Something’s (Still Up) with the Unemployment Numbers

About a month ago I posted that something was goofy with the latest employment numbers. Others have caught on.

Using Okun’s Law, here’s what The Wall Street Journal came up with:

The top left shows the plain scatterplot, the top right adds a regression line capturing Okun’s Law, and the bottom panel shows the size of the residual (the vertical distance to the line, not the whole red dashed line).

The conclusion; somehow we’ve gotten too much decline in the unemployment rate to be plausibly explained by the real GDP growth we’ve experienced.

I’ve never liked Okun’s Law much: it’s too flexible to be a “law”.

Anyway, what we’re seeing here isn’t impossible … just worthy of attention.

Okun’s Law also got goofy on the way up:

Christina Romer, President Barack Obama's former chief economic adviser, is watching this very closely. That is because Okun's Law also broke down in the other direction a few years ago when she was head of Mr. Obama's Council of Economic Advisers. She predicted the unemployment rate would rise to a little less than 8%, but instead it went to 10%. Some of the miss was because the downturn turned out worse than expected and much of it was because unemployment rose more than Okun's Law predicted.

I find her explanation plausible:

Ms. Romer has a theory for why the jobless rate rose more than Okun's Law predicted during the recession and why it has fallen more than the law predicted since the recession: She believes that company managers were so shocked by the financial crisis in 2008 and 2009 that they fired workers more aggressively than they would in a conventional downturn.

"Firms were terrified," she says. "We had just had the first financial crisis in 70 years. The world looked like it was falling apart …

But there’s a possible downside:

A less sanguine explanation could be a dangerous productivity slowdown. It might be the case that the workers being hired aren't improving their productivity as much as workers had before. If they aren't as productive, companies need more of them.

The hiring sounds nice, but a productivity slowdown would be bad in the long run for everyone. Less productivity means slower growth in the long run, an economy more susceptible to inflation shocks, slower growth in inflation-adjusted incomes and less government revenue to work down big deficits.

Robert Gordon, a Northwestern University professor who tracks productivity closely, says he sees "clear signs everywhere" that a productivity slowdown is happening. …

Pay attention to Gordon; he’s one of those scary-good people on understanding business cycles.

Read the whole thing, entitled “Piecing Together the Job-Picture Puzzle” in the March 12 issue of The Wall Street Journal.