Dumb trend spotting

An article in today's NYT's entitled Redefining Business Casual reminded me about another frequent annoyance I have about my chosen profession; the anecdotal observation that passes itself off as a behavioral trend based on some kind of data. In recent weeks, I've had to deal with more than my share of these unsubstantiated claims, expressed with great passion and authority from: "empty-nesters are moving back into the city," to "married people have more sex than people that just live together" to the claim in this article:

“I see a return to more traditional business wear...People dress up more in times of financial uncertainty and intense competition. It helps their sense of stability.”

The source of this information: the highly objective president of the Marcraft Apparel Group, which champions suits and ties.

All the evidence I've looked at lately suggests the opposite: a gradual downward slope in men's formal business wear (suits and ties) that's been going on for over a decade. This isn't to say that some fashion-forward types aren't reclaiming the tie or jacket in certain situations, but we are a long way from returning to 5 suits a week for most men in most industries.

Nevertheless, journalists and market researchers of a sort are drawn to these claims because while they seem to be superficially surprising (really!) they are also comforting. They are almost always deployed to confirm a some deeper truth about human nature It's illustrated above in the remark from the president of Marcraft Apparel Group. Men are wearing more suits. Really? Why? Because we dress up to give us a sense of stability in uncertain times. But of course. That does ring true. Except for the fact that suits are really expensive and most of us don't require them for work. So no matter how comforting a nice three-piece might be: they'd be a big waste of money

So while these facts are interesting and make us feel capable of penetrating into the secret motives driving human behavior, they usually mark the exceptions rather than rules. it's not that they're untrue. They're just not trends, at least in the traditional definition:

A statistically significant change in performance of measured data which is unlikely to be due to a random variation in the process.

The truth is I'm not that interested in trends to begin with. As I see it, my job isn't to predict the future, but to describe the present, which is hard enough.