U.S. failed to investigate GM despite troubling data

U.S. regulators looking into General Motors Co.'s Chevrolet Cobalt noticed a surprising statistic: warranty claims over the car's air bags were four times higher than for competing vehicles.

They also had customer complaints, crash reports and GM's own descriptions of the fault.

Even with all that data, a review group at the U.S. National Highway Traffic Safety Administration overruled an investigator in 2007 and determined there wasn't enough of a pattern to open a formal probe, according to documents released yesterday by the House Energy and Commerce Committee.

While the e-mails and memos add to evidence GM dragged its heels on responding to concerns over defects, they also shed new light on the government's decision not to act on a flaw now linked to 13 deaths in accidents.

"Everything in this memo suggests they should have opened a defect investigation," said Clarence Ditlow, executive director of the Center for Auto Safety, a Washington-based watchdog group. "The numbers are off the charts."

Congress and the Justice Department are investigating the chain of events that led GM to wait more than a decade to recall 2.59 million small cars. The ignition switch in the Cobalt, Saturn Ion and other vehicles could shut off when jarred, cutting power to the engine and deactivating air bags.

Warranty Claims

NHTSA's acting administrator told a U.S. Senate panel April 2 that it didn't recommend starting a probe in 2007 because the Cobalt and Ion didn't stand out compared to peers. The U.S. Transportation Department's inspector general, Calvin Scovel, is reviewing whether the agency acted properly.

GM Chief Executive Officer Mary Barra has apologized and ordered an internal investigation. Barra's name surfaced in only one document released by the committee, a memo to her providing brief context on a news story about a previous Cobalt recall for steering issues.

While the papers don't show why NHTSA decided against probing the claims, GM officials said they saw "no specific problem pattern," according to a memo from Gregory Magno, chief of NHTSA's Defects Assessment Division, who recommended the probe before being overruled.

NHTSA's acting administrator, David Friedman, told senators the agency doesn't have complete records to show why it overruled the recommendation.

"Frankly, it is something that is currently hamstringing our ability to fully pull together all of what happened," he said.

No Standout

What is known is that the agency studied data on the Cobalt and Ion, beyond points Magno raised, and decided the cars weren't unusual when it came to air-bag failures, Friedman said.

"They were a little bit above average, but they did not stand out," he said.

Nathan Naylor, a spokesman for the auto-safety regulator, said today he couldn't comment beyond what Friedman said at the hearings.

NHTSA received 29 complaints about air-bag failures in the Cobalt and Ion for model years 2003-2006, according to the documents released by the House panel. That included 25 crashes with injuries and four fatalities.

The rate of air bag-related warranty claims for the Cobalt in 2006 was more than four times higher than for Honda Motor Co.'s Civic, Toyota Motor Corp.'s Corolla and Ford Motor Co.'s Focus, according to the documents. After the Cobalt, the next highest claims rates were with two other GM vehicles, the Saab 9-5 and the Cadillac CTS.

Front Crashes

Magno perceived a pattern of air-bag nondeployments in the Cobalts "that does not exist in their peers," according to one memo. He'd earlier asked for a decision to start an investigation within two weeks, according to a Sept. 5, 2007, e- mail to senior officials in the agency's Office of Defects Investigation.

Magno, in the e-mail, said the agency knew of cases in which frontal air bags didn't deploy "in severe front crashes under circumstances where they would be expected to function and reduce injury levels."

He also presented agency officials with so-called early warning data, compiled from customer complaints and injury reports, four crash investigations into air bags that didn't work and four GM technical-service bulletins. These company communications with dealers describe repairs that should be offered to customers while stopping short of a publicized safety recall.

The Defects Assessement Division did what it's trained to do in this case, and that's to raise red flags, Friedman told reporters April 2.

"That's what we tell them to do," Friedman said. "We dig deeper into the data and the into the cases to understand if there's sufficient information of a specific defect or of a trend that allows us to move forward."

GM Tension

The documents released yesterday also pointed to a history of contentiousness between GM and its regulator that continued beyond the automaker's bailout by U.S. taxpayers.

In July 2013, less than a year before GM's recall, the head of NHTSA's Office of Defects Investigation, Frank Borris, complained to Carmen Benavides, GM's director of product investigations.

"The general perception is that GM is slow to communicate, slow to act, and, at times, requires additional effort of ODI that we do not feel is necessary with some of your peers," Borris said.

Borris listed six points of contention between regulators and the company, including an attempt to handle an air-bag fix for the Chevrolet Malibu as a "customer satisfaction campaign" rather than a safety recall.

NHTSA 'Bolt'

GM also angered agency officials by trying to handle a "fairly obvious" defect involving module connectors with a non-public technical service bulletin to dealers in 2012, Borris said.

The note got a rapid reaction from GM senior executives, including Mike Robinson, vice president of sustainability and global regulatory affairs.

"This note from NHTSA, both the content and the tone, comes like a bolt out of the blue," Robinson said. "We need to address this immediately and I would like to discuss. We worked way too hard to earn a reputation as the best, and we are not going to let this slide."

NHTSA also has come under its share of criticism for delays.

Scovel's review will be the fourth by his office into the effectiveness of NHTSA's defects-screening process since 2002. In 2011, a report found the office didn't meet its own timeliness goals in more than half the cases.

The agency has made changes to address the criticisms, Scovel told senators April 2. How effective those changes have been won't be known until the current probe is complete, he said.

For his part, Friedman said NHTSA has realized it needs to improve.

"We're in the middle of our due diligence effort," he said. "We're applying lessons learned along the way."