Simon Phippard asks whether there are clear skies ahead for air carriers' liability.

By 1966 the Warsaw System governing international air carriers' liability was close to collapse. It had been in place since the Warsaw Convention had been signed in 1929 with certain modifications agreed at the Hague in 1955. However, the United States had refused to ratify the Hague Protocol on the grounds that the limit of $20,000 for loss of life of a passenger was too low. Under threat of denunciation by the United States an IATA-sponsored compromise was reached in 1966 in the form of the Montreal Agreement between carriers flying to or from the United States which raised the limit for passenger damages to $75,000.Thirty three years later we have a new Convention which bears many similarities to the Warsaw regime. The intervening period has seen the Montreal Additional Protocols of 1975, of which only three have come into force, and those only since 1996; the IATA Kuala Lumpur inter-carrier agreement of 1995 and the ensuing Montreal Implementation Agreement; and the European Regulation on air carrier liability of 1997. Whether or not the proliferation of liability régimes, carrier initiatives to waive or raise limits, or the diversity of judicial means to avoid imposing low limits of liability in fatal accident cases should be regarded as anarchy1#, the uniformity which the Warsaw system aimed to achieve was undoubtedly once again under threat.

One of the new provisions in the Montreal Convention is that contracting states must require their air carriers to carry insurance in respect of their third party liabilities#.2 Given the extent to which the worldwide insurance and reinsurance industry provides cover against air carriers' liability exposure it is appropriate to examine the changes to be brought about by the Montreal Convention and the effect of certain issues that have not been resolved.

The Montreal Convention
On 28 May 1999, 52 states signed the Montreal Convention at the conclusion of an inter-governmental conference convened by ICAO. Since then one further state, Uruguay, has added its signature. The new Convention takes as its starting point the texts of the Warsaw Convention as amended at the Hague, the Guadalajara Supplemental Convention of 1961 and certain additional elements from the Montreal Additional Protocols. The Convention will enter into force sixty days after the deposit with ICAO of the thirtieth instrument of ratification. It seems likely that the Convention has widespread international support and may come into force within about two or three years.

Passenger claims
The principal change in the Convention lies in its treatment of the liability “limit” for injuries or death to passengers. This was the politically controversial issue which posed the greatest risk to the Warsaw system. Under Warsaw/Hague a carrier was able to limit its liability to 250,000 poincaré francs in the absence of intention to cause damage or recklessness with knowledge that damage would probably result#.3 If that test was satisfied, the carrier faced unlimited liability.Although the calculation in modern monetary terms of 250,000 poincaré francs had caused numerous problems over the years since the official and free market gold rates began to diverge substantially in the late 1960s, this problem has been resolved in many jurisdictions. This came about by a variety of means: the entry into force of MAPs 1 and 2 denominating those limits in Special Drawing Rights (SDR), and legislation or case law defining the limit in contracting states' own currencies. Nevertheless there remain many states where there is uncertainty over the value of the limit, so the adoption of an all-encompassing convention with all limits denominated in SDR - as distinct from a series of protocols to existing conventions - should result in a simpler series of convention relationships and in due course an end to this uncertainty.

Where the Convention departs from the existing Warsaw structure is in the treatment of “limits”. The carrier is now not liable for damages in excess of 100,000 SDR per passenger if it can show that the damage was not due to its negligence (or that of its servants or agents) or if it was solely due to the fault of a third party. The carrier therefore faces strict liability for death or bodily injury of a passenger, if the accident occurred on board the aircraft or while embarking or disembarking, up to 100,000 SDR. Over that threshold the carrier faces a presumption of liability, which it can discharge if it can show it was not at fault. Under the previous regime, if the carrier was at fault, but not to such a degree that it fell foul of the Warsaw Article 25 test, the passenger could only recover the applicable liability limits. Conversely, the carrier can no longer rely under the new Convention on the (absolute) defence that it had taken all measures necessary to avoid the damage or that it was impossible to do so#.4

The fifth jurisdiction
For passenger claims a “fifth jurisdiction” has been added#.5 This was one of the more controversial issues. It was heavily backed by the United States whose original aim was that United States passengers, wherever ticketed, would recover damages by reference to United States standards, alternatively that they would be able to bring claims in the United States. It raised the spectre of claims being pursued in the United States' courts by the “wandering American” against carriers based all over the world, even though the only link between the carriage in question and the United States was the domicile of the passenger. It also risked large variations in levels of compensation awarded due to differing treatment of similar claims by different courts.

