A single-vehicle rollover reads like a story with one character, the driver who left the road. But a rollover is a loss of control that began somewhere, and the first question is not who was driving. It is what the tire met when it tried to come back. Often the answer was waiting at the edge of the pavement.
The pattern. A right front tire drifts a few inches off the lane and onto the shoulder. Where the asphalt ends it can stand above the dirt in a vertical lip, and the driver who steers back is asking the tire to climb it. The side of the tire rubs along the lip instead, the driver turns the wheel harder, and when the tire finally mounts the pavement the vehicle goes where the wheel is pointed: across the lane, or into a slide. The crash report gives all of that one word, overcorrected. This page names no one, is pegged to no crash, and decides nothing about any of them. It is here because the way a vehicle came back into its lane is often the whole case. The Swerve told it with an animal in the road, Rolled is the rollover itself, and The Clear Zone is what waits past the shoulder.
The edge. The Federal Highway Administration does not treat this as driver folklore. “Exposed vertical pavement edges can cause vehicles to become unstable and prevent their safe return to the roadway,” it says, and “Rural road crashes involving edge drop-offs are 2-4 times more likely to include a fatality than other crashes on similar roads,” a figure it takes from a 2006 study published by the AAA Foundation for Traffic Safety. No one has to build the lip for it to be there: “Over time, regardless of the edge type, the edge may become exposed due to settling, erosion, and tire wear.” The agency’s answer is a shape, a paving method that “shapes the edge of the pavement at approximately 30 degrees” and so “gives drivers the opportunity to return to their travel lane while maintaining control of their vehicle,” and it credits that shape with an 11 percent reduction in fatal and injury crashes and a 21 percent reduction in run-off-road crashes, figures it draws from a 2017 study of two-lane rural roads (FHWA, SafetyEdge). The rubbing has a name, tire scrubbing, “a phenomenon that contributes to the loss of vehicular control,” and the agency’s summary of the research is one sentence long: “Studies have shown that severe crashes often occur when drivers lose control attempting to traverse a near vertical pavement edge to reenter the paved roadway.” The other half of the finding is the hopeful one: “Research has shown virtually all drivers can recover, even at high speeds, if the edge shape is 30 degrees rather than vertical” (FHWA, SafetyEdge: Frequently Asked Questions).
Why it rolls. The federal safety agency describes both ways the maneuver ends. One happens on the pavement: “NHTSA analysis indicates that an untripped rollover in a single-vehicle crash is likely to occur when a driver tries to regain the lane position after dropping two wheels off the roadway onto the shoulder,” untripped meaning that “the tire/road interface friction is the only external force acting on a vehicle.” The other happens off it, and it is the common one. By the agency’s analysis “approximately 95 percent of rollovers in single-vehicle crashes are tripped,” and “A tripped rollover event occurs when a vehicle runs off the road and is tripped by a ditch, soft soil, a curb or other object causing a vehicle to roll over” (NHTSA, DOT HS 812 444). A vehicle that darts across the road and off the far side has found the ditch. What the vehicle is matters too. The agency’s measure of top-heaviness is the static stability factor, built from “track width and center of gravity height,” and “A vehicle with a higher SSF value is considered to be more geometrically stable and less top-heavy.” Over model years 2004 to 2013, it reports, “the weighted average of SSF in pickup trucks worsened from 1.21 to 1.18”; its table puts passenger cars at 1.41. And when a rollover throws a person out, the odds are the worst a crash offers, which is Thrown Clear and The Roll.
What the law asks. Overcorrected is where the report stops. The case opens three questions there.
The driver. Texas requires a speed that is “reasonable and prudent under the conditions and having regard for actual and potential hazards then existing,” and a reduced one when “the operator is traveling on a narrow or winding roadway” or “a special hazard exists with regard to traffic, including pedestrians, or weather or highway conditions” (Transportation Code Section 545.351). The reaction itself is measured as a reaction. The instruction Texas courts may give on a sudden emergency, when the evidence raises it, covers a danger “arising suddenly and unexpectedly, which was not proximately caused by any negligence on his part and which, to a reasonable person, requires immediate action without time for deliberation,” and it tells the jury the driver’s conduct is not negligence if afterward “he acts as a person of ordinary prudence would have acted under the same or similar circumstances” (Dillard v. Texas Electric Cooperative, quoting the pattern instruction). The limit is in the sentence, and it cuts both ways: the drift that put the tire on the shoulder has to be accounted for too, which is Sudden Emergency. And many of the people hurt in a rollover did not steer. A passenger’s claim is the passenger’s own.
