Changes to the Current Employment Statistics Survey Hours and Earnings SeriesBackground The Bureau of Labor Statistics (BLS) is in the process of adding new data series to the Current Employment Statistics (CES) survey to improve its relevance to the needs of data users, as well as its value as an input to other key economic statistics. CES data on employment, hours, and earnings by industry are among the most visible and widely-used economic indicators at national, state, and metropolitan area levels. CES data are also among the timeliest economic indicators, with their release by BLS in the Employment Situation news release shortly after the end of the reference month. Planned improvements to the CES: Add new data on the hours and earnings of all employees; these data eventually will supplant existing data on hours and earnings of production and nonsupervisory workers as the principal hours and earnings series from the CES survey. The planned improvements originally included adding new data on gross monthly earnings (GME). GME include bonuses and irregular payments and are broader in scope than regular CES earnings series. BLS began producing and publishing GME series concurrently with the all employee hours and earnings series. However, following the release of October 2008 GME series on January 9, 2009 BLS will discontinue the collection and publication of the GME data. See below for more information. The current status of the planned changes to the CES is described below. All employee hours and earnings series – The CES program currently publishes series on the average hours and earnings of production workers in the goods-producing industries and nonsupervisory workers in the service-providing industries. Production and nonsupervisory workers account for about 80 percent of all employment measured by the CES survey. The new all employee hours and earnings series will provide more comprehensive information than the present series for analyzing economic trends and for constructing other major economic indicators, including nonfarm productivity and personal income. A pilot test of the collection of all employee hours and earnings data with CES respondents found the data available from the payroll records of most employers. The new experimental data became available for a limited number of industry series on April 6, 2007. See www.bls.gov/ces/cesaepp.htm for more details. The industry detail available is as follows:
CES publishes the following data types as experimental:
Plan for production and nonsupervisory worker hours and earnings series – The current BLS plan is to reduce the amount of industry detail published after the new all employee hours and earnings series are well established. While the comprehensive all employee series will be more useful than the current limited scope series for analyzing overall labor market trends, the current production/nonsupervisory series will still be of interest to some users. However, the production and nonsupervisory worker hours and payroll data have become increasingly difficult to collect, because these categorizations are not identifiable in many payroll systems. Many survey respondents report that it is not possible to tabulate their payroll records based on the production/nonsupervisory definitions. BLS is exploring reduction of production/nonsupervisory series due to the declining sample sizes, and the respondent burden generated by continued collection of these data items. Discontinuation of experimental gross monthly earnings – Following the release of October 2008 GME series on January 9, 2009 BLS will discontinue the collection and publication of the GME data. The decision to discontinue these series stems from data quality limitations. Throughout the test period for collection and production of the GME series, BLS has experienced low response rates for the GME data item, response error issues, and difficulty developing an estimator that can adequately reflect the underlying volatility that exists in an earnings series that includes bonuses and other irregular pay. The CES GME series was designed to be very similar in definition to the total wages concept used in the BLS Quarterly Census of Employment and Wages program, though not precisely the same. Total wage definitions vary slightly across States due to differing Unemployment Insurance coverage while the CES GME was defined consistently across all States. The concepts are similar enough however that the CES GME could be compared against the QCEW total wages figures to help gauge the accuracy of the GME estimates. The QCEW total wages figures essentially represent a universe count, and thus serve as an informal benchmark against which to compare the GME estimates. Comparisons showed that the GME series are consistently biased downward by a substantial amount during the first and fourth quarter of the year, a time when end-of-year bonuses and other payments cause the QCEW total wages to spike upward. BLS research indicated that this problem is largely traceable to low item response rates for GME and incomplete reporting of total wages in the CES survey responses. Further follow back efforts with CES survey respondents indicated that many simply do not have timely access to the information needed to report comprehensive GME figures. Further compounding the problem is that many of the large year-end payments that underlie the QCEW total wages spikes in fourth and first quarters are concentrated in a relatively small number of firms; when even a few of these firms do not respond, or can not report accurately to the CES survey, the resultant GME estimates are not reliable. In addition to respondent follow back, BLS researched several different estimation techniques; however, none of them were able to overcome the inherent limitations caused by missing and incomplete survey responses. BLS has therefore concluded that producing estimates for GME from the CES survey is not viable; the data simply are not reliable enough for BLS to continue production of these series.
Last Modified Date: February 6, 2009 |
Tools |
Calculators |
Help |
Info |
