On 21 November 2013 protesters gathered in central Kiev to demand the resignation of President Yanukovich and his government following his sudden decision not to sign an Association and Free Trade Agreement with the European Union. By the end of January 2014 unrest had spread to other parts of Ukraine, leading to violent clashes, loss of life and damage to property.
This article considers the implications for All Risks and Political Violence insurers and reinsurers.
Free trade agreements rarely set pulses raising in the West, however for Ukraine the question of accession to an EU Treaty ignites an old, highly divisive and existential question: Is Ukraine a Western-facing European nation, or should it more properly be associated with its Russian neighbour? The recent unrest is merely the latest manifestation of this long unresolved question.
Ukraine is a land of contradictions. To Russia, Kiev is the ancient crucible of the Russian state, having been established in 882 AD as the capital of the Kievan Rus proto-state long before Moscow. Yet Ukraine is now not part ofFree trade agreements rarely set pulses raising in the West, however for Ukraine the question of accession to an EU Treaty ignites an old, highly divisive and existential question: Is Ukraine a Western-facing European nation, or should it more properly be associated with its Russian neighbour? The recent unrest is merely the latest manifestation of this long unresolved question.
the Russian state at all. On the other hand, much of West Ukraine was never part of the Russian Empire, but subject variously instead to Polish and/or Austro-Hungarian rule until 1918, before being subsumed into the Soviet Union after the civil war. West Ukraine tends to have more Ukrainian speakers and be more
Western orientated than the East, although it would be trite to reduce the present political divisions to simply that dichotomy. (Krushchev’s gift of Crimea from Russia to Ukraine in 1954 also complicated matters.) The recent protests have taken on broader anti-government dimensions, being driven also by concerns about corruption, state power and human rights abuses. Unrest has spread from Kiev to regions throughout Ukraine.
Demonstrations have been violent, people have been killed, and government and commercial buildings have been damaged and occupied. At times, demonstrations have attracted several hundreds of thousands of participants. Trade has been interrupted. Draconian legislation has been passed (and then repealed) and a state of emergency threatened. How will All Risks and Political Violence Policies respond?
All Risks Policies: Political Violence Perils and Exclusions
Many standard wordings regarding the exclusion of or cover for political violence were created in the London insurance market. With the globalisation of insurance and reinsurance, these wordings appear in contracts around the world, including Ukraine.
Whether considering exclusions from All Risks policies, or policies covering political violence, each loss must be carefully considered against the backdrop of unrest, and also having regard to its own very particular circumstances and wording. It does not necessarily follow that all losses occurring during a period of unrest can be attributed to a particular excluded political violence peril. Regard must always be had to:
- the nature and meaning of the peril;
- the causal connection required between a relevant peril and a loss; and
- who bears the burden of proof.
Critically, no wording can be divorced from the law that governs it. It should not be assumed that every jurisdiction will ascribe the same meaning to words as English law. Although phrases such as “civil commotion” and “usurped power” have settled and particular meanings in most common law jurisdictions, this is not necessarily the case elsewhere. In fact it can be difficult even to translate these words into some languages.
Nevertheless, because these words and phrases originate in London, and because the English courts have considered their meaning many times, English law provides a natural reference point for any analysis of political violence insurance issues. Ukrainian law will govern locally issued policies, but English case law might be cited to the tribunals. English and U.S. laws may be directly applicable in global policies.
Typical All Risks Exclusions
Political violence exclusions in Ukrainian policies mirror those applied elsewhere in the world, since they are usually insisted upon by international reinsurers.
A typical All Risk Exclusion (the NMA2918) excludes loss or damage:
directly or indirectly caused by, resulting from or in connection with any of the following regardless of any other cause or event contributing or in any other sequence to the loss:
1. War, invasion, acts of foreign enemies, hostilities or warlike operations (whether war be declared or not), civil war, rebellion, revolution, insurrection, civil commotion assuming the proportions of or amounting to an uprising, military or usurped power.
2. Acts of terrorism… [Discussed further below]
Part 1 of the clause has been applied in the London insurance market since the Spanish Civil War. Part 2 (terrorism) was added after September 11. Note that all of the words in Part 1 are left undefined, and so must be interpreted by the relevant tribunal.
Nearly all of the words in Part 1 have been subject to judicial interpretation in England. Some of the most relevant are considered here.
For the exclusion to apply, at least one of the excluded perils must be operating and there must be at least an indirect link between such peril and the particular loss. The efficacy of any exclusion must also be judged against the cover intended to be provided by any particular policy. Tribunals are reluctant to read exclusions so broadly as to entirely defeat the purpose of the coverage offered.
Usually, All Risk policies will cover (or at least not specifically exclude) riot and looting. During times of civic unrest, the coverage question that most commonly arises therefore is: when have losses caused immediately by rioting and looting become sufficiently connected to an excluded peril, such as “civil commotion assuming the proportions of or amounting to an uprising”, or “rebellion”?