The belt. “For more than forty years evidence of a plaintiff’s failure to use a seat belt has been inadmissible in car-accident cases,” the Texas Supreme Court wrote in 2015, and then it changed the rule: “relevant evidence of use or nonuse of seat belts is admissible for the purpose of apportioning responsibility in civil lawsuits.” The condition is proof. “The defendant can establish the relevance of seat-belt nonuse only with evidence that nonuse caused or contributed to cause the plaintiff’s injuries,” and “the trial court should first consider this evidence, for the purpose of making its relevance determination, outside the presence of the jury” (Nabors Well Services v. Romero). So an ejection has to be answered with evidence, not assumed, and what the evidence shows goes into the percentages: a claimant recovers nothing “if his percentage of responsibility is greater than 50 percent” (Civil Practice and Remedies Code Section 33.001), and below that line the award is reduced by the claimant’s share (Section 33.012). That arithmetic is Comparative Fault, and The Belt is the earlier chapter.
The road. A road has a keeper, the state, a county or a city, and Texas lets the keeper be sued for a death “caused by a condition or use of tangible personal or real property if the governmental unit would, were it a private person, be liable to the claimant according to Texas law” (Section 101.021). For an ordinary defect in a road the government owes “only the duty that a private person owes to a licensee on private property,” which means proving it actually knew of the danger. That limit “does not apply to the duty to warn of special defects such as excavations or obstructions on highways, roads, or streets” (Section 101.022), where it is enough that the government should have known. The Texas Supreme Court has pointed to a mark on each side of that line. Explaining that “Special defects unexpectedly and physically impair a car’s ability to travel on the road,” it pointed to a Beaumont decision “holding that ten-inch drop-off along shoulder that prevented car’s left wheels from reentering the roadway once they had slipped off was a special defect” (State v. Rodriguez, citing Morse v. State). Nine years later it held that “a two-inch variance in elevation between traffic lanes is not a special defect,” because “Ordinary drivers, in the normal course of driving, should expect these slight variations on the road caused by normal deterioration” (City of Dallas v. Reed). Between two inches and ten, the tape measure is the case, and Special Defect is the entry on how the courts read it. Two limits come with the claim. The design of the road is not it: “Design of any public work, such as a roadway, is a discretionary function involving many policy decisions, and the governmental entity responsible may not be sued for such decisions” (Rodriguez). What the road was allowed to become is. And the clock is short: the government “is entitled to receive notice of a claim against it under this chapter not later than six months after the day that the incident giving rise to the claim occurred,” and a city’s charter can require notice sooner, within a charter period “permitted by law” that the statute ratifies (Section 101.101).
What the proof looks like. It is at the edge, and the edge does not keep. The scuff along the face of the pavement where the tire rubbed, and the marks of the slide that followed. The height of the drop, measured along the path the tire took, before the shoulder is graded or the edge is patched. The road’s file: the maintenance records, the complaints and the earlier crashes on that stretch, because what the government knew or should have known is the element. In the two-inch case the court noted that “there were no reported accidents caused by the two-inch drop-off prior to” the crash “or any complaints about this road condition.” The vehicle’s recorder, for the speed and the braking in the last seconds. The belts and the seats, examined before the vehicle is sold for scrap. None of it waits, which is why the first letter goes out before the road crew does (Spoliation). The highway agency’s FAQ gives the reason this has to be done on purpose: “Crash data rarely detects pavement edge drop-offs.” The report will say one vehicle, and overcorrected.
No one is named here, and nothing on this page decides any one crash. A dispatch keeps the record of a pattern: a single-vehicle rollover is not a case with one character, and overcorrected is a description, not a verdict. The tire found the edge first. The question the law asks is who should have found it before.
Dispatch is the Night Docket’s news strand: what happened on Texas roads and in Texas courts, reported from credible outlets and agency releases, with no names and no conclusions about open investigations, then connected to the entries that explain the science and the law. General information, not legal advice.
