Evaluations of high school driver education have yielded surprising results. Vernick and colleagues (1999) reviewed studies that evaluated the effects of driver education on adolescents, included a control group of some sort, controlled adequately for potential confounding factors, and did not rely on self-report to measure outcomes. Studies that met these criteria included two randomized controlled trials and five ecological studies that assessed differences in outcomes across time or region according to driver education status. Based on the results of these studies, Vernick and colleagues concluded that there is little evidence that adolescents who participate in a driver education course are less likely to be involved in an MVC or to have traffic violations than are adolescents who do not complete a driver education course. Quite the opposite, completion of a driver education course in high school may allow many adolescents to secure a driver's license that they would not otherwise be allowed to obtain until an older age. A number of states permit adolescents to procure a license at an earlier age if they complete a driver education course. Thus, the decrease in MVCs, if any, associated with driver education in high school is not sufficient to offset the increase in the per capita accidents that results from having more adolescent drivers licensed and on the road.

One of the studies supporting this conclusion was a randomized trial in DeKalb County, Georgia (Lund, Williams, & Zador, 1986). In this study, 16,338 high school students who applied for driver education and indicated a desire to obtain their license as soon as possible were randomly assigned to one of three conditions: the Safe Performance Curriculum (SPC), Pre-Driver Licensing Curriculum (PDL), or control. (The ethnic composition of this study sample was not detailed in the published report.) SPC consisted of 32 hours of classroom instruction, 16 hours of simulation instruction, 16 hours of driving range instruction, and three hours of instruction in evasive maneuvers, as well as three hours and 20 minutes of on-road instruction, 20 minutes of which were at night. PDL was a minimal driver education program, designed only to teach the skills needed to pass a licensure test, which consisted of 20 total hours of instruction, with only one hour of on-road instruction. Control students did not receive any high school driver education, but rather were left to learn from their parents, from a private driver training school, or not at all. Using state data on licensure, crashes, and violations, the study authors found that students assigned to either driver education group obtained licenses at a more rapid rate than did controls, but driver education participants did not have a decreased risk of MVCs or traffic violations. Overall, then, participation in driver education inspired earlier licensure among high school students, leading to more adolescent drivers on the road and a higher population crash rate.

As indicated by the review cited above (Vernick, et al., 1999), other studies, based on various study designs, have suggested similar effects of driver education. These include, for example, a comparison of adolescent licensure and fatal crash risk in states with differing proportions of students who completed driver education (Robertson & Zador, 1978) and an evaluation of changes in adolescent licensure and MVCs following the elimination of driver education in some Connecticut communities (Robertson, 1980).

Interestingly, driver education may alternatively act as an obstacle to licensure in cases where the introduction of driver education does not lower the age at which adolescents might obtain a license. In Louisiana, adolescents historically have been able to obtain a driver's license at age 15, irrespective of whether they have completed a driver education course. In 1993, however, the state passed a new law requiring 36 hours of driver education, with 6 hours of on-road instruction, as a prerequisite to licensure at this young age. Ulmer, Pruesser, Ferguson, and Williams (1999) compared the change in crash rates in Louisiana before and after the implementation of this law with the trend over time in crash rates in Mississippi, where licensure is allowed at age 15, and in Florida, where licensure is allowed at age 16. The imposition of the new requirement was associated with a decrease in the number of Louisiana 15-year-olds with driver's licenses, but did not appear to affect licensure among older adolescents. This change in licensure was associated with an overall, per capita decrease in crash rates in this state.