0:00:00.000,0:00:05.000
Once upon a time, there was a dread disease that afflicted children.

0:00:05.000,0:00:09.000
And in fact, among all the diseases that existed in this land,

0:00:09.000,0:00:12.000
it was the worst. It killed the most children.

0:00:12.000,0:00:15.000
And along came a brilliant inventor, a scientist,

0:00:15.000,0:00:18.000
who came up with a partial cure for that disease.

0:00:18.000,0:00:22.000
And it wasn't perfect. Many children still died,

0:00:22.000,0:00:25.000
but it was certainly better than what they had before.

0:00:25.000,0:00:31.000
And one of the good things about this cure was that it was free,

0:00:31.000,0:00:33.000
virtually free, and was very easy to use.

0:00:33.000,0:00:36.000
But the worst thing about it was that you couldn't use it

0:00:36.000,0:00:40.000
on the youngest children, on infants, and on one-year-olds.

0:00:40.000,0:00:42.000
And so, as a consequence, a few years later,

0:00:42.000,0:00:44.000
another scientist -- perhaps maybe this scientist

0:00:44.000,0:00:48.000
not quite as brilliant as the one who had preceded him,

0:00:48.000,0:00:50.000
but building on the invention of the first one --

0:00:50.000,0:00:53.000
came up with a second cure.

0:00:53.000,0:00:56.000
And the beauty of the second cure for this disease

0:00:56.000,0:01:00.000
was that it could be used on infants and one-year-olds.

0:01:00.000,0:01:05.000
And the problem with this cure was it was very expensive,

0:01:05.000,0:01:06.000
and it was very complicated to use.

0:01:06.000,0:01:10.000
And although parents tried as hard as they could to use it properly,

0:01:10.000,0:01:14.000
almost all of them ended up using it wrong in the end.

0:01:14.000,0:01:17.000
But what they did, of course, since it was so complicated and expensive,

0:01:17.000,0:01:20.000
they only used it on the zero-year-olds and the one-year-olds.

0:01:20.000,0:01:23.000
And they kept on using the existing cure that they had

0:01:23.000,0:01:24.000
on the two-year-olds and up.

0:01:24.000,0:01:26.000
And this went on for quite some time. People were happy.

0:01:26.000,0:01:29.000
They had their two cures. Until a particular mother,

0:01:29.000,0:01:34.000
whose child had just turned two, died of this disease.

0:01:34.000,0:01:38.000
And she thought to herself, "My child just turned two,

0:01:38.000,0:01:42.000
and until the child turned two, I had always used

0:01:42.000,0:01:47.000
this complicated, expensive cure, you know, this treatment.

0:01:47.000,0:01:48.000
And then the child turned two, and I started using

0:01:48.000,0:01:51.000
the cheap and easy treatment, and I wonder" --

0:01:51.000,0:01:53.000
and she wondered, like all parents who lose children wonder --

0:01:53.000,0:01:55.000
"if there isn't something that I could have done,

0:01:55.000,0:01:59.000
like keep on using that complicated, expensive cure."

0:01:59.000,0:02:02.000
And she told all the other people, and she said,

0:02:02.000,0:02:04.000
"How could it possibly be that something

0:02:04.000,0:02:07.000
that's cheap and simple works as well as something

0:02:07.000,0:02:09.000
that's complicated and expensive?"

0:02:09.000,0:02:11.000
And the people thought, "You know, you're right.

0:02:11.000,0:02:13.000
It probably is the wrong thing to do to switch

0:02:13.000,0:02:16.000
and use the cheap and simple solution."

0:02:16.000,0:02:19.000
And the government, they heard her story and the other people,

0:02:19.000,0:02:22.000
and they said, "Yeah, you're right, we should make a law.

0:02:22.000,0:02:24.000
We should outlaw this cheap and simple treatment

0:02:24.000,0:02:27.000
and not let anybody use this on their children."

0:02:27.000,0:02:29.000
And the people were happy. They were satisfied.

0:02:29.000,0:02:32.000
For many years this went along, and everything was fine.

