A $30 Part Was Missing From 9 Million Cars. FARS Data Shows the Fallout.

An engine immobilizer costs roughly $30 at the component level. It’s a chip in the key that talks to a chip in the car. If they don’t match, the engine won’t start. By 2015, 96 percent of automakers included one as standard equipment.^[1] Hyundai and Kia were not among them. Only 26 percent of their vehicles had the feature.

That gap sat dormant for a decade. Then in July 2021, a Milwaukee teenager posted a TikTok showing how to steal a Kia with a USB cable and a screwdriver. Peel off the plastic steering column shroud. Jam the USB into the exposed ignition cylinder. Turn. Drive away.^[2] The video spread under the hashtag “Kia Boys.” Theft insurance claims for affected Hyundai and Kia models exploded from roughly 1 per 1,000 insured vehicles in early 2020 to 11.2 per 1,000 by the first half of 2023 — a 1,000-percent increase, per the Highway Loss Data Institute.^[3]

NHTSA’s official tally: 14 crashes and 8 fatalities linked to stolen Hyundai and Kia vehicles.^[4] That number is almost certainly an undercount. FARS — the federal database that logs every fatal crash in the United States — doesn’t reliably flag whether a vehicle was stolen at the time of collision. The real signal hides in the model-year death curves.

The Anomaly in the Data

Normally, fatal crash involvement follows a predictable depreciation curve: new model years accumulate deaths slowly (fewer vehicles on the road, lower mileage), peak around age 5–8, then decline as vehicles are scrapped. The affected Kia and Hyundai models break this pattern.

Take the Kia Forte. Its 2021 model year — which should be in its safest period, with the newest safety technology and lowest accumulated mileage — logged 76 deaths in FARS. That’s higher than its 2019 model year (56 deaths) and 2020 model year (50 deaths). The only Forte vintage with more fatal involvements is the 2017 (81 deaths), which had six additional years of road exposure.^[5]

The Kia Optima shows the starkest shift. Model years 2005 through 2010 averaged 12.7 fatal crash involvements per vintage. Model years 2015 through 2020 — squarely in the immobilizer-absent window — averaged 58.3. That’s a 4.6x increase, the largest model-year escalation of any vehicle in the FARS database.^[5] The 2015 Optima alone accounted for 110 deaths.

The Kia Soul follows a similar trajectory: 78 deaths for the 2016 model year, and the 2020 vintage still registering 53 — abnormally high for a two-to-three-year-old subcompact.^[5]

The Math Behind the Missing Part

The calculation is straightforward. Affected models span 2011–2021 Kia (Optima, Soul, Sorento, Forte, Sportage, Rio) and 2015–2021 Hyundai (models equipped with traditional steel keys rather than push-button start). That covers approximately 9 million vehicles on American roads.^[2]

Across all Kia and Hyundai models in FARS, model years 2015–2020 accumulated 3,303 fatal crash involvements. For comparison, Toyota — with a substantially larger fleet — accumulated 3,440 for the same model-year bracket.^[5] On a per-vehicle basis, Kia/Hyundai are punching well above their fleet share in the fatality column.

An immobilizer wouldn’t have prevented every one of those deaths. Most fatal crashes involving these vehicles had nothing to do with theft. But the anomalous model-year death spikes — newer cars dying at rates typically reserved for vehicles 5–10 years older — suggest a meaningful fraction traces back to that missing $30 chip.

What NHTSA Didn’t Do

Safety Research & Strategies petitioned NHTSA to revise the compliance test for Federal Motor Vehicle Safety Standard 114, which governs theft protection. The agency declined.^[6] Attorneys general from 18 states requested a formal recall of the affected vehicles. NHTSA determined the issue did not meet recall criteria.^[4]

Instead, Hyundai and Kia issued what they called a “customer satisfaction campaign” — a software update that extends the burglar alarm from 30 seconds to one minute and adds an ignition kill feature. The catch: it only activates when the alarm is armed, meaning the doors must be locked with the key fob. Thieves who enter through a broken window — the primary method — bypass this entirely.^[6] Owners of vehicles ineligible for the software fix received a reimbursement for a steering wheel lock.

Limitations

FARS does not track whether a vehicle was stolen at the time of a fatal crash. The anomalous model-year death spikes we identify could reflect confounders: higher sales volumes in certain years, different buyer demographics, or geographic distribution patterns. The 4.6x increase for the Kia Optima almost certainly includes a fleet-growth component — Kia sold substantially more Optimas in the 2015–2020 window than in 2005–2010. We cannot isolate the theft-attributable fraction from the overall fatality signal without per-year registration-weighted rates, which FARS does not provide at the model-year level.

Counterargument

The strongest case against this thesis: Kia and Hyundai’s market share grew dramatically between 2010 and 2020. More cars on the road means more fatal crashes, regardless of theft. The Optima’s 4.6x death increase may simply reflect a 3–4x increase in sales volume, with theft contributing a marginal additional bump. NHTSA’s own analysis reached a similar conclusion — they found the theft-to-fatality pathway, while real, was statistically small relative to the overall crash population. The 14 confirmed theft-related fatalities, if accurate, represent 0.04% of total Kia/Hyundai fatal involvements in FARS during this period.

That framing only holds if you accept NHTSA’s confirmation methodology. FARS records over 36,000 fatal crashes annually. The agency’s 14-crash count relies on police reports that specifically noted the vehicle was stolen — a field that is inconsistently coded across jurisdictions. The true theft-related death toll almost certainly falls somewhere between NHTSA’s conservative 8 and the inflated signal in the raw model-year curves.