Friday, March 12, 2010

The Labor Market is on the Mend

The labor market in the United State is clearly on the mend.  The number of job losses per month is now negligible (probably—see my post on the effects of the snow storm here). and unless the economic outlook deteriorates substantially we are likely to see outright gains in employment in the months to come.  Nonetheless, detailed labor market continues to point to a tepid recovery in terms of overall growth rates.

The latest JOLTS data from the BLS confirms some healing in the labor market.  The black line in the figure below shows the separation rate adjusted for quits through January 2010.  (See my description of this data here.) The separation rate has now returned to its post-2000 average.  With fewer employees losing their jobs each month, the economy is likely to continue on the path to recovery.
Lower fire rates signal economic growth for two reasons.  First, the decline indicates firms are comfortable with the current size of their labor force.  These firms, at the least, foresee stability in future output.  Clearly, the actions of firms are far more important for assessing their outlook than the responses they give to the various confidence surveys.  Second, the reduction in the fire rate indicates positive consumption growth.  The growth comes through two important channels. 

First, the reduction in the fire rate indicates lower aggregate employment risk.  The lower risk decreases the precautionary motive of households, lowers savings rates, and increases consumption.  Second, and far more important, the number of households involuntarily loosing their jobs declines.  Households with a sharp reduction in income consume less.  So, with fewer households losing their jobs this important drag to growth diminishes.  [Note:  unemployed households do not consume appreciably more.  But the drop in their consumption has already been incorporated in the National Income Accounts.  The fact that they are still not consuming has zero effect on the growth rate of consumption.]

That’s the good news.  The labor market is stabilizing and growth should resume. 

The bad news:  Weak is not strong enough of a word to describe the likely recovery.

The picture below adds the hire rate to the previous graph.  The hire rate has shown no indication of improvement.  The lack of improvement in this rate is quite interesting and is in conflict with the data from initial claims, which have improved from 700 thousand jobs per week to a little under 500 thousand per week.  The current gap between the hire rate and the average hire rate since 2000 represents a shortfall of 1.2 million jobs per month.  This is an unbelievably large number. 

To put this in perspective, the jobless recovery following the 2001 recession was characterized with the separation rate near average and the hire rate near average.  The previous “jobless recovery” had job growth of between 50 and 200 thousand jobs per month.  This leaves room for the current jobless recovery to occur in an environment with only small average job losses per month.  (Small average losses per month imply that the variability in the survey will allow some positive months.)
The low hire rate also implies that firms do not see a particularly strong recovery.  While firms are content with their current workforce, they do not, as yet, see a sufficient increase in demand to warrant an expansion of their labor force.  This bodes ill for investment going forward as well.  Firms that do not need new people also, likely, have little need for new equipment. 

So, the low hire rate implies subdued growth for two reasons.  First, the data implies little or no investment growth going forward.  Typically, during a recovery, investment grows robustly contributing substantially to growth.  Second, the data implies that movement from unemployment/out-of-the-labor-force will be extraordinarily slow relative to typical recoveries.  People who move from unemployment to employment (low income to high) increase their consumption substantially.  This channel is a drag on consumption growth. 


The labor market is healing but is far from normal, even when normal is compared to the weak job market of the last ten years.  The U.S. economy faces serious structural problems going forward.  The number of unemployed people is likely to stay at depression levels for a long period.  [The unemployment rate is still likely to fall.  People will exit the labor force rather than continue to report themselves as unemployed.]

The current social safety network is ill-equipped to handle large numbers of semi-permanently unemployed people.  The United States is going to have to make some hard choices on how to handle this social problem.  The solutions must balance the welfare of the unemployed against the need to maintain incentives to seek employment.  Further, these decisions are going to have to be made in an environment of slow revenue growth.

Saturday, March 6, 2010

Jobs and the Jobs Bill (304,000 jobs over the next 12 months)

The economy lost an additional 36,000 jobs in February.  Interpreting the numbers is difficult because of the record breaking snow storm on the East coast during the reference period for the establishment survey.  Almost every analyst I heard on topic predicted that the storm subtracted anywhere from 30 to 80 thousand jobs from the labor report.  

It’s amazing how confident people are in their statements.  They might be right but I am less sure.  Keith Hall, the Labor Commissioner and one of the most forthright people around, said that, while it was likely the storm had a substantial impact on the jobs report, it was not possible to tell whether the jobs number was biased upwards or down as a result.  He did not know how much the storm effected the hiring and firing decisions of firms nor did he know how much the storm impacted temporary hiring, as businesses and schools hired temporary workers to help remove the snow—temporary help rose 48,000 in February.

I think there is a more general issue.  The storm effectively shut down firms in the mid-Atlantic region for a full week in February.  It seems to me that the first order effect of the storm depends on whether the impacted firms would have hired or fired in that week.  For some companies, the storm may have served as a temporary furlough, pushing back firings and perhaps delaying them indefinitely.  For others, the storm may have prevented them from hiring.  The net effect depends on the aggregate state of hiring and firing.  I don’t know the answer and neither does anybody else. 

The Jobs Bill:  Adding 230 to 370 thousand new jobs

For those of you who are regular readers, you know my opinion of economic stimulus:  I am not a believer in big multipliers.  But I like the jobs bill.  If you want to increase employment this bill is exactly the right medicine.  I suggest exactly this program in December 2008 (here scroll down to “What can the government do?”)  My plan went farther, in line with spending almost $1 trillion, but the incentives are the same.

The jobs bill works on the margin.  New hires are exempt from the firm’s share of social security plus the firm receives and additional thousand dollars if the employee is kept on the payroll for 1 year.  For the average worker in the United States, this is a $3,600 payment per qualifying hire.  The maximum payment is close to twice that. 

This is not enough money to change the hiring decision of the average firm; fortunately, economics does not work on the average; it works on the margin.  Because the jobs bill is only allocates between $12 and $15 billion, the bill is perfectly designed to work on the margin. 

I estimate that the jobs bill (the employment portion only) will add between 230 and 370 thousand jobs over the next year.  These jobs are new additional jobs in the United States. 

Why is the jobs bill effective and ordinary stimulus not?  Because the jobs bill changes the real relative price of labor.  (Stimulus does too but with the wrong sign.)  Labor demand curves are downward sloping:  cheaper equals more.  The jobs bill is just like Cash-for-Clunkers and the Home Buyer’s Credit.  The change in real relative prices induces a real change in demand.  [Note, this does not necessarily boost GDP; it is only expected to change the demand for the particular product.  The bills have costs.]