0:02:32.000,0:02:37.000
But then along came a lowly economist, who had children himself,

0:02:37.000,0:02:44.000
and he used the expensive and complicated treatment.

0:02:44.000,0:02:46.000
But he knew about the cheap and simple one.

0:02:46.000,0:02:48.000
And he thought about it, and the expensive one

0:02:48.000,0:02:51.000
didn't seem that great to him. So he thought,

0:02:51.000,0:02:54.000
"I don't know anything about science, but I do know something about data,

0:02:54.000,0:02:56.000
so maybe I should go and look at the data

0:02:56.000,0:03:00.000
and see whether this expensive and complicated treatment

0:03:00.000,0:03:03.000
actually works any better than the cheap and simple one."

0:03:03.000,0:03:05.000
And lo and behold, when he went through the data,

0:03:05.000,0:03:08.000
he found that it didn't look like the expensive, complicated

0:03:08.000,0:03:11.000
solution was any better than the cheap one,

0:03:11.000,0:03:13.000
at least for the children who were two and older --

0:03:13.000,0:03:16.000
the cheap one still didn't work on the kids who were younger.

0:03:16.000,0:03:20.000
And so, he went forth to the people and he said,

0:03:20.000,0:03:22.000
"I've made this wonderful finding:

0:03:22.000,0:03:25.000
it looks as if we could just use the cheap and simple solution,

0:03:25.000,0:03:28.000
and by doing so we could save ourselves 300 million dollars a year,

0:03:28.000,0:03:30.000
and we could spend that on our children in other ways."

0:03:30.000,0:03:34.000
And the parents were very unhappy, and they said,

0:03:34.000,0:03:36.000
"This is a terrible thing, because how can the cheap and easy thing

0:03:36.000,0:03:40.000
be as good as the hard thing?" And the government was very upset.

0:03:40.000,0:03:43.000
And in particular, the people who made this expensive solution

0:03:43.000,0:03:45.000
were very upset because they thought,

0:03:45.000,0:03:48.000
"How can we hope to compete with something that's essentially free?

0:03:48.000,0:03:50.000
We would lose all of our market."

0:03:50.000,0:03:53.000
And people were very angry, and they called him horrible names.

0:03:53.000,0:03:56.000
And he decided that maybe he should leave the country

0:03:56.000,0:04:00.000
for a few days, and seek out some more intelligent,

0:04:00.000,0:04:03.000
open-minded people in a place called Oxford,

0:04:03.000,0:04:06.000
and come and try and tell the story at that place.

0:04:06.000,0:04:10.000
And so, anyway, here I am. It's not a fairy tale.

0:04:10.000,0:04:12.000
It's a true story about the United States today,

0:04:12.000,0:04:15.000
and the disease I'm referring to is actually

0:04:15.000,0:04:18.000
motor vehicle accidents for children.

0:04:18.000,0:04:24.000
And the free cure is adult seatbelts, and the expensive cure --

0:04:24.000,0:04:28.000
the 300-million-dollar-a-year cure -- is child car seats.

0:04:28.000,0:04:30.000
And what I'd like to talk to you about today

0:04:30.000,0:04:33.000
is some of the evidence why I believe this to be true:

0:04:33.000,0:04:35.000
that for children two years old and up,

0:04:35.000,0:04:40.000
there really is no real benefit -- proven benefit -- of car seats,

0:04:40.000,0:04:45.000
in spite of the incredible energy

0:04:45.000,0:04:48.000
that has been devoted toward expanding the laws

0:04:48.000,0:04:50.000
and making it socially unacceptable

0:04:50.000,0:04:55.000
to put your children into seatbelts. And then talk about why --

0:04:55.000,0:04:56.000
what is it that makes that true?

0:04:56.000,0:04:59.000
And then, finally talk a little bit about a third way,

0:04:59.000,0:05:02.000
about another technology, which is probably better than anything we have,

0:05:02.000,0:05:05.000
but which -- there hasn't been any enthusiasm for adoption

0:05:05.000,0:05:07.000
precisely because people are so enamored

0:05:07.000,0:05:10.000
with the current car seat solution. OK.

