This story originally appeared in Corporate Counsel.
On the day he died, Mark Saylor was doing what he did for a living: driving on a California highway. Only he wasn’t driving his state-issued highway patrol car that day in late August 2009. Nor was he driving his own Lexus 250, which was at the dealer for servicing. He was driving a loaner with his wife, daughter, and brother-in-law on a leisurely family outing northeast of San Diego—until suddenly the car inexplicably took off. And no amount of braking could slow it down. As Saylor frantically tried to gain control, his brother-in-law called 911. "Our accelerator is stuck!" he told the dispatcher. "We’re going 120!"
It wasn’t just the speed that made this so dangerous. He read the sign they were passing: "End freeway one-half mile." The car was barreling toward a T-shaped intersection. When it got there, it hit another car, flew through a fence, rolled into a field, and burst into flames. The last word before the screams was, "Pray!"
This wasn’t the first time that someone driving a Toyota had experienced sudden unintended acceleration. And it’s not a problem that’s unique to Toyota. But this was the event that Toyota cites as the beginning of its ongoing crisis.
How has it responded? The company has moved aggressively to contain the damage. Shifting floor mats were identified as a primary cause of many of these episodes. The company found that the loaner that the 45-year-old Mark Saylor was driving was equipped with mats that had never been intended for that car. Later, the company fingered accelerator pedals manufactured by a third party as prone to sticking. And Toyota says that many accidents are caused by drivers who inadvertently step on the gas instead of the brake.
As the crisis mounted, the company seemed overmatched. Critics charged that Toyota had sacrificed quality—its traditional strong suit—in a rush to rack up sales. The National Highway Traffic Safety Administration (NHTSA), which had been criticized for years for its willingness to pin sudden acceleration on driver error, suddenly got tough. Toyota recalled more than 8 million cars and paid fines totaling more than $50 million. Litigation, which had slowed down before the Saylor crash, roared back to life, fueled by the recalls and new complaints. And the political pressure, coupled with a Democratic Congress, led to hearings in Washington that drew global attention. Toyota Motor Corporation president Akio Toyoda flew in from Japan to personally face the politicians’ angry questions.
But then everything seemed to calm down. As the company battled two large multidistrict litigation class actions (MDLs) in California, it quietly settled some of the smaller lawsuits, including the one brought by the Saylors’ survivors. The results of several investigations trickled in. Some had been commissioned by Toyota, and tended to include lots of technical data and to focus on floor mats and gas pedals. Then, in 2011, NHTSA concluded its own probe, which purported to be comprehensive, and Ray LaHood, secretary of the U.S. Department of Transportation (the parent agency of NHTSA), pronounced himself satisfied that Toyota’s cars were safe.
Not only had the public uproar subsided, sales rebounded. Following a slump that was probably attributable as much to the economic downturn as the bad publicity, last year the company regained its status as the world leader in car sales. For Toyota, the long ordeal seemed over.
But some leading automotive experts aren’t buying it. Last December, Toyota agreed to pay $1.3 billion to settle the MDL brought by car owners who claim that they suffered economic damages as a result of these events. Critics point out that it’s a pretty big number for plaintiffs who weren’t even directly affected. Beyond that, more than 200 personal injury cases remain to be resolved in the other MDL. The first bellwether trial had been scheduled for March, but it settled in January on confidential terms. At this writing, it’s unclear how the matter will play out; some lawyers expect another large settlement.
But putting aside the politics and litigation, these automotive experts simply don’t believe that the controversy has been put to rest. They acknowledge that some accidents are caused by drivers stomping on the gas instead of the brake, and some from defective floor mats and gas pedals. But the experts don’t believe that these explain the surge in complaints. Instead, they believe precisely what Toyota has for many years steadfastly denied: that the problem is rooted in electronics.
These experts have found some surprising support from insiders at the National Aeronautics and Space Administration who were close to the investigation NASA conducted into Toyota’s acceleration problems a couple of years ago (and which LaHood cited when he discounted problems with its electronics). And now the experts say they’ve found additional corroboration in the communications of Toyota’s own people. Corporate Counsel obtained scores of internal documents written by employees who were struggling to understand why cars were suddenly accelerating, and where the company could have gone wrong. Among the writers were executives, managers, lawyers, public relations specialists, and engineers.
