The fuzzy math of October’s ‘new jobs’ data

I think the employment report Friday will be stronger than the “experts” expect.

I always put the word “expert” in quotes when it comes to the jobs report because nobody is an expert. Everyone is a guess-pert. Me too.

Only I think I know a secret regarding this job report, which is crucial because it is coming out right before next Tuesday’s congressional elections.

First, let me tell you what the “experts” are guessing: 200,000 new jobs in October, give or take a couple of thousand.

Keep in mind what I said in Tuesday’s column. The seasonal adjustments to recent job figures have been suspiciously off. And when the Labor Department decided to revise August’s job figures upward by an unusually large amount, it was quietly guaranteeing that the September growth would be muted.

And it was. Only 134,000 new jobs were reported in September, which was well below what the “experts” were expecting.

So the October report will be building upon that muted September figure. And that alone might help October’s report look stronger than expected.

But that’s not the secret I know.

There are other things that will affect the October report. You might remember that there were hurricanes, which probably caused some companies in the South to close temporarily.

But, by the same token, those storms probably created a lot of jobs (and overtime) for people who worked to clean up the mess.

The Labor Department might have a guess as to whether the hurricanes were a pro or con for the job figures. A lot will depend on whether companies were able to report their job figures to Washington in a timely manner or, if not, whether the Labor Department took it upon itself to guess at the size of payrolls.

OK. I’ve kept you in suspense long enough. Now for the secret that I know.

It’s not really a secret at all to people who have been reading this column. And if the Democrats are smart, they will pay close attention because if the job figures on Friday are stronger than expected, they are going to need to have a quick explanation for voters, for whom the economy will be the main issue when they vote.

The secret is the Labor Department’s birth/death model. In short, each month the Labor Department guesses as to how many companies were “born” or “died” during a given month and how many jobs the new companies created.

And in October, for whatever reason, the Labor Department always guesses high. In October 2017, for instance, the Labor Department added 247,000 jobs for these newly created companies.

Only in April was the Labor Department’s guess slightly more generous.

This year’s estimates should be about the same.

You might be thinking that the estimates Labor makes are fair since jobs are created in October because of the upcoming holidays of Halloween, Thanksgiving and Christmas. But not so fast!

Those sorts of things should be picked up in the seasonal adjustments, which account for unusual events that affect the economy each and every year.

No these estimates are over and above the seasonal adjustments.

But I have to explain one other thing. The 247,000 birth/death jobs that were added in October 2017 were before seasonal adjustments. I know it’s confusing but it means that not all of those 247,000 jobs will show up in the headline number you’ll see at 8:30 a.m. Friday.

How much will this surge in estimated new jobs affect the total number reported? Nobody, including the people at the Labor Department, knows. It’s a secret.