96% of parents say their teen drives well - even the ones who've seen them text behind the wheel

Here is a number that should make parents uncomfortable: one in three say they worry their teen or young adult driver could cause a crash. Here is the number that should make everyone else uncomfortable: 96% of those same parents rate their child's driving as average or better than peers.

Both figures come from the same poll. The University of Michigan Health C.S. Mott Children's Hospital National Poll on Children's Health surveyed 1,780 parents of children ages 16 to 25 in February 2026. The results reveal a cognitive dissonance that traffic safety researchers have long suspected but can now quantify.

What parents are seeing

More than half of parents report witnessing at least one unsafe driving behavior by their child. Nearly half have seen aggressive driving - speeding, tailgating, weaving through traffic. One in four has observed distracted driving, including texting or multitasking behind the wheel. And 17% have seen their child drive while impaired - not just drunk or high, but also while sleepy or emotionally upset.

Parents of young adults ages 21-25 were more likely to report distracted or impaired driving than parents of teens ages 16-20 - suggesting these behaviors either increase with age and independence or become more visible to parents as driving situations diversify.

The Lake Wobegon effect on wheels

Despite witnessing these behaviors, the overwhelming majority of parents gave their child favorable driving ratings. Only 4% rated their child as a worse driver than peers. Even parents who had observed distracted or impaired driving tended to give their child the same ratings as those who hadn't seen such behaviors.

The most counterintuitive finding: parents who had witnessed aggressive driving were actually less likely to rate their child as a worse driver than peers. Mott Poll co-director Sarah Clark offered a possible explanation - many parents engage in these same behaviors themselves and may not view them as dangerous.

This is the pattern psychologists call the better-than-average effect, but applied to one's offspring rather than oneself. It suggests that for many families, risky driving has been normalized to the point where it doesn't register as a red flag.

Concern without consequences

The disconnect extends to action. While one in three parents expressed worry about their young driver, only one in four reported taking steps to address their child's behavior. Those who did act mentioned monitoring devices, restricted driving privileges, refusing car access, or threatening to stop paying for insurance.

That means roughly three-quarters of parents who observe unsafe driving behaviors do nothing about it - or at least nothing structured enough to report in a survey.

Clark emphasized that parental involvement remains critical for young driver safety, especially given wide variation in state driver education requirements. Some states allow online-only instruction. Some don't require formal driver's education at all. In many cases, parents are the primary - and sometimes only - driving instructors their children ever have.

Beyond texting - the distractions nobody regulates

While most states have hands-free laws and prohibit texting while driving, other common distractions remain unregulated: eating, grooming, managing navigation apps, or interacting with passengers. These behaviors are widespread among young drivers and carry real risk - any activity that diverts visual attention or slows reaction time increases crash probability.

Motor vehicle crashes remain a leading cause of death for teens and young adults, a statistic that has proven stubbornly resistant to improvement despite better vehicle safety technology and graduated licensing laws.

The poll's limitations

The survey relies on parental self-reporting, which introduces bias in both directions - parents may underreport their child's risky behaviors (because they don't want to admit it) or overreport their own vigilance (because they want to appear responsible). The poll also captures a snapshot rather than tracking behaviors over time.

There is no way to verify parental assessments against objective driving data, and the definition of 'unsafe behavior' may vary considerably among respondents. A parent who regularly speeds may not classify their child's speeding as a problem.

But the gap between concern and action, and between observed behavior and perceived skill, is too consistent to dismiss. Clark's recommendation is straightforward: families need ongoing conversations about safe driving, and parents need to take their concerns seriously before a preventable tragedy makes the point for them.