|Authors:||Arild Ragnøy, Lasse Fridstrøm|
During the winters of 1998 and 1999, an experimental study on temporarily reduced wintertime speed limits was performed, involving a number of different road links in rural Norway. The aim has been to assess the effect of reduced wintertime speed limits on speed, under varying conditions. A data set consisting of 16 786 observations was gathered on hourly mean and 85th percentile speed (dependent variables), as well as on traffic volumes, road surface conditions, luminosity and visibility (control variables). By means of a Box-Cox regression model, the partial effects on speed of reduced wintertime speed limits, as well as of the control variables, were estimated. Reduced wintertime speed limits are found to have a strongly significant effect on speed, especially when the road is not covered by snow or ice. Under such road surface conditions, however, road users voluntarily adjust their speed, regardless of the speed limit. Reduced wintertime speed limits are found to have a larger effect on speed the higher is the initial speed level on a given road link.