The FAA has a genuinely clever piece of regulatory reasoning buried in its child-safety rules: it allows children under two to fly on a parent's lap, unrestrained, because mandating a paid seat would push cost-sensitive families into cars — a far deadlier mode of transport. This is substitution-effect thinking at its best, and it has saved lives since the 1990s. But the same agency applies none of that logic to babywearing carriers. Under 14 CFR 121.311(b)(1), a lap infant must not "occupy or use any restraining device" during ground movement, takeoff, or landing. This means parents must remove a sleeping baby from a carrier and hold them in bare arms — the one configuration the FAA's own foundational 1994 study never actually tested. The study evaluated booster seats, car-seat-style carriers, harnesses, belly belts, and lap belts. It did not test unrestrained lap-holding, and it did not test modern soft-structured babywearing carriers. The 1995 proposed rulemaking justified the ban by arguing that the tested devices "may put children in a potentially worse situation than the allowable alternatives." The problem is obvious: they never measured the allowable alternative. The belly belt — the closest analog to a carrier — performed badly in crash tests, with the dummy making severe seat-back contact and the child at risk of being crushed by the adult's forward motion. But a belly belt is not a babywearing carrier; it's a strap that buckles a child's abdomen to an adult via the aircraft lap belt. Modern carriers distribute the infant across the wearer's torso with entirely different force dynamics. When the UK's Civil Aviation Authority and Europe's JAA pointed out that parents often cannot hold a lap child during turbulence or a crash, the FAA's 1996 final rule sidestepped the comparison. It acknowledged lap-holding risks, predicted parents would voluntarily buy a second seat, and reiterated its substitution argument against mandating seats — all without addressing whether a carrier might be safer than arms. The agency applied cost-benefit reasoning to one adjacent question and then abandoned it for the next. Rep. Bill Posey's HR 8972, introduced in 2024, would have directed the FAA to evaluate carriers. It died in the aviation subcommittee without a vote. The legislative path for niche child-safety regulatory reform is essentially nonexistent without either sustained advocacy pressure or an FAA internal review. The practical harm is small in absolute terms — flying is extraordinarily safe, and the main cost is inconvenience and woken babies. But the structural point matters: a rule premised on comparative safety analysis was built on a study that omitted the relevant comparison. That's not precaution; it's a gap in the evidence being treated as a conclusion. If the FAA's own framework says you must weigh alternatives, then not testing the most common alternative is a methodological failure, not a policy choice. The author's core argument is right: if the agency isn't going to run modern crash tests on babywearing carriers, the default should be permission, not prohibition. The FAA already accepts the risk of an unrestrained lap infant. Banning a device that almost certainly provides more restraint than bare arms, based on tests of a different device 30 years ago, is regulatory incoherence preserved by institutional inertia.
