Front airbags don’t inflate in hundreds of head-on crashes

A hit-and-run motorist slammed into the front of the 2005 Dodge Caravan she was driving, spinning it 180 degrees.

“It’s one that sticks with me,” Atlanta Police Officer Shane Keller said recently. The crash was so violent that rescuers needed the Jaws of Life to free Katz, 27, a Georgia wife and mother who had just buckled herself in to go to work.

Then they saw something “curious,” as the officer put it. The Caravan’s airbags had not deployed.

It’s a fatal mystery repeated in hundreds of traffic accidents, an investigation by The Kansas City Star found: front airbags that did not inflate in deadly front-end crashes.

Analyzing a National Highway Traffic Safety Administration database of all traffic fatalities over a six-year period, the newspaper discovered that far more people had died from wrecks where airbags didn’t deploy than all of those who died from injuries caused by airbags that fired too easily or too forcefully.

A decade ago, deaths blamed on overly aggressive airbags triggered congressional action, which brought about the “smart bags” of today. About 300 people have died from improper airbag deployments since 1990.

But The Star found those deaths are dwarfed by another body count just as disturbing. At least 1,400 drivers and front-seat passengers died from 2001 through 2006 in front-impact crashes involving vehicles whose airbags — smart or otherwise — never deployed.

“I have to say I’m shocked,” said Joan Claybrook, former chief of NHTSA and current president of the consumer advocacy group Public Citizen. “These airbags should deploy.”

To be sure, even when airbags work, people still die in serious accidents. In the six-year span analyzed by The Star, head-on crashes killed roughly 14,000 drivers and front-seat passengers, even though their airbags deployed.

But in that same period, the federal government has estimated, airbags saved 15,000 lives.

Nobody knows how many more lives could have been saved if the airbags had deployed in the cases reviewed by The Star. And because of insufficient data gathered by NHTSA, speeds for many of those wrecks also are unknown, raising questions as to whether those vehicles were going fast enough to activate the airbags.

For those reasons and others, current NHTSA officials disputed The Star’s findings and don’t consider uninflated airbags to be a significant problem.

“There is never an acceptable failure rate,” said agency spokesman Rae Tyson. However, he insisted that it is impossible to draw any conclusions about the lack of deployments based on the agency’s data.

In a written statement, NHTSA warned that “The Kansas City Star is doing a grave disservice to its readers, by implying — through an improper analysis of our own data — that air bags are not performing as intended. Nothing could be further from the truth.”

A spokesman for the auto industry, Charles Territo of the Alliance of Automobile Manufacturers, said that “no two accidents are alike” and that conclusions can’t be drawn from the analysis without knowing more about each accident.

However, another former NHTSA administrator disagreed, calling the database the “best in the world.”

“Yes, it’s flawed. But it is the best you’ve got,” said Ricardo Martinez, who led the agency during the Clinton administration.

He said the newspaper’s findings point NHTSA “to an area that warrants investigation to see if there is an opportunity to improve safety.”

In its database analysis, The Star eliminated thousands of fatalities in an attempt to produce as conservative a finding as possible. The newspaper’s analysis focused on head-on crashes into the front ends of other vehicles and objects such as trees and embankments.

If the newspaper includes front-end crashes into the side or rear of other vehicles — the type of crash that killed Brooke Katz — the number of deaths climbs to 1,900.

The Star also did not include fatal crashes identified in the database that involved principal impacts to the left or right fenders. Nor did it include victims who were ejected or died when their vehicles rolled over, crashed and caught fire or were submerged in water. And the newspaper excluded victims in vehicle models where airbags were not standard equipment.

Those steps eliminated at least 3,000 fatalities.

Finally, the newspaper consulted with automotive safety researchers, statisticians and other experts — including former NHTSA officials — in formulating its methodology for the analysis. All found it acceptable. And the National Institute for Computer-Assisted Reporting in Columbia, Mo., examined the same database and confirmed The Star’s numbers.

Moreover, the NHTSA database reveals that the annual number of deaths in nondeployment crashes grew dramatically — about 50 percent — since 2001. In 2006 alone, the deaths rose by 14 percent from the previous year — despite a 2 percent drop in all traffic fatalities.

