Probe: GM OK’d Faulty Ignition Switches
Washington — Congressional staff investigating the widening General Motors ignition switch recall of 2.2 million vehicles said Sunday there are indications GM approved the design of the switches in 2002 even though the company knew they did not meet specifications.
Officials with Delphi, the Troy, Mich.-based supplier of the part, told investigators that GM accepted the switches now under recall despite knowing they did not meet the company’s specifications. That disclosure, along with others in a memo to committee members Sunday, raises more questions about why GM and federal regulators didn’t act sooner to address a long-standing problem.
The defective ignition switches have been linked to 13 deaths and 31 crashes.
Sunday’s memo from investigators for the House Energy and Commerce Committee comes as GM chief executive Mary Barra is to testify before the Oversight and Investigations Subcommittee at a hearing Tuesday on Capitol Hill. GM is recalling Chevrolet Cobalts and HHRs, Saturn Ions and Skys and Pontiac G5s and Solstices to replace switches that can be inadvertently jostled out of position, potentially causing air bags not to deploy in the event of a crash.
But the recall was ordered more than a decade after the first indications of a problem. As early as 2001, during preproduction of the Ion, GM knew there were issues with the ignition switches, according to the company’s time line of events leading up to the recall.
Greg Martin, a spokesman for GM, said the company is cooperating fully with congressional investigators to ensure they have a “full understanding” of GM’s decisions. He did not directly address the claim that GM may have known the switches didn’t meet specifications, however.
Delphi officials did not return calls and e-mails for comment.
As part of the investigation, committee staff collected 235,000 pages of documents from GM and its regulators at the National Highway Traffic Safety Administration. Investigators held closed-door briefings with officials from GM, NHTSA and GM’s suppliers, Delphi and Continental, which produced the sensing module triggering air bag deployment.
U.S. Rep. Tim Murphy, R-Pa., who chairs the Oversight and Investigations Subcommittee, said the information collected through documents and briefings “paints an unsettling picture.”
NHTSA is under fire for not moving more quickly to force a recall. Federal regulators had received hundreds of complaints of cars stalling over the years and, as early as 2005, ordered an investigation into a crash where a 16-year-old girl died in a Cobalt, and the air bags did not deploy.
According to e-mails, in September 2007, a top official in NHTSA’s Office of Defects Investigation (ODI) proposed a deeper look into “frontal air bag non-deployment in the 2003-2006 Chevrolet Cobalt/Saturn Ion,” saying there was “a pattern of reported non-deployments,” Sunday’s memo said.
GM, however, said it did not see a pattern and ODI officials determined later that year that there was no discernible trend to the incidents and decided against a more formal investigation.
Barra and David Friedman, NHTSA’s acting administrator, will be questioned this week about the recall and why it took so long. Subcommittee members want to know whether GM’s internal procedures are robust enough to catch defects and why the company would have approved switches that did not meet its own specifications.
Officials from Delphi told committee staff in a briefing that GM signed off on what’s known as a Production Part Approval Process, or PPAP, document in February 2002 for the switch “even though sample testing of the ignition switch was below the original specifications set by GM.”
It was not immediately known why Delphi would have provided a part that did not meet specifications, why GM would have accepted it or whether it was considered a potential problem at the time.
Committee members are expected to pursue answers from GM as to why it would have accepted the part — especially since Cobalt engineers, in court depositions made public since the recall, have said themselves that vehicles shouldn’t have been sold if parts didn’t meet specifications.
Delphi officials also told committee staff that it was GM which finally requested a change to the part in 2006, which largely addressed the problem in some 2007 model year vehicles and those thereafter. However, because some of the original switches were sold as after-market parts, GM can’t be sure how many of those may have eventually ended up in other model year cars — leading to last week’s expansion of the recall.
Even that decision to fix the part, however, appeared to run into difficulty, according to congressional investigators. Records indicated that in 2005 — after GM opened an internal probe into issues reported with the switches — the company considered addressing the problem with a design change. But a GM engineer said it would be “close to impossible to modify the present ignition switch.”
Even so, GM personnel signed off on a change largely addressing the problem just the following year. It wasn’t until 2013 when internal investigators at GM even figured out that the part had been ordered changed by at least one of its own engineers.
Delphi officials told the committee that although that 2006 change led to “a significant increase in ... performance” it was still below GM’s original specifications.