Motor vehicle rollovers involving passenger cars, vans, pickup trucks, and sport utility vehicles (SUVs) result in approximately 10,000 deaths and 27,000 serious injuries each year in the United States. Although rollover occurs in fewer than 1 in 10 tow-away crashes involving light vehicles,1 these crashes account for almost one-third of light-vehicle occupant fatalities.
The National Highway Traffic Safety Administration (NHTSA) has developed a five-star rating system to inform consumers about the rollover resistance of passenger cars and light multipurpose passenger vehicles and trucks. This system has been incorporated into the New Car Assessment Program. The ratings derive from a correlation between measured values of static stability factor (SSF)2 for a range of vehicles and corresponding rollover rates determined from single-vehicle crash data. Among the 2001 model vehicles currently rated by NHTSA for rollover resistance, most SUVs received two- or three-star ratings, while most passenger cars received four-or five-star ratings (five stars indicates the best vehicle performance and one star the worst).
Congress requested this study of NHTSA’s rollover resistance rating system. Public Law 106-346 required the U.S. Department of Transportation to fund a study “on whether the static stability factor is a scientifically valid measurement that presents practical, useful information to the public, including a comparison of the static stability factor test versus a test with rollover metrics based on dynamic driving conditions that may induce rollover events.” Particular emphasis was to be placed on the potential role of consumer information on vehicle characteristics, in particular SSF, in achieving a reduction in the rollover crash rate and in related deaths and injuries. In response to a request from NHTSA, the Transportation Research Board of the National Research Council established a 13-member committee to conduct the study.3,4
Automobile crashes are complex events involving three main contributing factors and their interactions: the driver, the driving environment (e.g., weather and road conditions, time of day), and the vehicle. The crash data files used by NHTSA to develop its rollover resistance rating system5 include information characterizing the driver and road conditions associated with the crash. This information defines different crash scenarios that can be associated with markedly different risks of rollover. For example, scenarios involving young drivers (under age 25) or those who have been drinking carry a relatively high risk of rollover, as do scenarios involving inclement weather or curves in the road. The important question addressed by this report is the extent to which the vehicle—and in particular its SSF value—affects the risk of rollover for different drivers and driving environments.
As noted, in accordance with the requirements of Public Law 106-346, this study focuses on the potential for reducing the rollover crash rate, as well as resulting deaths and injuries, by providing consumer information related to vehicle characteristics, specifically SSF. It is important to remember that other approaches may be as or more effective. For example, a change in driver behavior leading to increased seat belt use also could result in a reduction in rollover-related deaths and injuries; NHTSA estimates that belted occupants are about 75 percent less likely than unbelted occupants to be killed in a rollover crash. Furthermore, it is essential to ensure that changes in vehicle design leading to a reduction in one contributor to overall vehicle risk—such as rollover—do not compromise other aspects of vehicle safety. Many complex risk trade-offs need to be considered in pursuing the ultimate goal of improved road safety.
The committee’s findings regarding SSF and NHTSA’s star ratings for rollover resistance are presented below.
Static Stability Factor
NHTSA, vehicle manufacturers, and others have used various static measures and driving maneuvers to characterize the rollover behavior of vehicles. In developing its consumer information on rollover, NHTSA selected SSF as an indicator of rollover propensity in single-vehicle crashes. This decision resulted in part from the ability to measure SSF directly for vehicles, and in part from
the statistical correlation between observed crash outcomes (rollover or no rollover) and SSF.
In developing its rating system, NHTSA undertook statistical studies to investigate the relationship between measured values of SSF for a range of vehicles and corresponding rollover rates determined from crash data. The agency reviewed crash frequencies and rollover rates, and used data from six states, selected as representative of national trends, for regression analyses using an exponential statistical model.6 At the request of the committee, NHTSA conducted additional analyses of these crash data using a logit statistical model. The agency computed separate rollover curves and associated confidence bands for different crash scenarios.7 These curves indicate that an increase in SSF reduces the probability of rollover in the event of a single-vehicle crash, although this trend is less pronounced for lower-risk scenarios.
Through a rigid-body model, SSF relates a vehicle’s track width, T, and center of gravity height, H, to a clearly defined level of the sustained lateral acceleration that will result in the vehicle’s rolling over. The rigid-body model is based on the laws of physics and captures important vehicle characteristics related to rollover.
Analysis of crash data reveals that, for higher-risk scenarios, SSF correlates significantly with a vehicle’s involvement in single-vehicle rollovers, although driver behavior and driving environment also contribute. For these scenarios, the statistical trends in crash data and the underlying physics of rollover provide consistent insight: an increase in SSF reduces the likelihood of rollover.
Metrics derived from dynamic testing are needed to complement static measures, such as SSF, by providing information about vehicle handling characteristics that are important in determining whether a driver can avoid conditions leading to rollover.
NHTSA’s Star Ratings for Rollover Resistance
NHTSA derived its star ratings for rollover resistance using an exponential statistical model and regression analysis correlating SSF with crash data. The agency based the ratings on an average rollover curve calculated using a dataset comprising single-vehicle crash data from six states. This average rollover curve gives the rollover risk—defined as the probability of rollover in the event of a single-vehicle crash—for each value of SSF, assuming an average scenario. Data on driver and environmental variables were used in estimating the curve. When developing its ratings, NHTSA did not consider the uncertainty in the average rollover propensity curve as reflected in the associated confidence bands.
