Teen driving restrictions don't have as big an impact as expected

My friend Jim captured this excellent moment in science reporting this morning. Thankfully, as I check Google News now, the headlines are drifting more towards the real story, which is fairly interesting. Turns out, deadly car accidents aren't so much a function of driver age as they are a function of driver experience.

Basically, over the past few decades, several states have placed stringent limits on teenage drivers—usually when they can drive, and who they can drive with. The idea was to separate first-time drivers from risky driving situations, and a lot of people assumed these measures were saving lives. Instead, we now know, the rules merely shifted when the deadly accidents happened. Some lives were saved. But, in general, the results were pretty much a wash.

The researchers found that states with the most restrictive graduated licensing programs — such as those that required supervised driving time as well as having night-driving restrictions and passenger limitations — saw a 26% reduction in the rate of fatal crashes involving 16-year-old drivers compared with states without any restrictions.

But the rate of fatal crashes among 18-year-old drivers in those states jumped 12% compared with the states without restrictions.

A similar trend was seen when comparing drivers in states with strong graduated licensing programs with those in states with weak programs: The rate of fatal crashes among 16-year-old drivers was 16% lower but was 10% higher among 18-year-old drivers.

Overall, since the first program was enacted in 1996, graduated programs were linked to 1,348 fewer fatal crashes involving 16-year-old drivers and 1,086 more fatal crashes involving 18-year-old drivers.

The speculative response: You can place restrictions on new drivers that limit their exposure to situations where mistakes are likely to happen. But, eventually, they'll have to navigate those situations on their own. And when they do, the mistakes creep back in. So maybe we need to look for a better way to mitigate the mistakes than simply instituting age-dependent restrictions. Personally, I wonder what the results would be if driving education included time to practice driving (either virtually or on a test course) with the distractions they're likely to encounter in real life. I know I learned how to drive and talk at the same time, and how to know when to shut everybody up, by experience. Maybe there's a way to do that in a safer environment.


  1. Kids who wait until 18 can often waive the driver’s ed requirement and other restrictions like having teen passengers, and driving at night. If these were compulsory for all drivers that might change the statistic.

    1. I did exactly that. Only had to take the written test and a short drive with an instructor and got my operator license.

      One thing, one simple little thing, would make everyone a better driver. PAY ATTENTION TO YOUR SURROUNDINGS!

      Ok, two things… AND USE YOUR TURN SIGNALS!

      1. And don’t tailgate.  Two second rule.
        And stay out of the fast lane if people keep passing you on the right.  i.e., keep right except to pass.

  2. I’d be interested to see how kids who take drivers ed at Mesa High School in Mesa, AZ compare to those how take drivers ed in the other schools in the district. Mesa High actually has a range at the school where students can practice driving, with other students in the car. Of course, given budgets, etc. they may have idled the range now.

  3. Almost every state has requirements that 16 and 17 year olds driver with their parents or guardians for 30 to 60 hours before they receive their license.  That number should actually be upwards of 100 hours.   That time should be spent in a variety of circumstances and conditions.  Teens should not experience a new driving condition on their own.  Drivers ed is limited in time and scope.  If teens are to become good drivers, the responsibility falls on the shoulders of the parents

  4. 2 small observations: perhaps training works (supervised driving time)? Also, about “inexperienced drivers of any age”: my (layman’s) understanding is that the brain’s prefrontal lobe doesn’t reach full maturity until age 22 or so, at which point executive functions like deciding on risky behavior based on a realistic assessment of possible consequences is finally in place. Any data on people who wait til their mid 20’s to start driving?

    1. The first counter-argument that came to mind is the success of professional drivers younger than that in racing.  Looking at the success of some teen/early-twenties drivers like Vettle, Hamilton and Alonso in F1 speaks to ‘experience’ as trumping “maturity” as far as driving skill.  These guys had hundreds, if not thousands, of hours of seat time before they got onto world stage, and it shows in their driving records (well, on track, anyway…).

      But you don’t need to go to the pros to see a benefit from simple experience and training. Take a look at http://www.accidentavoidance.com/AA%20Study%20by%20Bill%20Scottpdf.pdf, especially Figure 1, comparing teens who attended BSR’s Accident Avoidance class to their peers who did not.  The differences are striking.  (Figure 2, dealing with rookie West Virginia State Troopers is also interesting, if not as germane.)

      Yeah, the sample size in the first group is too small to say much, but BSR’s report is harder to dismiss.  (Note that while what they have up is their report, I believe they wrote it up after a story was published in the Winchester, VA paper about the teenager comparison, and after one of the local law enforcement groups noted their benefits as well.  I know I was told about similar stats back in 1998, before the data they use in their report existed.)

  5. Toss new drivers into one of those fancy Lexus driving simulators and throw dangerous situations at them. Or heck, if you’re cheap, sit them down with an xbox and a copy of Forza. Simulation still isn’t real life but there’s no penalty for failure in one either, and you can keep trying until you get it right.