What this demonstrates, in the age of YouTube and Wikileaks, is how hard it is for multinationals and their in-house counsel to keep a lid on their companies’ internal data.
Many of the documents are marked "secret" and "confidential." They were provided by Betsy Benjaminson, a translator who has worked for several agencies that translate Toyota documents from the Japanese (and who translated several of those quoted in this article). She says that these shops work for law firms hired to assist the company in litigation.
Benjaminson provided these and many more documents last year to Senator Charles Grassley (R-Iowa), the ranking member of the Judiciary Committee, who then wrote a letter to NHTSA expressing his concern that questions about electronics have not been resolved. Corporate Counsel showed Toyota the complete documents from which quotes were excerpted for this article; read the company’s response here.
Benjaminson is revealing her identity for the first time here. She decided to go public because lives are at stake, she says. "Up to now," she adds, "the corporate PR megaphone has completely drowned out the victims."
Four experts agreed to review the documents independently and share their impressions. Keith Armstrong, Antony Anderson, and Brian Kirk are based in the United Kingdom; Neil Hannemann lives in California. All of them have decades of experience. The documents they reviewed date from as early as 2000; the most recent were written a few months after the congressional hearings in February and March 2010. They include many emails along with spreadsheets, flow charts, and diagrams.
On one important point the experts agree: There is no smoking gun that shows that Toyota identified and concealed an electronic defect that was responsible for crashes. But numerous documents, they say, undermine the corporation’s repeated attempts to reassure the public, as exemplified by the testimony of Jim Lentz, the CEO of Toyota Motor Sales U.S.A. Inc. In February 2010 Lentz told a House subcommittee: "We are confident that no problems exist in our electronic throttle systems in our vehicles." He went on to testify, "We have done extensive testing on this system, and we have never found a malfunction that caused unintended acceleration."
The documents seem to tell a different story. An email written by Hiroshi Hagiwara, a Toyota vice president in Washington, D.C., and sent to executives in Japan a month before the hearings hints at the turmoil beneath the surface. Hagiwara and Chris Tinto, a V.P. for technical and regulatory affairs and safety, had been talking about the U.S. investigation and an earlier one in Europe that also involved unintended acceleration (UA).
"Tinto is extremely pessimistic," Hagiwara wrote, "and is saying (public hearings, someone will go to jail, I can’t completely take care of the pedal problem, etc.)." Tinto’s primary concerns (according to Hagiwara): "For NHTSA, we said that our investigations in Europe found that the pedal return is a little slow at a slightly open position, and that there were no accidents, but this is not true. Last year’s situation in Europe (many reports of sticking pedals and accidents, and a TI TS9-161 was filed on October 1, 2009) was not reported to NHTSA." That failure, Tinto said, "may be a violation of the TREAD Act"—the federal law that requires car manufacturers that conduct recalls in foreign countries to report these to U.S. regulators.
Still speaking of Tinto, who worked for NHTSA in the 1990s before he was hired away by Toyota, Hagiwara continued: "He appears to question how Toyota has grasped and handled the overall UA problem (mat, accelerator pedal, ECU [electronic control unit], and electronic throttle systems, etc.)."
Hagiwara reminded the executives to be careful what they put in writing. He asked them to fax any investigative reports related to Europe. "It is OK to write various things to me in emails written in Japanese," he advised, "but as much as possible only send materials that would not be controversial if disclosed (namely, things that have been reviewed), and it is best, I think, to discuss things orally."
The documents make it clear that in-house lawyers and public relations personnel worked together to craft a strategy. Christopher Reynolds, the U.S. general counsel, advocated defending the electronic throttle control by seeking "validation" by a panel of experts. Using the Japanese word for building consensus to act, he wrote in December 2009 that he hoped "we can finalize a nemawashi plan this week and begin to implement it."
