Teenage drivers are a danger to themselves and to others. In their first year of driving, 15–30% of teen drivers experience a crash (McCartt et al. 2009, Curry et al. 2015), and traffic accidents are teenagers’ leading cause of death in most countries (Kyu et al. 2016). These sobering statistics highlight the need for effective safety measures. At the same time, overly stringent restrictions often reduce the benefits of driving for teenagers, leading to large reductions in licensing (Gilpin 2019, Balia et al. 2023) and teenage employment (Argys et al. 2019, Bostwick and Severen 2024).
In a recent paper (Moore and Morris 2024), we show that a targeted driving law significantly enhanced road safety without compromising teenagers’ ability to gain experience on the road. Our study investigates a policy introduced in 2007 in New South Wales (NSW), Australia’s most populous state. The policy banned first-year drivers from carrying two or more passengers under the age of 21 between 11:00pm and 4:59am, a specific type of driving that is extremely dangerous and leads to many deaths (Tefft et al. 2013). Many jurisdictions impose broader restrictions on new drivers that limit independence, such as all-day passenger limits and banning any driving late at night. In contrast, this particular policy retains most benefits of driving, such as providing teenagerss with access to employment and education activities.
Using administrative data on drivers’ license records linked to detailed crash data, we use a differences-in-differences Poisson model that uses daytime crash outcomes (8:00am to 7:59pm) to account for other factors affecting nighttime crash risks. After showing similar pre-treatment trends in first-year drivers’ daytime and nighttime relative crash rates (Figure 1), we observe a sharp and sustained reduction in nighttime multi-passenger crashes between the restricted hours. We estimate that the restriction reduced reported multi-passenger crashes between 11:00pm and 4:59am by 57%, casualties by 50%, and hospitalisations and fatalities by 58%. An indication of the extent of the change is that late-night multi-passenger crashes represented 18.3% of first-year fatalities before the restriction was introduced, and 4.3% afterwards.
Figure 1 Multi-passenger crash rates of first-year drivers in NSW by time of day
Spillovers and persistence
One concern with this targeted restriction is the potential for changes in driving behaviour that increase other types of crashes. However, we find no evidence of such effects. Rather, we find that reduction in first-year drivers’ multi-passenger crashes actually extends beyond the restricted period, with a statistically significant decline in crashes between 8:00pm and 10:59pm (Figure 2). This suggests that the restriction reduced trips for activities spanning several hours (e.g. driving to the cinema or parties).
Figure 2 Estimated changes in multi-passenger crash rates of first-year drivers in NSW by time of day
A further concern relates to the long-term impacts: do teenage drivers compensate by taking on additional risks after the restricted period ends, or are their driving skills compromised? Interestingly, we find no evidence to support either of these concerns. In fact, we estimate reductions in nighttime multi-passenger crashes in the second and third years of driving, after the restriction was lifted. There are no discernible differences beyond the third year, but by that time the underlying crash rates are just one-fifth those of first-year drivers (Figure 3). There is no evidence that the restriction delays the development of driving skills, affects the likelihood of obtaining a license, or changes the transition rate to the next license class (where drivers are no longer subject to the restriction).
Figure 3 Implied effect on late-night multi-passenger crash rates by year of driving
Overall effects
Per 100,000 drivers affected, the direct and indirect effects of the policy were 5.9 fewer deaths, 41 fewer hospitalisations, 98 fewer minor injuries, and 164 fewer crashes with property damage. The estimated value of these improvements was approximately A$412 million (US$287 million), or A$738 (US$513) per first-year driver. Notably, the restriction's impact on reducing harm to passengers and other drivers accounted for nearly all of these gains, indicating that it substantially reduced the negative externalities associated with teenagers’ driving.
Remarkably, the fatality reductions from this targeted restriction are comparable to much more stringent regulations in the US:
- The estimated decrease in traffic deaths per teenager is 45-90% of the overall impacts of effective and comprehensive licensing systems that consist of many stringent rules (Gilpin 2019).
