Wednesday, August 11, 2010

Examining Unemployment Numbers for Policy Impacts vs Structural Problems

The notion that this recession has suddenly spiked structural unemployment, and that unemployment will stay at this level for a long time, may be overblown. I'd like to examine whether the unemployment numbers we're seeing are the product of a structural problem (in which case, retraining, education and welfare are important) or more temporary policy problems and uncertainties (in which case, a) Obama and the Democrats have done just as bad a job with the economy as Bush and the Republicans did with the war, and b) unemployment should drop as soon as there's a degree of policy clarity on what the results of healthcare reform, finance reform, stimulus, etc, will be)
Reason number 1 this may be true comes from behavioral investing: every single time, people believe that "this time, it's different". usually it's not so different. Other recessions ended up with a reasonable employment recovery. But let's at least acknowledge the possibility of "this time it's different".
Let's look at some facts.
Firstly, unemployment has spiked to 9.5% from mid 5%s.
underemployment has spiked up to as high as 17% or 18%.
The ratio of unfilled jobs to unemployed is ~4x higher than in previous recessions, and if it were comparable to previous recessions, unemployment would be 6.8% (source:
Temp hiring has been off the charts - temp jobs are up 20% year over year, while permanent private sector jobs are down 1%. (source:
Complementary and corroborating facts from today's Times:
-"The drop in hiring has been steeper than the rise in layoffs."
-"Compare the current slump with that of the early 1980s, which was similar in severity. Over the course of 1980, 18.1 percent of the labor force was unemployed at some point. In 2008, the first year of this slump, only 13.2 percent was, according to the Labor Department's most up-to-date data. That number surely rose in 2009, but it is unlikely to have come close to the 1982 peak of 22 percent. "

-"In the deep economic slump of the mid-1970s, the average hourly pay of rank-and-file workers — who make up four-fifths of the work force — fell 6 percent, adjusted for inflation. In the early 1980s, the average wage fell 3 percent. Even in the mild 1990-91 recession, it fell almost 2 percent... But since this recent recession began in December 2007, real average hourly pay has risen nearly 5 percent...This time around, nominal wages — the numbers people see in their paychecks — have risen throughout the slump."

-"The first two years of the downturn were disproportionately blue collar. In 2008 and 2009, the construction industry shed 25 percent of its jobs, and manufacturing lost 16 percent.

This year has been different. Manufacturing — especially of so-called durable goods, like computers and machinery, many of which are purchased by companies — has been adding jobs.

Many white-collar fields, meanwhile, are losing jobs: state and local governments, publishing, telecommunications. Retailers have added jobs, but more slowly than the rest of the economy, as consumer spending has remained tepid. "


-"The unemployment rate for college graduates is still just 4.5 percent, and the gap between their pay and everyone else's is larger than it has ever been. For most college graduates, the Great Recession has not lived up to its name. "




The part that most clearly indicates policy is the problem is the temp hiring. If businesses are hiring temps off the bottom of a recession, it means that they're seeing demand pick up but they're too uncertain to actually hire someone full time. This is usually followed by permanent hiring, but in this recovery, it has not been. The only obvious culprits here are a) higher required wages and healthcare costs (minimum wage, housing interventions and Obamacare are prime candidates here) and b) general lack of confidence in the economy and policy (the political climate in general is the prime candidate here).

The underemployment numbers are also reasonably telling here. If people can get jobs but not for enough hours, it means that firms are afraid to hire full time. It's easy to blame full unemployment on a structural shift, but hiring someone to do similar work as they had been doing but for less time is a confidence issue.

The drop in hiring being larger than the rise in layoffs tells you something also - the biggest problem isn't that people are being fired from jobs that are no longer structurally sound, the problem is that nobody will hire. That doesn't indicate a structural problem, that indicates a confidence and policy problem.

Finally, the shift from blue-collar layoffs to white-collar layoffs, and the rehiring of a lot of manufacturing jobs, indicates again that this shift isn't one pool of workers structurally forced out of industries, it's a garden variety (deep) recession that can (garden variety) recover in a situation where businesses feel comfortable. This isn't trivial.


So what evidence supports a structural shift?

The ratio of unfilled jobs to unemployed, the rising real wages and smaller share of the population who faced unemployment are all indicators of a structural shift.

However, the most telling number may be the college-graduate vs non-graduate gap, which is the largest it's ever been.

People treat everything structural as the same, but in reality, I can think of two different types of structural unemployment that would lead workers to have a low marginal product.

1) Industry permanent overcapacity or irrelevance, leading to mass layoffs who are never rehired because no matter how productive a worker is, it's hard for him or her to actually create incremental value in that industry.

2) Workers who don't have enough skills to produce value in excess of their wages, even in industries that aren't structurally overbuilt.


The shift from blue-collar to white-collar, the big gap between college graduates and non-graduates, the ratio of unfilled jobs to unemployed, the smaller share of the population who face unemployment (who initially were in concentrated industries but now are no longer), etc. are all compatible with the second explanation. The first explanation is incompatible with the blue-collar/white-collar observations, and to a lesser extent with the temp hiring/underemployment observations.

What does this end up looking like?

A garden variety recession exacerbated by terrible economic policy and 20 years of crappy public schooling leading to a small number of workers with no real employable skills when the economy is at anything less than full capacity (when they get hired because there aren't any other workers worth employing).

This suggests that

a) the Obama administration and the Democrats have been a disaster (spoiler alert! see here:

b) they may see significant improvements when they lose the Senate and House, because they won't be able to do as much (a benign inertia) and Republicans seem hell-bent on incinerating their own party through social issues instead of focusing on their strengths on fiscal ones.

c) Allowing immigration could actually increase structural unemployment because immigrants may be hired ahead of the 5% least-skilled members of the US population (this isn't an argument against immigration, which I am largely in favor of de-restricting, but an argument that immigration needs to be accompanied by better unemployment policy and better adult educational opportunities)

d) our public schools have been run into the ground. I'm no education expert, but I'd bet you state politicians, teacher's unions, teen pregnancy and anti-abortion policy leading to worse parents, and an anti-intellectual culture spread through a lot of demographic groups all share a lot of blame in that one.








1 comment:

  1. Hi, I Saw Your Link in I Love Your Blogs Kindly See my Link Here a New Alternative Money Making Network alternative to adbrite here