WASHINGTON - By all accounts, they had done everything right.
Setting out from Osburn, Idaho, at the end of the Memorial Day weekend three years ago to return home to Portland, Ore., Elizabeth Sanders and her cousin Linda McCathern were thinking safety.
Seat belts buckled. The two kids, ages 1 and 3, in safety seats. Reg-ular stops to avoid fatigue.
And, in a crucial decision made before the weekend trip, the women had decided to take Sanders' 1994 Toyota 4Runner instead of McCathern's Honda Accord in the belief that the sport-utility would be safer.
'Compared to mine, it just seemed like a safer vehicle,' McCathern would say later - from her wheelchair.
That belief was shattered in a millisecond on a two-lane highway in eastern Washington state. Forced to swerve suddenly to the right to avoid an oncoming vehicle that had drifted into their lane - a reaction known in safety-speak as a 'radical avoidance maneuver' - Sanders managed to keep the 4Runner upright and on the roadway.
But on the counterswing back to the left, the 4Runner rolled. In a flash, McCathern, a 36-year-old mother of three, became a quadriplegic and one of the more than 165,000 Americans killed or injured in rollover accidents every year.
Of that total, according to federal data, more than 10,000 are fatalities.
Government, industry and safety groups have wrestled with the rollover issue for decades. But as the McCathern case demonstrates, the need for a solution has escalated along with the number of people who are choosing sport-utilities for all-around family transportation.
TENDING TO TIP
While the focus has been on car-truck 'compatibility' - that is, whether sport-utilities are a threat to smaller vehicles on the roads - rollover, too, has become a major issue because of the trucks' higher center of gravity.
Indeed, some statistics show that McCathern and her cousin were in fact less safe in the 4Runner, not more safe, because of its relatively higher tendency to tip. According to federal data as crunched by the Insurance Institute for Highway Safety, five people are killed in rollover accidents in mid-sized sport-utilities like the 4Runner for every two who die in mid-sized cars like the Accord.
But in all kinds of accidents, the death rate is slightly lower for the sport-utilities than for the sedans.
Galvanized by the rising toll of rollover accidents, safety groups and some politicians are increasingly pressing government regulators to set a firm standard that would prevent the sale of vehicles that are prone to roll.
Jeffrey Foote, who represented McCathern in a suit against Toyota over the 1994 4Runner, said the absence of a rollover standard makes it harder for victims because 'defendants hammer home that they meet every government regulation.'
Still, an Oregon jury awarded McCathern $7.6 million last year. Toyota has appealed the decision.
Larry Setchell, a veteran product liability lawyer in Washington state, believes most juries are smart enough to see through the no-government-regulation defense, but he'd still like to see a standard.
'That would help everybody,' said Setchell. If the bar were raised a little higher, manufacturers would rise to the challenge. Fewer people would get hurt or killed. Attorneys could move on to other problems, he said.
But a quick resolution appears unlikely.
NHTSA TRIES AGAIN
The National Highway Traffic Safety Administration, which has repeatedly tried - and failed - to come to grips with the rollover issue by setting an objective standard for vehicles to meet, is trying again.
The agency last year began new research aimed at finding or developing a reliable, repeatable, dynamic test for rollover tendencies. But already the effort is under strain, especially from automakers, which say it is off the mark.
'You can roll over any vehicle - if you turn fast enough and abruptly enough,' says Barry Felrice, director of regulatory affairs for the Big 3's American Automobile Manufacturers Association.
No safety issue has bedeviled manufacturers, regulators and outside safety advocates the way rollovers have. In fact, the modern safety movement was born more than three decades ago in part because of an alleged rollover problem.
Ralph Nader's landmark 1965 book, Unsafe at Any Speed, said the rear-engine compact Chevrolet Corvair was prone to flip.
NHTSA, which grew out of the federal safety agency that was created the next year, has tried repeatedly to grasp and resolve the rollover issue. It also has failed repeatedly.
'The agency has probably spent 25 years on this problem,' acknowledged Gary Woodford, NHTSA's group leader for the rollover research program. 'I think we've learned a lot.'
NHTSA's latest rollover research, begun last year, has the trappings of a do-or-die effort.
Steve Kratzke, director of crash avoidance for NHTSA, said the research will result in proposals for a new safety standard or consumer information program - or else, research will end.
