People are Finding Jobs -- More are Losing Them

Every Thursday we see the same discussion of the weekly jobless claims reports.  It is not helpful, even though this is a good real-time indicator.

Some commentators note that initial claims, usually measured by the four-week moving average, have stabilized and even show some decline.  Economist Robert Gordon, a long-time member of the NBER dating committee, believes that this indicates an imminent end to the recession.  Look here for a nice summary by the Good News Economist, and here for the full Gordon article.  Check out the convincing charts and tables.  Gordon writes as follows:

If we refine the NBER weekly trough date to be the third week in the NBER trough month, then in four of the past five recessions the new claims peak leads the NBER weekly trough by a range of only four to six weeks, and in the fifth recession the new claims peak lags the NBER weekly trough by two weeks. Since new claims have recently reached a peak in the week ending 4 April 2009, it is tempting to conclude that the monthly trough of the US recession could come as early as the middle of May 2009 – a date earlier than most analysts appear to expect. 

Meanwhile, many observers highlight the scary chart showing continuing claims.  As we noted last week, this indicator is helpful in showing individual pain from the recession, but not for predictive purposes.  Continuing claims are not a consistent series because of changes in the duration of unemployment benefits.  Those presenting this chart have an obligation to address this issue.

Instead, most point to it as evidence that the recession will be prolonged.  It is not good evidence for this argument.

An Exercise

Continuing claims are nearly 6.8 million, up 110,000 from last week.  It is a deeply disturbing fact, representing plenty of distress.

Now try this.  Take the initial claims level of about 620,000.  Multiply this by the number of weeks of benefits.  In normal times this is 26 weeks, but it is now as long as 72 weeks in some states.  (It is still not long enough for some).  Let us take 52 weeks as a conservative number for many states.  Now multiply.

This shows that more than 30 million people lose jobs in a year.  It also shows that 80% find new jobs.

This conclusion is supported by various reports from the BLS.  The most authoritative sources use the business dynamics series, citing data from state unemployment records.  In data released last week, but ignored by everyone, the BLS reported on the most recent available data, from the third quarter of 2008.

The number of job gains was 6.8 million.  That is over 100,000 jobs for every business day.

The problem is that job losses were 7.8 million, a net loss of a million jobs in the quarter.  The fourth quarter data will be much worse.

Getting an Accurate Picture of Labor Dynamics

There are several important conclusions.

• Looking at changes of 100K or so in continuing claims is misleading.  This is statistical noise compared to the overall impacts.
• The critics of the BLS on job creation measurement (including the many birth/death model critics) are completely wrong.  There are many jobs being created.  The problem is that job creation is not keeping up with job losses.
• The extent of the impact is better represented by the 30 million or so people who have lost jobs in a year, and also their colleagues who feel threatened.  This is a big damper on consumption.
• The Obama Administration efforts to "count" job gains from stimulus do not capture the story.

Anyone who looks at the actual data gets a more accurate picture.  There is a sea change of job losses and gains.  It helps to understand why the initial claims series is so important.

To our continuing surprise, none of the mainstream media sources have reported this story.  It is almost as if they are following the bloggers rather than doing their own research.