A 'Genetically Modified' Liability Insurance Contract

Abstract:     
This article uses the new technique of genetic programming to discover liability insurance contracts that, in theory, would substantially reduce the effective cost of accidents faced by potential insureds. Specifically, it suggests using formulas containing statistics about the distribution of damages in lawsuits brought against the insured as a way of customizing the per occurrence limit on a case-by-case basis. It further suggests permitting the insurer to modify the conventional duty to settle that American judges have implied into most liability insurance contracts. This modification would permit statistics regarding the distribution of damages expected in a lawsuit materializing against the insured to control the details of any duty to settle. It shows the advantage to the insured of letting lawsuit-specific statistics determine the settlement offer needed to trigger the duty and the maximum amount owed by the insurer in the event it were to breach the duty to settle. The article concludes, however, that capture of the significant savings available in theory from any of these reforms will be extraordinarily challenging due to the absence of procedural mechanisms that would be needed to resolve disputes arising thereunder. Moreover, one should expect potential victims to object that these "genetically modified" insurance contracts reduce their ability to be fully compensated for injuries through the tort system.

Keywords: Liability insurance, genetic programming, insurance regulation, duty to settle, per occurrence limit, insurance, statistics

JEL Classifications: C02, C49, C61, G22, K12

Working Paper Series