What the Secretary of Transportation Has to Say About My Car Seat Research

On his blog, Secretary of Transportation Ray LaHood dismisses my research (see here and here) on car seats.

My favorite quote from the secretary:

“Now, if you want to slice up the data to be provocative, have at it. As a grandfather and as secretary of an agency whose number one mission is safety, I don’t have that luxury.”

Reading the Secretary’s blog post, it strikes me just how differently he is reacting to a challenge than Arne Duncan (now the Secretary of Education) did when I first told him about my work on teacher cheating when Duncan was in charge of the Chicago Public Schools. I expected Duncan to do what LaHood did: dismiss the findings, circle the wagons, etc. But Duncan surprised me. He said that all he cared about was making sure the children were learning as much as possible, and teacher cheating was getting in the way of that. He invited me into a dialogue, and we ultimately made a difference.

Here’s what LaHood might have written on his blog if the ultimate goal is really child safety:

“For a long time we’ve been relying on car seats to keep our children safe. The existing academic literature up until recently confirmed the view that car seats are very successful in that goal. But in a series of papers in peer-reviewed journals, Steven Levitt and his co-authors have challenged that view using three different data sets collected by the Department of Transportation, as well as other data sets. I’m no data expert, and I have an agency to run, so I don’t have the luxury of analyzing the data myself. But I am a grandfather and my agency’s number one mission is safety, so I’ve asked the researchers in my agency to do the following:

  1. Take a close look at the data sets we collect here in my agency, which are the basis for Levitt’s work. Is it really the case that in these data there is little or no evidence that car seats outperform adult seat belts in protecting children ages 2 and up? Our benchmark for measuring the effectiveness of car seats has always been versus children who are unrestrained. Maybe we need to rethink this going forward?
  2. Demand that the physicians at the Children’s Hospital of Philadelphia, who have repeatedly found that car seats work, make their data publicly available. It is my understanding that these physicians have refused to share their data with Levitt, but in the interest of getting to the truth, other researchers should have the chance to review what they have done.
  3. Carry out a series of tests using crash-test dummies to determine whether adult seat belts do indeed pass all government crash-test requirements. In SuperFreakonomics, Levitt and Dubner report on their findings with a very small sample of tests; we need much more evidence on the data.
  4. Try to understand why, even after 30 years, the great majority of car seats are still not properly installed. After all this time, can we really blame it on the parents, or should the blame be put elsewhere?
  5. After exploring all these issues, let’s figure out the truth, and let’s use it to guide public policy.

And if Secretary LaHood has any interest in pursuing any of these avenues, I stand at the ready to offer whatever help that I can.


kudos on taking a stand, Levitt!


Mary Roach's excellent book on death "Stiff" points out that the crash test dummy designers need actual human corpses to test first in order to design the dummies accurately. Since parents are understandably reluctant to donate their children's bodies to science, children are in fact the least reliably tested segment of the population in car safety trials, because the child dummies don't have the depth of research behind them that adult versions do.


While I am no Obama hater, this administration has done more to dismiss people who disagree with them than any other (including the Bush administration). I'm wondering if this is a trend that started with Clinton, continued with Bush and is continuing still with Obama. The next President will not need a cabinet, because everything he/she will do will be right.

I'm not sure that's going to happen. LaHood has the luxury of the cop-out, "I don't have that luxury". A lot of people don't have that cop-out.

My guess is, and I'm just shooting from the hip here, they aren't interested in further exploration of the research. Just like there are a number of people who aren't interested in continuing to question "why?". One thing we can learn (as my daughter asks "why" she has to clean her room) from our children is to always question.

I hope that the Secretary reconsiders. But it's unlikley.



Isn't insecurity terrible? It's especially problematic when it's someone in a power position.


6. Review the finacnial performance of child car seat producers and investigate their relationship with supposedly independent testers.

Wouldn't that be nice.


Maybe we should print this blog and send it to him. Because this does seem reasonable and maybe he isn't an unreasonable guy, maybe he is just misinformed.


What does Levitt believe LaHood's goal actually is? Is he saying LaHood doesn't care about children's safety? Levitt's phrase, "Here's what LaHood might have written on his blog if the ultimate goal is really child safety" implies that.

I believe you'd have a hard time proving that he has a goal other than child safety in mind here. Why aren't we asking why he's reacting that way instead?


Why respectfully disagree and commit to finding the truth when you can marginalize your putative adversary? Worked for years for Ashcroft et al.


Sec LaHood, ball's in your court, we await your reply


The "way of things" anymore, at all levels of public policy discussion, seems to be that we dismiss out of hand and with prejudice anything that at first blush doesn't mimic our own way of thinking. I haven't read these particular papers of yours yet, and their very titles make me uncomfortable. But you are a rigorous researcher, and I wouldn't even imagine dismissing the results until such time as I've read them and your suggested actions could be undertaken. Suspending judgment until evidence is in - what a concept!


How far off ideal installation qualifies as not "properly installed"? Does it mean the 250 pound state trooper at the car seat safety check station can wiggle it an extra 1/8 inch, or it's a couple degrees off level? I find it hard to believe that most people have the straps in the wrong place, etc.

People going on about how 90 plus percent of car seats are in wrong just makes the worrywarts worry more, while the people who drive around in a pickup truck with a toddler on their lap aren't going to change.


