Traffic crashes, the leading cause of teenager deaths in the US, accounted for 26% (2,163) of 2013 deaths among 16- to 19-year olds.1 Florida implemented graduated driver licensing in 1996, and by January 2012, all states and the District of Columbia had implemented some form of graduated driver licensing.2,3 It typically requires drivers under 18 to proceed through phases: extended learner permit (supervised driving for 3 to 12 months); intermediate (unsupervised driving in daylight, but restricted nighttime driving and, in many states, limited passengers); and full licensure (unlimited unsupervised driving).4
State-3,5–13 and national-level14–20 studies have reported 10%–40% reductions in crash rates for 16-year-old drivers after the implementation of graduated driver licensing. However, few have considered injury rates of 16- and 17-year-old passengers, bicyclists, pedestrians, and bus riders, as a result of graduated driver licensing.10,18,21 Restricted driving may increase use of alternative transportation. Riding with parents or on buses may be safer;22,23 however, walking and biking may be more dangerous.24 Previous national studies used fatalities;14–20 however, injuries, more common than fatalities, produce a sample size adequate to detect moderate associations. We conducted longitudinal analyses of traffic injuries in 43 states to estimate the association of graduated driver licensing with injuries among 16- and 17-year-old passengers, bus riders, bicyclists, and pedestrians; and with total injuries (including in drivers).
Counts of Visible Injuries in Crashes
We used crash data from 1995 to 2012 from 43 states to obtain counts of injuries (eTable, https://links.lww.com/EDE/B55, which presents crash data year and graduated driver licensing status by state). Data were obtained from police accident reports from National Highway Traffic Safety Administration’s State Data System25 and from states. Data from the remaining seven states were not available because of no response, cost to acquire, and data format/quality issues. Police accident reports record crashes involving at least one motor vehicle traveling on a road customarily open to the public that result in injury or major property damage.25 To ensure consistent reporting across states, we included visible, incapacitating, and fatal injuries, and excluded possible injuries and property-damage-only crashes.
Classifying Graduated Driver Licensing
The presence of graduate driver licensing was determined for systems with a learner permit phase ≥3 months, plus an intermediate phase restriction on nighttime driving or number of young passengers.2
Midyear population estimates by state, age, and year were obtained from the US Census Bureau; quarterly values were linearly interpolated.26
Other Traffic Laws
The presence of other traffic safety laws may confound the relationship between graduated driver licensing and outcomes. We obtained states’ impaired driving laws (minimum blood alcohol concentration, zero tolerance, and mandatory suspension for driving under the influence) and maximum speed limits.
We obtained state-specific quarterly per capita income estimates from the US Bureau of Economic Analysis.27 The national annual consumer price index was used to adjust for inflation.28 Each state’s annual unleaded gasoline price was obtained from the Energy Information Administration,29 then adjusted for inflation.
We estimated per person-year injury rates for passengers, bus riders, bicyclists, pedestrians, and drivers by dividing counts of visible, incapacitating, and fatal injuries by population estimates.
The adjusted rate ratio (aRR) for injury under graduated driver licensing compared with no graduated driver licensing was estimated using Poisson mixed regression. The outcome was the quarterly count of injuries with an offset equal to the log of the population for that quarter. Random intercepts were used to account for the repeated measures from each state over the study period, and robust variance estimation was employed to account for potential model misspecification (e.g., overdispersion).30 Regression models included terms for year (linear spline terms with knots at 2002 and 2005), quarter (three indicator variables for four quarters), age (0 for 16, 1 for 17), and graduated driver licensing (0 for absence, and 1 for presence), as well as an interaction for age and graduated driver licensing to examine age variation. We further adjusted for traffic laws and state economic factors. In sensitivity analysis of model choice, we examined whether inclusion of random time trends for each state would change estimates for drivers and passengers. The study was approved by West Virginia University’s IRB.
