Are SUVs safer than sedans? One dimension has a published number

Almost every answer to this treats safety as a single quantity that one body style either has more of or less of. It is not, and that is why the argument never ends: the people saying SUVs are safer and the people saying sedans are safer are usually talking about different crashes.

There are at least three distinct questions inside the popular one. How likely is the vehicle to lose control and roll. How well does it protect occupants when it is struck. And what does it do to the occupants of whatever it strikes. The first has a published per-vehicle number that almost nobody quotes. The second and third are genuinely contested. Separating them is most of the work.

NHTSA’s 5-Star Safety Ratings include a rollover result, and unusually it is published as a probability as well as a star count: the chance that the vehicle rolls over in a single-vehicle crash. It reflects the thing a buyer intuits about an SUV — a higher center of gravity relative to track width — and turns it into a figure you can compare between two specific vehicles.

Our extract carries that figure for 111 rated configurations across 73 model years. Grouped by body type, with any body type having fewer than 5 rated rows excluded, the pattern is not subtle.

The SUV median sits roughly 65 percent above the sedan median on this measure, and the two ranges do not overlap at all — the least rollover-prone SUV in the set is still above the most rollover-prone sedan. The star ratings say the same thing more bluntly: sedans in this set earn five stars for rollover and SUVs do not, while some SUVs sit at three.

4 further body types are present in the extract with fewer than 5 rated rows — compact, coupe, hatchback, wagon — and they are left out rather than shown, because a median over one or two vehicles is a number pretending to be a finding.

Stated plainly, because a clean-looking table invites more weight than it can carry.

• It is not a fatality rate. It is the published probability of rollover in a single-vehicle crash, which is one crash type among several.
• It is not a national sample. These are the model years we hold detailed records for, weighted toward mainstream vehicles from roughly the last two decades.
• It is not a trim-level figure. NHTSA rates body configurations, so one model can carry more than one row and a specific car may differ from its family's median.
• It says nothing about what happens in a two-vehicle collision, which is where the opposite argument lives.
• Unrated vehicles are simply absent. An older car with no rating is neither safe nor unsafe here — it is untested.

Everything above is a statistic. There is a physical version of it, sitting in the vehicle you are looking at, and you can check it in the time it takes to open the driver’s door.

Federal regulation requires manufacturers of utility vehicles to alert drivers that these vehicles have a higher possibility of rollover than other vehicle types. The mechanism is a label — under 49 CFR § 575.105(d), permanently affixed to either side of the sun visor, at the driver’s seating position, with a yellow heading area carrying black text and an alert symbol, a white message area, and sizing that keeps it legible, visible and prominent to the driver.

That is a striking thing to know while weighing an argument about whether SUVs are safer. The government did not merely publish a probability in a database. It concluded the risk was material enough that the driver should be told about it every time they flip the visor down.

Federal Motor Vehicle Safety Standard 126 requires electronic stability control on light vehicles, and its statement of purpose names the target directly: to reduce deaths and injuries from crashes in which the driver loses directional control, including those resulting in vehicle rollover.

That is the mitigation. Stability control intervenes in the loss of control that precedes a great many rollovers, and it applies to passenger cars, multipurpose passenger vehicles, trucks and buses rated at 10,000 pounds or less. What it cannot do is change the shape of the vehicle — the center of gravity is still where it was.

The case for the SUV is not folklore and should not be treated as such. In a collision between two vehicles of substantially different mass, the occupants of the heavier vehicle are better protected — momentum does not negotiate. That is the basis of the much-quoted comparisons showing worse outcomes for the smaller vehicle’s occupants in car-versus-SUV crashes, and it is real.

Two things are worth noticing about it. It is a statement about a two-vehicle collision, which is a different event from the single-vehicle loss of control the rollover figure describes. And it is a relative advantage that depends on what you collide with, so it shrinks as the vehicle fleet around you gets heavier.

We are not going to adjudicate that here, and pages that claim to usually do it by quoting one study and ignoring the rest. What we can do is put the checkable number next to it and be clear about which question each one answers.

They are not actually in conflict once you stop treating safety as one number. The honest summary is that a tall vehicle is measurably more likely to roll in a single-vehicle crash and measurably better placed in a collision with something lighter, and which of those matters more depends on the driving you actually do.

1. 1If most of your miles are rural, fast and unlitSingle-vehicle loss of control is a larger share of the risk you are exposed to. The published rollover figure is the one to weigh, and stability control is the equipment to insist on.
2. 2If most of your miles are dense, urban and slowMulti-vehicle collisions dominate, and the mass argument carries more weight — as does the visibility and crash-avoidance equipment that has nothing to do with body style.
3. 3Either way, compare the two cars in front of youThe category argument is unresolvable and the vehicle-level figures are published. A well-rated SUV and a poorly-rated sedan exist, and arguing about body styles will not tell you which is which.

• Look up the rollover figure for the exact year, make and model, not the nameplate across all years — the rating attaches to a body configuration.
• Confirm stability control is actually fitted on a vehicle from the phase-in window rather than assuming it from the model year.
• Check the model year's complaint record, because a stability-control system that owners report as faulty is not a mitigation.
• Run the free recall lookup — restraint and stability systems are among the most frequently recalled components, and the repair is free.
• Treat an unrated older vehicle as unknown rather than as either reassuring or alarming.