Strengthening Driver Licensing Systems for Teenaged Drivers

Graduated driver licensing programs are designed to reduce crashes involving teenagers by delaying full licensure while allowing beginners to obtain initial driving experience under lower-risk conditions. Graduated driver licensing strengthens the traditional learner phase by adding elements such as certification of a minimum number of practice driving hours and a minimum learner's permit holding period. It also adds an intermediate license between the learner phase and full licensure that restricts unsupervised driving in risky situations (eg, driving at night and transporting teenaged passengers). Although Florida is credited with having enacted the first graduated driver licensing program in the United States in 1996, the benefits of delaying licensure and restricting nighttime driving were demonstrated in the 1980s and early 1990s.^{1 - 4} Currently, all states and the District of Columbia have enacted some form of graduated driver licensing.

As states adopted graduated driver licensing, per capita rates of fatal crashes among teenaged drivers decreased substantially. Between 1996 and 2009, there were large declines in these crash rates for 15-year-olds (69%), 16-year-olds (68%), and 17-year-olds (53%) (the ages directly affected by graduated driver licensing) and smaller but substantial declines for 18-year-olds (43%) and 19-year-olds (39%).^{5 - 6} These decreases in rates of fatal crashes were greater than declines among 30- to 59-year-old drivers (33% during this period). However, in 2008 the rate of police-reported crashes per mile driven for 16- to 19-year-olds was 4 times the rate for older drivers, and their rate of fatal crashes was twice that for older drivers.^{7 - 8}

In jurisdictions that have adopted elements of graduated driver licensing, overall crash rates among young teenagers have declined by 20% to 40%.⁹ Restrictions on nighttime driving and transporting teenaged passengers have proven effective in reducing relevant crashes.¹⁰ National studies have found that 3-stage graduated driver licensing programs (with learner’s, intermediate, and full licensure stages) and laws that impose more licensing components, or more restrictive components, are associated with decreases in per capita rates of fatal crashes or fatalities among young teenagers.^{11 - 16}

The study by Masten and colleagues¹⁷ in this issue of JAMA adds new evidence on this important topic and reports a 26% lower per capita rate of fatal crash incidence for 16-year-olds associated with stronger graduated driver licensing programs. Masten et al also report a 12% higher rate of fatal crash incidence for 18-year-olds and a nonsignificant, slightly lower rate for 16- to 19-year-olds combined.

Implementation of the 3-stage concept of graduated driver licensing has differed among states, and licensing practices vary widely.¹⁸ The Insurance Institute for Highway Safety (IIHS) evaluated the strength of state licensing systems, rating them as good, fair, marginal, or poor. Criteria include the strength of the learner and intermediate phases (entry age and duration for both phases, a practice driving requirement in the learner phase, and strength of restrictions on nighttime driving and allowable passengers in the intermediate stage). The learner's permit and licensing ages are not inherent to a staged-in approach to licensure but are fundamental aspects of licensing laws and may be affected by graduated driver licensing provisions. Therefore, a comprehensive approach is needed to understand fully how licensing laws affect safety.

The findings reported by Masten et al differ from those reported in 2 other studies using the IIHS ratings to measure the strength of licensing laws.^{14 - 15} McCartt et al¹⁴ evaluated the association between stronger licensing laws and teenagers' per capita rates of fatal crashes. That study found a strong association between better IIHS ratings and lower rates of fatal crashes among 15- to 17-year-olds. Associations between the IIHS ratings and rates of fatal crashes for 18- and 19-year-olds were weaker and generally not statistically significant. Compared with laws rated poor by the IIHS, those rated marginal, fair, or good were associated with a significantly lower rate of fatal crashes for 15- to 19-year-olds combined, including a 16% lower rate for laws rated good. In another report, Trempel¹⁵ evaluated the association between stronger licensing laws and the rates of insurance collision claims per insured vehicle year; only licensed drivers are insured, and collision claims primarily involve low-severity crashes. There was a strong association between better IIHS ratings and lower collision claim rates among 16 to 17-year-olds and consistent, statistically significant associations between the better ratings and lower rates of collision claims for drivers aged 18, 19, and 16 to 19 years. Additional analyses in these reports^{14 - 15} examining individual licensing components revealed that earlier nighttime driving restrictions and restrictions allowing fewer teenaged passengers were associated with lower rates of fatal crashes among 15- to 17-year-olds and lower rates of collision claims among 16- to 17-year-olds; that more practice hours were associated with lower rates of collision claims, whereas longer permit-holding periods were associated with higher rates of collision claims; and that raising the permit age and the intermediate license age were associated with lower rates of fatal crashes.

Differences between the findings of these studies^{14 - 15} compared with the findings of the study by Masten et al¹⁷ may reflect different analytical approaches and differences in how graduated driver licensing programs were defined and rated. For instance, the IIHS classification of licensing laws differs from the classification used by Masten et al.¹⁷ In their study, Masten et al considered a mandatory learner's permit–holding period of at least 3 months a necessary component of graduated driver licensing. It appears that the state laws that were coded as not having graduated driver licensing programs in the study by Masten et al included 9 states with restrictions on nighttime driving and 2 states with restrictions on both nighttime driving and allowable passengers, whereas the 2 other studies^{14 - 15} found no independent benefit associated with a minimum holding period but reported benefits associated with restrictions on teenaged passengers and nighttime driving. Although minimum permit and licensing ages were also found to be important in the other study of rates of fatal crashes,¹⁴ Masten et al did not incorporate age requirements in their classification of graduated driver licensing programs or account for them in their statistical analyses.

However, taken together, the studies strongly support the benefits of graduated driver licensing for the youngest drivers. In the United States, depending on state law, graduated driver licensing programs have been directed primarily at 16-year-olds and, to a lesser extent, 15- and 17-year-olds, and most evaluations have focused on these ages. Moreover, as discussed by Masten et al,¹⁷ there are various ways in which graduated driver licensing may negatively affect 18- and 19-year-olds. There currently is no empirically validated explanation for effects of graduated driver licensing, positive or negative, on older teenagers. To the extent that some of the positive effects at earlier ages may be blunted, this is a serious issue deserving attention by researchers and policy makers. It is likely that further reductions in crashes involving young drivers can be achieved by strengthening individual components of licensing laws. Some proposed changes target older teenaged drivers, who were not shown to benefit from graduated driver licensing in the study by Masten et al. For example, in the state of New Jersey, the minimum intermediate licensing age is 17 years, and graduated driver licensing restrictions apply to all initial license applicants younger than 21 years. If other states adopted these provisions, younger teenagers would be unable to avoid graduated driver licensing by waiting until age 18 (the age at which new drivers are not subject to graduated driver licensing restrictions in most states) to apply for a driver's license. New Jersey's approach has been associated with significant reductions in the crash rates for 17- and 18-year-olds and virtually eliminates crashes among 16-year-olds without adversely affecting crash rates for 19-year-old drivers.¹⁹

In summary, a large body of evidence has found that graduated driver licensing is associated with significantly lower crash rates of young teenaged drivers. Additional crash reductions in many states likely could be achieved through strengthening current graduated driver licensing programs. Whether and how graduated driver licensing has affected older teenagers and whether these programs should be extended to include older teenagers merit further study.