The full rigour of this approach has been somewhat mitigated by the formula which has been adopted which requires a commercial connection with a jurisdiction for a carrier to be liable to suit before the courts of a “fifth jurisdiction”. A death or injury claim may now be brought where the passenger has his principal or permanent residence at the time of the accident (in addition to the jurisdictions prescribed under Article 28 of Warsaw/Hague) so long as the carrier operates services there and conducts its air transport business there, either itself or through a commercial agreement with another carrier. An agency agreement (such as with a general sales agent or an IATA ticket agent) would not suffice for these purposes.

This brings jurisdiction against air carriers more in line with traditional conflicts of law rules which usually require, as a minimum, a presence within a jurisdiction to be liable to service of proceedings. Of course that presence is not necessarily a pre-requisite of jurisdiction under the existing Article 28 (now Article 33) rules, which stand. One should also note that United States lawyers advise that - contrary to the general approach in most other jurisdictions - traditional doctrines of forum non conveniens may be employed in the United States to stay or dismiss cases, even though the state in question may be one of the “mandatory” Warsaw régime jurisdictions. This tactic should also be available where the United States is the “fifth jurisdiction” under the Montreal Convention.

Cargo and baggage
A carrier faces liability for loss, destruction or damage of checked baggage or cargo while in its charge. Liability for unchecked baggage is dependent on proof of fault. This may have been the case under proper construction of the previous law, but it is useful to have clarification.

The carrier can limit its liability in baggage cases to a composite limit of 1,000 SDR per passenger for both checked and unchecked baggage, unless the passenger is able to show that the carrier breached the “old” Warsaw/Hague Article 25 test.

In the case of cargo, the carrier can invoke an unbreakable limit of 17 SDR/kg. Contrary to expectations the absolute limit in cargo cases introduced in MAP 4 was retained in the Montreal Convention. This will be important for cargo insurers in particular in relation to high value consignments such as bullion, precious stones or bank notes where the limit of liability on a weight basis is only a very small proportion of the true value of the shipment. In these cases cargo interests and their insurers will be unable, in relation to carriage to which the new Convention applies, to recover anything like their total loss from the air carrier unless the full value has been declared.

The impact in the short term may be limited - due initially to the relatively small number of states party (only carriage between two states party to the Montreal Convention will be subject to the new régime) - but significant. Cargo interests will lose the leverage that the possibility of substantial financial compensation through legal means gives them over carriers to ensure adequate security levels. In due course it will be interesting to see whether more claims are pursued against other entities: airports or security services, for instance, or litigation in non-Convention states. Alternatively we may see cargo insurers requiring consignors to declare the full value of the consignment for carriage: in that event the carrier has no defence#6 to claim for loss, destruction or damage while the goods are in its charge, with possible implications for cargo rates for valuable shipments.

A similar régime applies as previously in the event of delay in the carriage by air of passengers, baggage or cargo. In these cases the carrier retains the “all necessary measures” defence with the clarification that the measures to be taken to entitle the carrier to rely on the defence are those reasonably required to avoid the damage.

In passenger delay cases there is a limit of 4,150 SDR and the same composite limit of 1,000 SDR per passenger applies to delayed baggage whether checked or not. In all these cases the carrier is unable to rely on the limits if the “old” Article 25 test is satisfied. In cargo cases the same unbreakable limit of 17 SDR/kg applies to delay cases as it does for loss, damage or destruction.

Issues for the future?
Retention of a substantial proportion of the existing text was obviously necessary both to secure agreement to a new instrument and to enable preparation of and debate on the draft document in the time available. Nevertheless there are a number of issues which could perhaps usefully have been considered further and which might give rise to difficulties of interpretation or application in the future.

