Should traffic enforcement be unpredictable? The case of red light cameras in Edmonton

▶ We develop a conceptual framework that can be used to optimally rotate a smaller number of red light cameras around a larger number of locations. ▶ We then conducted a simple experiment to determine the effectiveness of three different rotation schemes: random, cyclical and fixed or no rotation. ▶ We found that not moving the cameras and keeping enforcement predictable yields a lower average violation rate than moving the cameras around and making enforcement uncertain.