The number to hold in your head is 53.
A 67-year-old British pensioner who rides a rental scooter daily in Thailand carries an annual road-death exposure approximately 53 times the United Kingdom home-country baseline. That is not a doomer scare. It is a multiplicative model built from four independently sourced factors — the WHO 2023 country baseline, the foreigner overrepresentation factor triangulated from UK FCDO consular returns, the 65-plus age vulnerability multiplier from the published age-speed and odds-ratio literature, and the mode-of-travel factor for someone on two wheels. The central band for a 65-plus Western expat who rides a scooter in SE Asia is roughly 15 to 25 times home-country exposure. The worst case approaches 50. The conservative case is still over 15.
The dream-sellers do not show you the arithmetic. They show you the morning ride to the warung, palm fronds blurring past the helmet, a USD 4 lunch waiting on a beach. They do not show you the slope of the line that connects a 70-year-old’s reaction time to a 25 mph impact to a Thai DDC inpatient register where 84 of every 100 motorcycle casualties were not wearing a helmet. So this piece is the arithmetic, set out factor by factor, with every coefficient sourced.
The country baselines
The WHO Global Status Report on Road Safety 2023, published December 2023 against data year 2021, sets out the floor. Thailand at 25.4 road deaths per 100,000 population, with 83.8% of fatalities motorcyclists. Vietnam at 17.7, Cambodia at 18.8, Indonesia at 11.3, Malaysia at approximately 22.5, the Philippines at 9.7. The regional WHO South-East Asia average runs 14.4 per 100,000. Asia-Pacific 15.2. The European Region 9.3. The global rate 15.0.
The Western baselines you measure against come from the same year-class. The UK Department for Transport’s Reported Road Casualties Great Britain 2024 records 1,602 deaths against roughly 60 million population: 2.4 per 100,000. The US NHTSA 2022 State Traffic Data gives 12.76. Australia’s BITRE Road Trauma 2024 release gives 4.78. The EU average runs 4 to 5.
The first multiplier writes itself. Thailand to the UK on baseline alone is 25.4 against 2.4. Already more than 10×, before a single coefficient lands. Indonesia to Australia is 11.3 against 4.78, 2.4× on baseline. Philippines to the US is 9.7 against 12.76: a country where the headline US citizen actually moves into a lower-rate environment, until the next two factors land.
| Country | Deaths per 100k | Motorcyclist share | Latest national-police total |
|---|---|---|---|
| Thailand | Deaths per 100k 25.4 | Motorcyclist share 83.8% | Latest national-police total 17,447 deaths (Thai DDC 2024); 81% motorcyclist |
| Malaysia | Deaths per 100k ~22.5 | Motorcyclist share 65% | Latest national-police total 6,443 deaths (MOT 2023); 65% motorcyclist |
| Cambodia | Deaths per 100k 18.8 | Motorcyclist share ~80% | Latest national-police total ~1,590 deaths (2023) |
| Vietnam | Deaths per 100k 17.7 | Motorcyclist share ~58–60% | Latest national-police total Per WHO 2023 profile |
| Indonesia | Deaths per 100k 11.3 | Motorcyclist share 74.35% involved | Latest national-police total ~27,000 deaths (Indonesian police 2023) |
| Philippines | Deaths per 100k 9.7 | Motorcyclist share majority VRU | Latest national-police total ~11,096 (PSA 2021) |
| United States | Deaths per 100k 12.76 | Motorcyclist share ~14% | Latest national-police total NHTSA 2022 State Traffic Data |
| Australia | Deaths per 100k 4.78 | Motorcyclist share ~22% | Latest national-police total BITRE Road Trauma 2024 |
| United Kingdom | Deaths per 100k 2.4 | Motorcyclist share ~20% | Latest national-police total 1,602 deaths (DfT GB 2024) |
Source: WHO 2023 country profiles, data year 2021; national-police 2023/24 data where indicated · checked 2026-05-30
The foreigner factor
The second coefficient is the one no published study sets in stone. It is triangulated.
UK FCDO consular data — the same source read at length in how many expats die abroad — surfaced in a Hansard written answer of 2 February 2024, records 597 new death cases of British nationals in Thailand in 2022, and 430 in the period 1 January to 16 October 2023. Roughly 50 a month, sustained. The published “accidental” subcategory shows 8 in 2022 and 11 in 2023, but FCDO suppresses cells with fewer than five cases per row, so most road deaths fall outside the published “accidental” reportable category and remain invisible in the totals. The 8 and 11 are a floor, not a ceiling. Set against approximately 800,000 UK visitor arrivals to Thailand per year, against a long-stay UK expat denominator in the tens of thousands, the per-stay British death rate is several multiples of the Thai national per-capita rate at equivalent exposure.
Bali tells the same story sharper. The Bali Police 2024 release recorded 142 traffic crashes involving foreign tourists, 21 deaths, 171 minor injuries (a 35% year-on-year increase). In 2022 the equivalent figure was 68 crashes and 4 deaths. The trajectory of foreign-tourist crash involvement is rising in the year-on-year double digits. The same 2023 Bali Police data showed 632 total Bali road deaths against 7,224 accidents, with 86 to 87% of fatal crashes involving motorcycles.