    It would also help if the US driver certification process wasn’t such a joke. Require some real driving education and maybe some of these kids would set out on their own with some understanding of the physics of driving, rather than what the default speed limit is in a neighborhood.

  6. I wonder if this is a result of the “Woot!  I can have passengers now!” effect.  Like how a 16-year-old American would react if 16-year-olds could suddenly drink legally, compared to how they act in places where it’s already legal for them to drink.

  7. New test:  On an oily skidpan, excessive speed, 5 frat brothers in the back, doobies and guns on the passenger seat, cop in pursuit, AC/DC on at 11, get through the course and parallel in the safe house.  Sweat it out, kids.

    Also, some kids just don’t have spacial awareness – I remember the hysterical laughter being a passenger in a friend’s car, he kept bumping things, “we’re all going to die” was a common refrain.  Not even any booze.  Poor him.  Couldn’t avoid things, and had a band of laughing monkeys everywhere he went (he was the first to have regular access to a vehicle).

  8. In Canada, at least in BC, it’s set up this way. It doesn’t matter if you’re 40 or 17, you get a New Driver classification – complete with restrictions. 

  9. Turns out, deadly car accidents aren’t so much a function of driver age as they are a function of driver experience.

    Maggie, given the proximity of the datapoints, that’s not a particularly safe inference to make: the same rationale when measuring the height of babies at 16 and 18 days would lead you to conclude that they wouldn’t get significantly taller in later life.

  10. This is why I loved living in the country.
    Pops took me out to a field at 14 and just had me start driving around getting used to the car and basic principles of driving. Then we moved up to country roads, then city streets and then the interstate.
    By the time I got to Drivers Ed it was a snap.

    Nothing beats experience hands down.
    You want your kids to live longer? Then don’t just rely on the schools (my driver’s ed teach couldn’t have cared less as long as someone was able to get him to the donut shop in one piece) or the state to teach them. Spend some time with them yourself and teach them how to be responsible behind the wheel.
    Unless you’re a crappy driver yourself and then maybe you both could use some lessons.
    (Of course most stupid drivers don’t realize that they’re stupid drivers)

    Nearly 40 now and have only had two accidents and neither were my fault.

  11. Seems like the studies totally ignore simple statistics – reduce the percentage of a driver’s lifetime that they spend on the road and their chances of being in an accident are reduced as well. Not too complicated.

  12. I can’t image only being allowed to drive with my parents in the car.  That was the most stressful time when I was 16.  I was sweating, checking everything constantly.  Going the speed limit, did I signal, stay in lane, stop ahead…need to slow down gently, ect.. ect…

    Even now at 31 I drive better when I’m alone then when I have someone in the car.  I don’t have to listen to them make gasping/screaming/OMG we are going to DIE noises when something happens…

    I think a lot of small things could be done to help teen drivers.

    1. Install a cell jammer into their car, automatic $500 fine if a cop pulls them over and it’s not active.  (Which it should be by default if the car is in gear.)

    That’ll at least solve the problem of paying more attention to your phone than controlling the 3000+lb vehicle you are piloting around the roads.

  13. Why the negative spin?

    The post says”the results were pretty much a wash”
    and the linked article says
    “While the number of fatal crashes among 16- and 17-year-old drivers has fallen, deadly accidents among 18-to-19-year-olds have risen by an almost equal amount.”

    Nearly equal amount?

    If fatal accidents are down 26% among younger kids and up 12% among older kids, that sounds like a net win.  Apparently some people are going to die as new drivers, and these regulations seem to be reducing the numbers.  Apparently something like half of the effect is due to age and half due to inexperience.  Okay, it’s not quite that simple, but certainly the numbers don’t indicate that we should focus on inexperience and ignore age.  Why can’t we address both?  The Times article gives the impression that these laws have needlessly killed 18 year olds.

  14. What I want to know is: how many drivers responsible for fatal crashes are involved in ‘minor’ accidents first? Knowing the way some of the people I know drive, I would guess that number would be high. It’s carelessness and it seems to reflect the way their parents drive. Fender benders shouldn’t be viewed as a common part of learning to drive. They should be seen as a possible symptom of inattentive driving.

  15. Without knowing the relative size of the 16-17 and 18-? populations, it’s hard to interpret this article. Assuming that the size of the populations measured are roughly equivalent, however, isn’t a 26% decrease in 16 year olds followed by a 12% increase in 18 year olds who have accidents a net reduction in accidents?

  16. I’m with the others: this data says very little without controlling for *when they started driving*.

    Odds are that people just delayed the expense of a test until they could drive properly. In this case, the fix (giving a beginner’s license for a certain amount of time, no matter your age) seems both as simple as others have said, and as necessary as the numbers suggest.

  17. Why do we only have driver’s ed classes until we get our licenses? What if refresher classes in High School were required where you got experience driving, refreshing knowledge of driving laws, and “higher end” training, driving with distractions, breaking in emergency situations, eating a burger and driving with your knees and maybe how to do a couple of stunt turns.

