Car Seats: When to Turn, When to Stop

No COVID here today… Sometimes we need a reminder (at least me!) that there are many other parenting problems on our plate. Today, I visit a question I get a lot which somehow didn’t make it into any of my books (so far!): Car Seats

I am not so old that I do not remember car seats as a child. However, I definitely do not recall them as a permeant fixture in the car. There were many, many car trips with three children and my grandmother in the back seat, the smallest child held (buckled!) in her lap. I doubt my family ever used a booster seat in the car, and by 8 or 9 I was permitted in the front seat. To be clear: my family was on the risk averse end. Seatbelts at all times, defensive driving, and so on.

Car seat rules have evolved over time. In the present moment: kids under 2 are in infant seats, usually facing backwards. Two to four brings the giant five-point-harness car seat (when we finally donated the last of these I can only describe my husband as gleeful). And, in many places, kids up to age 8 are still required to be in booster seats. Recommendations often suggest continued booster us up until the age of 12.

These regulations have accelerated since the 1980s, and especially since the mid-1990s. In a somewhat surprising fact, a recent paper argued that these regulations have impacted fertility. The idea is that people do not want to have a third kid because they cannot fit in another car seat. Seriously. That’s what the paper says. Almost a 1 percentage point drop in third births! Once you hit kid 3, you’ve got to go to a minivan. Maybe it’s not worth it?

For the most part, car seat behavior is dictated by these regulations. You should follow the law, period. But that doesn’t mean we cannot ask why these regulations are in place, and what the data says. And, in particular, I want to focus on two questions.

  • First: what’s the evidence on front versus rear facing for kids, especially those above a year.

  • Second, what is the evidence on booster seats for older kids.

Transition #1: Front Versus Rear Facing

Infants sit in an infant seat, and they face backwards. This is understood, accepted and, indeed, they will not let you leave the hospital until you show your child is in the car seat correctly. Recommendations have long been that children until the age of 1 year face backwards. In more recent years, this has moved to the age of 2. Some people push it further. Why two years? If rear-facing is safer, why should there be an age cut-off at all?

The idea is driven by the biomechanics of what would happen if you were in a car accident. Tests with crash test dummies sized between 1 and 3 years (i.e. this paper) suggest less risk of injury (to the dummies) with rear facing than forward facing car seats. These tests are very convincing, since they are run in controlled environments, with dolls equipped with fancy accelerometers.

Translating this to data in the real world has been more of a challenge. In 2007, a very influential paper was published comparing injury rates in crashes between children using forward versus rear facing car seats. You can see the paper here. Using data from the National Highway Traffic Safety Administration, the authors argue that children aged 12 to 23 months are five times as likely to be injured in a side crash if in a forward facing seat, and twice as likely overall. This paper played a large role in the push to staying rear facing for longer.

However: it has since been retracted.

In 2017, the journal was contacted by an outside researcher who had been unable to replicate the analysis. The original paper writers then revisited their analysis and found they were also unable to replicate it. The issue, according to the journal, was thought to be a mishandling of survey weights, which caused their sample size to seem bigger that it was (Stata peeps: I think maybe “fweight” instead of “aweight” but they do not give details).

The authors retracted the paper and redid the analysis, now with more years of data, publishing an update here. Their conclusion was that while the direction of the effect was in favor of rear facing seats, there was no significance and the effects were small (0.5% of children 12-23 months in forward facing seats were injured, versus 0.2% of those in rear facing. But the actual number of injuries was only 14 and 3, much too low to get any significance).

This is all consistent with a world in which the the biology/mechanics/theory shows that rear facing has advantages, but the real world data suggests these may be quite small. So small that we’d need more than the 20 years of data we have to figure them out.

You might ask, given this: what is the point of turning kids around, if there is any suggestion it is safer, why not have them ride that way forever? One reason is that as kids get bigger it gets very cramped and uncomfortable. Another is that they often prefer riding forward and for kids who are carsick (I am very attuned to this concern), forward-riding is often better.

Transition #2: Out of Booster

Second big question: Does your kid need to be in a booster until they are 8 or 10 or 12? My older kid is quite tall and by the time she moved into a real seat she was as tall as some short adults. Was there any real point in maintaining the booster for so long?

(I’m focusing here on “boosters” in general, although most kids spend some time in five-point harness car seats, which combine a booster seat with a harness structure. I’ll try to dig into this transition another time, but data is limited. Here, I’ll tackle the question of transition out of any booster option.)

The arguments behind a booster seat are also based in biomechanics. Kids are short. Seatbelts are built for taller people. This means they are less well suited to protect little kids in an accident than to protect adults. Booster seats (or “belt positioning boosters” as they are sometimes called) are designed to get kids tall enough to optimize seatbelt functioning.

A number of years ago, Steve Levitt (of Freakonomics fame, but also a friend of mine) wrote several papers arguing that, in fact, booster seats didn’t protect kids from serious injury or death. This paper was subject to a lot of criticism and scrutiny. Notable to me, from a parenting standpoint though, is that they do find a reduction in less serious injuries. I would leave my kid in a booster even for that.

But even if you do not care about less serious injuries, follow-up work suggests larger safety improvements, even for older kids. One set of example papers are this one and this one which use data from State Farm Insurance on younger and older kids to examine the impact of booster seats on injuries. The authors argue that use of booster seats reduces injuries, even for kids 4 to 7. An obvious complaint with both papers, though, is that they use parent-reported injuries and parent-reported seat use. Variations in honesty or memory could bias results.

Better, I think, is this paper which relates variation in state booster laws to estimate the relationship between boosters and accident fatality rates. They find that fatality rates among 6 and 7 year olds dropped in places that introduced booster laws for those age groups, relative to those that did not. (They find similar results for 4 and 5 year olds but there is less variation in those groups so the results are weaker). I have some quibbles with the paper (I do not like their figure presentation) but this is a good approach to the problem and an important finding.

There isn’t any good data I can find on booster seats for kids beyond the age of 8, partly because they aren’t required which makes it hard to have any good variation (more careful drivers are likely to keep kids in boosters for longer, for example). But important to note is that it’s not so much age as size. A very tall kid may be ready to be out of a booster before a short one.

Bottom line here is that even though they are annoying, my read is that even in the absence of regulations it is a good idea to keep kids in boosters until they are fairly old. If you’re looking for a travel-friendly version, might I suggest the Bubble Bum?

Two Final Words:

First: for any car seat transition, it’s not really age, but size. Especially on the issue of boosters, the concerns arise because kids are shorter than adults. This means that some of the choices about boosters should reflect the size of your kid.

Second: The most dangerous thing is for kids to ride unrestrained. No seatbelt or restraint is much worse than an incorrectly used one. This might be obvious but in case it is not: Click it or Ticket!

Weigh in!

Keep the thoughts coming. I don’t always write back, but I read everything.

Write to Me!

Where to Find Me

COVID School Dashboard

My Website

Family Firm


Expecting Better