A safety standard would have the force of law. Consumer information would be similar to or even part of NHTSA's New Car Assessment Program, which puts ratings on vehicles for their performance in frontal and side-impact crash tests.
NHTSA, if it succeeds, would be plowing new ground.
No government in the world currently uses dynamic testing to regulate against rollover tendencies by cars and light trucks, industry and federal officials said. NHTSA's goal is to reach a conclusion on a dynamic test for 'untripped' rollovers - those not caused by hitting an object or running off the road - by the end of the year.
Evidence is mounting, however, that the undertaking is headed for a controversial conclusion. Almost certainly, response to the outcome will be far different from the near-universal support that greeted the agency's grand compromise for defusing last year's airbag hysteria.
Even Kratzke acknowledges, 'We don't think there is any single test that's going to be the silver bullet.'
Largely, that was the original goal of the project: to find a repeatable, reliable test that could be used on all kinds of vehicles to determine which, if any, are unsafely prone to roll over.
THE HUMAN FACTOR
At the heart of the controversy is the role of humans in rollover crashes, raising the question of whether vehicle handling can be evaluated without driver involvement.
NHTSA officials are using an automated steering controller - a virtual robotic arm - to study various handling maneuvers.
'We are trying to take the best of what's out there and not invent a new test,' Kratzke said.
Once the field of options has been narrowed, researchers intend to try the tests on 12 vehicles, three each from among cars, pickups, minivans and sport-utilities. The selection criteria were high sales numbers, handling characteristics that differed from one vehicle to another, and no significant design changes in three years.
Testing is to begin this month.
The NHTSA researchers have talked with engineers at the Big 3, Mercedes-Benz and Toyota about the tests they use in designing vehicles. They have met with Con-sumers Union officials and staff members and examined that company's now-infamous double-lane-change emergency-avoidance test.
It's the one that led to the 1995 and 1996 Isuzu Trooper II and the 1988 Suzuki Samurai being labeled 'unacceptable.'
Carmakers consider information about their handling tests to be proprietary and don't reveal specifics to reporters, but engineers say that manufacturers subject their vehicles to batteries of tests for handling, steering and ride.
Fred Heiler, spokesman for Mer-cedes-Benz of North America Inc., said there is no shortcut in the development process that would work better than on-the-road testing.
But he also acknowledged that the 'moose test' - a severe double-lane-change maneuver used in Europe - has become a de facto rollover standard there.
After a Swedish auto magazine rolled one of Mercedes' new A-class cars during the moose test, parent Daimler-Benz AG spent some
$200 million recalling and re-engineering the cars to enhance their stability.
Robert Lange, General Motors' executive director of safety engineering, said some of his company's routine tests eventually push vehicles beyond their design limits. But he added, 'I can tell you we don't do a deliberate test for rollover propensity.'
He said engineers all over the world include double lane changes among their tests, but he believes the maneuvers are too driver-dependent to be incorporated by NHTSA in its research.
In fact, the agency appears headed in other directions.
Kratzke confirmed NHTSA researchers are focusing - though apparently not exclusively - on the so-called J-turn and on the fish-hook maneuver used by Toyota. The names accurately describe the paths vehicles follow in each test.
The NHTSA focus troubles the people at Consumers Union, who have made rollovers their cause celebre in the arena of vehicle safety.
'I worry that the test needs to be able to perform a task that a consumer may have to go through, and at the moment I can't think of who would go through a J-turn or a fish hook,' said David Pittle, vice president and technical director of Consumers Union, publisher of Consumer Reports magazine.
Robert Knoll, the semiretired auto testing director for Consumers Union, criticized the J-turn because it does not include a reversal in steering. It is the sharp reversal that creates the danger of rollover in everyday driving, he explained.
'Our position is that no vehicle should roll because of steering input alone,' Knoll said.
Pittle said Consumers Union's double-lane-change test 'presents the same challenge as a real situation,' such as a child running into the street or a box falling off a truck.
Consumers Union officials also said NHTSA is wrong to be obsessed with eliminating test drivers from the research. The agency still uses people to test brakes and other vehicle functions, they noted.
At a former drag strip on a still-rural mountain plateau in central Connecticut, Consumers Union puts all the vehicles it evaluates through a full range of tests and inspections.
On the emergency-avoidance double lane change, Consumers Union officials say they negate human influence on results by having several drivers take a vehicle through the course as many as 15 times each before drawing a conclusion.
