How many ignition switch deaths? Postscript

   On Monday I posted a note on a New York Times story (from the preceding Friday) about the GM ignition switch disaster, the gist of which was that the particular accident the Times focused on in that story didn't seem to be caused by the defective switch.

   The following day the Times did a follow-up reporting that the ignition-switch death toll, as determined by Kenneth R. Feinberg, who administers the compensation fund GM set up, had risen to 23. Once more the Times illustrated the story by pointing to a particular victim, an infant rendered a paraplegic in an accident in which his great-grandmother and 13-year old aunt died. But through some evil chance, the Times managed to again alight on an accident that also seems unrelated to the switch problem.

   The particular accident is not mentioned in the massive Valukas report commissioned by GM, but was described in a Times' July 16 story: "Less than a mile from the Matthews family's home, another car swerved into their lane and crashed head-on into the Cobalt." The air bags did not deploy.

   As I briefly described the situation in Monday's post, and at greater length in posts referred to there, GM's theory that connects the ignition switch to the failure of the airbags is that there is an initial bump, as when a car goes over a curb or hits a bush or small tree, followed by a more serious crash into another object, such as a tree. The initial bump causes the defective switch to jump to the ACC position, which (after a delay of 0.15 seconds) turns off the airbag sensor. Almost all the accidents described in the Valukas report conform to that scenario, but that is clearly not what happened in the case described in Tuesday's Times. There was no initial bump that flipped the ignition switch, just a head-on crash.

   The Times has now unfortunately spotlighted the only two reported fatal accidents that don't seem connected to the ignition switch. (The Valukas report describes one other non-fatal non-conforming case.) It's not clear why the airbags didn't deploy in these cases, and the ignition switch could ultimately turn out to be the culprit. But the GM theory not only accounts for most of the accidents, it's the only theory we have; no one else has conducted engineering tests to determine if there are other reasons for the airbags not to deploy.  Perhaps a certain number of airbags fail to deploy for unknown reasons across all car makes and models. Or perhaps there is a deeper problem with the GM airbags that is as yet unappreciated. In any case, the Times coverage casts doubt on the extent of their reporters' understanding of what is going on.

   --Stan