Abstract
This paper describes two experiments with a mobile roadside speedometer board. In contrast to other versions, this board only displays the speed and does not urge drivers to slow down. Speed reductions of 2.0-6.8 km/h have been noted. Reductions are largest at the actual speedometer site. Statistical analysis shows that reduction lasts over 400-1000 m. Once the speedometer is removed, speed reductions cease. It seems reasonable to attribute this to the possibility that: (1) drivers may associate speedometer board with police enforcement even though no enforcement activity or signs are seen; and (2) some kind of social control: drivers might be identified as speeders by family members etc.