So I open up my LA Times this morning, and the front page of the business section greets me with this headline:
Steep drop in jobless rate has some Obama foes crying foul
And I thought: This is amazing. One tweet from Jack Welch sets off a wingnut firestorm that actually makes the front page of the Times. The power of these guys to set the news agenda is pretty spectacular.
But there may be an unappreciated irony at work here. Although the Bureau of Labor Statistics (BLS) didn’t cook the unemployment books, there’s no question that the headline number, which is derived from a telephone survey of households, can be fairly noisy from month to month. There was a big spike upward in September’s employment figure, and that could be real or it could be a statistical outlier.
Or there might be a third option: In a little-noticed part of yesterday’s report, BLS announced that it had systematically undercounted jobs by 386,000 from April 2011 through March 2012. So maybe it’s continued to undercount jobs since then, as Karl Smith suggests here. If so, then not only is the September number accurate, it’s making up for an undercount over the past six months. That’s the shaded portion under the red line in the chart below, which is a simple trend line that runs through the revised March 2012 figure and extends it through September. It suggests that the September employment number is right where you’d expect it to be if the economy were continuing a steady but modest recovery — which seems like a reasonable bet.
So here’s the irony: If BLS really has been undercounting, it means that the jobs picture has looked overly gloomy during the first half of the year, which is exactly when it hurt President Obama the worst. What this means is that the wingnuts might be more than merely wrong. They might have things 180 degrees backward. It’s quite possible that far from being unfairly favorable toward Obama, the BLS numbers have been unfairly hurting him. September’s spike corrected that, but probably too late to do him very much good.