There were 2,400 state quarters from 43 states; graduated driver licensing was absent for 37% of quarters. Two states (Indiana and Wyoming) did not adopt graduated driver licensing during the study period (eTable, https://links.lww.com/EDE/B55, which presents crash data year and graduated driver licensing status by state). Among 16- to 17-year olds, 383,724 passengers, 3,526 bus riders, 24,808 bicyclists, 35,003 pedestrians, and 565,602 drivers were injured from crashes (Table). Graduated driver licensing was associated with reduced total injuries for age 16 (aRR: 0.90, 95% confidence interval [CI]: 0.87, 0.93), but not age 17 (aRR: 1.03, 95% CI: 0.98, 1.08). It was associated with reduced injury rates for 16-year-old but not for 17-year-old drivers (age 16: aRR: 0.87, 95% CI: 0.83, 0.92; age 17: aRR: 1.05, 95% CI: 0.98, 1.12). It was associated with reduced injury rates for 16-year-old passengers (aRR: 0.93, 95% CI: 0.89, 0.97) but not 17-year olds (aRR: 1.02, 95% CI: 0.97, 1.07). The associations between graduated driver licensing and bus riders, bicyclists, and pedestrians were all near null (Table). There was a distribution of state-specific estimates of association between graduated driver licensing and bicyclist, pedestrian, and bus rider injuries. This distribution was centered on the null, and no geographic region or state characteristic was particularly common among states at the low or high end of the distribution. The inclusion of random time trends for each state in the model did not change the aRR for passengers. The aRR was 0.85 (95% CI: 0.81, 0.89) and 1.02 (95% CI: 0.96, 1.09) for 16- and 17-year-old drivers, similar to the main analysis.
Graduated driver licensing was associated with reduced driver and passenger injuries for adolescents particularly those aged 16 years. A New York study reported the rate ratio of fatal and incapacitating passenger injuries before and after graduated driver licensing was 1.19 (0.77, 1.84) for age 16 and 0.93 (0.61, 1.43) for 17;10 CIs were wide due to small sample sizes. A study of 1995–2012 Fatality Analysis Reporting System data reported an aRR of 0.96 (0.90, 1.03) for age 16 and 1.00 (0.93, 1.09) for age 17.31 With 191,619 injuries, relative to 6,047 fatalities among 16-year-old passengers, the injury analysis had an adequate sample size to detect moderate effects. Restricted driving under graduated driver licensing could increase instances of adolescents riding with parents, and subsequently increase passenger fatalities. However, riding with parents is much safer than riding with adolescents.22 In addition, graduated driver licensing restrictions likely reduce the number of passengers riding with high-risk adolescent drivers. Some evidence exists that graduated driver licensing may not cause adolescents to ride more frequently as passengers.32 California adolescents surveyed said that they drove earlier (58%), rescheduled an event (45%), or had parents or other adults as passengers (59%), to adapt to the nighttime driving restriction. Others walked or rode a bus or bicycle (31%). Forty-four percent reported that they drove at night despite the restriction.32
In this study, graduated driver licensing did not appear to be associated with more injuries to adolescent bicyclists or pedestrians. In another study of Upstate New York adolescents post- and pregraduated driver licensing, rate ratios of incapacitating injuries were 1.53 (0.89, 2.62) and 1.75 (0.91, 3.34) for 16- and 17-year-old pedestrians and bicyclists, respectively,10 although the study population might not be representative of 16- to 17-year olds in the 43 states studied. Potential reasons for no increase in bicyclist or pedestrian injuries after graduated driver licensing may include adolescents forgoing or rescheduling trips;32 walking rather than riding a bus or bicycle, or simply violating the law.32
