The Unemployment Problem Is a Little Bit Structural

Looking at sectors where jobs have begun to grow this year helps to show that most will return in time

At 9.8%, unemployment is still a serious problem in the U.S. So serious, that some economists believe that the problem is structural -- not cyclical. That would essentially means that there has been a paradigm shift in certain industries, and the jobs that existed there before the recession are gone forever. That contrasts with merely cyclical unemployment, which is when firms lay people off during a recession, only to rehire them when the economy improves. In a New Yorker column this week, James Surowiecki argues that the problem is cyclical, and that economists who say that it's structural are misguided. In reality, both camps are sort of right.

Here's what Surowiecki argues:

The structural argument sounds plausible: it fits our sense that there's a price to be paid for the excesses of the past decade; that the U.S. economy was profoundly out of whack before the recession hit; and that we need major changes in the kind of work people do. But there's surprisingly little evidence for it. If the problems with the job market really were structural, you'd expect job losses to be heavily concentrated in a few industries, the ones that are disappearing as a result of the bursting of the bubble. And if there were industries that were having trouble finding enough qualified workers, you'd expect them to have lots of job vacancies, and to be paying their existing workers more and working them longer hours.

Now that private sector jobs have been growing for about a year, we have more data to begin to test the plausibility of the claim that some sectors' jobs may be gone for good. So let's go to the data.

Here is a chart for ten major sectors. There are two bars shown for each. The first is for the period from 2008 through 2009, a time during which the U.S. private sector lost about 8.5 million jobs. The second bar is 2010, when private sector jobs began to slowly grow again, by about 1.2 million. First, let's look at the number of jobs added/lost for each sector during these periods:

Since these sectors are all different sizes, it's a little hard to draw conclusions about this chart. So it's helpful to turn to another that shows the percentage change in employment for each of these sectors during these periods:

Now we can see what's really going on here. Let's start by throwing out education and health care entirely. They actually grew throughout this three-year time period, so they are aberrations. That leaves eight sectors. Five endured job losses during the first period, but began to grow in the second. Three had losses in both periods.

For those five where the labor market has begun to grow this year, it's hard to argue that any structural problem could be present. After all, they're improving. That leaves three: construction, financial activities, and information.

In fact, construction and financial activities are two of the sectors precisely where you might expect to see structural problems. After all, the real estate market was the epicenter of the recession, which was driven by a financial crisis. Are these jobs gone forever?

It sort of depends on how you define forever. It's safe to say that the 1.9 million construction jobs lost aren't coming back anytime soon. But construction is, by nature, a cyclical industry. Eventually, as the economy and U.S. population expand, more business structures, homes, and infrastructure will be needed. Construction jobs will inevitably follow. But construction jobs may never make up the same portion of private sector jobs they once did during the peak of the real estate bubble.

So let's look at the portion of the private sector each of these industries made up during each time period:

You can see that there's been relatively little change in the share of jobs for information and financial activities over this three-year period. But construction's share has clearly fallen.

Construction jobs may have experienced unsustainable growth compared to even these other two sectors that had trouble growing this year. Is this a true structural problem? It sure looks like it. While construction jobs will eventually grow again, the industry may never be as large a portion of the private sector as it was during the real estate bubble.

But considering that this is just one industry, where jobs will inevitably begin to grow again eventually, it's still pretty hard to call the overall unemployment problem structural. After all, construction only accounted for about 20% of the jobs lost during the recession. So the unemployment problem is a little bit structural, but mostly cyclical.