Making Expectations More Realistic

I was in college during, and graduated into, the first of the “jobless recoveries” – that economic expansion following the 1990-91 recession.  Even after the recession officially ended, the economy continued to lose jobs.  It wasn’t until the first quarter of 1993 that employers made up for all the losses.  I remember reading newspaper articles and watching news stories about the labor market.  Every time the monthly payroll figures came in lower than expected, investors would express their concern.  I remember thinking, “Who cares if the number is off by a little bit, as long as the economy is creating jobs?”  It is with a similar perspective that I look at the world today.

 

Why A Jobless Recovery?

 

The recession in the early 1990s changed what we previously thought about recessions.  In the first quarter after the recession ended, the economy continued to shed jobs of a considerable magnitude – 252k positions lost in the second quarter of 1991.  (Calculations are based on BLS data.)  This was not the first recession in the post-WWII era in which employers kicked people to the curb after the recession officially ended.  In the three months following the 1953-54 recession, approximately 13k jobs were lost.  And, in the first quarter following the 1973-75 recession, the economy lost 130k jobs. 

 

(Certain aspects of the labor market are a lagging economic indicator.  A perfect example of this is the unemployment rate.  But we are focusing here on payrolls, which are a concurrent indicator.  Thus, one would expect these numbers to be positive once a recession has ended.)

 

What made the period in the early 1990s a “jobless recovery” was not just the fact that jobs were shed in the months after the recession ended, but also that they took much longer than usual to come back.  For instance, following the 1948-49 recession, the economy required three quarters to make up for the jobs that were lost and in that third quarter the rate of job growth was impressive (more than one million jobs added).

 

Job growth following the 1953-54 recession was a touch weaker, taking a year to make up for the jobs lost during the downturn.  The situation was relatively similar following the 1957-58 recession.   By comparison, three quarters after the recession in the early 1960s, the economy was net positive for job creation.

 

Jobs lost during the 1969-70 recession were more than made up six months after the recession ended, while the 1973-75 recession required a bit longer: three quarters.  During the 1980 recession, the economy did not even need a full quarter after the recession ended to make up for jobs lost.  The hit in 1981-82 necessitated a year-long recovery.

 

That brings us to our modern era of recessions and the jobless recoveries.

 

During the 1990-91 recession, employers shed about 1.3 million jobs.  Ballpark.  Two years later, the economy was up net only 9k positions.  This recession saw less than half the jobs lost than during the previous recession (1981-82), but it took twice as long to make up for it.

 

Although the tech bubble burst in early 2000, the economy did not officially enter recession until the first quarter of 2001.  The downturn lasted only one year, and the impact on the economy was negligible relative to other recessions in the post-WWII era (based on BEA estimates of GDP).  Still, even though GDP continued to advance, payrolls were slashed.  When the recession ended, approximately 1.8 million jobs were lost.  The economy continued to shed positions for another six quarters, losing nearly another 1 million slots.  (The BLS non-farm payroll data is monthly, but I look at quarterly sums to more easily compare with the quarterly GDP figures.)  The economy took nearly two years (7 quarters) after that to make up for the lost positions.  In all, more than three years (13 quarters) had passed after the recession ended before the economy was net positive job creation.

 

Current Dynamics

 

The most recent recession ended in the second quarter of 2009 with massive job losses: approximately 7.2 million positions shed.  More than 1.1 million jobs were lost in the two quarters that followed.  It has been nearly two years since the recession ended, and the economy is still net negative nearly 7 million jobs (a little over 6.9 million positions).  Ballpark.  It is reassuring to see growth in certain areas, particularly in manufacturing.

 

Yesterday, the ADP payroll figures were released and they were less than expected, contributing to market weakness.  The ADP numbers don’t always track well with the government statistics, but they still provide some valuable insight.  Tomorrow, we have the government’s release covering April.  Most likely, we are going to see that the economy continues to add jobs at a tepid clip.

 

Many factors can impact short-term numbers.  We can develop all sorts of models to try to capture the stochastic processes around the trend of job growth (I know a little about that stuff, I wrote my master’s workshop thesis on forecasting oil prices, where I used all sorts of statistical models).  But what is more important to getting this economy back on its own feet than employment, particularly long-term job prospects?  That said, the general trend is more important than trying to perfectly model the impact of some temporary, exogenous shock (say, oil prices).  With that in mind, I’m going to revert back to my old self and say, “Who cares if the actual number is off the estimate by a little bit, as long as the economy is creating jobs?” 

 

As indicated in a recent Market Watch article (http://www.marketwatch.com/story/applications-for-us-jobless-benefits-surge-2011-05-05?siteid=yhoof), some economists estimate that the economy will have to add approximately 300,000 jobs, on average, each month to really absorb new entrants to the labor market (congratulations to all those graduating college later this month) as well as the number of unemployed persons.  In looking at the monthly figures from BLS, we know that the economy has been adding nowhere near that level.  Does it really matter if the job growth in a given month was 210k or 175k?  Either way, it is suboptimal.  Again, we are down nearly 7 million jobs. 

 

It is difficult for anyone to realistically say that the US economy is completely out of the woods, particularly when you consider that job growth remains sluggish even with all of the fiscal and monetary stimuli that are still in place.  We will likely see continuation of this dynamic in tomorrow’s BLS report.  Until growth is more broad-based, I’m not really going to care if some economist’s statistical model indicates that the economy is going to add 225k jobs in a month, but only 205k are added.  At this point, I’m happy that the economy is adding jobs, and I’ll be happier if we see evidence of continuation of the upward trend in the monthly pace of improvement.