Letter to Dr. Runge: The Temporal Recall Problem
June 21, 2002
Dr. Jeffrey Runge, Administrator
National Highway Traffic Safety Administration (NHTSA)
400 7th Street SW
Washington DC 20590
Dear Dr. Runge:
On May 15, CAS wrote you about how outrageous geographic recalls have become when NHTSA says, "It's not hot in Death Valley and it doesn't snow hard In Buffalo." CAS pointed out how NHTSA attempted to cover up the inadequacy of geographic recalls by withholding the city and state information on the public domain website and from the Vehicle Owner Questionnaires (VOQs). CAS has now learned that GM is using confidential agreements with consumers who are victimized by geographic recalls to help NHTSA cover up the inadequacy of such recalls.
CAS has also learned that NHTSA is accepting temporal as well as geographic limitations to safety recalls to the benefit of manufacturers and the detriment of consumers in dereliction of its Congressional mandate to protect the public. Although it is well know that a recall may affect only a portion of a model year, it is virtually unheard of for a manufacturer to claim that only skipping, random months of production are affected by a defect. In recalling the 1996-97 4-wheel drive Blazer, Jimmy and Bravada, (T-trucks) GM not only limited the recall to "Salt-belt States," but also limited it to the months of November 1995, June 1996 and October 1996 claiming that the warranty claims rate was high in those months but not others. (Attachment A, .PDF, is GM's recall report and NHTSA's memo accepting it.) By NHTSA and GM's own position, this is a corrosion defect where failures occur only after exposure so that warranty data are not a good indicator of a defect. It is very likely that many, if not most, of the failures will occur after the 3 year/36,000 mile warranty expires. The scattering of these limited data is nothing more than an artifact of the randomness of small samples.
As shown by the enclosed redacted letter (Attachment B, .PDF) from a consumer who experienced a ball joint failure and personal injury crash in a 1997 GMC Jimmy, GM is requiring consumers to sign a confidential agreement not to disclose the settlement terms to any third party. Fortunately, before accepting the settlement this very sharp consumer notified the Maine Attorney General who in turn notified CAS. Note that NHTSA had no crashes and only 4 injuries when it closed the investigation, presumably in the three months covered by the recall. This one outside the recall scope crash obviously changes NHTSA calculus. NHTSA promised to monitor this recall to verify its effectiveness. Has the agency done so because we have found dozens of reports of ball joint failures on 1996-97 GM T-models in the agency's own complaint data base?
At a minimum, NHTSA must ask GM for all reports of crashes, failures, repairs and parts sold regardless of whether covered by warranty.
Clarence M. Ditlow