The employment report or employment “situation” for each month is released the first week following the end of that month by the Bureau of Labor Statistics in the Department of Labor (www.bls.gov) and is one of most important economic reports tracked by market watchers. The report is part of the Current Employment Statistics (CES) program that surveys some 160,000 businesses, so there is considerable detail by industry in this report. Unfortunately, revisions to the previous month’s estimates due to late reporting can be large at times, so it is best to view changes in employment over a three-month moving average to be better able to see trends.
One of the ways to gain advanced insight into what this report will reveal is to watch changes in the U.S. Department of Labor’s weekly release of the “Unemployment Initial Claims” report. Weekly filings for unemployment can be impacted by weather events, holidays and other factors (for example, Hurricane Katrina delayed filings for several weeks in some areas), so again a moving average (four-week in this case) is reported and is a better indicator of trends in employment. A rising trend in the four-week moving average in initial applications for unemployment insurance suggests a softer employment report in the coming month while falling initial claims would suggest a stronger employment report.
The fact that the last four weeks of initial claims showed falling numbers would be consistent with a stronger employment report.
The key numbers reported in this release are (1) the number of new jobs created and (2) the unemployment rate. However, there are some unique aspects of this report that require careful interpretation.
The data reported in the employment situation report are derived from two distinct surveys: a survey of households (60,000) and a survey of establishments (160,000 businesses and governments). The household survey provides information on the labor force, employment and unemployment, information which is used to estimate the unemployment. The establishment survey (also called the “payroll” survey) also derives information on employment, but adds average weekly hours worked and average hourly earnings. While both surveys contain information on employment, they differ in a number of important ways. Two major differences are (1) the household survey contains no duplication of individuals while the establishment (payroll) survey reports persons in those establishments on the payroll even if they have more than one job, and (2) the establishment survey excludes self-employed, agriculture and private household workers.
These differences result in different estimates of job creation in the household survey (-52,000 in November 2005) than in the establishment survey (+215,000 November 2005).
The establishment survey is viewed as a more reliable indicator of non-farm job creation in the economy and is the number generally reported in the economic press.
Interpreting the job growth number
The civilian labor force in the United States is currently 150.2 million (November 2005) and is expected to grow at a rather slow rate of 1%-1.5% annually. If we seek to keep the number of unemployed from growing (assuming no change in the labor force participation rates), we must create about 1.5 million jobs a year or 125,000 per month to absorb these new entrants to the labor force. If unemployment rates are to fall, job growth should exceed 125,000 per month, while slower job growth would result in an increase in unemployment. The size of the labor force is also subject to change as individuals move in and out of the labor force. Thus a growing labor force can mitigate some of the job growth and leave unemployment rates less impacted. For example, in November of 2005, the Department of Labor reported that 215,000 nonfarm jobs were created, far exceeding the need for a minimum of 125,000. This was interpreted as a fairly good report, but the unemployment rate remained unchanged at 5% from October and the number of unemployed remained unchanged at 7.6 million. The explanation for this lies in changes in the size of the labor force, which grew by 97,000 over the same period.
The range of monthly changes in total employment thus far in 2005 has been from a low of 126,000 in May to 300,000 in February with an average of 167,000. Revisions, which are made during the next two months, can also be significant as noted by the September estimates which were revised from an initial -8,000 jobs to plus 17,000 in the November revisions.
Because of the volatility of monthly changes and the revision factor, it is best to look at a three-month moving average as a better indicator of future employment growth.
The importance of this job growth number lies in what it portends for consumption growth. Since consumption accounts for about two-thirds of spending in the economy, it is easy to see why this report is closely followed by the financial markets.
However, the report also contains average weekly hours worked which can be tracked as an indicator of the relative tightness of labor markets and future job growth. When weekly hours worked continue to grow above 40 hours, it suggests that at some point additional jobs will be created to relieve the overtime pressures. This is consistent with recent declines in average weekly hours worked as employment has grown. If average weekly hours worked were increasing, it could provide evidence of building wage pressures that may lead to potential inflationary pressures.
Next time, we’ll consider consumer confidence and consumer sentiment.
Dr. William D. Gunther is professor of economics and director of the Bureau of Business and Economic Research in the College of Business at the University of Southern Mississippi. Contact him at firstname.lastname@example.org.