A Sampling-Based System of Civil Liability

Abstract

To achieve more cost-effective deterrence of unreasonable risk-taking through civil liability, I propose and demonstrate previously unrecognized benefits of using simple random sampling to resolve multiple claims against a business or government defendant in the aggregate. I show that counter to intuition and prevailing assumptions and practice, simple sampling will enhance, not compromise, deterrent results regardless of the number of claims and the variety and significance of differences among them. Indeed, it can be used to resolve multiple claims that bear no resemblance to one another except for targeting the same defendant. The proposal can thus be employed to increase the efficiency of resolving relatively similar claims in class and consolidated actions and, by extending the application of such collectivizing processes, lower the cost of resolving all other claims that would be adjudicated as separate actions. I close by sketching a design for a reformed civil liability system that fully integrates and exploits the law enforcement benefits of sampling to better achieve the primary social objectives of accident risk deterrence and insurance.