0:05:10.000,0:05:13.000
So, many times when you try to do research on data,

0:05:13.000,0:05:17.000
it records complicated stories -- it's hard to find in the data.

0:05:17.000,0:05:20.000
It doesn't turn out to be the case when you look at seatbelts versus car seats.

0:05:20.000,0:05:22.000
So the United States keeps a data set

0:05:22.000,0:05:25.000
of every fatal accident that's happened since 1975.

0:05:25.000,0:05:28.000
So in every car crash in which at least one person dies,

0:05:28.000,0:05:30.000
they have information on all of the people.

0:05:30.000,0:05:33.000
So if you look at that data -- it's right up on the National Highway

0:05:33.000,0:05:35.000
Transportation Safety Administration's website --

0:05:35.000,0:05:37.000
you can just look at the raw data,

0:05:37.000,0:05:41.000
and begin to get a sense of the limited amount of evidence

0:05:41.000,0:05:44.000
that's in favor of car seats for children aged two and up.

0:05:44.000,0:05:48.000
So, here is the data. Here I have, among two- to six-year-olds --

0:05:48.000,0:05:50.000
anyone above six, basically no one uses car seats,

0:05:50.000,0:05:56.000
so you can't compare -- 29.3 percent of the children who are unrestrained

0:05:56.000,0:06:00.000
in a crash in which at least one person dies, themselves die.

0:06:00.000,0:06:05.000
If you put a child in a car seat, 18.2 percent of the children die.

0:06:05.000,0:06:07.000
If they're wearing a lap-and-shoulder belt, in this raw data,

0:06:07.000,0:06:12.000
19.4 percent die. And interestingly, wearing a lap-only seatbelt,

0:06:12.000,0:06:14.000
16.7 percent die. And actually, the theory tells you

0:06:14.000,0:06:17.000
that the lap-only seatbelt's got to be worse

0:06:17.000,0:06:18.000
than the lap-and-shoulder belt. And that just reminds you

0:06:18.000,0:06:20.000
that when you deal with raw data, there are hundreds

0:06:20.000,0:06:23.000
of confounding variables that may be getting in the way.

0:06:23.000,0:06:28.000
So what we do in the study is -- and this is just presenting

0:06:28.000,0:06:31.000
the same information, but turned into a figure to make it easier.

0:06:31.000,0:06:34.000
So the yellow bar represents car seats,

0:06:34.000,0:06:38.000
the orange bar lap-and-shoulder, and the red bar lap-only seatbelts.

0:06:38.000,0:06:40.000
And this is all relative to unrestrained --

0:06:40.000,0:06:41.000
the bigger the bar, the better. Okay.

0:06:41.000,0:06:43.000
So, this is the data I just showed, OK?

0:06:43.000,0:06:46.000
So the highest bar is what you're striving to beat.

0:06:46.000,0:06:50.000
So you can control for the basic things, like how hard the crash was,

0:06:50.000,0:06:54.000
what seat the child was sitting in, etc., the age of the child.

0:06:54.000,0:06:56.000
And that's that middle set of bars.

0:06:56.000,0:06:59.000
And so, you can see that the lap-only seatbelts

0:06:59.000,0:07:01.000
start to look worse once you do that.

0:07:01.000,0:07:03.000
And then finally, the last set of bars,

0:07:03.000,0:07:06.000
which are really controlling for everything

0:07:06.000,0:07:08.000
you could possibly imagine about the crash,

0:07:08.000,0:07:11.000
50, 75, 100 different characteristics of the crash.

0:07:11.000,0:07:14.000
And what you find is that the car seats and the lap-and-shoulder belts,

0:07:14.000,0:07:18.000
when it comes to saving lives, fatalities look exactly identical.

0:07:18.000,0:07:22.000
And the standard error bands are relatively small around these estimates as well.

0:07:22.000,0:07:25.000
And it's not just overall. It's very robust

0:07:25.000,0:07:27.000
to anything you want to look at.

