This Jobs Report Is Probably Wrong, So Let's Look at the Last 20, to Be Safe

Every number in today's jobs report is almost certainly wrong. Every single one.

That's not an indictment of the Bureau of Labor Statistics. It's just the nature of statistics. The numbers you've seen quoted in every analysis of the jobs report today will change, possibly dramatically, over the next 60 days. "Over the past three years, the employment report has understated how many jobs were created by as much as 99,000 and overstated it by as much as 86,000," Zachary Goldfarb writes in the Washington Post. That means we might have created anywhere between zero and 200,000 jobs in August.

With the knowledge that this jobs report is probably wrong, I collected and summed up the last 20 jobs reports, going back to the beginning of 2011, to get a clearer sense of things. Over the last year and a half, this is what a perfectly typical jobs report would look like:

"Last month, we created about 140,000 jobs. Government shrunk by 17,000 workers. Health care added the most jobs -- about 24,000, followed by food services. Mining and logging grew the fastest, followed by food services. Construction and retail were basically flat."

As it turns out, that's awfully close the month we just had. So, to the popular suggestion that last night's jobs report spooked the president and compelled him to give a more somber address ... let's just say: It's possible! But it's not likely. This jobs report didn't change everything. In fact, it was, depressingly, normal. Like so, so, so, so many jobs reports before, it confirmed a very clear picture of the labor economy: Things are getting ever-so-slightly better, and you need a microscope to see the rate of change.

Two graphs for the road. The first shows the percent change in jobs in some major sectors over the last 20 months. The second shows total jobs added (or, for government, lost) over the last 20 months (in thousands).

Screen Shot 2012-09-07 at 3.27.55 PM.png
Screen Shot 2012-09-07 at 3.37.59 PM.png