Of 66 vehicles tested, only two - the Samurai and the Trooper II - have been rated 'unacceptable.' The number is three if the Trooper's twin, the 1996 Acura SLX, is counted.
Only a handful of others were called 'poor.' They were the 1989 Bronco II, the 1990 Nissan Axxess and the 1998 Chevrolet Blazer.
'A BAD CROWD'
Still, Pittle broad-brushes sport-utilities as 'a bad crowd,' citing unsatisfactory handling characteristics. But with enough focus, he contends, government regulators could find a way to discriminate the better handlers from the worse handlers.
'We did not ask (NHTSA) to adopt our test,' Pittle said. 'We did not say, 'Do a double lane change.' We just said, 'Do a test that will discriminate among SUVs in particular.' '
After the Samurai tests in 1988, Consumers Union petitioned NHTSA for a rollover standard. But following six years of study, NHTSA said it could not come up with one.
Now, Consumers Union officials say an information program is all they want, on the grounds that this would stand a better chance of succeeding than a firm standard.
'We know consumers care about safety information. Then the information alone could be a driver for getting manufacturers to make safer cars,' Pittle said.
Gerald Donaldson, senior research director at Advocates for Highway and Auto Safety, a group that still wants a rollover standard, predicted a consumer information program is the most that can be expected.
'This is a White House that really doesn't want to regulate vigorously,' he said. And Vice President Al Gore is even less inclined, he added. 'They will be throwing a bone to the dogs. We are the dogs, the safety community,' Donaldson said.
Lange of GM said that, based on hard facts, NHTSA probably should terminate the research. But that is not likely to be feasible politically.
He said it is unlikely but not inconceivable that NHTSA will try a safety standard, but it might have trouble meeting requirements of the federal safety act. Those requirements say that a rule must have an impact on a problem that exists in the real world and otherwise poses an unreasonable safety risk.
'I think they are in a tough spot, I really do,' Lange said. 'What we hope for is a rational review of the research and a scientific response. If we get half a loaf, I hope that half-loaf is not half-baked.
'If there is a consumer information solution, it ought to emphasize the role of drivers, driver behavior, in rollovers,' Lange said. It should include the message, 'You the driver control your own destiny,' he said.
'All kinds of people drive tall vehicles with narrow track widths safely all the time. Others drive low vehicles with wide tracks unsafely all the time,' he said.
Indeed, insurance industry records show the low-slung, wide-track Chevrolet Corvette has one of the industry's worst rollover death records. In an analysis of 1991-95 federal crash data, the Insurance Institute for Highway Safety says that 1.2 rollover deaths occur per year for every 10,000 Corvettes registered.
But, some sport-utilities have even worse records. According to the insurance institute analysis, the Toyota 4Runner and the Nissan Pathfinder registered death rates of 1.3 per 10,000, while the two-wheel-drive Geo Tracker led with a rate of 1.6 deaths per 10,000. (There were no comparable data for the Tracker's twin, the Suzuki Sidekick.)
GM believes one answer to the rollover problem is to encourage customers to equip their vehicles with automatic stability systems and to develop more advanced systems.
RULES, POLITICS, LAWSUITS
Despite the either-or tone of the debate, NHTSA has taken several steps toward mitigating the effects of rollovers and is considering others.
Rules are pending to require interior protection for motorists' heads, and the agency is looking at tougher rules on roof crush resistance, side-door requirements and even better windows to keep people from being ejected, Woodford said.
And NHTSA has proposed new rules for warning labels about sport-utilities' rollover potential. But, said Consumers Union's Knoll, 'You can't sticker over a safety problem.'
Likewise, trial attorney James Butler of Columbus, Ga., is not impressed by NHTSA's efforts. He said of the agency's rollover research: 'It's political cover, plain and simple.'
He contends 'political economics' make it impossible for the federal government to regulate away the vehicles' popularity.
Foote, who represented McCath-ern, said the principal onus of correcting the problem is on manufacturers. Even though they have known that sport-utilities, with higher centers of gravity and narrower tracks, are inherently less stable than other vehicles, he told jurors hearing McCathern's suit, companies make a business decision to produce and promote them.
'They have marketed this vehicle as a reliable and dependable family vehicle,' he said. '(But), unfortunately, the capabilities of the Toyota 4Runner were such that it could not respond to (Sanders' emergency) steering, and she ended up in a rollover accident.'