NHTSA declined to make top agency officials available for interviews.

Although its spokesman discounted The Star’s findings, he added that 1,400 deaths, if true, were not that alarming.

“If it’s a real number, it’s not a surprise to us,” Tyson said.

Indeed, the problem of noninflating front airbags cannot come as a surprise to NHTSA officials. In complaints filed with NHTSA by the public following injury and fatal accidents, uninflated airbags outranked any other complaint about components. In fact, The Star found that nondeployments represented nearly one-fourth of the thousands of gripes lodged with the agency over severe accidents.

A former NHTSA official said the analysis raises serious questions. After reviewing NHTSA crash records at the newspaper’s request, George Washington University engineering professor Kennerly Digges said, “You see things here that can cause you a lot of worry.”

Especially in a “terrible crash … you don’t understand why you didn’t get a deployment,” said Digges, a former NHTSA director of vehicle safety research.

Claybrook, who led NHTSA during the Carter administration, said the agency and Congress needed to investigate the nondeployment problem.

Problems in most models

The analysis of NHTSA fatality data showed that airbags didn’t deploy in virtually every make of auto that Americans drive, both foreign and domestic. And the nondeployments ranged from aging models to new cars fresh off the lot.

For instance, at least 100 drivers and front-seat passengers died while traveling in model-year vehicles 2004 through 2006. They included Hilton Thompson and Pansy Evans, elderly siblings from southeast Missouri. They died in 2004 in a Cadillac that Thompson had bought new just the day before.

GM declined to comment on that accident or any other specific crash.

“The conventional wisdom is that airbags save lives. Do they save every life? Unfortunately, no … (but) airbags are helping” avoid fatalities, said Alan Adler, GM spokesman for product safety.

Among other findings by The Star:

•About 25 percent of the 1,400 deaths where airbags didn’t deploy resulted from vehicles smashing into trees or poles. Automakers and some owner’s manuals acknowledge that the airbags might not deploy in crashes into trees and poles.

Some critics have blamed that problem on too few crash sensors. But auto industry spokesman Territo disagreed that there could be too few sensors, adding that “manufacturers are working to enhance safety in all types of vehicle accidents.”

•NHTSA has paid relatively little attention to airbags that don’t deploy. Since 1996, investigators have launched about 600 inquiries into how well activated frontal airbags performed. But probes into nondeployments have totaled only about 50, according to records that NHTSA gave The Star, plus documents the newspaper found.

Indeed, records show that agency investigators examined at least 20 nondeployment crashes in which 12 people were killed and found initial signs of airbag failure. But records show only three of those crashes became part of in-depth agency investigations, and only one led to a recall.

In a written statement, NHTSA said inspectors have focused instead on new airbag systems to make sure the safety devices meet updated regulations. The agency said it monitors hundreds of safety issues that never lead to in-depth investigations.

•In the database of fatal accidents, an important factor — crash speeds — was difficult to nail down. Estimated speeds, from police reports, were available for just under 40 percent of the accident victims in The Star’s analysis. The average speed was 55 mph. Experts said airbags should deploy at speeds of 10 mph to 16 mph in head-on crashes into an immovable barrier.

However, because so many of the actual speeds were not recorded, NHTSA maintains that many cars may not have been traveling fast enough for the airbags to deploy, even if all of the impacts were severe enough to kill.

NHTSA spokesman Tyson also noted that airbags are only considered supplemental safety devices. They are thought to save one-sixth the number of lives that fastened seat belts save.

If you’re not belted, “the airbag is meaningless,” Tyson contended. “I am dismissing half of your (1,400 fatality) number anyway because the occupants were not belted and, as far as I’m concerned, bets are off.”

Experts agree that seat belts are the best and primary safety system for drivers and passengers. But they said NHTSA’s own data prove that airbags are far from being meaningless and have saved thousands of lives. In fact, federal regulations require airbags in crash tests to protect both belted and unbelted occupants.

“Failure to wear your belt should not mean a death sentence,” Claybrook said.

Regardless, nearly half of the people killed in The Star’s examination were wearing seat belts when their airbags didn’t deploy.

They included the Ohio driver of a 2004 Ford F-150 pickup truck that last year sheared off a pole, hit a culvert, struck a small tree and finally smashed into a bigger tree. Police estimated his speed at 60 mph.

