The effects of reflector posts on driving behaviour and accidents were studied in an experiment including 20 pairs of two-lane rural road sections. One road in each pair was randomly assigned as an experimental road, the other remaining as a control road. The experimental roads were furnished with reflector posts at 60 m intervals for the whole length of the road in the autumn of 1987. The total lengths of experimental and control roads were 548 km and 586 km respectively. Measurements of driving speeds and lateral positions of cars on selected sites were made before the introduction of reflector posts in autumn 1987 and repeated three times during the following year. Studies on the effects on accidents were based on police reported accidents in 1982-1986 (before period) and 1988-1989 (after period). The reflector posts increased driving speeds by 5 km/h in darkness on roads with 80 km/h speed limit, when there was no snow in the ground. On broader roads with higher geometric standards and 100 km/h speed limit the posts had no significant effects. On roads with reflector posts, cars drove nearer to the hedge line. The lateral shift was about 60 cm in the wintertime on roads with 80 km/h speed limit. On roads with 100 km/h speed limit and in the snow-free time of the year on the roads with 80 km/h speed limit the shift was 15-40 cm. On roads with 80 km/h speed limit reflector posts seemed to increase the number of injury accidents by about 43% and all accidents by 20%. The number of accidents in darkness increased most. These effects were statistically significant. The results on these roads were greatly affected by exceptionally high accident numbers on three control roads in the before-period. If accident rates on these three roads had been near the group average, the estimates of the effects of reflector posts would have been only about half those mentioned above. On roads with 100 km/h speed limit, reflector posts decreased the number of injury accidents by 14% and the number of all accidents by 6%. The number of injury accidents in darkness, however, increased by 31%. The effects, however, were not statistically significant. It is concluded that on roads with comparatively low geometric standard and 80 km/h speed limit, reflector posts have adverse effects on driving behaviour that significantly increase driving risk. On wider roads with a higher geometric standard and 100 km/h speed limit, such deteriorating effects were not found and reflector posts can improve road safety, except in darkness.
Abstract