Several studies document the success of graduated driver licensing (GDL) systems in reducing young teen crash rates, but it is not yet clear whether any portion of the crash reduction is achieved by producing more capable drivers. The purpose of this study was to determine whether young teen drivers licensed under the North Carolina GDL system remain crash-free longer than those licensed prior to GDL, independent of the crude reductions in exposure (i.e., decreasing and delaying licensure) that may be responsible for most documented effects of GDL. Survival analysis was used to compare retrospective cohorts of 16-17 year olds before (n=105,569) and after (n=327,054) the North Carolina GDL system was implemented. The crash incidence of GDL-licensed 16-17 year olds (combined) was 10% lower than that for pre-GDL teens for at least 5 years after being licensed to drive independently (hazard ratio [HR]=0.90; 95% confidence interval [CI]=0.89, 0.91). However, more refined analysis revealed the reductions to only be among females (7%; HR=0.93; CI=0.91, 0.94) and males (15%; HR=0.85, CI=0.84, 0.87) licensed at age 16 and not among females (0%; HR=1.00; CI=0.95, 1.06) and males (0%; HR=1.00; CI=0.92, 1.09) licensed at age 17. Sixteen-year-old drivers licensed under the North Carolina GDL system experienced lower first-crash incidence during the first 5 years of unsupervised driving than did those licensed under the previous system. The benefits are greater for males, who tend to have higher crash rates. The findings contradict conventional wisdom that the entire benefit of GDL results merely from decreasing or delaying licensure among young drivers.
2010 Elsevier Ltd. All rights reserved.