How Do The Long-Term Unemployed Compare To The Rest Of The Labor Market?

by: Rortybomb

The situation for the long-term unemployed looked significantly better after last week's jobs report. The average duration of unemployment dropped from 38.1 to 35.3 weeks over that month, which included statistical rebalancing for the population. A year ago, 43% of the unemployed were out of a job for more than 27 weeks; now that number is down to 38%.

This is a good development, though it intensifies two of my larger worries about how people will view the economy. The first is that we'll think the economy is doing too well. If we think the economy is healthy, then the Federal Reserve and Congress will put the brakes on too fast, killing the possibility that full employment, the best social program we have, will really happen. There is already evidence of this happening. The sequestration, which will kill a million jobs, looks increasingly likely to happen, even though there is little long-run justification for premature austerity.

The other, oddly, is that we'll think the labor market is so weak that it can no longer be helped by emgerency stimulus. Neil Irwin wrote an overview, as a result of the Scarborough and Krugman back and forth, of what I'll call "pundit macroeconomics." It's a theory of why pundits care about cutting social insurance and deficits even though the economic logic is missing. The missing part of this argument is that many elites feel that while there are too many unemployed, they are uniquely unqualified for the jobs that are available.

Let's update one of my favorite graphs around, which shows how likely it is that the unemployed will find jobs by the duration of their unemployment. I just got new data from the BLS that gives us these numbers through October 2012. Is there a relatively healthy short-term labor market, with a collapsed long-term one? Let's compare 2007, 2011, and 2012:

As you can see, no matter how long you've been unemployed, the labor market in 2012 is weaker than it was in 2007. It was less likely that those unemployed for less than five weeks could find a job in 2012 than they could in 2007. The same goes for the long-term unemployed.

This pushes back against recent research by Rand Ghayad and William Dickens of the Boston Fed. They dissagregate the Beveridge Curve by duration, arguing that our problems are primarily concentrated among the long-term unemployed. However, they are likely just picking up on changes in the long-term distribution of the unemployed (which, as noted above, has been collapsing since June 2012, when their data ends), rather than strictly structural elements. Looking at the labor market through the graph above, we can see that it is generally weak, which is not just a function of the long-term unemployed.

Is duration falling because the unemployed are simply dropping out of the labor force? Here's the transition from unemployment to no longer in the labor force, or the liklihood of the unemployed simply dropping out, comparing the pre-crisis time period and today:

Compared to before the recession, the long-term unemployed are less likely to drop out of the labor force. People are still looking for jobs, though a little less in 2012 than in 2011. That said, there wasn't a large pickup in this rate in 2012, so it is unlikely to be the primary driver in the drop of unemployment duration.

Rob Valletta of the Federal Reserve Bank of San Francisco just put out an economic letter on the long-term unemployed. He does the actual work of parsing out weekly transitions from the CPS data and finds this transition, the same dyanmic noted above.

If you break it down by month and look at it over a longer timeframe, it still has the same result. The labor market is depressed for everyone, not just a select group.

Notice the bump out at month 20 for the recovery period, where it actually goes above the expansionary period. Though it isn't clear what is driving this, it is likely both a function of an improving job market as well as people no longer qualifying for unemployment insurance. Unemployment insurance pulls in several directions then. It increases duration both through encouraging longer searches with better matches by providing liquidity. It provides stimulus to the economy, while also keeping people from leaving the labor force and giving up on their searches entirely.

Valletta also finds that "for most categories of workers, the share of long-term unemployment differs little from the share of short-term unemployment." There are some exceptions, notably younger workers. However, the long-term unemployed aren't a dumping ground for certain types of workers; it reflects a general malaise in the labor market.

This isn't to downplay the serious issues of long-term unemployment. The long-term unemployed do have a harder time finding jobs. But the best cure for this situation is to broadly boost the economy through fiscal and monetary stimulus while dealing with the housing market, rather than transitioning to either targeted job policy or deficit reduction.