Today’s Jobs Numbs: Can We Handle the Nuance?

Friday, August 3, 2012 at 9am
By John Friedman

On the first Friday of every month, at 8:30 am East Coast Time, the Bureau of Labor Statistics releases a bunch of numbers — no, let’s call that a trove of numbers —  about employment during the previous month. That bunch of numbers, via media echo-chamber, gets winnowed down to but two numbers, which today are 8.3 and 163,000.

The similar numbers from a month ago were 8.2 and 80,000.

Those numbers represent, using the lay terms, the unemployment rate and the number of jobs created last month (experts and the like use different, more precise, less accessible language, such as the U3 rate, but this blog is about lay ideas, and is intended for lay people with lay lives).

So today we wake to the news that unemployment went up (which is bad) and so, too, did job creation (which is good).

*  *  *

Most non-lay people view the jobs creation rate as the more interesting number. It is, like most statistics, susceptible to a lot of criticism because (a) in seeking to divine the number of jobs created last month in the entire country, it must necessarily rest on a sample of the entire thing, and, because (b) the BLS, like statisticians everywhere, believes that making adjustments for measurable phenomenon, such as seasonality, are not only do-able but an important part of generating statistics that measure what they are intended to measure, and therefore adds seasonal adjustments into the mix.

In JFF parlance, then, the BLS definitely sprinkles on some secret sauce which, upon a quick search of their site, is only described generally (though my hunch is that, if you’re a wonk and you want to find out how they do it, there is a decent amount of info out there to be had on the topic). The good news is that, overall, and judging by the measly pickings from a search of BLS numbers-rigging scandals online, BLS’s reputation for objectivity appears to be strong.

*  *  *

So what does the population do when the two main numbers that come out are both good and bad? Hmmmmmm . . .

Well, in the political realm the answer is quite clear: you accentuate the number that best supports your story. So the Repubs will stress the unemployment figure going up, and the Dems will stress the jobs-creation figure going up.

In the economics and investing realm, though, you’d have to give the nod to the jobs number. For instance, as I write this the stock market is up 2%, so clearly today’s numbs-trove leads market participants on the whole more hot-to-trot on the buying side rather than on the selling side, as in I’ll buy that from you at x, followed by, Well, you can have it at a penny more than x, after which, after a bunch of similar back and forths, it’s the seller who is mostly setting the price, which means up.

And what of the lay people of the land? What do they hear?

*  *  *

In 2004, just about exactly four years ago, we learned that nuance was a bad word. As someone who really, really enjoys his nuance, this was disheartening.

It’s safe to say, though, that my heart had and has no say in the matter, and that an awareness of subtle distinction is not where we, as a people, are coming from these days. We often just can’t handle the nuance.

For many people, then, only one number came out today. The one that supports the story they like.

 *  *  *

Barry Ritholtz, via his blog The Big Picture, used to be my go-to guy for a first uptake of the jobs numbers in the morning. These days he does less writing on his blog, and he certainly doesn’t jump online to post numbers first thing in the morning on release day like he used to, so now kudos go to Meteor Blades, a regular at DailyKos, whom I thank for his early-morning wake up (he is a West Coaster) and for doing this table, which I think really screams out about what’s what in terms of overall job creation for the last ten Julys (scraped directly from his posting today):

July 2003: +  25,000
July 2004: +  46,000
July 2005: +374,000
July 2006: +209,000
July 2007: –  40,000
July 2008: -210,000
July 2009: -339,000
July 2010: –  58,000 (worsened by Census layoffs)
July 2011: + 96,000
July 2012: +163,000

Yes, my friends, over the last ten years, only two Julys — 2005 and 2006 — saw better job growth numbers than we saw today.

*  *  *

Do these numbers surprise you? I myself was surprised at (a) the weakness of 2003 and 2004, and (b) how 2012 was the third best July out of the last ten, and (c) what is a very clear trend over the last four years.

Am I only seeing what I want to see? Am I ignoring nuance?

Hope not. Hope I — we — are better than that.

746 words

◄ Back to Blog
Categories:
Tags:

Leave a Comment

Name
Email
Website
Message