands, technological advances, and malicious intentions – a third factor that was not present in the WCS.