Risk compensation denotes offsetting behavioral responses to safety improvements. Theoretical arguments suggest that, when drivers are required to drive safer cars or drive in a safer manner, they will tend to increase their driving speed or drive in some other risky manner.

The purpose of this paper is to review critically the theory and evidence on risk compensation. Our general conclusion is that the application of risk compensation theory, especially to some types of regulations, is questionable, and the empirical support for significant offsetting behavior is weak. Specifically—(1) the role of limitations in processing information is not appreciated, especially regarding risk perceptions and the learning component associated with new regulations;​ (2) the types of regulations and types of driving behavior are not adequately distinguished;​ and (3) the empirical studies haves mixed results and are subject to important limitations.