GM Killed the Spark and Sonic. The Replacement Cuts Fatal Crashes 83%.

The Chevrolet Spark was the cheapest new car you could buy in America at $13,600, a 2,246-pound city car with a crash lethality of 74.4% in the FARS database.^[1] Three out of four fatal crashes involving a Spark killed someone, which is less a safety rating than a coin flip with a thumb on the wrong side.

GM axed the Spark after 2022 and the Sonic after 2020. Both subcompact sedans carried lethality rates above 74% in NHTSA's fatal crash data.^[1] Sonic: 494 deaths, 306,000-unit fleet, 1.40 per 100 million VMT; Spark: 517 deaths, 350,000 vehicles, rate of 1.28. Combined: 1,011 dead in two of the lightest production cars on American roads.

Enter the second-generation Chevrolet Trax, a subcompact crossover that now sits in GM's cheapest slot at $21,700.^[2] It weighs 2,983 pounds. Over 200,000 sold last year.^[3] FARS death rate: 0.22 per 100 million VMT, with 53.8% crash lethality.^[1] Double the fleet of the Spark, 63% fewer deaths, and an 83% improvement in the death rate.

Now for the part that should bother every engineer who touched the Spark's crash structure: impairment was not the problem. FARS toxicology shows Spark drivers tested positive for alcohol or drugs in just 16.2% of fatal crashes, lowest of any GM small car in the dataset.^[1] Cavalier sat at 22.4%, Cobalt at 22.4%, and even the Trax itself runs higher at 21.5%. Spark buyers were the most sober drivers in GM's small-car stable, and they died anyway, because 2,246 pounds of sedan meeting a 5,000-pound pickup at highway speed is not a negotiation.

GM's small-car lineage tells the same story across three decades of escalating body counts: Cavalier (1,225 dead, 85.7% lethality), Cobalt (1,540 dead, 80.8%), Aveo (138 dead, 89.0%), HHR (534 dead, 80.1%).^[1] Each replacement inherited the price bracket and the kill rate. Only the Cruze broke pattern, dropping to 56.3% with a larger unibody. Now the Trax adds 740 pounds of crossover architecture, a higher seating position, and geometry that spreads impact loads across more metal. GM never advertised any of that as a safety feature. An 83% death-rate reduction was a side effect of selling entry-level buyers a vehicle heavy enough to survive the American fleet.

But 350,000 Sparks and 306,000 Sonics remain registered.^[4] MY2021 Sparks already show 22 FARS fatalities, and production may be closed, but the ghost fleet rolling America's highways is very much not.

If you own a Spark or Sonic: check your VIN at nhtsa.gov/recalls for outstanding safety recalls. If you are shopping for a replacement, the Trax, Hyundai Venue, and Nissan Kicks all occupy the same price bracket with dramatically better crash outcomes per FARS data. An extra $8,000 at purchase buys something no extended warranty covers: surviving a crash.

Limitations

FARS captures only fatal crashes. A vehicle with low fatal-crash involvement may still have elevated injury or total-loss rates not reflected here. Death rates use VMT estimates from the National Household Travel Survey, introducing approximately ±15% uncertainty for lower-volume models. The Trax's second-generation redesign (MY2024+) has limited FARS exposure; its current numbers largely reflect the first-generation model. Fleet registration estimates are point-in-time and may not match current DMV registrations. Lethality comparisons across vehicle classes are influenced by collision partner mix: heavier vehicles in multi-vehicle crashes tend to show lower occupant fatality rates partly because they transfer kinetic energy to lighter collision partners.

Strongest Counterargument

The Spark and Trax serve different buyer demographics and are driven in different environments. Spark buyers skewed younger and more urban; the Trax attracts suburban families. These populations face different crash exposure profiles regardless of vehicle. The 83% rate difference likely overstates the pure vehicle-design contribution because it conflates who drives these cars with what happens when they crash. A controlled study matching driver age, geography, and driving patterns would likely show a smaller gap. That said, the sheer weight and geometry differences between a 2,246-lb sedan and a 2,983-lb crossover are real structural factors in crash survivability, supported by decades of IIHS research on vehicle size and weight.^[5]