TangleBones

There’s an ongoing debate about how much the public can trust the information they get from bloggers. The gold standard, of course, to which people compare bloggers are journalists. Journalists, the occasional Jayson Blair aside, have a pretty good reputation for fairness and accuracy. And bloggers certainly can’t compete against that standard.

The problem is that the reputation, earned by a few reporters, is evenly applied to all of them. Obviously, this is just as unfair as painting all bloggers as partisan douchebags because of… well, you know can fill in that list as easily as I can. Heck, stick me on that list.

Which brings me to the point of this whole diatribe: I was reading Jalopnik the other day for my dose of snarky automotive news (vitamin SAN!), when I came across an article about recalls. Now, as a consumer obsessed with quality, reliability and safety, this is the kind of thing I’m fascinated with. You see, recalls are often seen as a sign that a company makes lousy products. But the way you measure the quality of companies that make children’s safety equipment (e.g. car seats, strollers, high chairs) is by how many recalls they order. You see, companies in that segment that don’t give a damn about your kid cracking his skull open don’t do much safety testing, especially on shipping products, and they don’t issue many recalls.

Automotive recalls are a bit stickier. Some recalls are initiated by the automaker. Others are forced on them by government regulators. Some are the result of lawsuits brought by consumers. It’s almost impossible for the consumer to know which is which.

People who saw Fight Club know about the recall coordinator’s formula:

Take the number of vehicles in the field “A”, multiply it by the probable rate of failure “B”, then multiply the result by the average out-of-court settlement “C”. A times B times C equals X. If X is less that the cost of a recall, we don’t do one.

In the book, the narrator goes on to describe how this means that they do a lot of recalls on silly things like windshield washers, but tricky electrical problems that are expensive to correct and hard to prove get settled by lawyers.

So when you see that a company has a lot of recalls, you have to determine the cause of the problem, the cost of the components, and the reason a recall was initiated to determine if the automaker in question is a scummy manufacturer of unsafe vehicles or a responsible company that is concerned with your safety and well-being.

Needless to say, this isn’t what most people do. It’s hard. So they just count the recalls and stay away from companies that have a lot of them.

And sadly, even that is too hard for some journalists.

Back to the story: Jalopnik’s post was nothing more than a link to an article by BusinessWeek Online called The Most Recalled Cars 2006. I read the article, saw that they provided a small amount of context to the recall issue, and felt disappointed that they didn’t try to do any real reporting. You see, their article was just a pointer to some information they dug up on Edmunds.com about which cars had the most recalls.

As the owner of a 2005 Honda Accord, I was a little surprised to see the 2006 Accord listed in their top 12. I’ve been checking on Honda recalls pretty regularly (I said I was obsessed) and didn’t remember seeing 3 for the Accord. So I jumped over to the National Highway Traffic Safety Administration: Office of Defects Investigation Recall Search (I have it bookmarked), and sure enough, there were 3 recalls listed for the Accord. Upon closer inspection, I realized that one of them wasn’t from Honda at all – it was for an after-market manufacturer of replacement headlights. Let’s be clear: these aren’t Honda parts. They aren’t made by Honda, sold by Honda, or recommended by Honda. These headlights are the ones preferred by the Kids These Days™ because they’re whited out and don’t contain the little amber bits that will ruin their rep.

Even closer inspection made me realize that this didn’t just affect Honda Accords: this recall covered several Hondas and Chevys and Fords and Toyotas. And then I remembered this paragraph from the article:

Moreover, as parts sharing has increased, recalls often encompass a variety of brands and vehicle types. One 292,000-unit recall due to faulty headlamps from a third-party supplier, Walnut (Calif.)-based Anzo USA, for instance, affects Hondas (HMC), Toyotas, Fords, and Chevrolets alike.

The recall that they’re talking about is the one I found listed for the Accord – but that paragraph makes it sounds like a) the headlamps are original equipment, b) the same part is used on all of those cars and c) that the part was failing. In fact, this is a whole line of completely different after-market parts that were shoddy by design.

So I dug a little deeper. The top vehicle on their list with 4 recalls was the Chevy Silverado 1500. But two of those recalls were for after-market headlamps. The second vehicle was the Dodge Durango with 4 recalls. 1 minute of searching turned up 6 actual recalls, all of them real, all of them affecting more vehicles than either of the 2 recalls affecting the Accord, and several of them very serious. The funny thing was that the last of these recalls was initiated two weeks before the BusinessWeek article went to print, but even throwing out that one, they were still off.

So, being the Web 2.0 collaborative kind of guy that I am, I left comments on both Jalopnik and the BW site letting them know that the data was faulty. Here’s what I got back from the author of the article:

Hello readers, this is Matt Vella, the author of the story. As was mentioned in the text, recalls are hard to get a handle on for several reasons, including multi-manufacturer parts sharing and the complexities of manufacturer-initiated reporting. In reference to Jemaleddin’s points, we included the head lamps recall to illustrate how some recalls range far beyond one brand or manufacturer. The chart within the article was created with research in conjunction with Edmunds.com’s database. But as you point out, these number fluctuate constantly. This piece is intended to take a broad look at the trend. Thanks for reading!

Of course, my point wasn’t that the numbers fluctuate. It’s that they got them wrong. And guess what: they still didn’t correct the problems. Blaming Edmunds.com is a nice touch, but doesn’t solve the problem.

So I commented again, giving them all the necessary data to fix the story. That one they didn’t post. Maybe it was the part where I questioned their commitment to the facts. Whatever. Here’s the real data:

To sum that up: half of their data is wrong. And what are they planning to do about it? Evidently nothing. Meanwhile, people visiting their page, and pages that have copied their lousy table like Jalopnik, will get not only an incomplete picture of the recall situation, but an incorrect one.

Now, will this cost any of those automakers sales? Probably not enough to hurt them (though GM is on the brink) since the pain is spread around so many companies. But the fact that BusinessWeek would rather throw Edmunds.com under the bus than fix the data that I was able to provide them with less than 10 minutes of research says something pretty sad about the state of journalism. The days of big time bloggers being able to delete valid criticisms and dodge hard questions are over. Maybe it’s time we hold journalists to the same standards.