Employment News – A Muddle Part 4

The government monthly “Employment Situation Report” from BLS on the first Friday isn’t the only game in town for understanding what’s happening to employment. There are alternative data series that makes the news that can help give us some idea of what’s happening in employment. None of these data series is as comprehensive or reliable as the BLS series, but they do offer alternative data for reasonableness.  After all, as we say in part 1 and part 2 of this series, the BLS data is far from straightforward and can often give fuzzy or contradictory signals.

In addition to the BLS series there is:

  • The ADP Private Sector Employment report. ADP is a private data processing company whose core business is processing payroll for thousands of companies in all types of businesses. Because of their large market share in this area, they are privy to exact counts of workers (payroll checks) each month.  They have models that extrapolate from their client firms to the entire private sector as a whole.  This data series is usually released on the Wednesday before the BLS release on first Firday. An example is report for January 2011 is here.
  • The  U.S. Department of Labor provides a weekly release of the Unemployment Insurance Weekly Claims Report. An example for last week is here. This is the sum of all the individual state unemployment claims data. Note: BLS monthly report is unrelated to any state-level unemployment compensation claims.  This report, while there is a Seasonally Adjusted series as well as NSA series, is notoriously “noisy”, meaning there’s a lot of variation from week-to-week. It is extremely sensitive to things like bad weather causing newly unemployed workers to wait until Monday to file instead of doing it on Friday. Nonetheless, looking at the 4-week moving average is sometimes helpful.  Here’s a graph of recent behavior in this series, courtesy of Calculated Risk:

The DOL reports on weekly unemployment insurance claims:

In the week ending Jan. 29, the advance figure for seasonally adjusted initial claims was 415,000, a decrease of 42,000 from the previous week’s revised figure of 457,000. The 4-week moving average was 430,500, an increase of 1,000 from the previous week’s revised average of 429,500.

Weekly Unemployment ClaimsClick on graph for larger image in new window.

This graph shows the 4-week moving average of weekly claims for the last 10 years. The dashed line on the graph is the current 4-week average. The four-week average of weekly unemployment claims increased this week by 1,000 to 430,500

  • Finally, there’s a relative newcomer in town, the Gallup Poll series. Gallup, the famous polling organization has in recent years been conducting surveys of households to determine an unemployment rate.  The definitions and techniques are not strictly comparable to the BLS CPS survey each month, but it provides another checkpoint.  Here’s a recent example:

PRINCETON, NJ — Unemployment, as measured by Gallup without seasonal adjustment, remained at 9.6% in mid-January, the same as at the end of December. This marks a one-percentage-point improvement from 10.6% in mid-January 2010.

Gallup's U.S. Unemployment Rate, January 2010-January 2011 Trend

Employment News, A Muddle – Part 3

This is the follow-up to my first two posts on the January Employment Situation Report and the muddled picture it paints.  The first post here recounts how there’s some apparent contradiction with the headline unemployment rate declining from 9,4% to 9.0% (yeah!) but only 36,000 new jobs created (not good at all).  The second post explains some methodology behind the numbers to try to resolve the contradictions. Now I’m going to offer what I think it tells us.

Many commentators, particularly the talking heads on the financial news channels but also some leading economists and analysts have decided this was a pretty decent report.  James Hamilton as well as Nomura Securities has concluded it’s good news. Calculated Risk seems to think that underlying the report is good news but that winter distorted the report. I beg to differ. Now I have no great computer models or crystal balls like many analysts, but I’m more in the camp with Mark Thoma who emphasizes the level and scale of unemployment and concludes we simply aren’t improving fast enough.

Let’s look at this graph from Calculated Risk and compare this “recovery” to all the post WWII recessions:

Percent Job Losses During RecessionsClick on graph for larger image.

This graph shows the job losses from the start of the employment recession, in percentage terms – this time aligned at maximum job losses.  In the previous post, the graph showed the job losses aligned at the start of the recession.
In terms of lost payroll jobs, the 2007 recession is by far the worst since WWII, and the “recovery” for payroll jobs is one of the slowest.

Yes, I will agree that positive jobs growth is better than no jobs growth or jobs elimination. Yes, the red line in this graph is going up. But it’s rising at an unacceptably slow rate. My screen isn’t wide enough to project out when that line is going to show us getting back to number of jobs we had when this mess began – back before Wall Street blew up the economy. But it seems to me that too many analysts are concluding something along the lines of “well, the snow and winter distorted the numbers. Without the snow, there probably would have been another 100,000 jobs and that’s OK.”  I’m sorry, even if we grant their arguement that that snow caused us to underestimate new jobs by 125,000, it is still too little too slow. We need 150,000 jobs a month just to stay even with a growing population.  For this to be a true recovery that includes all of America, not just the lucky ones on Wall Street or in Washington, we need to see a lot faster increases in growth – increases we haven’t seen since 1984.

Just to repeat here’s our real situation:

  • 6.21 million people have been unemployed for more than 26 weeks.
  • 8.4 million workers have part-time jobs but want full-time work and cannot get it.
  • 16.1% of our workforce wants to work full-time and our economy cannot put them to work fully
  • There are still 5.5% fewer jobs in the economy than there were 3 years and 1 month ago, despite a growing population.
  • The present rate of “recovery” will have us get those jobs back after another 3+ years.

The issue I think most analysts have missed in this report is the effect that the “99’ers” are having on the report. The “99ers” are unemployed people who have exhausted the maximum 99 weeks of unemployment compensation. Let’s face it, after 99 weeks of looking and not finding a job, you are most likely to quit looking until it appears like a real recovery in your area.  When Congress decided in December 2010 to not create another tier of unemployment benefits and extend benefits for these people, the fell off the unemployment compensation rolls. That also means there’s no reason to keep “actively” searching. So they have started to drop out of the labor force. That’s why labor force participation rate has continued to decline.  This is not good. This is treating the unemployed as if they are invisible – the disappeared workers.

I’m going to add one post to this series. In the next one I’ll look at alternative data series that makes the news that can help give some us some idea of what’s happening in employment.