0:07:27.000,0:07:30.000
One thing that's interesting: if you look at frontal-impact crashes --

0:07:30.000,0:07:33.000
when the car crashes, the front hits into something --

0:07:33.000,0:07:37.000
indeed, what you see is that the car seats look a little bit better.

0:07:37.000,0:07:39.000
And I think this isn't just chance.

0:07:39.000,0:07:40.000
In order to have the car seat approved,

0:07:40.000,0:07:43.000
you need to pass certain federal standards,

0:07:43.000,0:07:48.000
all of which involve slamming your car into a direct frontal crash.

0:07:48.000,0:07:50.000
But when you look at other types of crashes, like rear-impact crashes,

0:07:50.000,0:07:53.000
indeed, the car seats don't perform as well.

0:07:53.000,0:07:55.000
And I think that's because they've been optimized to pass,

0:07:55.000,0:07:57.000
as we always expect people to do,

0:07:57.000,0:07:59.000
to optimize relative to bright-line rules

0:07:59.000,0:08:03.000
about how affected the car will be.

0:08:03.000,0:08:04.000
And the other thing you might argue is,

0:08:04.000,0:08:06.000
"Well, car seats have got a lot better over time.

0:08:06.000,0:08:09.000
And so if we look at recent crashes --

0:08:09.000,0:08:11.000
the whole data set is almost 30 years' worth of data --

0:08:11.000,0:08:13.000
you won't see it in the recent crashes. The new car seats are far, far better."

0:08:13.000,0:08:17.000
But indeed, in recent crashes the lap-and-shoulder seatbelts,

0:08:17.000,0:08:20.000
actually, are doing even better than the car seats.

0:08:20.000,0:08:23.000
They say, "Well, that's impossible, that can't be."

0:08:23.000,0:08:25.000
And the line of argument, if you ask parents, is,

0:08:25.000,0:08:28.000
"But car seats are so expensive and complicated,

0:08:28.000,0:08:31.000
and they have this big tangle of latches,

0:08:31.000,0:08:34.000
how could they possibly not work better than seatbelts

0:08:34.000,0:08:36.000
because they are so expensive and complicated?"

0:08:36.000,0:08:39.000
It's kind of an interesting logic,

0:08:39.000,0:08:42.000
I think, that people use. And the other logic, they say,

0:08:42.000,0:08:44.000
"Well, the government wouldn't have told us [to] use them

0:08:44.000,0:08:46.000
if they weren't much better."

0:08:46.000,0:08:48.000
But what's interesting is the government telling us to use them

0:08:48.000,0:08:50.000
is not actually based on very much.

0:08:50.000,0:08:53.000
It really is based on some impassioned pleas of parents

0:08:53.000,0:08:56.000
whose children died after they turned two,

0:08:56.000,0:09:00.000
which has led to the passage of all these laws -- not very much on data.

0:09:00.000,0:09:04.000
So you can only get so far, I think, in telling your story

0:09:04.000,0:09:06.000
by using these abstract statistics.

0:09:06.000,0:09:11.000
And so I had some friends over to dinner, and I was asking --

0:09:11.000,0:09:14.000
we had a cookout -- I was asking them what advice they might have for me

0:09:14.000,0:09:18.000
about proving my point. They said, "Why don't you run some crash tests?"

0:09:18.000,0:09:20.000
And I said, "That's a great idea."

0:09:20.000,0:09:22.000
So we actually tried to commission some crash tests.

0:09:22.000,0:09:27.000
And it turns out that as we called around to the independent

0:09:27.000,0:09:30.000
crash test companies around the country,

0:09:30.000,0:09:32.000
none of them wanted to do our crash test

0:09:32.000,0:09:36.000
because they said, some explicitly, some not so explicitly,

0:09:36.000,0:09:38.000
"All of our business comes from car seat manufacturers.

0:09:38.000,0:09:42.000
We can't risk alienating them by testing seatbelts relative to car seats."