US State Department non-natural death data for the first half of 2023 (532 deaths total across non-natural categories abroad) ranks Thailand in the top 7 with 14 deaths, motorcycle crashes the largest single category. Vehicular accidents account for 28.8% of US non-natural deaths abroad; homicide 21.1%.
The triangulated factor: 1.5 to 3.0× the local population baseline, with a central point estimate of 2.0×. That is the foreigner factor. It is what unfamiliarity with road conventions, left-hand traffic, mixed two-wheel and four-wheel flow, rental scooter unfamiliarity, helmet noncompliance, and tourist-corridor exposure compounds into. It is not a single published constant — it is a triangulated band from three independent consular data streams.
The age vulnerability factor
Tefft, working for the AAA Foundation for Traffic Safety in 2011 and publishing in Accident Analysis & Prevention in 2013, gave the literature its central instrument. Pedestrian death risk against vehicle impact speed: 10% at 24.1 mph, 25% at 32.5 mph, 50% at 40.6 mph, 75% at 48.0 mph, 90% at 54.6 mph. The age-adjusted finding, which is the one that matters here: the risk of severe injury or death for a 70-year-old pedestrian struck at 25 mph is similar to the risk for a 30-year-old struck at 35 mph. A 10 mph age penalty.
EU SafetyCube D6.1, the project deliverable on vehicle-related risk factors, sets the population-level number. Elderly pedestrians (65+) account for approximately 45% of all pedestrian fatalities in EU data. Odds ratios for severe pedestrian injury versus the 60–64 baseline run 1.18 for 65–69 and 1.42 for 70–74. Pedestrians struck in 50 km/h or lower zones die at approximately 25%; in 70 km/h or higher zones, the death rate exceeds 40%.
The driver-cognition literature gives the third leg. Reaction times of drivers aged 65–75 lengthen approximately 30% faster with task load than those of young drivers; the 50–60 age group shows about 10% faster degradation. Low-contrast visual sensitivity declines materially from the sixties; mesopic (twilight and night) contrast detection particularly impaired. 65+ drivers show elevated fatal-crash rates at night, slower glare recovery, poorer night road-sign recognition. SE Asia compounds this: poorly lit secondary roads, mixed traffic with motorcycles weaving through stopped cars, no shoulders, painted lines that vanish in monsoon rain.
The bone-fragility overlay closes the model. An impact that produces a recoverable femur fracture at 35 produces a fatal pelvic ring disruption at 70. The energy threshold for fatal injury shifts down through the seventh decade.
The composite 65+ age vulnerability factor: 2.0 to 3.0× the baseline rate, with a central 2.5×. The upper bound captures the rider who is also not wearing a helmet, also riding at night, also riding on the wrong side of an unfamiliar road system.
The model
Run the arithmetic for a 67-year-old British pensioner who rides a rental scooter daily in Thailand:
local baseline 25.4 per 100,000 × foreigner factor 2.0 × age vulnerability 2.5 × mode factor 1.0 = approximately 127 per 100,000 per year
Against a UK home-country baseline of 2.4 per 100,000, that is approximately 53× home-country exposure. A conservative read (foreigner 1.5×, age 2.0×) gives 76 per 100,000, still roughly 32× UK home rate. A maximum-vulnerability read (foreigner 2.5×, age 3.0×) gives 190.5 per 100,000, roughly 79× UK home rate.
The central, defensible band for a 65+ Western expat in SE Asia who rides a scooter is approximately 15 to 25× home-country exposure. The upper end approaches 50×. The conservative case is still over 15.
| Scenario | Arithmetic | Per 100k / yr | vs home |
|---|---|---|---|
| UK 67yo, Thailand, daily scooter (central) | Arithmetic 25.4 × 2.0 × 2.5 × 1.0 | Per 100k / yr 127 | vs home ~53× UK |
| UK 67yo, Thailand, weekly scooter (conservative) | Arithmetic 25.4 × 1.5 × 2.0 × 0.5 | Per 100k / yr 38.1 | vs home ~16× UK |
| UK 67yo, Thailand, daily scooter, no helmet (worst) | Arithmetic 25.4 × 2.5 × 3.0 × 1.0 | Per 100k / yr 190.5 | vs home ~79× UK |
| US 67yo, Philippines, daily scooter | Arithmetic 9.7 × 2.0 × 2.5 × 1.0 | Per 100k / yr 48.5 | vs home ~3.8× US |
| AU 67yo, Bali, daily scooter | Arithmetic 11.3 × 2.5 × 2.5 × 1.0 | Per 100k / yr 70.6 | vs home ~14.8× AU |
| UK 67yo, Thailand, no scooter (rare passenger only) | Arithmetic 25.4 × 2.0 × 2.5 × 0.2 | Per 100k / yr 25.4 | vs home ~10.6× UK |
Source: WHO 2023, DfT GB 2024, NHTSA 2022, BITRE 2024, triangulated FCDO/Bali Police · checked 2026-05-30
What the contract pays
The contract that sits behind these numbers is the Thai compulsory motor insurance, Por Ror Bor, which every registered vehicle in Thailand carries. The death benefit is THB 300,000 (about USD 8,800 at current rates) if the deceased was not at fault. If at fault, THB 35,000 (about USD 1,030). A single-vehicle scooter crash, the most common foreigner crash pattern, defaults to at-fault. The compulsory cover pays the family of a deceased foreigner roughly USD 1,000.
