On Uncontrolled Acceleration And Black Boxes

Last week, big business shill Theodore H. Frank wrote an op-ed drawing on data from a Los Angeles Times article reviewing the fifty-six fatalities attributed to sudden uncontrolled acceleration problems with Toyotas. Frank noted that, in about half of the car crashes, the driver’s age could be ascertained from the LAT‘s compilation and the ages of the drivers skewed to the elderly.
The next day, blogger Megan McArdle tracked down the ages of “all but a couple” of the drivers involved in the Toyota crashes and revealed that the “overwhelming majority” were over fifty-five years old.
A lot of people have hypothesized that the sudden uncontrolled acceleration accidents involving Toyota might be caused by a computer or electronic bug in the cars’ throttle. Since there’s no reason to believe that Toyotas with a computer bugs would discriminate against older drivers, Frank and a host of other bloggers* trumpeted the results as proof that there is no electronic problem with Toyota’s computerized engines and that, in fact, the blame lay with older drivers’ driving skills (or lack thereof). (Question(s): McArdle used a cutoff age of 55 and up. Are 55 year olds, in today’s world, frail or senescent? Most research does not show a significant decline in driving ability until a couple of decades after 55 and I know many people in their sixties who are in far better physical shape than I am. What would her findings have been if she included only drivers 70 and up?).
Ted Frank and a bunch of his colleagues from the (shallow end of the) think tank business used the findings to question the honesty of drivers who reported uncontrolled acceleration problems, likening them to frauds like “balloon boy.”
So what should we conclude? Should we conclude that the whole “Toyota panic” is merely a media-driven phenomenon about routine errors committed by all elderly drivers?
I don’t think so. As I blogged over a month ago, in 2009 forty-one percent of complaints of sudden uncontrolled acceleration involved Toyotas, while Toyota only held sixteen percent market share – a fact that was lost on a lot of people. Since the time I posted that blog, NPR’s Robert Benincasa did something that the government does not do – track reports of sudden uncontrolled acceleration by make and model – and found that, since 2002, Toyota has seen a troubling rise in complaints of sudden uncontrolled acceleration. The problem doesn’t seem to be old people and driving; the problem seems, if anything, to be old people and Toyotas specifically.
In addition, the “older driving theory” doesn’t account for the most spectacular Toyota crash of all – the (physically fit) California state trooper whose recorded conversation with a 911 operator details his efforts to get his Lexus to brake.
Ultimately, I think we – whether as consumers or jurors or simply concerned citizens – need to come to grips with the fact that there may be a problem with Toyotas that we may never directly explain. A lot of people have theorized that Toyota’s problems may lie with a computer bug inside its engines. (Competing explanations – floor mats, driver error, etc. – don’t seem to account for the disproportionate number of Toyotas involved in these crashes). If it’s the case that there’s a computer bug that plagues Toyotas, we may never find out precisely what it is and why, in some cases, it caused crashes. Toyota’s engines may forever remain to us a bit of a “black box” – a computerized system that we can’t see inside or fully understand.
People tend to assume that, if there’s a computer programming error, we can simply pore over the code and figure out if there’s an error. After all, computer programming is just logic and logic is supposed to be completely transparent. But, as science fiction writers like Isaac Asimov have shown us, you can start with a few logical principles that dictate the behavior of computers or robots and wind up with some completely unintended consequences.
We are all familiar with real life examples of this. One dramatic, and fairly recent example, was the Great Northeast Blackout of 2003 (which was caused in part by computers behaving in unexpected ways). Giant companies like Microsoft come out with products like Windows Vista that are so ridden with programming problems that they become unsalable.
Sometimes the bugs are never figured out. When a program that you’re running crashes, often the product’s designer has no reason why it crashed – that’s why, after the program returns to life, it asks you for permission to send a report to the manufacturer for analysis. My friends in computer programming tell me that, very often, software engineers are unable to untangle the reasons for these errors.
We may never get to the bottom of Toyota’s uncontrolled acceleration car crashes. But that does not mean the problem is not real. Or that Toyota should not be held accountable for its failure to investigate and address these issues.