From 1994 to 2004, 8,953 16-year-old drivers were involved in fatal crashes in the 43 States examined; 64 percent of the drivers were male and 36 percent were female. During this 11-year period, one-third (34%) of all fatal crashes involving 16-year-old drivers occurred in State-quarters in which GDL programs had been implemented (Table 2). The restrictions in effect for the smallest proportion of State-quarters were a requirement of at least 30 hours of supervised driving (19% of State-quarters) and restrictions related to carrying passengers (15%).
With adjustment for changes over time and differences among States that were unrelated to GDL, implementation of GDL programs was associated with an overall 11 percent lower fatal crash involvement rate for 16-year-old drivers [Incidence Rate Ratio (IRR) 0.89, 95% CI 0.80, 0.99] when compared to fatal crash involvement rates in States without GDL programs. This overall difference is based upon comparison of fatal crash involvement rates for State-quarters with and State-quarters without GDL programs. The difference reflects the combined results for all States, some of which have relatively weak GDL requirements, and therefore underestimates the impact of stronger programs. There was no significant change for drivers aged 20-24 (IRR 0.97, 95% CI 0.92, 1.03) and 25-29 (IRR 0.99, 95% CI 0.93, 1.05).
Only GDL programs (during State-quarters) with five or more components had significantly lower fatal crash involvement rates for 16-year-old drivers compared to State-quarters without GDL programs. This difference for 16-year-old drivers was 18 percent for programs with five components and 21 percent for those with six or seven components.
Figure 1. Percent Change in Annual Fatal Crash Involvement Rate in Relation to Number of GDL Program Components, Compared to State-Quarters With None of the Seven Components in Table 1, for Drivers Age 16, 20-24, and 25-29; United States, 1994-2004. Vertical Lines Represent 95-Percent Confidence Limits.
The lower fatal crash involvement rates associated with novice licensing systems varied with the provisions included in the laws. Compared to State-quarters with none of the seven specified GDL components, State-quarters with only age restrictions did not show an association with a significantly lower rate of fatal crashes involving 16-year-old drivers. This is shown in Table 3 below.
GDL programs B through H with various combinations of four major components (other than age components) were also compared to State-quarters having none of the seven GDL program components. Significantly lower rates (16-21%) in fatal crashes involving 16-year-old drivers were associated only with three types of GDL programs, all of which contained a minimum waiting period of at least three months after obtaining a learner permit and a nighttime restriction, plus either:
These GDL-associated lower fatal crash involvement rates of 16-year-old drivers were not seen for drivers ages 20 to 24 and 25 to 29, who were not affected by GDL. The percent changes for these three types of programs did not differ significantly from one another.
Table 3. Incident Rate Ratios (IRRs) and 95-Percent Confidence Intervals (CIs) for Fatal Crashes Involving 16-Year-Old Drivers in Relation to GDL Program Components, Excluding Age Requirements, Compared to Older Drivers in the Same States. United States, 1994-2004.
Reading Table 3: Program type F, for example, has three components and was in effect in 141 State-quarters. Sixteen-year-old drivers exposed to Program F were 18-percent lower (i.e., 1-0.82) in those 141 GDL State-quarters, compared to State-quarters with none of the seven components in Table 1. For drivers ages 20 to 24 and 25 to 29 the differences were only 8 percent and 6 percent, respectively, (comparing their fatal crash involvement rates in those 141 quarters with fatal crash involvement rates in State-quarters with none of the seven components) and were not statistically significant since the confidence intervals included.