In this paper, we make three central arguments. The first important argument questions the implicit assumption of the activity-level effect literature that there is a monotonically increasing relationship between the probability of an accident's occurring and the quantity of the underlying risky...