  18. http://www.streetsurvival.org/

    If someone you know is a teen driver, send ’em.

    (On a tangential note, until I started autocrossing at the age of about 35, I *thought* I knew how a car handles…. I can credit at least one successful accident avoidance on the road to lessons learned on the course.)

    1. Well, I started autocrossing around 20 or 21, and while it helped, getting out on a race track for an advanced driver training course 20 years later was still an eye opener.  Track time and autocross teach different, complementary skills, and both can help make you a safer driver on the street.

  19. My driver education teacher in the 1980’s was a bit of a nut-job. I recall him using his passenger side safety brake to stop our car on some rural train tracks, at which point he turned on a mini tape recorder playing the sound of a locomotive whistle. Of course, being the 80’s, you heard the whir of the cassette and weren’t at all convinced by the crude recording. This is the same guy that had us practice anti-skid training in the iced over school parking lot. Which was fine until he added two elements to the scenario: we were to pretend that the turn signals and main brakes weren’t working. Should this situation ever occur, however, I am trained to roll the window down and signal with my hands, all the while reaching down to work the parking brake to control the skid. The confidence this teacher instilled in me was enormous, and two months later I crashed the IGA food delivery van at work.

  20. 22 year old, here. I’m going to chime in.

    When I got my driver’s license, it involved a written test, a driving test, and then 3 months where I could only drive with others in the car if one of those others was over 21. There were no driving logs, the restrictions weren’t in place for 3 years, and the age for a license was 16 years exactly.

    Today, though, kids who should be going through what I went through instead have to wait until they’re 16.5 years, they have to have 80 hours or something of driving logs, and I have no idea when they can drive with others in the car. Oh, and if they fuck up? More driving classes.

    Here’s a quote that all of you cranky old people might be familiar with:  won’t somebody please think of the children??

    It is not sane, and it is not healthy to be sheltering everyone from everything indefinitely. Kids nowadays can’t even climb a god damn tree without the cops being called and having a psychological evaluation. Give kids their freedom back, stop relentlessly doing things “for their safety.” You know what? Some kids will die. And if you don’t let kids drive, then some adults will die. You can’t prevent accidents. What if it’s your kid that dies? Well that’s damn tragic, but you shouldn’t force everyone to live in bubble wrap because you had something traumatic happen to you. You can improve car safety, though, and the auto companies are working on it.

  21. Even if we take the study at face value, isn’t 1086 less than 1,348?  That’s 262 teenagers that didn’t die.  Seems like a win, if the statistics hold up to review.

    Started driving at 16, didn’t die, but probably had a few close calls/risks that I wouldn’t take now.

  22. very, very, very tentatively:

    a) There’s still a small net win (i.e. the early fatality reduction exceeds the later fatality increase)

    b) Might we want to look at keeping the restrictions for longer (some till 21, esp in urban areas?) and check again? We know the male brain is still developing in terms of ability to evaluate consequences at this age, so we might see a significant improvement in the net win?

  23. Maybe they should restrict the horsepower and weight of the vehicle based on the miles driven and other driving history (like speeding)?

  24. Here’s another vote for a Street Survival session or three. The kids learn a remarkable amount about how they and their car can avoid trouble in the first place, and actively manage it if it comes. And they learn this stuff in a safe environment where it’s next to impossible to hurt themselves, and pretty hard to break the car.
    There’s lots of places to get this kind of training, true, but Street Survival is stunningly cheap, widely available, and the kids get to use the car they normally drive. I have no relationship with this organization apart from having volunteered as a track worker many times, and instructed once. But I have been consistently surprised (and delighted) at how much and how quickly the kids learn. For most of them, it’s white knuckle terror in the morning, relaxed, confident, and much better in the afternoon.

    Automated safety systems are wonderful. Right up to the point where you exceed their abilities, physics takes over, and you just became a passenger. On your way to becoming a news item. Then what do you do? Street Survival and similar programs address that question.

    A year ago, we had waiting lists-I just learned that our last warm-weather session for the year here is Minnesota was cancelled due to low enrollment. No idea why.

  25. If the overall deaths remained the same, one could argue that earlier is better, from a society-making-an-investment point of view. But it sounds like restrictions succeeded in lowering the number of deaths quite a bit.

  26. For comparison: in Victoria, Australia:
    You can get your learner’s permit when you’re 16.
    Edit: Whilst on your learner’s permit you can only drive with a fully licensed driver in the front seat.
    You must log 120 hours (soon to be changed to 160 hours) of driving practice and be at least 18 years old before you can do your test.
    After you pass your test you spend 1 year with peer passenger restriction (only 1 passenger can be between the ages of 16 and 21, unless there is a fully licensed driver in the car) and must display a red “P” plate on your car, designating the level of your probationary licence.
    After this you must display a green “P” plate on your car for 3 years, without passenger restriction.
    If you have a probationary licence your blood alcohol content must be 0.00%.
    Then you get your full drivers licence.

Comments are closed.