Almost three years have passed since some of the most serious riots of the last stages of the Indonesia's Suharto administration, but despite numerous insurance claims the only legal authority to be found is Lexington vs D P Mann (2000), a High Court decision dealing with the meaning of “events” in reinsurance contracts in the context of the riots. It leaves commentators wondering whether the Indonesian experience (and similar happenings in West Africa, the Middle East, and South America) have led to new lessons being learned, lessons which could assist re/insurers in both interpreting and drafting contracts which will be in force during times of unrest.

An underlying problem when the riots erupted was that many of the businesses affected had no insurance in place to cover the losses. Most businesses that were covered for riots had obtained the protection by way of a special endorsement triggered only when losses do not emanate from politically-motivated events. Naturally political, social, and economic upheavals such as those preceding Indonesia's change of government are invariably subject to great controversy. Interested parties in or close to conflict areas have a lot to lose (or, indeed, to gain) from the way such incidents are described or categorised. One man's terrorist is another's liberator. Somebody's political rioter is someone else's malicious delinquent.

It is perhaps not so surprising that from the outset the Insurance Council of Indonesia was inclined to the view that the riots fell within the definition of “riot, strike and malicious damage” used in clauses included in some of the local policies. Reinsurers' representatives, however, took the view that the events preceding the ousting of the former president Suharto were in the nature of “civil commotion or popular uprising” and in fact not covered by the policies in question. The arguments are not dissimilar to re/insurance disputes going back to the civil wars in Liberia, and before them in Lebanon.

The ensuing disputes over coverage were emotive. Parties did not always concentrate on the plain meaning of the words used. Discussions about indemnification of losses sometimes proceeded before it was even established if the applicable wordings had any relevant extensions or exclusions. This is understandable where the original intention of the parties might have been different to the words actually used to describe the cover; it is not uncommon for reinsurers and insurers worldwide to find themselves arguing as to which “riot” or “explosion” or “malicious damage” is covered, even in situations where the contract itself may make no distinction between, for example, one type of riot and another, and therefore, on the face of it, covering all riots or all explosions.

Prevention before cure

Preventing such disputes demands precise definitions. The parties should make all relevant distinctions, and be clear about what their contracts are intended to cover. Moreover, however sophisticated a definition is provided, various risks covered or excluded are incremental in nature and not capable of being clearly delimited. For example, one can try to define the specific riot covered, but it may be a question of degree as to whether an event transcends mere rioting and is promoted to civil commotion or insurrection. If events during civil unrest are grey (and they invariably are) a thorough investigation into the facts on both a micro and a macro level will be required to establish the evidence necessary to pursue or defend a claim. There may be a fine line between an event covered and one that is not, or is specifically excluded.

As with problems arising from conflicts in Lebanon, Liberia, and Kuwait, it is because of the sensitive and fragile position of evidence that re/insurers in Indonesia were immediately encouraged to investigate not only the specifics of the riots, but their background, utilising expert groups of lawyers and forensic investigators as well as academic, political, economic and other commentators to establish with clarity how re/insurance law would interpret the events. Inconsistencies in approach appeared immediately.

Some took the view that coverage had to be considered on a case-by-case basis, with reference to the extent to which the actual terms of their re/insurance contracts covered the losses. Others assumed that they had a back-to-back cover, which they somewhat optimistically interpreted as meaning that reinsurers would indemnify them for any payments made. Important issues about obligations, for example to follow the settlements, in light of the different terms and conditions in contracts, as well as applicable law and jurisdiction problems, were perhaps not always fully explored. Had they been, parties might have realised that terms which afforded cover in insurance contracts sometimes varied slightly in the reinsurance contracts, thus affording no cover. Different applicable laws meant different results in otherwise similar cases.

Classified unrest

One way to prevent such problems recurring, whether in Indonesia or other trouble spots around the world, is for re/insurers to redefine in more detail the terms and conditions of policies sold in such volatile markets, by classifying unrest according to a scale of events. In the Indonesian context, for example, there were suggestions that the riot clause should be split into two categories of cover: the first for damage occurring during “riots, strike, lockouts, malicious acts...” and the second a more expensive cover which would extend to damage caused by “terrorism, sabotage, civil commotion, popular uprising and revolution without the use of firearms... ”

The logic of such distinctions may be that with the more precise definitions, riots of the Indonesian type would fall under the latter of the two covers, and make disputes easier to deal with. Not necessarily! The full wording in this example would need to delimit the meaning of “riot” or the type of riots covered. If a contract covers riots, then, unless there is an operative exclusion or delimiting definition, all riots are encompassed. An appropriately drafted civil war or unrest exclusion could deal with such circumstances by explicitly excluding undesirable riots such as politically motivated ones, or by indirectly excluding them through excluding civil commotion, for example, where the riot forms part of it.

Notwithstanding the benefit of redefining terms, clear evidence about circumstances surrounding losses remains of paramount importance. If a tribunal wished to apply a clearer definition of riots, civil commotion, or popular uprising, it would still need clear, impartial, analytical evidence to determine the appropriate heading of damage. Specific evidence on the events provides the necessary nexus between them and the defined terms. Such evidence will determine, for example, whether a specific riot is part of a wider civil commotion, or indeed whether an incident is part of a greater event or one of many events for re/insurance purposes.

In Lexington vs D P Mann (2000) Mr Justice Walker had to deal with these problems. The reinsurers attempted to obviate the need to fully investigate the Indonesian riots, and applied to resolve as a preliminary issue the question as to what English law would classify as one occurrence. The judge, perhaps not unexpectedly, refused to decide the preliminary issue as to whether the riots necessarily constituted more than one occurrence. He said such a question can only be understood within the factual circumstances of the case and the contractual context. In other words, full evidence in the light of specific clauses.

There are, therefore, no shortcuts available to insurers and reinsurers. They cannot obviate the need to define contractual terms properly, or to obtain full evidence. It is, of course, open to parties to decide on whom the evidential burden may fall in such circumstances, thus alleviating the pressure on a reinsured to prove the loss, or a reinsurer the applicability of an exclusion clause. English courts would, depending on the wording, allow a re/insurer to delimit its obligations by clearly defining the meaning of an insured risk and “loss and/or one event,” or by inserting specific conditions expanding or limiting a back-to-back cover.

It remains only to reiterate the basic, relevant lessons which should have been learned by the re/insurance market long before the Indonesian riots:

  • Be clear about the exact risks covered and how they are delimited in relation to one another.

  • Define conditions and exclusions fully to provide for relevant eventualities.

  • Provide for a mechanism for proof of loss, causation, and evidential burden whenever possible.

  • Consider the impact of the governing law and jurisdiction, which may have the serious result of redefining the meaning and effect of the relevant terms.

  • Investigate fully the circumstances of losses, including the social, political, and economic state of the country where the events take place.

    Hermes Marangos is a Partner in the law firm, Elborne Mitchell.