Japan (metropolitan rail network)
#00154
Case study of
#00155 Install blue LED lighting on platforms as a low-cost calming measure
Implementer
Japanese rail operators (evaluated by Matsubayashi, Sawada & Ueda)
Timeline
Jan 1, 2000 – Dec 31, 2013
Location
Description
Blue LED lights were installed at platform ends across Japanese stations. Before-and-after panel data studies (14 treated stations, neighbouring and control stations, 2000–2013) reported large decreases in suicides and no displacement to neighbouring stations. A separate 2014 reconsideration disputed this, noting that only ~14% of attempts occur both within station premises and at night — the only conditions platform lighting could plausibly affect — making the headline reductions implausibly large as a true causal effect; estimates rest on a single company's data with very wide confidence intervals. Replicators should treat evidentiary status as contested.
Metrics
3Lessons learned
- Reported reductions (−74% to −84%) are heavily contested and cannot be treated as established causal effects; wide confidence intervals and reliance on a single company's data limit generalisability.
- Only ~14% of attempts occur both on-station and at night, so even a genuine effect would apply to a small fraction of total suicides — a critical constraint when sizing expected impact.
- Two 2014 papers on this intervention are often conflated: one found no displacement to neighbouring stations, the other argued the causal effect is minimal; replicators should read both separately.
Sources
2Documented Jul 13, 2026