We Solved Child Car Seat Safety Decades Ago. 70% of Kids in Fatal Crashes Still Aren't Properly Restrained.

Lurie Children's Hospital of Chicago pulled a decade of FARS data and counted 50,000 children under 13 involved in fatal crashes between 2011 and 2021. Nearly 70% had suboptimal restraint use.^[1] Not 70% of kids riding around without seatbelts, and not 70% of reckless parents. Seventy percent of the children whose crash was bad enough that somebody in it died were strapped in wrong, strapped in too loose, graduated to the next seat too early, or not strapped in at all.

Car seats are the single most studied, most proven, most universally available child safety intervention in the history of consumer products. Rear-facing infant seats cut fatality risk by 71%.^[2] Forward-facing harness seats reduce death by 54% for toddlers aged one to four. Child restraint systems overall cut mortality risk by 28% compared to seatbelts alone, according to Durbin et al.'s landmark 2006 study in Pediatrics.^[3] The science isn't ambiguous. Every hospital in America requires a car seat before discharging a newborn. Every state mandates them by law, and this problem was solved before most of the parents in the dataset were born.

Three kids die in car crashes every day, and four hundred twenty-nine get injured.

The Kozhumam et al. study, published in Traffic Injury Prevention, broke down what "suboptimal" actually looks like, and the taxonomy of failure is its own kind of horror show.^[1] Thirty-six percent of the children were prematurely transitioned to a less protective restraint type, which is the cruelest category because it means the parents bought the next seat up, installed it, buckled their child into it, and drove away believing they were doing the right thing. A three-year-old in a forward-facing harness seat who should have still been rear-facing. A five-year-old in a booster who needed another year in a five-point harness. A seven-year-old riding without a booster at all because she looked big enough, and "big enough" is a judgment call that kills children at a rate the data can now quantify.

Twenty percent were completely unrestrained, and fifteen percent were riding in the front seat when they belonged in the back. Nine percent managed to combine both failures simultaneously: front seat, no restraint, under thirteen years old, in a crash that killed someone.

Kozhumam's team identified 75 county-level hotspots where suboptimal restraint use concentrated, and they correlated strongly with the Child Opportunity Index, which measures a neighborhood's access to education, health resources, and economic stability.^[1] Under-resourced communities had worse outcomes. States with stricter child restraint laws had significantly lower rates of suboptimal use. Policy works. It is also, apparently, optional in enough jurisdictions that the national average sits at seventy percent wrong.

The counterargument writes itself: FARS captures only fatal crashes, which represent a vanishingly small fraction of the roughly 6.7 million annual crashes in the United States.^[4] Maybe children in fatal crashes are a biased sample. Maybe unrestrained or poorly restrained kids are overrepresented precisely because improper restraint contributed to the crash's lethality. That is exactly the point. The 70% figure describes what happens when it matters most, when the margins between a child who walks away and one who doesn't are measured in millimeters of harness slack and months of premature seat graduation. The fatal-crash denominator is not a limitation of the data. It is the data.

Ages four through twelve carry the highest risk. Infants generally ride in their hospital-issued carriers, and toddlers sit in the seat their pediatrician recommended. But somewhere between preschool and middle school, parents start eyeballing fit instead of reading the weight chart on the side of the seat. The transition from harness to booster is where the American Academy of Pediatrics' guidelines and American driving culture diverge most violently: the AAP says booster until 4'9", which for most children means age eight to twelve.^[5] Many parents graduate their kids at five or six because the child complains, because their friends aren't in boosters anymore, because the seat looks silly in a third-grader's hands. Peer pressure as a mechanism for pediatric traffic fatality is not a sentence anyone planned to write, but the data forces it.

What You Should Do

Check your child's current seat against NHTSA's guidelines at nhtsa.gov/car-seats. Rear-facing until at least age two, or until the child exceeds the seat's maximum height and weight limits. Forward-facing with a five-point harness until the child reaches the harness seat's upper weight limit, typically 40 to 65 pounds. Booster seat until 4'9", rear seat until age thirteen, no exceptions. If a child looks too big for their seat, they are almost certainly safer staying in it for another six months than graduating early. NHTSA offers free car seat inspections at certified stations listed at nhtsa.gov/get-car-seat-checked.

Limitations

FARS records only crashes involving at least one fatality, which excludes the vast majority of motor vehicle incidents. The 70% figure may overstate the national restraint-misuse rate for all crashes since improper restraint increases injury severity and thus the likelihood of FARS inclusion. The study period ends in 2021, predating potential improvements from state law changes enacted after that year. Restraint status in FARS is coded by investigating officers at the scene, introducing observer-reliability variance, particularly for distinguishing "correct restraint type, improperly installed" from "correct restraint type, properly installed."