The voluntary cover that might pay more is mostly unavailable. Allianz and AXA standard travel policies exclude motorbike riding from comprehensive and basic plans; Allianz Adventure Pack is an opt-in add-on. Most travel and expat health policies require a valid motorcycle license plus an International Driving Permit with motorcycle endorsement for any motorcycle-incident cover. Rental shops in Thailand and Bali routinely do not check licenses. Riding a 125cc rental scooter on a car-only home-country license is, in insurance terms, riding uninsured — and the crash that voids the policy is the crash that produces the bill.
The bill that arrives behind the death notice is the repatriation. Body repatriation from Thailand to the UK runs GBP 3,000–6,000 at floor, GBP 8,000–12,000 with embalming, consular paperwork, certificates, and freight class. Body repatriation to the United States or Australia runs USD 10,000–20,000 on long-haul. Cremation locally and repatriation of ashes runs GBP 1,000–3,000. If the underlying incident voids the policy, the family pays out of pocket. That is the standard pattern in FCDO consular case files: a death notice that arrives with a USD 10,000–20,000 bill behind it. The rider who survives instead meets the private-room and ICU day rate, for as many days as it takes.
Contest the model
The model is multiplicative; it lives or dies on the coefficients. Swung one at a time to the ends of their published ranges, with the rest held at centre, the two contested coefficients move the result less than the plainest input in the model — whether there is a daily scooter under the rider at all.
▼ driver at the low end of its range · ▲ at the high end · rust = worse (higher than base) · verdigris = better · one driver moved at a time, the rest held at base
Source: Model composition on the Thai baseline of 25.4/100k (WHO 2023): foreigner factor triangulated from FCDO Hansard 2024 and Bali Police 2024; age factor from Tefft 2011 and SafetyCube D6.1 · checked 2026-05-30
Every assumption, every direction it moves the result:
- Local baseline. WHO Global Status Report 2023, data year 2021. Slow-volatility, country-level published. The Thai 2024 national-police update (17,447 deaths) is consistent with the WHO 25.4/100k against the Thai population. Direction: stable; the central figure may understate Vietnam (where reporting has tightened) and Indonesia (where underreporting is documented).
- Foreigner factor 1.5–3.0×, central 2.0×. Triangulated from FCDO Hansard February 2024 and Bali Police 2024–25. Not a published constant. Lower bound captures the careful, helmeted, daytime-only rider; upper bound captures the tourist-corridor exposure with no helmet and night riding. Direction: doubling from 1.5 to 3.0 doubles the result.
- Age vulnerability 2.0–3.0×, central 2.5×. Composed from Tefft 2011 age-speed (10 mph penalty), SafetyCube D6.1 odds ratios (1.18 at 65–69, 1.42 at 70–74), reaction-time literature (~30% slower under load 65–75), mesopic vision degradation. The upper bound captures helmet non-use compounding (Thai DDC 84% non-use inpatient benchmark). Direction: each decade above 65 pushes the factor toward the upper bound.
- Mode factor. For someone who rides a scooter, set to 1.0; the population rate already includes motorcyclists at 80%+ in Thailand. For someone who does not ride, set to ~0.2 (the non-motorcyclist residual share of road deaths). The factor collapses the difference between a daily rider and an occasional passenger.
- Helmet wear. 84% non-use in the Thai DDC inpatient register implies an additional ~3× fatality risk for the non-helmeted versus helmeted rider. Helmet status, not in the model as a separate coefficient, sits inside the age factor’s upper bound.
- License and insurance status. Riding without a proper motorcycle license voids most travel and expat insurance. This does not change the per-100k probability of death; it changes the post-event cost the family carries. The model arithmetic and the insurance arithmetic are independent layers.
- Home-country baseline. DfT GB 2024 for UK 2.4; NHTSA 2022 for US 12.76; BITRE 2024 for AU 4.78. The ratio is sensitive to which year is read; the central figures are within ±10% of the 2020–2024 average for each.
What the numbers actually argue
They do not argue that you should not ride. They argue that the ride is not the same ride it was at 30. The rental shop hands the same key to the same scooter, but the body that takes it is not the body that took the key in 1992 in Phuket. The road is the same road, more or less. The driver and the bones inside the driver have changed. The actuarial frame the helmet-and-license disclaimers will not put on the rental form is the frame this piece puts on it.