0:09:42.000,0:09:46.000
Now, eventually, one did. Under the conditions of anonymity,

0:09:46.000,0:09:49.000
they said they would be happy to do this test for us --

0:09:49.000,0:09:54.000
so anonymity, and 1,500 dollars per seat that we crashed.

0:09:54.000,0:09:56.000
And so, we went to Buffalo, New York,

0:09:56.000,0:09:58.000
and here is the precursor to it.

0:09:58.000,0:10:00.000
These are the crash test dummies,

0:10:00.000,0:10:03.000
waiting for their chance to take the center stage.

0:10:03.000,0:10:05.000
And then, here's how the crash test works.

0:10:05.000,0:10:08.000
Here, they don't actually crash the entire car, you know --

0:10:08.000,0:10:11.000
it's not worth ruining a whole car to do it.

0:10:11.000,0:10:12.000
So they just have these bench seats,

0:10:12.000,0:10:14.000
and they strap the car seat and the seatbelt onto it.

0:10:14.000,0:10:16.000
So I just wanted you to look at this.

0:10:16.000,0:10:18.000
And I think this gives you a good idea of why parents think

0:10:18.000,0:10:20.000
car seats are so great. Look at the kid in the car seat.

0:10:20.000,0:10:23.000
Does he not look content, ready to go,

0:10:23.000,0:10:25.000
like he could survive anything? And then, if you look at the kid in back,

0:10:25.000,0:10:28.000
it looks like he's already choking before the crash even happens.

0:10:28.000,0:10:31.000
It's hard to believe, when you look at this, that

0:10:31.000,0:10:33.000
that kid in back is going to do very well when you get in a crash.

0:10:33.000,0:10:35.000
So this is going to be a crash

0:10:35.000,0:10:38.000
where they're going to slam this thing forward into a wall

0:10:38.000,0:10:41.000
at 30 miles an hour, and see what happens. OK?

0:10:41.000,0:10:43.000
So, let me show you what happens.

0:10:43.000,0:10:46.000
These are three-year-old dummies, by the way.

0:10:46.000,0:10:48.000
So here -- this is the car seat. Now watch two things:

0:10:48.000,0:10:50.000
watch how the head goes forward,

0:10:50.000,0:10:52.000
and basically hits the knees -- and this is in the car seat --

0:10:52.000,0:10:57.000
and watch how the car seat flies around, in the rebound, up in the air.

0:10:57.000,0:10:59.000
The car seat's moving all over the place.

0:10:59.000,0:11:01.000
Bear in mind there are two things about this.

0:11:01.000,0:11:04.000
This is a car seat that was installed by someone

0:11:04.000,0:11:07.000
who has installed 1,000 car seats, who knew exactly how to do it.

0:11:07.000,0:11:08.000
And also it turned out these bench seats

0:11:08.000,0:11:11.000
are the very best way to install car seats.

0:11:11.000,0:11:14.000
Having a flat back makes it much easier to install them.

0:11:14.000,0:11:17.000
And so this is a test that's very much rigged in favor of the car seat,

0:11:17.000,0:11:20.000
OK? So, that kid in this crash fared very well.

0:11:20.000,0:11:22.000
The federal standards are

0:11:22.000,0:11:24.000
that you have to score below a 1,000

0:11:24.000,0:11:26.000
to be an approved car seat on this crash,

0:11:26.000,0:11:30.000
in some metric of units which are not important.

0:11:30.000,0:11:33.000
And this crash would have been about a 450.

0:11:33.000,0:11:35.000
So this car seat was actually an above-average car seat

0:11:35.000,0:11:37.000
from Consumer Reports, and did quite well.

0:11:37.000,0:11:40.000
So the next one. Now, this is the kid, same crash,

0:11:40.000,0:11:45.000
who is in the seatbelt. He hardly moves at all, actually,

0:11:45.000,0:11:48.000
relative to the other child. The funny thing is,

0:11:48.000,0:11:51.000
the cam work is terrible because they've only set it up

0:11:51.000,0:11:53.000
to do the car seats, and so, they actually don't even have a way

0:11:53.000,0:11:55.000
to move the camera so you can see the kid that's on the rebound.