I agree with you that this could have been a more open response...in an ideal world. But we live in a world filled with morons who believe everything they hear without taking care in analyzing things first...I understand he might want to go ahead and stop any parents from wanting to jump the gun on car seats....so something you might not understand Steven, is that a lot of people will read what you write and not understand the subtlety in it, are you advising parents to go ahead and eliminate car seats for kids over 2? i don't think so, correct me if I am wrong but I think you are basically saying that there is enough evidence in the data you analyzed to warrant a much closer look at the effectiveness and need of certain carseats amd booster seats by researching it further... as a public official I would assume you don't want to create confusion especially when it is in regards to child safety...



Unfortunately, while most people have a stated mission, such as car safety, their real mission is preserving their own egos. It's not their fault; they evolved that way, and we humans have to put a lot of effort and conscious thought into operating in any other way.

But the result is that in the pursuit of ego preservation, people will reject or embrace damn near anything, no matter how little sense it makes, to prevent the cognitive dissonance that comes with admitting you are wrong.

Ray LaHood is an old man, and also works for the government (the most ego-preserving-at-all-costs entity around). So the chances of him putting forth a response like the one outlined above are virtually zero.



More is at stake than impact damage when it comes to child car seats, I believe.

I remember, on sweltering summer days in Florida, trying to find, then link together, the multiple seatbelts on my son's car seat. That was bad enough...but then I though of having to try to untangle him from that if I was in water...upside down...at night. Or trying to retrieve him from burning wreckage.

It was clear to me that the simplicity of regular seatbelts was superior to the complexity of the car seats.

Next, the notion of facing a child AWAY from you is just crazy. What if the baby is choking? What if the baby is in some sort of distress--do you have to stop and pull over to figure it out? It seems to me that front-facing is the only way to go, and that whiplash can be dealt with in other ways.

When my son was very young, on one of those long, lonely Florida backroads, we--GASP!!!--faced his seat forward! HE LOVED IT! He could see and communicate with us in his way. I could see his eyes in the rearview mirror and see his delight as the wind whipped through the car.

I think one way to fix this would be for factory seatbelts to come with an extra strap or so that would serve as a crotch strap for our children. Also a way to make sure the chest belt didn't cut into the face or neck of the child. Same simplicity and quick-release, but made for children.


reality check

Here in NY the law is currently that a kid needs to be in a car seat till they are either 8 years old or 80 lbs. As a parent I've done a lot of shuffling of seats and would love to see cars with seat belts that accommodate 4-8 year olds better without a booster seat.

Does law enforcement want to see this...no. No writing tickets to drivers of kids who are buckled in but not in seats.

Do the manufacturers of car seats want to see this? Of course not.

How about car rental agencies? I've paid more to rent a car seat than it cost to buy it. Last time I flew I turned my daughters car seat into a backpack so I could haul it around the airport easier.

If there is minimal advantage to car seats for older kids why have these laws been passed? It's easy to require everyone to buy something in the name of safety. It's probably easy to throw negative comments at someone (like a state legislator) who thinks before they vote against requiring folks to buy something for the (supposed) sake of safety.

As a parent (and an aunt and a hopefully future grandmother) I'd like to see cars with back seat restraints built in that can safely hold kids 4 and up. (younger than that you can't count on them staying in the belts)



Mom is going to blow a cow when she hears this.


I think Ed (commenter 7) is on the right track here. Its not by definition LaHood's charge to react to data in the same way Levitt et. al. do in their studies; rather, to reconcile statistical findings -- both government and academic -- with myriad interests: bureaucracy, media, public opinion, service to the President, etc.

Put differently, the public confusion that would likely come as a result of Ray LaHood going on CNN and telling everyone car seats don't work would probably cause more problems than sticking to conventional wisdom until -- assuming the research is sound, which I have no reason to dispute -- the administration can formulate a way to socialize these findings in a way that accounts for unintended consequences.

I think this more or less parallels Levitt and Dubner's position in the climate change debate: its not that they're wrong, its just that whiplashed disruptions of conventional wisdom are sometimes as damaging as conventional wisdom being wrong in the first place. The scientific/environmental/advocacy folks are hyper-aware of that, which probably accounts for that community's strong negative reaction to that part of Superfreakonomics.



you know what i have to say about it?

'thank you'.


I would be interested in your interpretation of the following paper mentioned in the LaHood's blog post you linked:


In particular, they say:

In a more recent series of analyses, Levitt and Dubner and Levitt used FARS data from 1975 to 2003 and, by various methods, directly compared the mortality with child restraint systems vs seat belts in children aged 2 to 6 years and could not demonstrate a difference in effectiveness. Levitt and Porter restricted the FARS data set to 2-vehicle crashes in which someone in the other vehicle (ie, the vehicle without the index child occupant) died, under the assumption that the distribution of restraint use among children in potentially fatal crashes is independent of whether someone in the other vehicle dies, after adjusting for various crash-related characteristics.

The precision of these previous effectiveness estimates and their relevance to today's fleet were limited by their reliance on FARS (the limits of which were presented in the introduction), by the inclusion of older vehicles and restraints, and by failing to control for the severity (potential unsurvivability) of the crashes. Our analysis based on a modern cohort of children involved in crashes in which the car was rendered undrivable addressed many of the limitations of these previous analyses and found effectiveness estimates for child restraint systems to be greater than that for seat belts alone, with estimates greater than those of both Levitt and Hertz.


Brian DeMarzo

I'm sure that they can come to the conclusion that "car seats are BETTER than seat belts." Even the slightest bit 'better' is still 'better'.

Which reminds me of the phrase, "Better is the enemy of Good Enough."