An administrative policy set up by the National Swedish Police Board provided the possibility of conducting an experiment on the effect of traffic surveillance on traffic accidents. 86 stretches of road, or 10 per cent of the highway network, were surveyed. The methodology was a before (1977-1978) and after (1979-1980) study with a control. The magnitude of the effect is difficult to express, but the decrease in accidents was considerable (12-19 per cent). The long term effect of increased surveillance on speeding behaviour was investigated on 50 stretches of road with a length of 10-20 km. After two years, at three times the normal level of surveillance, there was a statistically significant effect on speeds above 100 km/h compared with drivers on roads with half the normal level of surveillance. The subjective risk of detection with respect to different traffic violations, mainly speeding, was also investigated. Ten road-sections of swedish highways were selected for study. Statistics were available of traffic enforcement activities on these road sections. A total of about 3500 drivers familiar with the roads were interviewed and an evaluation was carried out.
Abstract