Can Targeting Messages to High-Risk Drivers Prevent Future Citations and Crashes? Evidence from a Randomized Pilot in Washington, D.C.

In 2023, Washington, D.C. fielded a pilot program that targeted risky drivers with personalized messaging interventions designed to reduce risky driving behavior. Risky drivers were identified by a machine learning model that predicted the risk of a vehicle getting into a crash based on past automated traffic enforcement citations and crashes. Over 90,000 drivers at high risk of being involved in a crash were randomly assigned to receive a personalized letter, text message, both, or no information. We found no measurable impacts on total citations or citations for risky driving behavior for drivers receiving any of the interventions after 3 or 12 months. We also found no measurable impact on crashes 12 months after the interventions. This study is one of the first to test the effectiveness of personalized messaging interventions on a large sample of risky drivers and evaluate impact based on actual driving behavior rather than intended or perceived behavior.