Saturday, June 05, 2010

May Adjusted Jobless Claims Fall to 21-Month Low

I haven't featured this pair of charts for awhile, showing: a) jobless claims vs. the labor force and b) jobless claims as a share of  the labor force, both updated through May (BLS data here and here).

The top chart shows why unadjusted jobless claims are meaningless: the size of the U.S. labor force has almost doubled over the last 42 years, from 77.57 million in 1968 to the current level of more than 154 million.  The bottom chart shows jobless claims adjusted for the size of the U.S. labor force.  Jobless claims averaged 455,250 in May, which is 0.2948% of the May labor force of 154,393,000, and represents a 21-month low (lowest since August 2008). Jobless claims as a percent of the labor force have declined in 13 out of the last 14 months, starting last April. 

This measure of initial jobless claims, adjusted for the increasing size of the U.S. labor force over time, shows that jobless claims peaked during this recession above the levels of the last two recessions (1990-1991 and 2001), but were never anywhere close to the levels of the previous three recessions in the mid-1970s and early 1980s, and about the same as the 1969-1970 recession. The sharp reduction in adjusted jobless claims from the March 2009 high of 0.415% follows the same pattern of sharp reductions at the end of each of the last six recessions.  Call it an upside-down V-shaped recovery. 


At 6/05/2010 11:23 AM, Anonymous gettingrational said...

Very interesting that this is not a headline in most media right now. The upside down "V" recovery might be called the "caret" recovery!

At 6/05/2010 12:27 PM, Blogger bix1951 said...

Thankyou MP for continuing to tell the truth about the economy. Things are not nearly as bad as big media says. In fact, things are pretty good.

At 6/05/2010 12:38 PM, Anonymous Anonymous said...

I don't know how many college students are graduating with offers, but if they don't find a job, they won't show up in initial claims for unemployment, but will count for unemployed.

At 6/05/2010 1:09 PM, Anonymous morganovich said...

longer unemployment benefits shade these statistics somewhat. they also get reduced by temporary census hiring. there is every reason to expect this number to pop when the census ends.

take a look at labor force as a % of population and you get a different story. civilian labor force as a % of adult population dropped in may.

from BLS:

"In May, the civilian labor force participation rate edged down by 0.2
percentage point to 65.0 percent"

take out the census hires and unemployment was heading up.

At 6/05/2010 1:52 PM, Anonymous Anonymous said...

Actually, it's the top chart that reveals the troubling truth. The number of people dropping out of the labor force is massive by historic standards. The bottom chart hides that data.

At 6/05/2010 2:42 PM, Anonymous Anonymous said...

Things are not nearly as bad as big media says. In fact, things are pretty good.

The "big media" has been doing everything possible to downplay the depth of this crisis for fear that the fallout will land on their "golden boy". Every negative statistic is "unexpected", every cloud has a sliver lining. While it's true that the economy is showing some signs of life - it's amazing how little economic activity a trillion dollars of stimulus will buy - things are not "pretty good". Worse still, the country is fast approaching an entitlement and debt crisis that our current leaders have only made worse through their incompetence and recklessness. Here are a few things to think about:

50 Statistics About The U.S. Economy That Are Almost Too Crazy To Believe

Read 'em all.

At 6/06/2010 2:02 AM, Anonymous Torred said...

And that drop in the labor force at the end doesn't concern you?

See the big vertical spikes in 2000 and 1990 and the smaller ones in 1980 and 1970? What caused those? How do you suppose we should adjust for that?

At 6/07/2010 11:02 AM, Blogger Junkyard_hawg1985 said...


You mixed statistics with this chart. You took the weekly statistics and adjusted them to total workforce for both sets of data. Unfortunately, the weekly unemployment numbers only account for a subset of the workforce. The % of the workforce covered by Unemployment Insurance has shifted considerably since 1975. In 2008, only 36% of the unemployed was covered by unemployment insurance. From the BLS (

"Because of these and other limitations, statistics on insured unemployment cannot be used as a count of total unemployment in the United States. Indeed, during 2008, only 36 percent of the total unemployed received UI benefits. The weekly data on UI claims do have important uses, however, and provide a timely indicator on labor market conditions."

This data is also evident when comparing the "insured unemployment rate" from the weekly statistics to the unemployment rate as calculated the the BLS from the monthly survey. Here are some reference points:

The week of May 31, 1975, the weekly unemployment rate based on UI was 7.0%. The May 1975 unemployment rate from the BLS was 8.3%. An offset of 1.3% (or 18.5% increase in the unemployment rate between the two sets of statistics). In May 2009, the insured unemployment rate based on weekly statistics was 5.1% compared to an unemployment rate of 9.1% according to the BLS. This is an increase of 4% (or 78.4% increase in the unemployment rate between the two sets of statistics). Today, the unemployment rate is 9.7% with 3.6% unemployed by the initial claims report (+169%). The current numbers are badly skewed by UI that extends for 99 weeks today. In the past, those who wanted to be on UI had to be rehired, then re-fired to get their unemployment check. Now they can avoid the hiring and refiring (lowers initial claims).


Post a Comment

<< Home