Terrorism Losses and All Perils Insurance † Journal of Insurance Regulation Summer 2005

This paper deals with the role of insurance for damages from catastrophic events. Prior to the attack of September 11, 2001, insurers typically provided coverage for losses from terrorism and those from almost all other causes. These causes were thought by the industry to be of sufficiently low probability and were not a reason for making changes in premiums, underwriting, or reserves. In times past, property-casualty insurance had provided protection only for losses from named causes, but over time contracts had evolved into ones that paid for damages from all causes except for specifically named exclusions. Terrorism was not one of these exclusions, so many insurers (and re-insurers) had to pay claims resulting from the September 11 attacks. After these events, coverage for losses caused by terrorism was explicitly excluded from many commercial insurance policies. This paper argues that the “all-cause” or “all perils” contract was useful in providing protection against unusual events, but that it created problems for insurers and policyholders when a catastrophic “unnamed peril” event, such as September 11, occurred. We provide a behavioral explanation for both the offering of all perils insurance for seven decades and the subsequent changes in contracts to exclude terrorism. We then suggest public policy interventions that could improve the functioning of the property-casualty market.