May 28, 2017

Economix Blog: Understating Job Growth



Dollars to doughnuts.

One of the many bright spots in the June jobs report was that April and May had their employment gains revised upward by a combined 70,000 jobs. In other words, job growth in those months was actually better than originally reported.

Which brings me to a peculiar trend, first brought to my attention in May by Justin Wolfers: Most months during the recovery — 37 of the 47 months for which we now have a third estimate of employment — the Labor Department initially understated job gains. And most months during the recession — 13 of the 18 months — the bureau initially understated job losses.

Source: Bureau of Labor Statistics, via Haver Analytics. Source: Bureau of Labor Statistics, via Haver Analytics.

That is to say, the revisions have been largely pro-cyclical. Job changes were better than we originally thought when they were good, and they were worse than we originally thought when they were bad.

I’m not sure what explains these patterns. Revisions in previous cycles do not seem to have been as heavily pro-cyclical; there was more of a balance between upward revisions and downward ones in both recessions and expansions. (If you have any potential explanations for the change — other than pure chance — please share your thoughts in the comments.)

I also wonder what effect the repeated understating of monthly job growth might be having on the recovery.

Would consumers, employers and investors be more optimistic (and willing to spend more money) if the original jobs estimate each month (which gets the most press) reflected the stronger hiring that later became evident? Or am I overestimating how much the official numbers affect behavior?

Article source:

Speak Your Mind