Why Friday's jobs report should be taken with a grain of salt

Three times today I’ve been asked on media outlets about the likely effect on the presidential election of Friday’s jobs report, depending on what the Bureau of Labor Statistics announces.

Unfortunately, the BLS report is likely to sway some voters — and therefore have an impact on this tight race. But it shouldn’t.

The report surely will be used by one of the candidates to make a sweeping case for himself and against his opponent. An unemployment rate above last month’s 7.8 percent, or a number of new jobs below September’s 114,000, will be wielded by Mitt Romney as evidence the economy is slowing – while a rate below last month’s, and a number above it, will be used by Team Obama as evidence the economy is moving in the right direction. 

In truth, no one should base their vote on tomorrow’s labor report is because a single month’s report isn’t a reliable gauge of which way the economy is heading. A report that the unemployment rate is 7.8 percent, for example, means only that the Bureau of Labor Statistics is 90 percent sure the real rate lies between 7.6 percent and 8.0 percent. By the same token, an announcement that the economy created 100,000 jobs in October means the BLS is 90 percent sure the number of new jobs is between 90,000 and 110,000. 

A small shift one way or the other from September’s readings may have huge political significance but has no statistical significance, and therefore should have no political significance. 

So much uncertainty attaches to a single report that the BLS is continuously revising its job numbers as more information comes in. At the start of September the BLS announced that August’s job growth had slowed to 96,000 new jobs — a finding that had a noticeably negative effect on the President’s campaign and at a critical juncture. But a month later, in its October release, the BLS revised the August number upward to 141,000.

Or look back on that awful year of 2008, when between July and August the Bureau was reporting relatively small declines of 20,000 to 60,000 jobs per month. We now know the true job numbers were falling off a cliff. BLS’s revised job numbers showed the economy was losing roughly 200,000 jobs a month, starting in April. 

My advice: Regard tomorrow’s BLS numbers with a good grain of salt. Month-to-month reports shouldn’t be taken nearly as seriously as they are. It’s the long-term trend that counts. And here the picture is encouraging regardless of what’s reported tomorrow.  Since the start of 2012, employers have averaged 146,000 new hires a month. It’s a slow rate of recovery, but a recovery nonetheless.