February Employment Report

In addition to summarizing the latest from the Bureau of Labor Statistics, you’ll note I have provided three graphs. The first is my usual update of the employment-population ratio (EP) and the labor force participation rate (LFP). The other two are interesting charts from AB reader Spencer (more on this later). First the latest news:

Nonfarm payroll employment continued to trend up (+97,000), and the unemployment rate (4.5 percent) was essentially unchanged in February, the Bureau of Labor Statistics of the U.S. Department of Labor reported today. Employment grew in some service-providing industries but declined sharply in construction. Manufacturing employment continued to trend downward. Average hourly earnings rose by 6 cents, or 0.4 percent, over the month.

CNN just claimed that this unemployment rate is a historic low. What can’t we have a better press? And how does such a low increase in the payroll employment number translate into a lower unemployment rate? Well, let’s also look at the household survey which reports that employment FELL by 38,000, which means that the employment-population (EP) ratio fell to 63.2%. So why did the unemployment ratio fall? Ah yes, the labor force participation rate (LFP) fell to 66.2%. And CNN reports this is good news?

My graph starts with January 1997 and goes through February 2007. Some AB readers have asked me why not go much further back. Well – Spencer has but he has in mind something quite different that the usual and silly argument that we should compare the current state of the labor market to some historical average of something like the unemployment rate. I have two problems with using a historical average of the unemployment rate to define something like the natural rate of unemployment rate.

One is that the unemployment rate is not really that informative as all does is to state the difference between LFP and EP. The other is that these series tend to change over time in ways that are more than transitional variations.
What Spencer has done is to take the trend line in EP and compare that to current EP. His trend line would suggest that the natural rate of the employment-population ratio is near 65.3%, which of course, is higher than my rough guess that it’s around 64%. While this is a neat contrast to those claims that we are beyond full employment, let me suggest this isn’t quite right. LFP, which had hovered around 59% from 1948 to the late 1960’s rose rather steadily from the late 1960’s to the latter part of the 1990’s and then started hovering around 67%. EP also tended to rise along with LFP and hit 64% during Clinton’s second term.

The great labor market debate since the 2001 recession has centered on whether we should expect LFP to return to 67% or to stay at its current level. I doubt anyone thinks that LFP would continue to trend upwards but is it now trending downwards? Or is the current level of LFP an indication of the discouraged worker effect? Economists may differ on this issue, but I don’t any serious treatment of the labor market would expect either the labor force participation rate or the unemployment rate to revert to some long-term average.