0:11:55.000,0:11:59.000
Anyway, it turns out that those two crashes, that actually

0:11:59.000,0:12:03.000
the three-year-old did slightly worse. So, he gets about a 500

0:12:03.000,0:12:07.000
out of -- you know, on this range -- relative to a 400 and something.

0:12:07.000,0:12:10.000
But still, if you just took that data from that crash

0:12:10.000,0:12:13.000
to the federal government, and said, "I have invented a new car seat.

0:12:13.000,0:12:16.000
I would like you to approve it for selling,"

0:12:16.000,0:12:19.000
then they would say, "This is a fantastic new car seat, it works great.

0:12:19.000,0:12:21.000
It only got a 500, it could have gotten as high up as a 1,000."

0:12:21.000,0:12:24.000
And this seatbelt would have passed with flying colors

0:12:24.000,0:12:26.000
into being approved as a car seat.

0:12:26.000,0:12:28.000
So, in some sense, what this is suggesting

0:12:28.000,0:12:31.000
is that it's not just that people are setting up their car seats wrong,

0:12:31.000,0:12:33.000
which is putting children at risk. It's just that, fundamentally,

0:12:33.000,0:12:35.000
the car seats aren't doing much.

0:12:35.000,0:12:37.000
So here's the crash. So these are timed at the same time,

0:12:37.000,0:12:39.000
so you can see that it takes much longer with the car seat --

0:12:39.000,0:12:41.000
at rebound, it takes a lot longer --

0:12:41.000,0:12:45.000
but there's just a lot less movement for child who's in the seatbelt.

0:12:45.000,0:12:47.000
So, I'll show you the six-year-old crashes as well.

0:12:47.000,0:12:52.000
The six-year-old is in a car seat, and it turns out

0:12:52.000,0:12:57.000
that looks terrible, but that's great. That's like a 400, OK?

0:12:57.000,0:12:58.000
So that kid would do fine in the crash.

0:12:58.000,0:13:02.000
Nothing about that would have been problematic to the child at all.

0:13:02.000,0:13:05.000
And then here's the six-year-old in the seatbelt,

0:13:05.000,0:13:07.000
and in fact, they get exactly within, you know,

0:13:07.000,0:13:11.000
within one or two points of the same. So really, for the six-year-old,

0:13:11.000,0:13:15.000
the car seat did absolutely nothing whatsoever.

0:13:15.000,0:13:18.000
That's some more evidence, so in some sense --

0:13:18.000,0:13:22.000
I was criticized by a scientist, who said, "You could never publish

0:13:22.000,0:13:24.000
a study with an n of 4," meaning those four crashes.

0:13:24.000,0:13:28.000
So I wrote him back and I said, "What about an n of 45,004?"

0:13:28.000,0:13:30.000
Because I had the other 45,000 other real-world crashes.

0:13:30.000,0:13:34.000
And I just think that it's interesting that the idea

0:13:34.000,0:13:36.000
of using real-world crashes, which is very much something

0:13:36.000,0:13:38.000
that economists think would be the right thing to do,

0:13:38.000,0:13:40.000
is something that scientists don't actually, usually think --

0:13:40.000,0:13:43.000
they would rather use a laboratory,

0:13:43.000,0:13:45.000
a very imperfect science of looking at the dummies,

0:13:45.000,0:13:49.000
than actually 30 years of data of what we've seen

0:13:49.000,0:13:52.000
with children and with car seats.

0:13:52.000,0:13:56.000
And so I think the answer to this puzzle

0:13:56.000,0:13:59.000
is that there's a much better solution out there,

0:13:59.000,0:14:02.000
that's gotten nobody excited because everyone

0:14:02.000,0:14:06.000
is so delighted with the way car seats are presumably working.

0:14:06.000,0:14:09.000
And if you think from a design perspective,

0:14:09.000,0:14:11.000
about going back to square one, and say,

0:14:11.000,0:14:13.000
"I just want to protect kids in the back seat."