One of the weaknesses in Toyota’s defense was flagged in an email sent by assistant GC Webster Burns the following April. Commenting on a demand letter from NHTSA that the company pay a $16.375 million fine for delaying its sticky pedal report, Burns wrote: "We need to keep in mind that we continue to find significant differences within Toyota about the significance of the sticky-pedal phenomenon which will be exploited by NHTSA in any litigation."
Some documents require translation by specialists. An undated spreadsheet showed test results of an engine’s electronic throttle control system, including numerous faults that the document said cause sudden acceleration. "My guess is they were fixed in development," says Hannemann, who has been hired by plaintiffs suing Toyota, and also by the defense in a suit against a Toyota dealer. "But this shows you have to find issues during testing. And how do you know you catch them all?"
Several documents illustrated what the experts describe as a propensity of Toyota employees to define problems as they wish them to be, regardless of the facts. One is Toyota’s analysis—performed three days after Saylor’s death—of car owners’ complaints received by NHTSA. Some drivers described their own harrowing experiences. Several were adamant that theirs had nothing to do with floor mats, yet that didn’t always matter to Toyota’s reviewer.
One woman riding in a 2006 Toyota Tacoma said that it was the third such experience she’d had with the car. "Two times previously Toyota has replaced the cruise control," she reported. "This is not a cruise control problem. This is a gas pedal issue. I was told previously the mat was under the gas pedal. This is hardly the problem." In the column provided for the cause, the reviewer wrote: "The mat catches (specifics unknown)." It was the most common cause listed on the chart, regardless of what the drivers had to say. Antony Anderson, an independent electrical consultant who specializes in electrical machine and control system failure investigations (and has provided independent expert testimony for plaintiffs who sued Toyota), says the document shows how Toyota’s "poor analysis" makes it appear that the incidence of stuck floor mats "is very much higher than it really is."
Another example of preemptive answers appeared in an undated email written by an engineer. He asked if acceleration can be caused by radio wave interference. Then he recounted his earlier experience with interference: "Previously, when I was in charge of Hilux [a truck model] in the Japan domestic service division, I experienced an engine stall malfunction due to radio wave interference from a nearby U.S. naval base in Yokohama. At that time I was told that it could absolutely never occur." Keith Armstrong, an expert in electronic circuit design as well as electromagnetic interference (EMI), says the idea that radio waves can’t cause electronic malfunctions is absurd: "I know of no expert in this field who doesn’t work for the auto industry (and some who do) who would ever make such a ridiculous claim." Armstrong has advised electronic suppliers on EMI safety issues, and he has also twice advised NHTSA, at its request. (He has not been involved in Toyota litigation.)
Anderson and Hannemann are even more troubled by an email exchange between Michiteru Kato, a general manager based in Japan, and Tinto in D.C. In messages dated October 11, 2007, the two were discussing television coverage of sudden acceleration in the Tacoma. Tinto wondered whether the mothership was looking into the situation. Actually, Kato replied, headquarters had not received any technical field reports from dealers or regional offices "because as you know, the sudden acceleration or surge issue usually can’t be duplicated by the dealer and they can’t find any abnormality on the vehicle. In those cases the dealer does not make the field report." Consequently, he added, "Toyota does not know what’s happening on the Tacoma vehicles and just started the investigation."
Hannemann found it more than odd that Toyota was, as the email makes plain, getting information about its own problem cars from NHTSA, the media, and Internet forums. "You should be telling NHTSA things, not the other way around," says Hannemann, who has worked as a product development engineer at Chrysler Corporation and as a chief engineer at the Ford Motor Company. (Toyota says that in April 2010 it established a new program to investigate all reports of unintended acceleration. It attempts to contact individuals within 24 hours to arrange a full analysis of their vehicles.)
The documents also revealed introspective moments during which executives considered where their company went wrong. One Japanese exec, identified only as Takimoto, wrote in March 2010: "All of the current problems were caused by the low level of completeness of vehicle development during the time period when I was in charge. I am really very sorry." Another executive ruefully admitted in February 2010 that quality was hurt by the fact that "the numbers of prototype vehicles, production vehicles, and quality assurance test vehicles were dramatically reduced," as were "test vehicles for evaluation and quality verification."
