Why didn't stalling alone trigger GM recall?
GM saw no safety defect, but rivals did
In 2004, 12,000 Saab 9-3's were recalled to fix glitches that caused stalling.
General Motors asserted for a decade that unexpected stalling was not a safety defect, even as its competitors issued about 90 recalls for similar symptoms during the same period and federal safety regulators opened 42 investigations into stalling.
It was only after engineers learned that flimsy ignition switches were causing airbags to fail in Chevrolet Cobalts and Saturn Ions that GM treated the matter as a serious safety concern, and regulators now say GM waited nearly five more years to issue a recall after company engineers definitively learned of that connection.
Until announcing a recall in February, GM’s only public response to accumulating complaints of stalling was a series of service bulletins sent to dealerships and a June 2005 statement saying “practically any vehicle” can shut off if the ignition is bumped inadvertently. A GM spokesman told The New York Times at the time that GM didn’t consider the issue to be a safety concern because the car remained controllable and could be restarted.
In contrast, an analysis of the government’s recall database by Automotive News found that American Honda, Nissan North America, Chrysler Group and other automakers voluntarily issued multiple recalls during the same period for stalling.
“The engine could stall without warning and a crash could occur,” Honda told the National Highway Traffic Safety Administration in 2007, when it recalled nearly 167,000 cars, pickups and minivans.
Nissan said in 2010, explaining a recall of more than 747,000 SUVs and pickups: “This could cause engine stalling, increasing the risk of a crash.”
Chrysler said in 2011, as it recalled about 250,000 minivans and Dodge Journeys: “Engine shutoff while driving could increase the risk of a crash.” Chrysler told NHTSA it had looked into the issue for about six months and traced the problem to the vehicles’ ignition switch being jarred or bumped into “accessory” mode.
The frequency of such recalls raises questions not only about GM’s longstanding position, but also about why NHTSA didn’t respond to the stalling issue.
Going back to the 1970s, “NHTSA litigated a series of defect cases in the federal courts that established loss of vehicle power on the road as a safety defect,” Clarence Ditlow, head of the Center for Auto Safety, wrote last week in a letter to NHTSA’s acting administrator, David Friedman. “NHTSA did not have to wait to establish a connection between ignition switch failure and airbags not deploying to open a defect investigation and obtain a recall.”
Ditlow: “NHTSA did not have to wait.”
NHTSA responded to the letter with a statement noting that GM, which it fined $35 million on May 16 for breaking federal law, “knew about the safety defect for years and did not take action to protect Americans from that defect until earlier this year.” A GM spokesman declined to comment, except to say, “We’ve recalled the vehicles.”
From 2004 through 2013, NHTSA opened 42 investigations related to stalling complaints, and the inquiries resulted in 31 recalls covering 5.1 million vehicles. About 60 recalls in the same period for stalling were not connected to a NHTSA investigation.
Two of the investigations involved Saab, which GM owned at the time. In October 2004, GM decided to recall 12,000 Saab 9-3 sedans because a software glitch could cause the cars to stall. “If the vehicle stalls as a result of this condition, it can be restarted immediately,” GM explained in the defect notice.
Ten days later, documents show, the program engineering manager for the Chevrolet Cobalt, Gary Altman, reported to his team that a car he was test-driving had unexpectedly turned off when he bumped the ignition with his knee.
Within two years of that incident, GM opened and closed at least four internal inquiries of the issue, bought back at least 11 Cobalts from owners who complained about stalling and approved a redesign to the car’s ignition switch. But it didn’t issue a recall, even though it announced two more recalls for the stalling issue affecting Saabs.
In addition to the three Saab recalls, GM issued eight other recalls from 2004 through 2013 for stalling, but all of them also involved either a risk of fire or an inability to restart the car.
In a June 2013 deposition for a lawsuit brought by the family of a woman who died in a Cobalt crash, Altman testified that he believed in 2004 that the tendency of the car’s ignition switch to rotate inadvertently didn’t amount to a safety defect requiring a recall — and that his opinion hadn’t changed nine years later. Altman is one of two engineers GM put on paid leave in April.
While GM was receiving and studying complaints of stalling, NHTSA officials appeared to be focused on crash reports that Cobalt airbags were failing. A September 2007 e-mail written by Gregory Magno, the head of NHTSA’s Defects Assessment Division, proposed that the agency investigate the crashes but made no reference to complaints about stalling.
More than three decades earlier, NHTSA’s chief counsel, Frank Berndt, argued that “if a defect causes failure of a critical vehicle component or of a major vehicle control system, it is safety-related,” according to a memo he wrote that is posted on the Center for Auto Safety’s Web site. Judges concurred with that theory in four cases involving GM and Ford Motor Co.
Federal law requires automakers to notify NHTSA within five business days of determining that a safety-related defect exists. They are allowed to issue service bulletins only for issues not involving safety concerns.
Reports analyzing black-box data from two Cobalt crashes suggest that GM engineers learned of the link between the ignition switches and airbags sometime between August 2007 and May 2009. The reports include nearly identical flow charts showing five conditions under which airbag algorithms can be disabled, except that the 2009 report includes a sixth reason — the ignition moving out of “run.”
NHTSA’s Friedman pointed to the 2009 report as the earliest evidence yet showing that GM knew of a “clear connection.” He said the report — an analysis by Continental Automotive of data from a September 2008 crash that killed two teenagers in Michigan — would have allowed NHTSA to take action sooner if it had known about it.
GM began investigating the Michigan crash about three weeks after it happened, according to a police report that logged a call from a GM investigator, and it reported the deaths to NHTSA soon afterward, classifying the crash cause as “unknown.” NHTSA asked GM for more data on two other fatal Cobalt crashes that GM reported that quarter, but not the Michigan crash.
By the time GM engineers met with representatives from Continental on May 15, 2009 — less than three weeks before the company’s bankruptcy filing — the automaker knew of 31 complaints to NHTSA about Cobalts and other cars with the same ignition switch stalling and had received at least 89 complaints about it, documents show. It also had black-box data from 14 crashes, four of which were fatal, showing that the ignition was in “accessory” in seven of the incidents and in “run” during the other seven.
What GM engineers, executives and lawyers knew about the problem in 2009 is important because a number of lawsuits accuse the company of committing fraud by not disclosing the issue as a potential liability before a judge agreed to shield the company from liability for pre-bankruptcy incidents.
“It is inconceivable that individuals within GM’s upper management and general counsel’s office did not know about the ignition switch defect in GM vehicles, or the attendant contingent liabilities, when GM entered bankruptcy in June 2009,” Alexander Schmidt and Jonathan Flaxer, lawyers representing eight consumers suing GM in bankruptcy court, wrote in a filing last week.
You can reach Nick Bunkley at email@example.com. -- Follow Nick on