Recessions and Unemployment
By now you’ve probably heard that the job market tanked last month. The number of jobs out-and-out declined for the first time in almost five years. But the unemployment rate is still really quite low. What’s going on?
Today’s report from the Bureau of Labor Statistics showed the unemployment rate falling from 5.0% to 4.9%, though it was essentially unchanged (to 4.93% from 4.98%). The real question is how can folks like me argue that a recession is imminent or perhaps even underway with unemployment this low?
Let me count the ways.
First, historically it’s not that unusual for a recession to start at low jobless levels. A recession began in Nov 1948 when the unemployment rate was 3.8%; July 53, 2.6%; Aug 57, 4.1%; April 60, 5.2%. And this isn’t all just ancient history: when the last recession started in March 2001, the unemployment rate was 4.3%.
Second, the recessionary indicator is not the unemployment level; it’s the change. The rate bottomed out at 4.4% last March and has been climbing since. An increase of the magnitude we’ve experience since then almost always signals recession.
Third, other related signs are also revealing weakness in the job market. Long-term unemployment—the share of the unemployed who’ve been jobless for a least six months—is unusually high given the low unemployment rate. Historically, when unemployment has been at or below five percent, the share of the long-term unemployed has been about 10%. Since October 2005, the jobless rate has also been 5% or lower, but the long-term rate has averaged 18%. In other words, there are a lot more people stuck in unemployment than you’d expect given the low rate.
Underemployment, a labor-underutilization concept described here (see U-6 in the table), was 9% last month, the highest since Sept 2005. And wage growth has slowed as well, another signal consistent with a job market wherein labor supply is starting to outpace labor demand.
It’s obviously a good thing to be starting this labor market downturn—if that’s where we are—with historically low joblessness. We’re extremely unlikely to see double-digit unemployment rates, like we did in the early 1980s downturn (though African-Americans, whose jobless rate was 9.2% last month, are already almost in double-digits).
But this is no time for complacency on the jobless front. The Senate did the right thing by including extended unemployment insurance benefits as part of their stimulus package, but word is that the White House is stonewalling against the extension, based exclusively on the low overall jobless rate. The numbers, especially the high share of long-term job seekers and the upward trend in the unemployment rate, underscore that this is precisely the wrong position. For once, let’s be ready for the storm before it arrives in earnest. If it’s not too late already.
Comments (10)
I'm not sure how to edit in the new format yet, but "December" in the first sentence should read "last month" ie January.
Sorry. JB
February 2, 2008 3:27 PM | Reply | Permalink
Hey, I fixed it! I'm a techno-genius!
February 2, 2008 3:31 PM | Reply | Permalink
BLS initial December 2007 jobs report (18,000 increase); BLS revised December 2007 jobs report (82,000 increase). WHOOPS.
BLS initial January 2007 jobs report (17,000 decrease); BLS revised January 2007 jobs report (? , ?). WHOOPS?
February 2, 2008 4:51 PM | Reply | Permalink
That's true, Ellen.
But see my analysis of the jobs report (link in piece--see figure at end of piece, eg) where I show that the job market's been weakening for a while now, and this longer term deterioration will outlast any monthly revision. Note also that last year underwent a major downward revision.
February 2, 2008 5:01 PM | Reply | Permalink
Yes; but you can't say what portion of this "longer term deterioration" is due to systemic causes and what portion to incidental ones.
Give it up, Jared. You're dancing in the dark.
February 2, 2008 5:33 PM | Reply | Permalink
The fact is the unemployment figures are just not an accurate reflection of the actual amount of unemployed American citizen workers.. those who have stopped collecting benefits fall off the unemployment rolls and are treated as though they don't exist. This doesn't even take into considered the problems of the underemployed.
I can not believe anyone who considers themself to be progressive or even liberal would accept these stats and ignore what is plain to see.. it's all across the country.
Also, the study out of Harvard by Dr. Borjas has proven that there is no worker shortage in the US, never has been. Let's see some articles on the true economic realties faced by poor and struggling American citizens please.
February 2, 2008 9:01 PM | Reply | Permalink
Re: Those who have stopped collecting benefits fall off the unemployment rolls and are treated as though they don't exist.
This is not true. The unemployment stats do not depend on who is collecting benefits. They are derived from a phone survey. There are certainly inaccuracies with this survey, though they have always been present and I see no reason why they should be greater now than 20 years ago.
In the main the problem with the US economy is not high levels of unemployment-- rather, as Jared notes, it is high levels of underemployment. Jobs as such are easy to get. But most do not pay enough and may underutilize the emplpoyee's skills and education.
February 2, 2008 9:16 PM | Reply | Permalink
Thanks for the info, Jon. I googled for info, and read the info on the CPS government site. I'm even more troubled. They survey 60 thousand households about all who are employed in each, including those 16 years old and up. They combine earnings for the entire survey group and provide an average.. so unless you are factoring how that might make things appear, the numbers can be fairly misleading.
That isn't a true reflection of actual unemployment, it also doesn't factor in who are newly come in to the US, whether illegal aliens or visa workers..
I have read articles by writers, and also by people who represent the long term unemployed and underemployed that refer to a more accurate counting of the unemployed as being closer to 24 to 26%. We also never hear local unemployment numbers being discussed in the media. For example, we heard about the raid on the Michael Bianco factory in New Bedford, MA, because they were using illegal alien labor. The owners of the factory claimed that they couldn't find citizen workers in the area to do the jobs.. which were leather stitchers. The truth was that New Bedford, an old mill town hit hard by the outsourcing of factory jobs since the '80s, had plenty of workers who were experienced in leather stitching.. and they were long term un/underemployed. The unemployment rate for New Bedford at the time of the raid was approx 26%, and the poverty rate was higher than 60% in the town. We see corporate interests and government misrepresenting and ignoring the realities and hurting innocent citizens, who have civil rights protections and human rights considerations.
We need something more thorough to determine the actual rate of unemployment. Surveys and polls might suit for some things, but they are too easily manipulated towards whatever the powers that be want them to reflect.
February 3, 2008 11:34 AM | Reply | Permalink
Note that while U3 is around 5%, U6 is substantially higher than that.
And from my experience, I think U6 is more accurate. For example, for the six weeks starting Christmas week, I got 2 days on Christmas Week and three days, starting Friday of the third week after Christmas week and ending the following Tuesday.
Under U3, I was employed for 3 weeks and unemployed for 3 weeks, for a weekly average unemployment rate of 50%.
Under U6, I was employed part time for economic reasons for 3 weeks ... 0.5 employed and 0.5 unemployed ... and unemployed for 3 weeks, for a weekly average unemployment rate of 75%.
In reality, excluding Christmas Day I had 29 days of labor on offer and 5 days of labor accepted, for an hours based unemployment rate of 82%.
Clearly, U6 is a better approximation for my situation, working at a warehouse in Northeast Ohio during a slide into recession.
Also, clearly I am hoping that economic instruction at community colleges is a counter-cyclical demand, because trying to get manual labor in NE Ohio during a recession is teh suck.
February 3, 2008 2:59 PM | Reply | Permalink
NE Ohio -- what she said about Oakland.
February 3, 2008 10:43 PM | Reply | Permalink