The experts who reviewed the documents offered their own assessments. Brian Kirk, the founding director of Robinson Systems Engineering Ltd, which specializes in safety critical software and systems for the transportation industry (and is not involved in Toyota litigation), says that the engineers "seem to be genuinely trying to understand the problems and provide practical solutions within the constraints of legacy and time pressure. However, there is no apparent safety engineering process forming a rigorous basis for understanding and solving the issues." Hannemann also finds a general lack of rigor. When technicians investigate complaints, they don’t seem to press to find the root cause. "It seems that their problem solving is focused on something that’s predetermined," he notes. And if they’re not going to rigorously test cars prior to production, then they need to listen carefully to complaints from consumers, who are essentially doing the testing for them. But the company wasn’t doing that either, he says.
The problem of sudden acceler­ation emerged after electronic controls were introduced into cars in the late 1970s and early 1980s. Before then the issues were driver error and mechanical problems—like a throttle return spring failing. Diagnosing electronic failures, on the other hand, is much more challenging. As Keith Armstrong puts it: "Electronics in its very nature is weak, unreliable, sensitive." And when a component fails, it doesn’t necessarily leave evidence. "Think of your PC," he says. "Sometimes it will crash and you’ll reboot it. And if someone then asked, ‘Where is the evidence?’ you may not be able to show them."
Sudden acceleration is still a very rare event, but unlike operating a computer, a malfunction in a car can cause serious injury or death.
Toyota isn’t alone in struggling with this problem. Ford has had its own sudden acceleration problems over the years. They were the centerpiece of the 2003 book Sudden Acceleration: The Myth of Driver Error, which was partly funded by a product liability award in an SUA case paid by the company.
For Ford and some of the other manufacturers, the big problems began with the introduction of cruise control. That was the function that was linked to early incidents, and some groundbreaking lawsuits, though causation wasn’t easy to prove. The companies preferred to blame the drivers, asserts Clarence Ditlow, the longtime director of the Center for Auto Safety and one of the authors of the book. The book estimated that only about 1 percent of SUA episodes actually result from "pedal misapplication," as the companies call it.
Ditlow reserves some of his harshest criticism for NHTSA. "NHTSA has been controversial from the beginning," he observed in the book. "It was criticized by the auto industry for being too aggressive in its regulation, and by consumer advocates for being too weak and responsive to the industry."
On one level Ditlow can commiserate with the agency, which is notoriously underfunded, he notes. But he can’t excuse its handling of SUA complaints. It has adopted the same attitude, Ditlow says, as the car manufacturers: If you can’t find a failure, it must be driver error. "I think that’s wrong," he said in an interview. He also sees the same posture in NHTSA’s 1989 report on sudden acceleration that has been cited many times by the defense in product liability lawsuits.
The main battleground has been the courts, where plaintiffs have made slow progress convincing juries and judges that electronic malfunctions are real. But translating the evidence into a winning formula hasn’t been easy. Proving a circumstantial case rarely is. "We have the burden of proof, and we should," says Molly O’Neill, who works with the dean of SUA trial lawyers, Thomas Murray of Sandusky, Ohio. "But you cannot open up the car and show what went wrong. That’s the nature of electronics."
Before one of Murray and O’Neill’s biggest wins, a trial judge ruled that their expert couldn’t mention EMI, which was an important part of their explanation of how the accident happened. In the Daubert hearing, the judge said she was concerned that the expert’s findings couldn’t be replicated in tests. They later won Jarvis v. Ford on appeal in 2002, in a decision written by Sonia Sotomayor before she was elevated to the U.S. Supreme Court. Murray says it was one of the first successful challenges to the electronics in a car’s cruise or throttle control. Yet, for EMI, it was another argument lost in translation.
Asked to respond to criticism of its company, Ford offered a full-throated defense of its regulator: "NHTSA has investigated alleged unintended accelerations many times over many years and has concluded that driver error is the predominant cause of these events. NHTSA’s work is far more scientific and trustworthy than work done by personal injury lawyers and their paid experts." The emailed statement concluded: "Ford has reviewed its own data and determined that its vehicles are not affected by the problems experienced by Toyota owners."