Even though Lloyd Holland, 63, wore both lap and shoulder restraints, he jackknifed into the steering wheel hard enough to bend the top flat — just inches above the inactivated airbag.

A NHTSA inquiry did not reach a conclusion about the nondeployment. Ford spokesman Daniel Jarvis said the company investigated and determined that the various impacts did not generate enough force to trigger the airbag.

Fairfield Township Police Chief Richard St. John remains unconvinced.

“The front end of the truck was destroyed,” he said. “I have no idea why (the airbags) didn’t deploy.”

Debate over ‘why?’

Automakers contend that the most likely reason that front airbags don’t inflate is that, in many wrecks, they’re not supposed to and could even injure you if they did.

Their task is tricky. Experts said sensors often located in the front of the vehicle are designed to detect the force of a crash and send a signal to a microcomputer that decides whether the wreck is serious enough to fire the airbags — all in less than the blink of an eye.

“Airbags may or may not deploy depending on how much angle you have in a frontal collision,” said Robert Yakushi, director of product safety for Nissan North America Inc. “Hitting a tree, or (a vehicle) going off road, doesn’t guarantee that the airbag will deploy every time.”

In a fatal airbag nondeployment lawsuit pending in Texas, Chrysler officials said front airbags were calibrated to fire when its 2006 minivans struck an immovable barrier at 16 mph or more, head-on.

“However,” the company acknowledged in a legal response to plaintiffs’ questions, “all deployment thresholds are probabilistic in nature.”

That is, not foolproof.

“Thus it is unlikely, but possible,” that airbags wouldn’t deploy at even higher speeds, the carmaker stated. It cited “factors such as the angle of impact, direction, size, weight, relative stiffness of the subject vehicle as well as corresponding attributes of the object struck …”

“This would be true for all vehicles, whether made by Daimler Chrysler Company LLC or others.”

Recall records reveal a variety of other reasons for airbags not deploying or not inflating sufficiently: Loose bolts. Faulty welds. Missing parts. Cracked components.

Since 1990, airbags at risk of failing have led to 45 recalls involving at least 3.5 million vehicles, The Star found. The recalls affected more than 100 models of cars and light trucks, from pricey Italian sports cars to environmentally friendly hybrids.

One recall of 240 Ford Escorts and Mercury Tracers cited a six-week period in 1994 when poorly welded inflator canisters were installed in the cars. Ford said in recall documents it happened only at an assembly plant in Mexico.

Ford quickly identified the problem, recalled the vehicles and worked with its supplier to correct it, Jarvis said. The process took less than three months.

Nissan and the federal government logged what one motorist called “an indicator light problem” with a 2002 Nissan Xterra. There were nearly 50 complaints about the airbag warning lights recorded through the end of August 2002, according to NHTSA records.

In fact, the light was indicating a potential danger. Nissan discovered that airbag-related wiring could come loose and told the government in early 2003 that the driver’s side airbag “will not deploy in the event of a crash.”

Still, Nissan initially didn’t treat the problem as requiring a safety recall because the light alerted motorists to a possible malfunction, said company spokesman Tony Pearson. Nissan at first issued only what’s called a “service bulletin.” But after a nearly yearlong NHTSA investigation, the automaker agreed to the safety recall of about 27,000 Xterras and 37,000 Altimas.

Nobody yet understands why the airbags stayed put in the practically new Caravan driven by Brooke Katz in March 2005.

DaimlerChrysler denied any link to past fatalities when it recently sent recall notices to 270,000 owners of 2005-model Caravan and Chrysler Town & Country vehicles. The company, now known as Chrysler, said a problem it found in airbag sensors would still allow bags to deploy.

“There are no confirmed reports of accidents or injuries related to any failure … in the subject minivans,” Chrysler assured NHTSA in an agreement that ended a federal inquiry into its minivan airbags earlier this year.

A key word in that statement is “confirmed.” A company spokesman, Michael Palese, told The Star that litigation has kept the carmaker from running tests on the event data recorders in the Katz minivan — or in another Chrysler minivan, this one a 2006 model, which also is tied up in court. Its airbags didn’t deploy in a 2006 head-on fatality in Texas.

