Key points are not available for this paper at this time.
OBJECTIVES: The objective of the current study was to quantify the effects of the strength of US state graduated driver licensing laws and specific licensing components on the rate of teenage driver fatal crash involvements per 100,000 teenagers during 1996-2007. The strengths of state laws were rated good, fair, marginal, or poor based on a system developed previously by the Insurance Institute for Highway Safety. METHODS: Analysis was based on quarterly counts of drivers involved in fatal crashes. Associations of overall ratings and individual licensing components with teenage crash rates were evaluated using Poisson regression, with the corresponding fatal crash rate for drivers ages 30-59 controlling for state- or time-dependent influences on crash rates unrelated to graduated licensing laws. RESULTS: Compared with licensing laws rated poor, laws rated good were associated with 30 percent lower fatal crash rates among 15- to 17-year-olds. Laws rated fair yielded fatal crash rates 11 percent lower. The longer the permit age was delayed, or the longer the licensing age was delayed, the lower the estimated fatal crash rates among 15- to 17-year-olds. Stronger nighttime restrictions were associated with larger reductions, and reductions were larger for laws limiting teenage passengers to zero or one than laws allowing two or more teenage passengers or laws without passenger restrictions. After the effects of any related delay in licensure were accounted for, an increase in the minimum learner's permit holding period showed no association with fatal crash rates. An increase in required practice driving hours did not appear to have an independent association with fatal crash rates. CONCLUSIONS: Graduated licensing laws that include strong nighttime and passenger restrictions and laws that delay the learner's permit age and licensing age are associated with lower teenage fatal crash rates. States that adopt such laws can expect to achieve substantial reductions in crash deaths.
McCartt et al. (Fri,) studied this question.