In an emailed statement, NHTSA also defended its enforcement, citing Toyota’s recalls and fines. And in this case, it said, "we went above and beyond our regular review process" and chose NASA to investigate. The results "made clear there was no evidence of any electronic cause of sudden, high-speed unintended acceleration in the Toyota vehicle models that were the subject of the study."
By 2009 the SUA front was quiet. Three years earlier, a Missouri appeals court had overturned an $80 million SUA verdict won by attorney Mark Evans against General Motors Company—the largest by far. Some of Murray and O’Neill’s best cases had settled, and NHTSA had closed its SUA investigations. Then came the Saylor crash—and suddenly everything changed.
It was the 911 tape that did it. Posted on YouTube, it seemed to make the danger real for a larger audience—maybe because no translation was required. And it also seemed to give people who had experienced similar sudden acceleration permission to talk about it—and the social media means to do so. "Without that crash that was caught on a 911 tape," says Ditlow, "no one would have paid any attention to [sudden acceleration]. And there would have been no controversy."
As the tempest grew, the pressure mounted on Congress, NHTSA, and Toyota. The company shifted into crisis mode, and the Japanese executives slowly began to listen to the American managers who understood the U.S. legal landscape. Reynolds, the U.S. general counsel, began calling the shots along with the public relations people. But it wasn’t always easy for them to convince the bigwigs in the motherland. When Congress began planning hearings, for example, the initial word from Japan was that Toyota’s president would sit them out. It took some time to bring him around.
The biggest challenge during the hearings came from a surprising source. David Gilbert was an automotive technology professor at the University of Southern Illinois in Carbondale, which, it so happens, receives resources and funding from Toyota. But he came to the hearing to talk about an experiment he’d cooked up that challenged claims Toyota made about its electronics. Toyota insisted that any electrical fault in its cars would trip an error code, which would immediately reduce power and send the car into "limp home mode." Gilbert decided to test this assertion by rewiring Toyota’s throttle in a way that would mimic a short circuit and send the rpms surging. Then he’d check for an error code.
At the hearing he revealed what he’d found (previewed the night before on ABC News). The cars he tested hadn’t produced the code, suggesting a vulnerability in the system. The politicians seemed deeply impressed, and Toyota was caught off guard.
Toyota arranged a multipronged "rebuttal." It had hired Exponent, a scientific and engineering consultancy, to work on technical issues in connection with the inquiry. It was now tapped to respond to Gilbert. Toyota also asked Chris Gerdes, who directs the center for automotive research at Stanford University, to review Gilbert’s experiment. In March the company uploaded a video.
The centerpiece was a replication of Gilbert’s experiment using cars manufactured by a half-dozen of Toyota’s competitors. And they all performed exactly as Toyota’s: The engines raced, and no error codes were detected. Mike Michels, a vice president of communications, asked the experts the significance of what they’d seen. Gerdes and two Exponent consultants suggested that the demonstration bore no relation to what happens in the real world. In essence it meant nothing, they argued.
Behind the scenes, Toyota played hardball with critics. A public relations manager named Masami Doi had spelled out the approach in a December email. "There are at most around 10 people who are the sources of negative tone communications. If they can be suppressed, I think we will be able to manage it somehow. Like you said, let’s go with an intention of destroying each individual person’s ability to oppose us, one by one. (To do or not to do is a separate question.)"
The individual who undoubtedly felt the most pressure was Gilbert. He did not respond to requests for an interview. But he did talk to The Huffington Post several months after these events, and the website reported the pressure Toyota applied in an apparent attempt to force him to recant. Toyota’s outside lawyers met with Gilbert and university officials "to discuss Gilbert’s use of donated Toyota vehicles." Gilbert was also pressured to fly to California to watch demonstrations prepared by Exponent. He returned unswayed. An alum who works for Toyota wrote an email to a university official suggesting that Gilbert be fired. But he survived—and claims that he has no regrets. It was the right thing to do, he says, "because that could be someone’s life that I could be saving."
In a statement, Toyota denied that it cut its support of the university. The employee who suggested that the school fire Gilbert sent a personal email "and in no way represents Toyota’s position on the matter."