- The reduction in fatalities is similar to the gains from raising the US minimum driving age by one year (Huh and Reif 2021).
Importantly, these improvements were achieved without reducing licensing rates among teenagerss.
Potential mechanisms
We conduct additional analyses to explore the mechanisms behind the success of the NSW restriction. Consistent with high levels of compliance with the restriction, we found that the crash reduction could be entirely explained by a decrease in nighttime driving with peers, rather than safer driving in such situations. The persistence of the road safety benefits over time suggested a role for habit formation, where compliance with the restriction during the first year of driving influenced longer-term behaviour. A positive correlation between the short-term and long-term effects on different subgroups added support for this hypothesis.
Our study also highlighted the ‘expressive value’ of the law, which may have influenced social norms around nighttime driving with peers. Over time, as the crash reductions grew larger and fewer first-year drivers were caught violating the restriction, it became clear that the law had a broader impact on teenage driving behaviour. These findings suggest that targeted restrictions, particularly those implemented at the start of a driver’s independent driving experience, can have significant and lasting effects on road safety.
Conclusion
Our findings align with another recently published study that analysed an Italian intervention: a vehicle-power limit for newly licensed teen drivers (Balia et al. 2023). This policy, which restricted new licensees from driving high-powered vehicles, resulted in a significant reduction in crashes, partly due to a decrease in the number of new drivers on the road. However, the study also found that those who did get their license were less likely to cause severe accidents, and these benefits persisted after the restriction was lifted. This parallels our findings in NSW, where the targeting of specific high-risk driving behaviours led to substantial and persistent safety improvements. Both studies highlight that effective road safety policies can achieve significant reductions in crashes without completely restricting teenage driving. They also emphasise the importance of focusing on specific risk factors, rather than implementing broad, less-targeted measures.
References
Argys, L M, T A Mroz, and M M Pitts (2019), “Driven from Work: Graduated Driver License Programs and Teen Labor Market Outcomes” Federal Reserve Bank of Atlanta Working Paper No. 2019-16.
Balia, S, R Brau, and M G Nieddu (2023), “Depowering Risk: Vehicle Power Restriction and Teen Driver Accidents in Italy”, Journal of Law and Economics 66(4): 765-795.
Balia, S, R Brau, and M Nieddu (2023), “Temporary restrictions on the use of high-power vehicles significantly reduce teen driver accident rates”, VoxEU.org, 3 August.
Bostwick, V and C Severen (2023), “Driving, Dropouts, and Drive-Throughs: Mobility Restrictions and Teen Human Capital”, IZA Discussion Paper No. 16183.
Curry, A E, M R Pfeiffer, D R Durbin, and M R Elliott (2015), “Young Driver Crash Rates by Licensing Age, Driving Experience, and License Phase”, Accident Analysis and Prevention 80: 243–50.
Gilpin, G (2019), “Teen Driver Licensure Provisions, Licensing, and Vehicular Fatalities”, Journal of Health Economics 66: 54–70.
Huh, J and J Reif (2021), “Teenage Driving, Mortality, and Risky Behaviors”, American Economic Review: Insights 3(4): 523–39.
Kyu, H H, C Pinho, J A Wagner et al. (2016), “Global and National Burden of Diseases and Injuries among Children and Adolescents between 1990 and 2013: Findings from the Global Burden of Disease 2013 Study”, JAMA Pediatrics 170(3): 267–87.
McCartt, A T, D R Mayhew, K A Braitman, S A Ferguson, and H M Simpson (2009), “Effects of Age and Experience on Young Driver Crashes: Review of Recent Literature”, Traffic Injury Prevention 10(3): 209–19.
Moore, T J and T Morris (2024), “Shaping the Habits of Teen Drivers”, American Economic Journal: Economic Policy 16(3): 367-393.