The scope or exclusivity of the Warsaw regime has often posed difficulties in the past. Precisely how far does the Convention govern the rights and liabilities of an airline and its passengers or cargo interests? The question has been considered in recent years by both the Supreme Court in the United States7# and the House of Lords in the United Kingdom#.8 Both courts have found the Convention to be exclusive: one in a case arising from a body search for security purposes and one in claims arising from the landing of an aircraft in Kuwait during the Iraqi invasion. On the other hand, there are decisions of the French Courts that the Convention is not exclusive#.9

Those cases raise their own issues in terms of application in a wider content, but other problems arise on a regular basis: for instance cases of denied boarding. Taken at its most literal, the House of Lords judgment in Abnett would indicate that the Convention would govern such cases. However, liability to passengers is predicated upon there being an accident within Article 17, or on delay. But do such instances constitute an “accident”? If so what bodily injury has been suffered? Or should they be characterised as delay? One view is that whether liability arises within or without the Warsaw régime depends on whether carriage had started or not. For instance if a passenger is denied onward carriage due to alleged passport irregularities observed at a transit stop the cause of action is different from that arising when boarding is refused at the point of origin for the same reason.

Hitherto the courts have not dealt much with these issues. For that reason some further guidance on the intended scope of the Convention would have been helpful. The original Warsaw Convention was aimed at the obvious accident situation. Since then the safety of the industry has improved enormously and the existing structure has been applied to circumstances which were not originally envisaged when the original text was established.In a similar way it may prove unfortunate that the nettle has not been fully grasped in relation to mental injury. Again public perceptions of recoverability have changed enormously since 1929, but case law varies as to whether the phrase “bodily injury” covers psychological trauma and the such like. The conference took the decision not to extend the definition specifically to cover the same in the absence of associated physical injury: no doubt this was the price of agreement on a delicate issue on which contracting states who referred to preserve the status quo and those who favoured an extension each felt that the position could be argued before the courts in due course if there was not already sufficient caselaw on the issue.

There are other issues which could have been dealt with more specifically. For instance, in an industry dominated by global alliances, codesharing agreements and the such like, the existing structure regulating the liability of actual and contacting carriers does not really reflect commercial realities, and problems may arise when a carrier is brought before the courts of certain jurisdictions purely by virtue of a codesharing agreement. While there was a case for starting with a blank sheet of paper and attempting to legislate for all the difficulties of application and interpretation that have arisen the task would have been a huge one. Moreover, that process was not essential to deal with the major political issue, namely the perceived low level of passenger liability limits, and would have run the real risk of losing much of the jurisprudence that has grown up around the existing Warsaw system over nearly 70 years. The retention of this structure should - with certain reservations - preserve much of that jurisprudence and enable the air transport industry to contemplate a stable liability régime for the foreseeable future.

Simon Phippard, partner, Barlow Lyde & Gilbert.

#1. See Awford. I., The IATA “Umbrella” Inter carrier Agreement 1995; (1995) Insurance Law Quarterly, Issue 24, p25.
#2. Montreal Convention 1999 Article 50.
#3. The “old” Warsaw/Hague Article 25 test: “The limits of liability . . . shall not apply if it is proved that the damage resulted from an act or omission of the carrier, his servants or agents, done with intent to cause damage or recklessly and with knowledge that damage would probably result; . . .”
#4. Warsaw/Hague Article 20.
#5. Warsaw Article 28 provided for jurisdiction before the courts of the domicile of the carrier, or of the carrier's principal place of business, or where the carrier has a place of business through which the contract was made, or the courts of the place of destination.
#6. The “old” Warsaw/Hague Article 20 defence (see footnote 4 above) will only be applicable to delay cases, in a modified form.
#7. El Al v Tseng [citation to follow]
#8. Sidhu and Abnett v British Airways [1997] AC 430. See also [1996-97] TAQ 394.
#9. See for instance Duffy v British Airways (1988) 42 REDA 204, British Airways v Mohamed (1997) RFDA 155.