All the attention put NHTSA in the hot seat. When Ditlow testified before a Senate committee after Gilbert had addressed one in the House, he called for a "fully public" investigation of Toyota’s electronic controls with "independent scientists and engineers with no ties to the auto industry."
Joan Claybrook had also been pressing NHTSA to seek outside help. Claybrook, who was NHTSA’s leader during the Carter administration, often works closely with Ditlow, and they both met with LaHood and NHTSA’s current administrator, David Strickland, as the investigation was gearing up. "We urged that they go outside of the agency" in choosing investigators, she says, "because the agency didn’t have a lot of expertise in this area."
Claybrook, a frequent NHTSA critic, was pleased when the agency chose NASA to investigate. "They were totally independent, and they have a lot of skill," she says. "Unfortunately, they don’t have a lot of experience with automobiles."
The worst thing that came out of the whole investigation, Claybrook and many others say, was the widely repeated sound bite from Transportation’s LaHood: "The jury is back. The verdict is in. There is no electronic-based cause for unintended high-speed acceleration in Toyotas. Period." That was not what NASA’s investigation concluded, Claybrook says. It was another one of those bad translations.
Some NASA insiders who were close to the investigation were equally dismayed by LaHood’s pronouncement. One NASA scientist who requests anonymity because he isn’t authorized to speak to the press puts it this way: " ‘We didn’t find anything’ and ‘there isn’t anything to find’ are not the same." In the relatively short time they had to investigate, he says, the NASA team didn’t see anything conclusive—that’s what the results said. The scientist continues: "I’m not clear on why the devil we got the job in the first place."
Perhaps the most frustrating aspect of the task, from what this man could tell, was working with people from the Transportation Department (including at least one from NHTSA). His colleagues complained to him: "Every time we find something, we’re told it’s not what we’re looking for." It had to be unintended acceleration, not erratic acceleration. And it had to involve brake failure. If they ran across data that involved UA but no brake failure, they were told to ignore it. This struck him as a strange kind of "independent" investigation.
What else would you not know from NASA’s report? They were expected to explain why certain cars had failed, but they were given no access to those that actually had, he says. The failure analysis team was given little more than a month to work on one accelerator pedal from one car that "misbehaved"—the only such part they ever saw—and it wasn’t even from a runaway. Some of the scientists on this team wouldn’t sign NASA’s final report. And other individuals balked, too, he adds. (A NASA spokesman referred requests for comment to the Transportation Department, which did not respond to detailed questions about the investigation—nor did NHTSA or LaHood.)
Like the NASA scientist, translator Betsy Benjaminson came to her work without preconceptions. Born in Cleveland and now living in Sderot, Israel, Benjaminson is a freelancer who works for translating agencies based in the United States. She’s translated all kinds of documents, including many for companies with legal cases. But the SUA litigation was different, she says.
The biggest difference: The story isn’t over. "I felt that I am not just translating past events, but that I actually saw that these things continue to happen," she says. "And lives of real people were potentially still at risk."
The feelings that led her to share the documents had nothing to do with a grudge against Toyota. As someone who had lived in Japan for four years and had long admired Japanese culture, she was favorably inclined toward the company. She also had fond memories of her first car, a "cute red Toyota Corolla." And during her earlier work for Toyota, she had enjoyed a bond with the U.S. project management team. "We were a tight-knit team that related to one another as friends," she says. But eventually she had to set that aside.
The change came slowly. She began working on Toyota litigation in 2010. Before then, she’d been "oblivious" to the events in the U.S., she says. Slowly she began to notice "odd things" in documents she saw in connection with her role as translator. Revised press releases sometimes obscured important details, she says. Emails among engineers "revealed facts that directly contradicted" Toyota’s public statements.
Then it got worse. She read reports about runaway cars, including survivors’ accounts of crashes that killed their companions. She was deeply affected. A "tipping point" came when she read a document the company had prepared based on complaints filed with NHTSA. "A summary of the injuries and deaths was attached," she recalls, "and it was cynically titled ‘Souvenirs from NHTSA.’" For her, that was it.
After all of Toyota’s strategizing, it could not have anticipated how those words would sound to a translator—or what she would do. It was one more thing lost in translation.