Does Driver's Ed. Lead to More Car Crashes?

The conclusion couldn’t be any starker:

Indiana lawmakers say the state’s driver education program isn’t working, citing a fractured system administered by three separate agencies and statistics that put the program’s usefulness in doubt.

Yes, it’s a bit of a political issue …

Public affairs director Sarah Meyer of the Indiana Bureau of Motor Vehicles told a group of state lawmakers last week that a study of current drivers under 18 showed those who took driver’s education had nearly four times the crashes that those who didn’t take the classes had.

But hey, there are statistics to back it up:

Nearly 5 percent of the 51,000 teens who took driver’s education had one or more reported accidents, compared with 1 percent of the 71,932 drivers without formal driver training.

Is driver’s ed. a menace to society? At least one politician asked the obvious question:

“Why do we even offer driver’s education?” asked Rep. Phil Hinkle, R-Indianapolis, after hearing the statistics.

Indeed, lots of government-run programs don’t produce the intended result. And the safety value of driver’s education has been challenged before. But is it really possible that sitting in a driver’s-ed. class makes you four times more dangerous?

Maybe the courses give young drivers false confidence that makes them more dangerous? Or maybe the instruction is really bad?

Or maybe there’s a selection bias at work here, whereby the drivers who take driver’s ed. (whether by choice or not) have different characteristics than the drivers who don’t.

Or maybe, more believably (to me at least), the statistics above obscure one salient fact: drivers who take a driver’s ed. course are allowed to drive at a younger age than drivers who haven’t taken a course.

There might be a lot of reasons to critique a driver’s-ed. program, but one of them should not be that you are shocked – shocked! – that turning more young drivers loose on the roads may result in more accidents.

Larry Karp

Or..... Maybe students who have taken Drivers Education understand the importance of reporting an accident and the consequences of not reporting. In addition, they are probably more prone to carry proper insurance (parents have more $$ to protect). Maybe some racial / income bias as well??? Thoughts??


Let's all agree going forward to only do studies that truly compare apples to apples. With the only change being what we want to study. If you can't do that then why bother with the study.

Remember half of all statistics are made up.

Nate C.

There could be a location bias. For example, young drivers who live in rural areas may be less likely than young drivers in urban areas to be involved in an accident. This may occur regardless of whether the driver has taken driver's education. Can we add a variable titled "population of area in which accident occurred" to the equation? There are a few other proxies - I suppose the overall accident rate, including adults, might be useful as well.

Lawrence Miller

There is a section in "Nurture Shock" about the failure of Drivers' Ed. In the studies the authors cite, it seems to make very little difference to safety, and they argue that it is because Drivers' Ed teaches driving technique effectively, but that accidents are caused, by and large, by poor decisions; driving while tired or distracted, showing off for friends, etc.


Perhaps NOT taking the class correlates with having an engaged parent who is willing to do the training himself/herself, and the presence of a parent (and the absence of classmates) in the car is a better way to learn how to drive safely.

Calvin Graham

And while on the subject of comparing like with like, what's the comparrison in distance travelled per year? This could be like the other driving classic that 'women make safer drivers' which overlooks the fact that, on the whole, male drivers tend to drive a vast number more miles every year


Dubner writes "Or maybe there's a selection bias at work here..." Bingo - we have a winner. As one of my colleagues likes to say, endogeneity is where you find it...


But following the link provided shows that in Indiana there is only 3 months difference between the age at which one can obtain a license with successful completion of a driver education program versus without.

I find it easy to believe that 16-year-olds are, on average, less skilled or careful drivers than 21-year-olds or 40-year-olds, but can there really be *that much* difference between 16 years 6 months and 16 years 9 months?

Think and thrive. Don't drink and drive.

I think your penultimate paragraph has it right.

My son has had his learning permit (driver must accompany him) for two years and will be a more mature person when he chooses to get tested, likely without the need for driver's ed, although that will get us lower insurance rates.


Agree, in addition to the possible explanations you posit, you first need to question the validity of the statistics themselves. Hopefully they are valid, but ...

Also, location is really important. Rural drivers have much lower accident frequency (because there's less traffic). Do urban kids take drivers' ed. more often?

There are other factors effecting frequency as well. You need to control for these factors before you can draw conclusions.


Driver's Ed is a dumb class that teaches nothing of what truly needs to be known about piloting a two ton mass down the roadway. Better our children go to a Performance Driving School at a road course. At least there they'll learn something about how their vehicle works, how the law of physics apply, and how human instinct works against them ala target fixation and such. Americans have been brainwashed to believe speed kills so we keep focusing on making cars and roads safer while simultaneously ignoring driver skill. A car as docile (read: wimpy) as a Toyota Prius suddenly turns into a death trap mainly because the driver has not the skill to deal with the slightest perceived malfunction. Our society puts way too much emphasis on technological externalities while blithely ignoring, or otherwise having no faith in, improving the human factor.


I have to agree with GLK's assertion that the class teaches little of value. My daughter took the class recently. She was bored. When she told me what they did in class, there was little of real value. The focus seemed to be, "As long as your rear end is in the seat the required number of hours, you're good to go."

L. F. File

Are the insurance companies really so dim as to give premium deductions for something that increases their exposure? What statistics are they basing their determinations on?

And when they say "...51,000 teens who took driver's education..." are they only including those that took classes before they had a license? Many of the "teens" may have taken classes after being required to due to a traffic infraction or accident.

E Olson

Post #5 shares my suspicion. If a teenager's only instruction is driver's ed, and there is no involved parent the result could be a driver with minimum skills. How much road time does a student get when there are many other students to share that time behind the wheel?

Makes me wonder about another question: why would insurance companies lower your rates for attenending drivers ed?


I think Indiana licenses drivers at a somewhat earlier age if they have taken a driver's education, at least at one time I'm reasonably certain that this was the case. While I am certainly willing to accept the idea that driver's ed is worse than useless, I do wonder if this might be a complicating factor.


There is a great difference between driver education courses and driver training. It is well-understood that most people learn better by doing than by observing alone. Let's see a proper study that counts accidents of various types among the 2 x 2 combination of course attendance and training classes.


An addendum: indeed, those taking a driver's ed course may get their license a minimum of 3 months earlier than those who don't. (I just checked.) And those who take driver's ed most likely do it to get the license earlier. So there are more opportunities available for the driver's ed graduate to have an accident by age 18 than for the non-driver's ed graduate. Plus there may be systematic differences between those who take driver's ed in terms of attitude, risk aversion, etc.

The Indiana Bureau of Motor Vehicles, the folks who produced the study, have no known research capacity.

Bottom line: there is an excellent chance that the study is pure garbage. If insurance companies stop giving a discount for graduates of recognized driver's ed programs, then I would be more inclined to believe the results.


I am also skeptical of the "reported accident" statistic. Could rural teens (or their parents) be more likely to settle-up with neighbors privately after a crash? In a larger city where you don't know everyone including your neighbors, getting the police involved could be far more common.


I know when I was 16 I got into 4 really minor accidents (I think the cars were moving under 10mph in all four), only one of them was reported (the rest were privately settled since the damages were all under $500). Since then I have not been in one.

I did take drivers ed, but the reason for the accidents was just general irresponsibility and immaturity, nothing to do with drivers ed (which was admittedly worthless).

Thomas Paine

My daughter took the after school drivers class in South Florida, and afterwards I found out our auto insurance company did not recognize the class for a discount. The insurance company required watching a DVD they supplied and signing a pledge. I also thought the school driving class was a waste after watching them drive a course in a parking lot at 3 mph and never going out on the open road.