Abstract
Much of the research on the relationship between young driver training and crash risks has been correlational. What remains unclear is whether driver training has a direct positive impact on reducing crashes among teen drivers. Answering this question will allow planners and policymakers who aspire to enhance teen driving safety to better understand the effectiveness of driver training on safe driving.
In the proposed study, we predict post-licensure traffic crashes among young drivers under 19 years old in Ohio based on whether they received formal driver training before obtaining a driver’s license. Our analysis is part of a larger research initiative to study teen driver safety, in collaboration with Children’s Hospital of Philadelphia (CHOP) and the State of Ohio. Data came from a licensing record database maintained by the Ohio Bureau of Motor Vehicles (BMV). The database contains detailed driver demographics, including date of birth, sex, and home address, as well as information on each driver’s interactions with the BMV, including licensing transaction dates and a driver training completion date. In accordance with data privacy agreements between Ohio and CHOP, a data operations team at CHOP will prepare a de-identified dataset that links young driver’s training records and demographics with their crash records and socioeconomic status variables that are associated with their home Census tracts such as median household income. This study is exempt from institutional review board oversight by CHOP owing to use of de-identified data.
Young drivers’ crash risks and whether they choose to take driver training are likely both affected by their safety awareness. Due to the difficulties in measuring safety awareness of young drivers, conventional statistical models such as logistic regression are unable to capture the effect of safety awareness. This shortcoming, known as endogeneity, will lead to inaccurate estimates of the relationship between young driver crash risks and their driver training status. To overcome this issue, we propose a two-stage logistic regression modeling framework. In the first stage, we predict young drivers’ likelihood of taking driver training using the travel time to the nearest driving school from the drivers’ home Census tracts and the median household income of their home Census tracts. Previous studies from the research team have found that both travel time and home Census tract’s median household income are significant contributors to teens’ likelihood of taking driver training (Dong, Wu, Jensen, et al., 2023; Dong, Wu, Walshe, et al., 2023). In the second stage, we use the young driver’s predicted probabilities of taking driver training from the first stage to estimate whether they have been involved in traffic crashes post-licensure. We hypothesize that young drivers who took driver training have lower crash risks than those who delayed licensure until 18 years old and forwent driver training.
Description
Timeline
Strategic Description / RD&T
This project relates directly to the US DOT’s top research priority of improving safety. Understanding the effectiveness of drivers education is critical to policy decisions targeted to improving safety outcomes for young drivers and the general public. The US is an outlier in terms of wealthy countries in terms of its overall fatality rates and safety trends of the past two decades. The research will also contribute directly to the US DOT’s third research priority of addressing equity issues. Traffic fatalities and their recent uptick disproportionately affect people of color, particularly Black residents, especially when controlling for geography. The uptick in fatalities has also disproportionately affected vulnerable road users such as pedestrians and bicyclists.
Deployment Plan
NA.
Expected Outcomes/Impacts
The findings will either confirm or reject the premise that drivers education improves safety outcomes for teens. This would be the first large scale study using a quasi experimental design.
Expected Outputs
We plan to publish our findings in a transportation safety journal.
TRID
Motor vehicle crashes are a leading cause of death for teens in the United States with almost seven crash-related deaths, on average, per day (Centers for Disease Control and Prevention (CDC), 2021; Miniño, 2019). To reduce young driver motor vehicle crash fatalities, all 50 U.S. states have adopted Graduated Driver License (GDL) laws that delay licensure, place driving restrictions on teen learner drivers and, in some states, require engagement in formal driver training prior to licensure for teens (Walshe et al., 2024; Williams, 2017; Williams et al., 2012). Formal training includes a required number of hours of behind-the-wheel (BTW) training with a certified instructor and classroom or online safe driving skills classes. Recent findings in Ohio indicate that new teen drivers licensed before age 18, who are subject to the combination of mandatory driving skills classes and behind-the-wheel training in addition to GDL restrictions, have better driving skills and lower crash rates than new drivers age 18 who are exempt from these requirements. (Chapman et al., 2014; Shell et al., 2015; Walshe et al., 2022).
Individuals Involved
| Email |
Name |
Affiliation |
Role |
Position |
| xiaoxiad@upenn.ed |
Dong, Xiaoxia |
University of Pennsylvania |
Co-PI |
Faculty - Adjunct |
| erickg@upenn.edu |
Guerra, Erick |
University of Pennsylvania |
PI |
Faculty - Tenured |
Budget
Amount of UTC Funds Awarded
$
Total Project Budget (from all funding sources)
$159602.88
Documents
Match Sources
No match sources!
Partners
| Name |
Type |
| City of Philidelphia |
Deployment & Equity Partner Deployment & Equity Partner |