A study of 1992–2002 Fatality Analysis Reporting System data reported that graduated driver licensing reduced total traffic fatalities (drivers, passengers, bicyclists, and pedestrians) by 6% among 15- to 17-year olds.18 Previous studies have suggested heterogeneity between ages 16 and 17.14,31 Our study showed that for total injuries (passengers, bicyclists, pedestrians, bus riders, and drivers), graduated driver licensing was associated with a 10% reduction for age 16, but not for age 17. Adult supervision and restrictions on nighttime driving and number of underage passengers are likely common among 16-year olds. As adolescent’s age and restrictions are removed, the influence of graduate driver licensing should decrease. An analysis of 2006–2009 Fatality Analysis Reporting System data found that the nighttime driving restriction was in effect for 84% of 16-year old, and 30% of 17-year-old drivers in fatal crashes.33 In 36 of 50 states, 17-year olds could have full licensure (unlimited unsupervised driving), particularly after taking a driver education course.4
Strengths of this study include examination of nondrivers and drivers, which increases understanding of the broader impacts of graduate driver licensing; longitudinal analysis of injuries from 43 states; examination of injuries, which produces a sample size adequate for analysis of injuries to bus riders; and adjustment for temporal trends, seasonal factors, traffic laws, and economic factors. Limitations include not having licensure status to estimate the effects of individual graduated driver licensing components; and calculations of adjusted economic factors included all ages, not just adolescents. Also, our use of population estimates to calculate rates did not account for exposure to driving, bus riding, bicycle riding, or walking.
We express appreciation to Michele Fields and Laurel Sims at the Insurance Institute for Highway Safety for their assistance in describing graduated driver licensing systems. We thank Herbert Linn at West Virginia University Injury Control Research Center for editorial assistance. These individuals received no compensation for their contributions.
1. Centers for Disease Control and Prevention. Web-based Injury Statistics Query and Reporting System (WISQARS). 2015.Atlanta, GA: Centers for Disease Control and Prevention, Available at: http://www.cdc.gov/injury/wisqars/
. Accessed April 17, 2015.
2. Insurance Institute for Highway Safety. Graduated Driver Licensing Effective Dates. 2015.Arlington, VA: Insurance Institute for Highway Safety, Available at: http://www.iihs.org/media/028244c2-87b1-4108-8122-c055877ea5da/-1425338212/Laws/gdl_effective_dates.pdf
. Accessed April 7, 2015.
3. Ulmer RG, Preusser DF, Williams AF, Ferguson SA, Farmer CM. Effect of Florida’s graduated licensing program on the crash rate of teenage drivers. Accid Anal Prev. 2000;32:527–532.
4. Williams AF, Tefft BC, Grabowski JG. Graduated driver licensing research, 2010-present. J Safety Res. 2012;43:195–203.
5. Males M. California’s graduated driver license law: Effects on older teenagers. Californian J Health Promot. 2006;4:207–221.
6. Masten SV, Hagge RA. Evaluation of California’s graduated driver licensing program. J Safety Res. 2004;35:523–535.
7. Rios A, Wald M, Nelson SR, Dark KJ, Price ME, Kellermann AL. Impact of Georgia’s teenage and adult driver responsibility act. Ann Emerg Med. 2006;47:369.e1–369.e7.
8. Kirley BB, Feller A, Braver E, Langenberg P. Does the Maryland graduated driver licensing law affect both 16-year-old drivers and those who share the road with them? J Safety Res. 2008;39:295–301.
9. Shope JT, Molnar LJ. Michigan’s graduated driver licensing program: evaluation of the first four years. J Safety Res. 2004;35:337–344.
10. Zhu M, Chu H, Li G. Effects of graduated driver licensing on licensure and traffic injury rates in Upstate New York. Accid Anal Prev. 2009;41:531–535.
11. Foss RD, Feaganes JR, Rodgman EA. Initial effects of graduated driver licensing on 16-year-old driver crashes in North Carolina. JAMA. 2001;286:1588–1592.
12. Willis DK. Fatal crashes involving 16 year-old Texas drivers pre- and post-GDL: who, when, where, and why? 2006.College Station: Texas Transportation Institute.
13. Fohr SA, Layde PM, Guse CE. Graduated driver licensing in Wisconsin: does it create safer drivers? WMJ. 2005;104:31–36.