Civil Commotion Assuming the Proportions of or Amounting to an Uprising
There is a good deal of case law on the meaning of “civil commotion,” a phrase that has been in use in non- marine policies since 1727, although there is no exhaustive definition. It has been said to be “rising of the people, for purposes of general mischief.” Cases also emphasise the need for “turbulence” and “tumult”.
The court in the famous English case of Spinney’s1 said that a “civil commotion … amounting to a popular uprising” should involve a “really substantial proportion of the populace, although obviously not all the population need participate, and that there should be tumult and violence on a large scale.”
In terms of motivation, there “must be some unanimity of purpose amongst those participating, and that the purpose must involve the displacement of the government,” although it can connote a lesser degree of organisation than “rebellion”.
The task for a tribunal is to assess the evidence and to make a judgment. This can be difficult as it may not be possible to precisely identify the persons involved, let alone ask them about their motivation. A decision will necessarily involve making inferences and some generalisations, albeit based upon the evidence presented.
Rebellion and Insurrection
In English law, “rebellion and insurrection have somewhat similar meanings to each other … each means an organised and violent internal uprising in a country with, as a main purpose, the object of trying to overthrow or supplant the government of that country.”2
Similarly, U.S. case law has said that the terms mean “a violent uprising by a group or movement acting for the specific purpose of overthrowing the constituted government and seizing its powers.”3
It is not necessarily the case that demonstrators calling for resignations or elections will amount to “rebels.” If that were so, many run-of-the-mill political demonstrations around the world would amount to rebellions.
Tribunals will need to consider evidence as to the motivation and degree of organisation of the participants and form a view as to whether their intention is to actually to “overthrow or supplant” the government. The presence, or otherwise, of violence, and the overall gravity of the situation will also be important.
A typical terrorism exclusion will exclude loss and damage directly or indirectly caused by or resulting from “any act of terrorism”, which is often be defined as follows:
… an act, including but not limited to the use of force or violence and/or threat thereof, of any person or group(s) of persons, whether acting alone or on behalf of or in connection with any organisation(s) or
government(s) committed for political, religious, ideological or similar purposes including the intention to put the public, or any section of the public, in fear.
The terrorism exclusion is notable in that it is the only excluded political violence peril to be specifically defined in All Risks contracts. This common exclusion was widely applied following the 9/11 attacks. The clause is very widely drafted indeed. Under the definition, terrorism need not involve any actual violence; a mere threat wold be sufficient. Also no group or organisation is needed. The key element for insurers to
prove is the motivation of the perpetrators. It is the element of “political”, “religious” or “ideological” motive that distinguishes acts of terrorism from ordinary criminality. For example acts perpetrated purely for private gain would not amount to “acts of terroism.”
The Ukrainian Prime Minister has blamed recent events on “terrorists.” This is a typical reaction by regimes under threat. (The Thai government reacted similarly when faced with demonstrations in 2010.) Naturally, any coverage analysis should focus upon the law and facts, and not upon the mere slogans of politicians or demonstrators.
- Spinney’s (1948) v Royal Insurance 1 LLR  406
- National Oil Company of Zimbabwe v Sturge  2 LLR 281
- Pan Am v Aetna – SDNY Sept 1973; 71 Civ 1118 - endorsed by US Court of Appeals 4
Evidence and Burden of Proof
In practice, the precise circumstances of a loss, and the motivation of the particular individuals who caused it, may never be known. Tribunals have to draw inferences from the information available, and consider if the parties have discharged their burden of proof as to the engagement of coverage, or of exclusions, as the case may be.
Under most laws, the burden of proving that a policy exclusion is engaged will, by default, lie upon the insurer. Exclusions in All Risk and Political Violence insurance policies seek to alter this presumption with a reverse burden of proof clause such as this:
If insurers allege that by reason of this exclusion, any loss, damage, cost or expense is not covered by this insurance, the burden of proving the contrary shall be upon the insured.
The effect of such a clause was considered by the English Court in Spinney’s. The court held that:
insurers … must produce evidence from which it can reasonably be argued that (a) a state of affairs existed or an event occurred falling within the exception, and (b) the excepted peril directly or indirectly caused the loss. It is only when an arguable case of this nature is made out that the insured is required to disprove it.
In other words, an insurer cannot arbitrarily allege an exclusion applies in order to reverse the burden; rather the insurer must first adduce evidence to show that it has an arguable case that the exclusion could apply. If it can do this, the burden shifts back to the insured to explain, with evidence, why an exclusion should not apply. In practice, it can be difficult for insureds to achieve this.
