A review of automobile mortality data among novice drivers: Do objective ratings suggest a best practice methodology for graduated drivers license programs

Motor vehicle fatalities (MVFs) in 2005 represent the number one cause of death for U.S. adolescents aged 16 to 24. All 50 United States and the District of Columbia have some form of graduated drivers license (GDL) program designed to reduce injury and fatality accidents among newly licensed drivers. Due to lack of over-arching federal regulation, different states experience varying levels of effectiveness. The Insurance Institute for Highway Safety (IIHS) has created a point system based on age restrictions, passenger restrictions, and night driving restrictions to rate individual GDL programs. The primary purpose of this descriptive, ecologic study is to describe the relationship between the objective quality of GDL component programs as described by the IIHS and MVFs among newly licensed drivers. This study utilized fatality data found in a population of 12,477,206 novice drivers aged 16 to 18 in the United States for the year 2005 as described by the Center for Disease Control and Prevention Web-based Injury Statistics Query and Reporting System. Aggregate fatality data for each state was ranked and compared to state groups ranked by the IIHS. Spearman's correlation analysis was employed to analyze the relationship between objective GDL program ratings and fatality data. The primary results reveal the relationship between fatality data and GDL ratings (r = 0.54, p < .05) as described by the IIHS. The results also highlight specific states and geographic regions that suffer unusually high fatality rates when compared to the national average. The size of the target population, factored by state, represents a third variable that may offer an additional look into fatality trends within this population. This study demonstrates the need for states to adopt an ideal GDL program as outlined by the INS. Further investigation into how population density affects GDL programs and the strength of their components is warranted.