14. Masten SV, Foss RD, Marshall SW. Graduated driver licensing and fatal crashes involving 16- to 19-year-old drivers. JAMA. 2011;306:1098–1103.
15. Baker S, Chen L, Li G. Nationwide review of graduated driver licensing. 2007.Washington, DC: AAA Foundation for Traffic Safety.
16. McCartt AT, Teoh ER, Fields M, Braitman KA, Hellinga LA. Graduated licensing laws and fatal crashes of teenage drivers: a national study. 2009.Arlington, VA: Insurance Institute for Highway Safety.
17. Lyon JD, Pan R, Li J. National evaluation of the effect of graduated driver licensing laws on teenager fatality and injury crashes. J Safety Res. 2012;43:29–37.
18. Dee TS, Grabowski DC, Morrisey MA. Graduated driver licensing and teen traffic fatalities. J Health Econ. 2005;24:571–589.
19. Fell JC, Jones K, Romano E, Voas R. An evaluation of graduated driver licensing effects on fatal crash involvements of young drivers in the United States. Traffic Inj Prev. 2011;12:423–431.
20. Morrisey MA, Grabowski DC. Gas prices, beer taxes and GDL programmes: effects on auto fatalities among young adults in the US. Appl Econom 2010;43:3645–3654.
21. Pressley JC, Benedicto CB, Trieu L, Kendig T, Barlow B. Motor vehicle injury, mortality, and hospital charges by strength of graduated driver licensing laws in 36 States. J Trauma. 2009;67(1 suppl):S43–S53.
22. Chen LH, Baker SP, Braver ER, Li G. Carrying passengers as a risk factor for crashes fatal to 16- and 17-year-old drivers. JAMA. 2000;283:1578–1582.
23. Chen LH, Braver ER, Baker SP, Li G. Potential benefits of restrictions on the transport of teenage passengers by 16 and 17 year old drivers. Inj Prev. 2001;7:129–134.
24. Beck LF, Dellinger AM, O’Neil ME. Motor vehicle crash injury rates by mode of travel, United States: using exposure-based methods to quantify differences. Am J Epidemiol. 2007;166:212–218.
25. National Highway Traffic Safety Administration. SDS overview. Available at: http://www.nhtsa.gov/Data/State+Data+Programs/SDS+Overview
. Accessed September 3, 2014.
26. Census Bureau. Population estimates. 2010.Available at: http://www.census.gov/popest/data/historical/2000s/vintage_2009/datasets.html
. Accessed September 19, 2012.
27. Bureau of Economic Analysis. Regional data: GDP and personal income. 2014.Available at: http://www.bea.gov/iTable/iTable.cfm?reqid=70&step=1&isuri=1&acrdn=3#reqid=70&step=1&isuri=1
. Accessed October 29, 2014.
28. Bureau of Labor Statistics. Consumer price idatabases. 2014.Available at: http://www.bls.gov/cpi/#tables
. Accessed October 29, 2014.
29. Energy Information Administration. State energy data system. 2014.Available at: http://www.eia.gov/state/seds/seds-data-complete.cfm?sid=AL%20-%20PricesExpenditures
. Accessed October 29, 2014.
30. White H. Maximum likelihood estimation of misspecified models. Econometrica. 1982;34:1–25.
31. Zhu M, Zhao S, Long DL, Curry AE. Association of graduated driver licensing with driver, non-driver, and total fatalities among adolescents. Am J Prev Med. DOI: 10.1016/j.amepre.2016.02.024.
32. Williams AF, Nelson LA, Leaf WA. Responses of teenagers and their parents to California’s graduated licensing system. Accid Anal Prev. 2002;34:835–842.
33. Carpenter D, Pressley JC. Graduated driver license nighttime compliance in U.S. teen drivers involved in fatal motor vehicle crashes. Accid Anal Prev. 2013;56:110–117.