0:14:13.000,0:14:15.000
I don't there's anyone in this room who'd say,

0:14:15.000,0:14:16.000
"Well, the right way to start would be,

0:14:16.000,0:14:19.000
let's make a great seat belt for adults.

0:14:19.000,0:14:21.000
And then, let's make this really big contraption

0:14:21.000,0:14:24.000
that you have to rig up to it in this daisy chain."

0:14:24.000,0:14:27.000
I mean, why not start -- who's sitting in the back seat anyway except for kids?

0:14:27.000,0:14:30.000
But essentially, do something like this,

0:14:30.000,0:14:32.000
which I don't know exactly how much it would cost to do,

0:14:32.000,0:14:33.000
but there's no reason I could see

0:14:33.000,0:14:35.000
why this should be much more expensive than a regular car seat.

0:14:35.000,0:14:39.000
It's just actually -- you see, this is folding up -- it's behind the seat.

0:14:39.000,0:14:41.000
You've got a regular seat for adults, and then you fold it down,

0:14:41.000,0:14:43.000
and the kid sits on top, and it's integrated.

0:14:43.000,0:14:47.000
It seems to me that this can't be a very expensive solution,

0:14:47.000,0:14:50.000
and it's got to work better than what we already have.

0:14:50.000,0:14:55.000
So the question is, is there any hope for adoption of something like this,

0:14:55.000,0:14:57.000
which would presumably save a lot of lives?

0:14:57.000,0:15:01.000
And I think the answer, perhaps, lies in a story.

0:15:01.000,0:15:05.000
The answer both to why has a car seat been so successful,

0:15:05.000,0:15:08.000
and why this may someday be adopted or not,

0:15:08.000,0:15:12.000
lies in a story that my dad told me, relating to when he was a doctor

0:15:12.000,0:15:15.000
in the U.S. Air Force in England. And this is a long time ago:

0:15:15.000,0:15:17.000
you were allowed to do things then you can't do today.

0:15:17.000,0:15:21.000
So, my father would have patients come in

0:15:21.000,0:15:24.000
who he thought were not really sick.

0:15:24.000,0:15:28.000
And he had a big jar full of placebo pills that he would give them,

0:15:28.000,0:15:31.000
and he'd say, "Come back in a week, if you still feel lousy."

0:15:31.000,0:15:32.000
OK, and most of them would not come back,

0:15:32.000,0:15:34.000
but some of them would come back.

0:15:34.000,0:15:38.000
And when they came back, he, still convinced they were not sick,

0:15:38.000,0:15:43.000
had another jar of pills. In this jar were huge horse pills.

0:15:43.000,0:15:45.000
They were almost impossible to swallow.

0:15:45.000,0:15:49.000
And these, to me, are the analogy for the car seats.

0:15:49.000,0:15:53.000
People would look at these and say, "Man, this thing is so big

0:15:53.000,0:15:55.000
and so hard to swallow. If this doesn't make me feel better,

0:15:55.000,0:15:58.000
you know, what possibly could?"

0:15:58.000,0:16:00.000
And it turned out that most people wouldn't come back,

0:16:00.000,0:16:03.000
because it worked. But every once in a while,

0:16:03.000,0:16:08.000
there was still a patient convinced that he was sick,

0:16:08.000,0:16:11.000
and he'd come back. And my dad had a third jar of pills.

0:16:11.000,0:16:13.000
And the jar of pills he had, he said,

0:16:13.000,0:16:16.000
were the tiniest little pills he could find,

0:16:16.000,0:16:18.000
so small you could barely see them.

0:16:18.000,0:16:20.000
And he would say, listen, I know I gave you that huge pill,

0:16:20.000,0:16:24.000
that complicated, hard-to-swallow pill before,

0:16:24.000,0:16:26.000
but now I've got one that's so potent,

0:16:26.000,0:16:28.000
that is really tiny and small and almost invisible.

0:16:28.000,0:16:31.000
It's almost like this thing here, which you can't even see."

