The Labor Market Is Worse Than You Think

Last week, the Bureau of Labor Statistics released its new official unemployment rate for June of 9.5%. I explained that, if you include discouraged workers, that number is actually 10%. And if you include "marginally attached" workers, that number is even higher. But what about workers being asked to cut back hours? Laura Conaway over at NPR's Planet Money blog (hat tip: Matthew Yglesias), tells us that the news gets worse.

She refers to the following chart:

And says:

Viewed a certain way, the drop in the average American workweek looks small -- very, very small. In June, the Bureau of Labor Statistics reports, production and nonsupervisory workers spent one-tenth of an hour less on the job than they had in May. Six minutes? How big a deal is that? Here's how big:


It's "the lowest level on record for the series, which began in 1964," the BLS says.

I dug into the historical BLS reports. In November 2007, before the recession began and unemployment was a measly 4.7%, the average weekly hours of production workers was up at 33.8 hours. Let's use that as a reasonable baseline for what average weekly hours should be. I think that's generous, if you look at the graph. In June it was 33.0 hours.

There are a number of ways to interpret how this change really affects employment. Here's how I'd do it. First, the percentage change between 33.8 and 33.0 is a decrease of about 2.37%. But let's multiply that by the June 2009 employment level of 90.0% (the level that takes discouraged workers into account). After all, if you're unemployed, you can't work less. That means in June, if hours were reduced from our baseline of 33.8 to 33, the subsequent decrease in employment would have been around 2.1%. Or unemployment leaps to 12.1%.

You can quibble about whether this quantitative experiment is really meaningful, but I think it is. The idea is that, if hours were not cut, then the unemployment level alone would be a good indicator of the labor market's health. But if you take those cut hours into account, you can better determine the true condition of the labor market. Through my methodology, that's quite a bit uglier.