NHTSA chose not to pursue an investigation.
The revelation came Sunday when a Congressional subcommittee that's reviewed tens of thousands of documents related to GM's recall fiasco released a memo that summarized its initial findings (which you can find here – warning, PDF). "Although we have had the documents for less than a week, they paint an unsettling picture," said Representative Tim Murphy (R-PA).
Hearings held by the House Energy and Commerce Subcommittee are scheduled begin Tuesday morning. General Motors CEO Mary Barra and NHTSA administrator David Friedman are expected to testify.
In a written statement Sunday, a NHTSA spokesperson said, "the agency reviewed data from a number of sources in 2007, but the data we had at the time did not warrant a formal investigation. Recent data presented by GM provides new information and evidence directly linking the ignition switch to the airbag non-deployment."
The House hearing is the first of several inquiries into General Motors conduct and its decade-long delay of a recall of more than 2.5 million cars saddled with a potentially deadly ignition-switch defect. Documents have revealed GM knew about the defect as early as 2001 – before cars event went on sale – but didn't act until February.
In the interim, at least 13 people have have died as a result of crashes related to the defect. One study commissioned by the Center for Auto Safety believes the number of deaths is much higher, that as many as 303 deaths could be tied to the defect, though the methodology of that study has come under fire.
A Senate panel holds its own hearings Tuesday. On Wednesday, GM is supposed to submit its response to 107 questions for NHTSA's own investigation. A Department of Justice investigation is pending.
Over the past month, Congressmen, safety advocates and industry experts have wondered why NHTSA and GM failed to connect the incidents – why an Early Warning Reporting system failed to set off alarm bells. As it turned out, it did, and no one acted. "The red flags were many, and yet those responsible failed to connect the dots," Rep. Fred Upton (R-Mich.) tweeted Sunday.
Another portion of the memo released Sunday noted that senior employees at Delphi, the maker of the defective ignition switches, told investigators that GM had approved use of the part, even though testing of the torque in the switches fell below set parameters. It also said that a desire to keep costs down may have played a key role in not fixing the defective switches before the cars went on sale.
In March 2005, the documents show the Cobalt's project engineer manager closed an examination of the switches with no action because, "the lead time for all solutions is too long," and "tooling cost and piece price are too high."
Among the questions the House subcommittee is set to ask Barra, according to the memo is, "Why did GM approve ignition switches that did not meet its specifications for torque performance? What was GM's assessment of the implications for performance and safety?"