0:16:31.000,0:16:33.000
And it turned out that never,

0:16:33.000,0:16:36.000
in all the times my dad gave out this pill, the really tiny pill,

0:16:36.000,0:16:39.000
did anyone ever come back still complaining of sickness.

0:16:39.000,0:16:42.000
So, my dad always took that as evidence

0:16:42.000,0:16:46.000
that this little, teeny, powerful pill

0:16:46.000,0:16:50.000
had the ultimate placebo effect. And in some sense, if that's the right story,

0:16:50.000,0:16:52.000
I think integrated car seats you will see, very quickly,

0:16:52.000,0:16:56.000
becoming something that everyone has. The other possible conclusion

0:16:56.000,0:16:59.000
is, well, maybe after coming to my father three times,

0:16:59.000,0:17:01.000
getting sent home with placebos, he still felt sick,

0:17:01.000,0:17:03.000
he went and found another doctor.

0:17:03.000,0:17:05.000
And that's completely possible. And if that's the case,

0:17:05.000,0:17:08.000
then I think we're stuck with conventional car seats for a long time to come.

0:17:08.000,0:17:09.000
Thank you very much.

0:17:09.000,0:17:13.000
(Applause)

0:17:13.000,0:17:15.000
(Audience: I just wanted to ask you, when we wear seatbelts

0:17:15.000,0:17:18.000
we don't necessarily wear them just to prevent loss of life,

0:17:18.000,0:17:20.000
it's also to prevent lots of serious injury.

0:17:20.000,0:17:24.000
Your data looks at fatalities. It doesn't look at serious injury.

0:17:24.000,0:17:26.000
Is there any data to show that child seats

0:17:26.000,0:17:29.000
are actually less effective, or just as effective as seatbelts

0:17:29.000,0:17:31.000
for serious injury? Because that would prove your case.)

0:17:31.000,0:17:34.000
Steven Levitt: Yeah, that's a great question. In my data, and in another data set

0:17:34.000,0:17:37.000
I've looked at for New Jersey crashes,

0:17:37.000,0:17:41.000
I find very small differences in injury.

0:17:41.000,0:17:43.000
So in this data, it's statistically insignificant differences

0:17:43.000,0:17:47.000
in injury between car seats and lap-and-shoulder belts.

0:17:47.000,0:17:48.000
In the New Jersey data, which is different,

0:17:48.000,0:17:51.000
because it's not just fatal crashes,

0:17:51.000,0:17:53.000
but all crashes in New Jersey that are reported,

0:17:53.000,0:17:56.000
it turns out that there is a 10 percent difference in injuries,

0:17:56.000,0:17:58.000
but generally they're the minor injuries.

0:17:58.000,0:18:00.000
Now, what's interesting, I should say this as a disclaimer,

0:18:00.000,0:18:05.000
there is medical literature that is very difficult to resolve with this other data,

0:18:05.000,0:18:09.000
which suggests that car seats are dramatically better.

0:18:09.000,0:18:11.000
And they use a completely different methodology that involves --

0:18:11.000,0:18:14.000
after the crash occurs, they get from the insurance companies

0:18:14.000,0:18:16.000
the names of the people who were in the crash,

0:18:16.000,0:18:17.000
and they call them on the phone,

0:18:17.000,0:18:18.000
and they asked them what happened.

0:18:18.000,0:18:21.000
And I really can't resolve, yet,

0:18:21.000,0:18:23.000
and I'd like to work with these medical researchers

0:18:23.000,0:18:26.000
to try to understand how there can be these differences,

0:18:26.000,0:18:29.000
which are completely at odds with one another.

0:18:29.000,0:18:32.000
But it's obviously a critical question.

0:18:32.000,0:18:35.000
The question is even if -- are there enough serious injuries

0:18:35.000,0:18:38.000
to make these cost-effective? It's kind of tricky.

0:18:38.000,0:18:40.000
Even if they're right, it's not so clear

0:18:40.000,0:18:41.000
that they're so cost-effective.