Abstract
Changeable Message Signs (CMS) provide motorists with real-time traffic information about traffic congestion, incidents, roadwork zones, speed limits ahead, and Amber alerts. Although they require substantial investments, little appears to be known about their effectiveness. The purpose of this paper is to investigate empirically whether CMS reduce the number of secondary incidents. Our unique dataset combines 2008 weather data with geometric information, road work information, and 2008 accident data for a 74 miles stretch of Interstate 5 from the Mexico-US border to Orange County, CA. This freeway, which has 4 to 6 lanes in each direction and a maximum AADT volume of 230,000 vehicles, is equipped with 11 CMS. Our data suggest that secondary crashes represent approximately 5.2 percent of all primary incidents. We estimate a simple logit model to analyze the factors that contribute to secondary incidents. We find mild evidence that CMS reduce secondary crashes; their influence extends approximately 22 miles downstream from their location, with a maximum at approximately 11 miles. These results have implications for investments in CMS to provide information to motorists, although inter-vehicle communication may soon offer a viable alternative to CMS.