This post has been corrected.
These monthly unemployment reports keep getting stranger and stranger. In January, the U.S. unemployment rate dropped to 9.0%, which is a big 0.4% decline from 9.4% in December, according to the Bureau of Labor Statistics. Yet, the report is unclear on the actual number of jobs created during the month. Through one surveying method, it finds 589,000 new jobs, but just 36,000 new jobs through the other. Due to this uncertainty, it's premature to celebrate stronger job growth at this time.
Let's start with the unemployment rate chart:
This shows how great January's progress appears to be. At 9.0%, the unemployment rate is the lowest we've seen since April 2009.
Yet the number of jobs added to the economy looks far less impressive:
In fact, 36,000 jobs added is downright anemic.It's the fewest jobs added yet in any month during the recession when there was a net increase in jobs. Looking at the total private sector jobs added helps to show the progress better, since the temporary Census jobs skew the numbers in the chart above for much of 2010. Here's how just the private sector has fared:
In January, 50,000 jobs were added, which is barely above May 2010's 48,000. This implies that January had a very weak performance.
Unless, that is, you look at the BLS's other way of estimating employment. The jobs numbers above were tabulated through its Establishment Survey. According to its Household Survey, 589,000 jobs were created in January, which is what led to the decline in the unemployment rate.
Unfortunately, it's hard to do a lot of direct comparison between January and December with the BLS's Household Survey numbers. The agency does a major revision related to population control effects in the January report and does not dynamically update its historical data to reflect the changes.
For some more strangeness, let's look at what happened with those Americans who want a job but are not considered a part of the labor force. Here's a chart showing the BLS estimates for discouraged and marginally attached workers (these are not seasonally adjusted like other statistics discussed):
Check out the change from December to January. Weird, right? In January we saw the number of discouraged workers plummet by a whopping 325,000. That's great! But is it? The number of marginally attached workers -- those who want a job but are not considered in the labor force for a reason other than discouragement -- increased by 516,000, or 40%. In other words, it looks an awful lot like the drop in discouraged workers is occurring because they're just switching in status to marginally attached. There's no genuine good news here.
How about another strange result? The broadest measure of underemployment, U-6, declined significantly as well in January -- to 16.1% from 16.7% in December. This statistic measures the percentage of Americans who are unemployed, discouraged, marginally attached to the workforce, or forced to work part-time because full-time work isn't available. This could appears to be good news. U-5, which is precisely the same measure as U-6, except it doesn't count workers who were forced to work part-time for economic reasons, only declined from 10.9% to 10.7%. In other words, U-6 declined by 0.4% more than U-5. That drop appears to be the result of more people who were forced to work part-time finding full-time work. It's the only variable that appears different between the two measures.
Finally, here how many jobs were added or lost from various sectors:
Interestingly, manufacturing led the way in January for these selected sectors. Government employment fell slightly, across federal, state, and local jobs. Meanwhile, the construction industry continues to bleed jobs. Possibly the most notable result here is the decline in temporary workers. Since overall professional and business services jobs increased, this could imply that more temp workers are finding permanent employment.
What's the big takeaway from this report? The labor market recovery is still moving forward, but there's little certainly about whether it's doing so slowly or at a more rapid pace in 2011. According one BLS survey, just 36,000 net jobs were created and only 50,000 for the private sector. Those are very weak numbers. But another survey says 589,000 jobs were created. That would indicate much, much stronger growth.
So beware of politicians and pundits claiming that the big drop to 9.0% suggests that jobs are suddenly growing at a rapid pace. That may be true, but it's hard to be comfortable with that assertion, given the wide data discrepancy in the January report. It does appear more likely, however, that more part-time and temporary workers are finding permanent, full-time employment. Meanwhile, however, there are still over 20 million Americans who want a job, but can't find one.
Correction: An earlier version of this post incorrectly stated that the BLS Household Survey indicated a large number of workers exiting the workforce between December and January. In fact, an annual correction to the survey complicates any comparisons between December and January, in ways that make our initial data analysis incorrect. We have accordingly revised the post and removed a chart on unemployment duration. We regret the error. For more information, see this post.