August 2, 2011
Among safety advocates’ most vociferous criticisms of NHTSA and NASA’s investigation into Toyota Unintended Acceleration were the copious black smears over key bits of data and text in their twin reports released last February. These redactions have kept independent scientists from knowing exactly what the investigators did, irrespective of assessing the quality of the research. (See: How NHTSA and NASA Gamed the Toyota Data)
Alice and Randy Whitfield of Quality Control Systems Corporation, ever the assiduous students of NHTSA’s statistical and informational folkways, went for broke. Shortly after the reports were released, they filed a Freedom of Information Act request for non-redacted versions of the reports and supporting material that was missing from the record. In response, NHTSA publicly released some of the information in the form of less redacted versions of Technical Assessment of Toyota Electronic Throttle Control (ETC) Systems and Technical Support to the National Highway Traffic Safety Administration on the Reported Toyota Motor Corporation Unintended Acceleration Investigation, but continued to withhold other information.
In April, NHTSA rejected a QCS FOIA request for an unredacted copy of the NESC report, its appendices and other unpublished reports mentioned in the footnotes. The agency subsequently denied the Whitfields’ appeal for some of the materials because “the Agency properly determined that the currently redacted materials are entitled to confidential treatment under FOIA Exemption 4.” Exemption 4 of the Freedom of Information Act is industry’s best friend, permitting the removal from the public record information containing confidential business secrets. NHTSA apparently is Toyota’s second best friend, because it has rarely denied the automaker’s requests for confidentiality on this basis – even when the information doesn’t qualify. The Whitfields noted that they had already found former redactions that had nothing to do with business secrets, like “production counts of Toyota Camrys which had been available on NHTSA’s own website for years.”
But what the government withholdeth, the Google giveth.
Alice and Randy Whitfield have posted a new brief report about their adventures in NHTSA information mining, Keeping Secrets about NASA’s “Toyota Study” of Unintended Acceleration. The Whitfields found via a Google search unredacted portions of the documents they sought. They were struck by the redaction in one sentence of the NASA report’s Appendix A about data losses on the Controller Area Network (CAN): “Occurrences of CAN data loss are recorded in SRAM and remain available until the battery is disconnected. According to Toyota Motor Corporation (TMC) 292 instances were reported of CAN data loss by dealers for cars brought in for any problem. REDACTION.”
So what was so critical to Toyota’s ability to maintain its competitive edge that neither its rivals nor the great unwashed could see? Here’s the rest of it:
“Possible correlation of these incidences with UA cases was not checked.”
This discovery is of a piece with the general Swiss cheese-like nature of the NHTSA-NASA reports. Scratch the surface and you find:
– NHTSA relied on Toyota’s defense litigation experts, Exponent, for a warranty analysis used to dismiss the significance of physical evidence of an electronic cause of UA in some Toyotas. This conflict of interest was not disclosed.
– NHTSA and NASA based analyses on miscoded data and unsupported assumptions while failing to record and maintain the original data it was based on, preventing replication.
– NHTSA/NASA withheld from public view pieces of their report that are not related to Toyota’s confidential business.
That’s strike three. How many are yet to be discovered? We’ll all keep scratching.