the helmet stops at the handlebars

Problem statement

Vietnam's December 2007 mandatory helmet law is one of road safety's celebrated successes: "various studies have found between 90% and 99% adult driver compliance in the years since." Children are the exception. WHO-affiliated observations in four cities in April 2008 found helmet use of only 15–53% among children aged seven and under and 38–53% among 7–14-year-olds, and "the impact ... of the more recent 2010 law for child motorcycle helmets has been far less pronounced." A national campaign in 2015 — billboards and television, school-based guidance, capacity-building for provincial officials, and a nationwide police enforcement blitz around schools on 6–9 April 2015 — pushed observed child helmet use in Ha Noi from 23.2% (March 2014) to 70.7% (April 2015); by December 2015 it was 36.8% and by May 2016 35.4%, essentially back to baseline. Ho Chi Minh City went 48.3% → 66.2% → 47.6% → 49.0%; only Da Nang held part of its gain (37.1% → 75.6% → 55.9% → 66.7%). The unsolved problem is that the parental behaviour — helmeting a child for the school run on the same motorbike on which the parent is helmeted — has proven resistant to the law-plus-enforcement formula that worked for adults, and nobody yet knows what makes it stick in one city and not another: the evaluators state they "were not able to determine why effects were sustained in some locations (eg, Da Nang), but not others (eg, HCMC)."

Why this matters

Motorcycles are the family vehicle across Vietnam and much of Southeast Asia, and children ride as passengers daily; head injury is the dominant cause of motorcycle death and helmets are the single most effective protection. The adult success shows the country can enforce; the child gap shows enforcement alone does not carry over when the person at risk is not the person being fined and when parents hold specific counter-beliefs. The 2009 WHO Bulletin study found "67% [of parents] said the fear of neck injury was the most important reason their children did not wear a helmet" — a belief with no clinical basis that continues to circulate — alongside perceived weak enforcement, cost (a standard helmet costs roughly US$7–13 against an urban per-capita monthly income of about US$192), and a perceived low likelihood of crashing. The same adult–child compliance gap is reported internationally, so a durable Vietnamese solution would travel to Cambodia, Thailand, Indonesia and beyond.

What’s been tried and why it hasn’t worked

Legislation came first: the 2007 law covered adults; because under Vietnam's administrative-sanctions rules "children aged < 14 years cannot be given sanctions," a 2010 decree shifted liability to the adult driver carrying an unhelmeted child, while children under six reportedly remained outside the requirement — a carve-out that secondary accounts link to the same neck-injury concern (see Source Notes). Enforcement then produced the classic blitz-and-decay curve: the April 2015 police operation coincided with the peak observation and the subsequent fall tracks the end of visible enforcement, not any change in parents' beliefs. Social marketing (parental-responsibility messaging, school helmet guidelines) has been layered on since 2008 by government with AIP Foundation and international partners, and helmet give-aways have addressed cost, yet the 2017 evaluation's baseline of 36.1% across the three largest cities shows how little of the earlier work had held. The evaluation "was not able to isolate which components of the campaign were most effective," observed only school journeys (children ride for many other trips), and covered only urban centres although "the majority of Vietnamese live in non-urbanised settings." What the pattern tells us: the adult law worked because the fine, the risk and the behaviour all attach to the same person; for children the parent weighs a small fine and a discounted risk against a live myth, daily hassle, and a norm they can see at every school gate — and short enforcement bursts move the norm only while they last.

What would unlock progress

Progress needs a behaviour-change design that makes child-helmet use self-sustaining between enforcement waves: durable school-gate norms (a school where every child arrives helmeted is a visible norm), direct and credible refutation of the neck-injury belief through trusted messengers (paediatricians, teachers) rather than billboards, and enforcement designed for persistence (predictable, school-linked, low-intensity) rather than for a peak. The Da Nang anomaly is the natural experiment: understanding what differed there — municipal ownership, school engagement, sustained enforcement — is the cheapest route to a scalable model. Adjacent fields: child-restraint and rear-seat-belt work in high-income countries confronted a comparable "adult complies, child doesn't" gap and is a natural source of comparison; the Vietnamese adult law itself is the precedent for how quickly a norm can flip when the design is right.

Entry points for student teams

The measurement is the contribution here, because the persistence window is the whole point: a team could run the observation arm alone — repeated, standardised roadside counts of child helmet use at the same school gates at the same times across a term, producing the decay curve nobody has published since May 2016 — and hand the intervention design, pre-registered and instrumented, to a follow-on team rather than trying to mount and evaluate a behaviour change inside one semester. That arm needs in-country access: a Vietnamese or regional partner school or university, permission to count from public sidewalks, and an ethics review that treats unidentified head-counts of children in a public street as minimal-risk — access normally held through an existing university partnership or an implementing NGO such as AIP Foundation, not obtainable cold. A behavioural-science team can design and pre-test the neck-injury-myth refutation and its candidate messengers, but a pre-test only means something with Vietnamese motorcycle-owning parents, so it needs the same partner or a recruited Vietnamese-language online panel, and the team should state plainly which population it actually reached. The document-only door needs no partner at all: reconstruct the Da Nang case from the published evaluation, provincial road-safety plans, AIP Foundation and WHO reporting and press coverage, and write it as a transferable playbook of what was done differently, by whom, and at what cost. Relevant skills: injury prevention, behavioural science, social marketing, school-community engagement, and observational field methods.