The Job openings and labor turnover survey: what initial data show: early results from these new data series show trends that are in line with other surveys, both private industry and government, and allow for a more complete picture of the labor market

Monthly Labor Review,  Nov, 2004  by Kelly A. Clark

• E-mail
• Print
• Link

Turnover estimates and other economic indicators

As stated earlier, quits tend to decrease during recessions because workers' outlook toward finding another job worsens with deteriorating economic conditions. (11) As economic conditions worsened throughout 2001 and 2002, consumer confidence plunged, and fewer people quit their jobs than at the same time the prior year. (See chart 5.) The seasonally adjusted quits series shows a decrease throughout the published series, and the consumer confidence index exhibits the same downward trend as the quits rate over the course of the series. The consumer confidence series shows something of a rebound in late 2003 and early 2004, perhaps signaling that quits may be expected to increase even further in late 2004. The correlation of quits and consumer confidence is 0.80, which is positive and significant.

[GRAPHIC OMITTED]

One of the only other data series providing a national turnover rate has been the Bureau of National Affairs (BNA) quarterly Job Absence and Turnover report. (12) This long-running series provides results from approximately 300 U.S. member companies surveyed. The JOLTS total separations data trend with the BNA turnover series, but at a higher level partly because BNA does not include layoffs, job eliminations, or departures of temporary staff, whereas JOLTS includes all types of separation during the reference month. (See chart 6.)

[GRAPHIC OMITTED]

Although the BNA report provides a long time series for turnover estimates, the JOLTS program provides a timely and nationally representative indicator of turnover for both hires and separations. In addition, with a much larger sample size and a more inclusive definition of turnover, the JOLTS statistics are more reliable and useful. With the larger sample size, JOLTS is able to publish more industry detail. However, the BNA report publishes turnover rates by establishment size class, which JOLTS may pursue in the future because turnover rates appear to vary by establishment size.

In mid-2003, BLS once again added to the national statistical framework with data series showing what underlies net employment changes, the Business Employment Dynamics (BED). (13) Quarterly statistics on gross job gains and gross job losses also prove an interesting comparison to hires and separations flows. (See chart 7.) These series track net employment changes at the establishment level. A preliminary analysis has shown JOLTS total private hires and separations, summed for each quarter, have outpaced the gross job gains and gross job losses, which is as expected. The gross job gains and gross job losses are computed by comparing the employment level of the third month of each quarter. JOLTS measures each individual hire and separation that occurs during every month, and thus the data series are, by definition, higher than the gross job gains and losses series. For example, if an establishment's employment level was 10 in the third month of the first quarter and 10 in the third month of the second quarter, there would be no employment change and thus no effect on the gross job gains or losses. However, there may have been three hires and three separations in between those two points, which JOLTS data would reflect.