NHTSA partitioned the average rollover curve into five regions, based on the rollover probability in the event of a single-vehicle crash. If a vehicle’s SSF corresponds to a rollover probability range of 0–10 percent, as defined by the average rollover curve, the vehicle is assigned five stars. If its SSF corresponds to a rollover probability range of 10–20 percent, it is assigned four stars, and so on. If the vehicle’s SSF indicates that—according to the average rollover curve—it has a rollover probability of greater than 40 percent in the event of a single-vehicle crash, it is assigned a one-star rollover resistance rating. Rollover curves generated from crash data represent an average over many different vehicle makes and models. Therefore, a data point representing the probability of rollover in a single-vehicle crash for a given vehicle make or model may fall above or below the curve.
NHTSA used two series of consumer focus group studies to develop and evaluate its star rating system. The first series addressed rollover and the effects of information about rollover on consumers. In the second series, consumer awareness and understanding of rollover problems were explored, and consumer comprehension of two potential texts aimed at explaining the agency’s rollover resistance ratings was evaluated. The rollover information on NHTSA’s website has attracted interest, as indicated by site use statistics. However, no empirical data on consumers’ use of the ratings were available to the committee.
NHTSA’s implementation of an exponential statistical model lacks the confidence levels needed to permit discrimination among vehicles within a vehicle class8 with regard to differences in rollover risk.
The relationship between rollover risk and SSF can be estimated accurately with available crash data and software using a logit model. For the analysis of rollover crash data, this model is more appropriate than an exponential model.
The approximation of the rollover curve with five discrete levels—corresponding to the five rating categories—is coarse and does not adequately convey the information provided by the available crash data, particularly at lower SSF values, where the rollover curve is relatively steep.
There is a gap between recommended practices for the development of safety information and NHTSA’s current process for identifying and meeting consumer needs for such information. In particular,
The focus group studies used to develop the star rating system were limited in scope.
The agency has not undertaken empirical studies to evaluate consumers’ use of the rollover resistance rating system in making vehicle safety judgments or purchase decisions.
The committee has synthesized its findings in the areas of vehicle dynamics, statistics and data analysis, and consumer information into two summary findings that respond to the congressional mandate for this study.
Summary Finding 1
SSF captures important vehicle characteristics related to rollover propensity and is strongly correlated with the outcome of actual crashes (rollover versus no rollover), as demonstrated by statistical analyses of crash data. Data from dynamic testing could provide important information on vehicle crash-avoidance metrics that would complement static measures.
Summary Finding 2
NHTSA’s star ratings for rollover resistance are likely to be of limited use in presenting practical information to the public because
There were shortcomings in the statistical methodology used to derive the average rollover curve.
The approximation of the rollover curve by five discrete rating categories is coarse and does not adequately convey the degree of resolution among vehicles provided by available crash data.
The limited procedures used by NHTSA to develop and evaluate the star rating system do not demonstrate with reasonable confidence the likely effectiveness of the system.
RECOMMENDATIONS FOR A FUTURE APPROACH
The committee concludes that consumer information on motor vehicle rollover can assist the public in choosing safer cars and encourage manufacturers to investigate ways of making vehicles less susceptible to rollover. To be comprehensive, such information needs to capture
The results of dynamic tests that assess a vehicle’s control and handling characteristics, and
Information from static measures indicative of a vehicle’s rollover propensity.
In accordance with the requirements of the Transportation Recall Enhancement, Accountability, and Documentation (TREAD) Act (Public Law 106-414), NHTSA is investigating several driving maneuver tests for rollover resistance. Challenges remain in developing the requisite dynamic tests, together with related consumer information that is technically accurate, as well as practical and useful to the public. Nevertheless, the committee has not identified any insurmountable engineering barriers to the development of a representative dynamic test (or tests) that would differentiate meaningfully among vehicles. Similarly, the development of consumer information based on static measures and dynamic tests appears feasible, particularly if NHTSA takes advantage of recommended development practices and proven techniques for communicating risk-based information to consumers.
Despite the absence of technical barriers to providing more comprehensive consumer information on rollover, the protracted history of NHTSA’s rulemaking initiatives on rollover suggests that the agency may encounter difficulties in obtaining support for its actions from all the major stakeholders. Furthermore, vehicle manufacturers, consumer groups, and others involved in vehicle testing are likely to incur additional costs when NHTSA introduces dynamic testing related to rollover. For these reasons, the committee concludes that consumer information on rollover that captures both static measures and dynamic test results probably will not be available in the near future.
The current rollover resistance ratings are likely to be of limited use to the public because of the way in which information on SSF is delivered. However,
SSF may form a reasonable initial basis for developing consumer information on rollover until additional measures based on both static metrics and dynamic testing become available.
NHTSA should vigorously pursue its ongoing research on driving maneuver tests for rollover resistance, mandated under the TREAD Act, with the objective of developing one or more dynamic tests that can be used to assess transient vehicle behavior leading to rollover.
In the longer term, NHTSA should develop revised consumer information on rollover that incorporates the results of one or more dynamic tests on transient vehicle behavior to complement the information from static measures, such as SSF.
NHTSA should investigate alternative options for communicating information to the public on SSF and its relationship to rollover. In developing revised consumer information, NHTSA should
Use a logit model as a starting point for analysis of the relationship between rollover risk and SSF.
Consider a higher-resolution representation of the relationship between rollover risk and SSF than is provided by the current five-star rating system.
Continue to investigate presentation metrics other than stars.
Provide consumers with more information placing rollover risk in the broader context of motor vehicle safety.