Palese said the company can’t verify the cause of the nondeployments or that the airbag systems even failed. Nonetheless, he said Chrysler is “confident the airbags worked exactly as designed” by not going off.

Judging from the worst area of damage to the Katz vehicle, he said the driver may have been thrown “out of position” to a degree that the airbag system decided against deploying. In many wrecks, Palese said an airbag that deploys too late or too close to the occupant can do more harm than good.

Gregg Katz, however, cannot imagine how his wife’s fate could have been any worse.

“How many people buy cars based on those five-star crash ratings they see advertised on TV, expecting them to perform the way they’re supposed to?” he said in an interview.

“And when they fail … how do you say to a 3-year-old that her mother is gone?”

In her princess-style bedroom, daughter Addisyn, now 6, points to framed snapshots of a clowning Brooke Katz and passively tells visitors, “Mommy died.”

Deadly delays

When a nondeployment problem does come to light, regulators and carmakers can spend months — even years — looking into possible defects before issuing a recall. Sometimes, people perish on the roads in similar accidents while investigations drag on.

NHTSA said that there is no “typical” time for ordering a recall but that it’s not unusual for investigations to last 18 months.

Consider the case of Phillip A. Howell, 40, whose pickup crashed on a slushy highway near Syracuse, N.Y.

It was Oct. 27, 2001. A teenage driver lost control of her Toyota truck and collided head-on with Howell’s 2000 Chevrolet Silverado. Both drivers were wearing seat belts, but only one survived.

The teenager’s airbag fired. Howell’s didn’t.

“The circumstance of the airbag failure to deploy remains unknown,” the accident’s investigating officer wrote.

Howell’s wife, Kelly, said she remembers watching him bleed in the hospital, a dent in his head the size of a peach, and sobbing to herself: “Why didn’t his airbag go off? Why didn’t his airbag go off?”

Earlier tests indicated to GM something may be wrong. In fact, the automaker suspected as early as October 1999 that an “anomaly” in the front airbags of many of its 2000-model pickups could potentially lead to nondeployments in some wrecks, records show

However, it wasn’t until June 2002 that GM ordered a safety recall.

A letter signed that month by GM’s director of product investigations, Lyndon R. Lie, revealed the snail’s pace of the recall. The letter informed federal regulators that “both front air bags failed to deploy” in a 1999 test of a 2000-model pickup hitting a barrier at 30 mph.

GM and its supplier, Delphi Corp., analyzed a possible “bouncing” problem in the sensor system. But the airbags had worked in a follow-up crash test.

For more than 30 months, engineers “continued to conduct extensive investigation and testing to determine the causes of this complex condition,” GM’s letter to NHTSA stated.

Just weeks before the company decided to recall its trucks, a Silverado driver died without an airbag going off in a head-on collision in south Texas. Daniel A. Farias, who was 17, lost control of his Silverado on a highway. As in Howell’s case, the airbags of the oncoming vehicle fired, police records showed.

Unlike Howell’s truck, however, Farias’ pickup was built shortly after GM had recalibrated sensor modules in newly built trucks in a factory effort to correct the anomaly, according to records.

Another Silverado driver was seriously injured east of Dallas in February 2002, also after GM had recalibrated the sensor modules. He sued GM over the nondeployment, and the carmaker ultimately paid a $500,000 settlement, records show.

In addition to the 2000-model Silverados, the recall involved airbag systems in Chevrolet Tahoes and Suburbans, and GMC Sierras and Yukons — some 525,000 vehicles in all.

The recalibration took only 15 minutes at the dealership.

GM declined to comment on the recall or on the crash that killed Phillip Howell.

“This was our first vehicle with airbags,” his widow tearfully recalled. “That was very important to us when we were shopping.”

Like many crash victims’ families whom The Star interviewed in its investigation, Kelly Howell hasn’t gone to court to try to prove what may have gone wrong.

Now a single mother, she couldn’t afford the tens of thousands of dollars that automotive experts wanted to analyze or replicate the crash and to appear in a courtroom. Just storing the truck as evidence, which she did for a year, cost $12 a day.

To this day, she said, “I keep getting recall notices addressed to my husband.”