For example, in the case of losses occurring in Kiev over recent weeks, it may not be difficult for insurers to provide some evidence that certain losses were arguably caused by excluded perils. The burden would then shift to the insured to prove that the loss was not caused by an excluded peril, but by a covered peril. This can present difficulties in situations where witnesses and evidence may be difficult to obtain.
Causation – Linking the Peril to the Loss
For an exclusion to bite, a sufficient causal connection must exist between the excluded peril and the loss suffered by the insured.
The default position under most legal systems is that a loss must be proximately (or immediately) caused by an excluded peril in order for the exclusion to be triggered. In light of the practical difficulty in showing the precise cause of, for example, each smashed window or each fire occurring during political violence, insurers have sought to displace the usual rule in the policy wording.
For example the NMA 2918 excludes loss:
directly or indirectly caused by, resulting from or in connection with any of the following regardless of any other cause or event contributing or in any other sequence to the loss.
The aim of the wording is to engage the exclusion when an excluded peril is only one of many causes of loss. A literal application of the clause would make it very easy for insurers to link excluded perils to particular losses by reaching back into the chain of causation as far as they wish. The English courts have however limited insurers’ ability to do this. In Spinney’s the judge held that, notwithstanding the words of the clause:
Plainly, there must be some limit on the application of the clause, for the chain of causation recedes infinitely into the past. The draftsman must have intended it to stop somewhere: and that point must be the point at which an event ceases to be a cause of the loss, and becomes merely an item of history … the task is to assess whether the particular act of violence simply takes place against the background of a “warlike” state of affairs, or whether it has itself (even in a rather remote way) a warlike aspect of its own.
In fact, in civil unrest situations, it is very often the case that the excluded peril is not the proximate cause of actual loss. For example, looting could take place several miles away from the centre of the disturbance, albeit against the backdrop of unrest and lawlessness. To trigger the exclusion insurers need to draw some credible causal link between an excluded peril and the loss, even if other causes are operative.
Political Violence Insurances
Political Violence Insurance is available to buy back some of the perils usually excluded from All Risks policies. Common products include:
- Standalone terrorism covers;
- Strikes, riots and civil commotion covers (SRCC); and
- War on Land covers.
A variety of products are available on the London market, and can be bought in various combinations. Some more recent wordings allow particular heads of cover to be selected from a “menu” of perils.
Although it is difficult to generalise, political violence insurances tend to have a number of common features:
- The perils covered are specifically and closely defined (in contrast to the exclusions in All Risks policies where most terms are left undefined);
- The wording will expressly require a close or direct link between the peril insured and the loss occurring. For example, a terrorism insurance will likely only cover losses directly caused by an act of terrorism; and
- Perils not specifically covered will be excluded in broad terms.
Ukraine has a terrorism pool, and so little terrorism cover is placed on international markets. SRCC cover is available from international insurers, although the number of local policies issued is thought to be modest. International insurers may however be exposed via global programmes.
Reinsurance: Coverage and Claims Control
Locally issued instance policies will likely be governed by Ukrainian law. As with any reinsurance, reinsurers will wish to check that claimed losses are covered by both the insurance and reinsurance.
In the case of facultative reinsurances, reinsurers will wish to exercise their claims control rights to control the claim. This is particularly relevant in circumstances where the local market has little experience of handling civil unrest claims.
Unrest in Ukraine has extended over two months now, and has impacted a number of cities. Questions may arise as to how many loss occurrences have taken place, for presentation to excess of loss insurances and reinsurances. Such issues depend very much on the particular wording used, and there is a good deal of case law in this area.
For example, Mann v Lexington4 concerned damage to 22 insured stores at locations throughout Indonesia, over two days of rioting. The deductible (in a retrocession) applied “each location any one occurrence.” The question for the court was whether the losses could be combined as a single claim, or should be presented as 22 separate claims. The court held that there was insufficient unity of time or place for the losses to be combined. There was some evidence that the riots had been orchestrated, however the court held that it: “is difficult to conceive of a situation in which if the properties were some distance apart, and if there was a lack of unity of time, there could still be one occurrence by virtue of some factor as ‘orchestration.’”
There are more aggregative wordings in use in the market, such as those seeking to combine losses arising from similar “causes”.
Some wordings contain “hours clauses” that will enable an insured or reinsured to present, for example, all riot or civil commotion losses occurring during 72 hours in a particular province as a single loss occurrence. The wording has the advantage of simplicity, although care should be taken to be sure that all the losses presented can indeed be attributed to the relevant time period and geographic area.
At the time of writing it remains to be seen how matters will develop in Ukraine. It is tempting to view such situations in a linear fashion, with demonstrations leading to riots, civil commotion, insurrection and
ultimately possibly to civil war. In reality, such situations are dynamic, with unrest and violence ebbing and flowing over time, and arising in differing locations with different purposes. Individual losses can therefore require very careful scrutiny.