What's New in the 2008 Current Employment Statistics Program

by Sharon Jans

In 2005 the Current Employment Statistics (CES) program began work to add a new data series on hours and earnings for all nonfarm private-sector employees. The All Employee Hours and Earnings series was developed to measure average hourly earnings, average weekly hours, average weekly earnings, average weekly overtime hours, and gross monthly earnings of all employees. The All Employee Hours and Earnings experimental series was first released in April 2007. This release includes national level estimates at the total private-sector level and limited industry detail from March 2006 through January 2007. Each subsequent month, the experimental data is updated. Starting in February 2008 with the release of January 2008 CES preliminary estimates we will be publishing this new All Employee Hours and Earnings data series for Minnesota.

Historically the CES program has published hours and earnings series for production and nonsupervisory workers. The new All Employee Hours and Earnings data series covers hours and earnings for all employees rather than just for production and nonsupervisory workers. It also has an additional component, the gross monthly earnings (GME) data series. While these changes may not be obvious to the occasional data user, they are significant and expand the breadth and depth of the CES hours and earnings data series.

In addition to the rollout of the All Employee Hours and Earnings experimental data, the CES program is revising its industry coding structure from the 2002 North American Industry Classification System (NAICS) to the 2007 NAICS structure. These data will also be published in March 2008 with the release of the January 2008 estimates. The NAICS conversions are minor and reflect changes in the service and information industries. More detailed information about the NAICS 2007 conversion is available at the end of this article.

Overview of the CES program

The Bureau of Labor Statistics (BLS) conducted the first monthly studies of employment and payrolls in 1915, covering only four manufacturing industries. Over the next two decades BLS developed and expanded the program and in 1933 published average hourly earnings and average weekly hours for all 90 manufacturing industries and 14 nonmanufacturing categories. With the onset of the Great Depression and no reliable measure of employment or unemployment, BLS produced comprehensive estimates of salary employment and total wages in nonfarm industries in 1936. During this same period interest in employment statistics for states and areas also grew, and by 1940 estimates of total nonfarm employment for 48 states and the District of Columbia were available. Since 1949 the CES program has been an integrated federal-state cooperative program providing employment, hours, and earnings data by industry on a national, state, and sub-state basis.

CES is a sample-based survey,¹ and at the national level the sample consists of approximately 400,000 business locations or about 160,000 businesses and government agencies. In Minnesota the sample consists of approximately 5,900 business locations. Each month data are collected on the number of nonfarm wage and salaried employees, production workers² and women workers, including both full-time and part-time for the reference week, the week that includes the 12th of the month. Additional data are collected on the number of production and nonsupervisory workers, including hours worked, overtime hours worked by production workers, and payroll.

In order for CES data to maintain its integrity and usefulness, once a year the CES estimates are benchmarked or revised using actual worker data from other sources that were not available at the time of estimation, such as unemployment insurance (UI) tax records. Through the benchmark process, actual employment counts are used to revise the estimated CES data while retaining its time-series structure.

Since all states use the same BLS guidelines to produce nonfarm employment estimates and maintain their historical data series, CES data can be compared across states and to national CES trends. One of the best ways to gauge a state’s or individual industry’s economic health is by looking at worker trends. The logic is simple: Businesses typically hire when their goods or services are in demand and reduce worker numbers as demand for their goods or services declines. CES data are one of the timeliest economic indicators with the release of national estimates by BLS shortly after the end of the reference month. State and sub-state area estimates are released two to three weeks later.

What data are in the current CES program?

The current CES program collects data on total employment with breakouts for production and female workers. It also collects hours and earnings data for production workers in the goods-producing industries (manufacturing, construction, and natural resources and mining) and nonsupervisory workers in the private service-providing industries (trade, transportation, and utilities; information; financial activities; professional and business services; educational and health services; leisure and hospitality; and other services), which account for about 80 percent of employment measured by the CES survey. The employment data collected are the total number of people employed full-time or part-time in a nonfarm establishment who received pay for any part of the pay period that includes the 12th of the month. Data exclude proprietors, the self-employed, private households, all agriculture production except logging, and unpaid family workers or volunteers.

Hours and earnings data series are based on gross payrolls and the corresponding paid hours for production, construction, and nonsupervisory workers. Gross payrolls include pay for overtime, shift premiums, payment for vacations and holidays, paid sick leave ,and other leave payments made directly by the employer to employees for the pay period reported. The payroll figures exclude bonuses, commissions and other lump-sum payments unless earned and paid regularly each pay period or month. The hours figures relate to the hours for which pay was received; this is different from scheduled or standard work hours in that it includes overtime. Overtime hours refer to the hours worked by production workers for which overtime premiums were paid.

What data are in the new CES program?

BLS has been collecting hours and earnings data for all employees from survey respondents since late in 2005 in order to publish the new All Employee Hours and Earnings data series. The data collected include total payroll and total hours which produce average hourly earnings, average weekly hours, average weekly earnings, and average weekly overtime hours³ for all nonfarm private sector employees. Hours collected for all employees who worked full- or part-time during the pay period including the 12th of the month includes paid hours and may include vacation, holiday, and overtime hours where worked. Gross payrolls or earnings data for the pay period that includes the 12th of the month are designed to measure the regular earnings of workers, which may include overtime pay, shift premiums, payment for vacations and holidays, sick leave, and other leave made directly by the employer to employees for the pay period reported. They exclude bonuses and other irregular payments.

Gross monthly earnings (GME) is the other new earnings series being added to the CES program, and it has a much broader scope than the all-employee payroll data. It is defined as the total gross wages paid during the entire calendar month before deductions and includes all non-wage cash payments in addition to regular wage and salary payments. GME includes items such as bonuses, lump sum payments, and exercised stock options. In addition the time period covered is more expansive. The all-employee payroll data item covers just employees working the pay period that includes the 12th of the month, while GME covers all workers who were paid at any time during the month. The GME has a definition similar to total wages used in the BLS Quarterly Census of Employment and Wages program.

The data changes actually are data additions to the current CES program. A new “all employees” data series on the hours and earnings will be added to the existing data series on hours and earnings of production and nonsupervisory workers. The GME addition of total earnings covers both regular and irregular pay for all employees. These series are more comprehensive in coverage, providing improved information for analyzing economic trends and improved input to productivity and personal income series.

Why the changes?

BLS has been considering a revision of CES data types for several decades but was unable to implement it until 2005. The major revision addresses “all employees” versus “production workers.” The production worker concept is a relic of the early 20th century manufacturing economy. Today’s economy is more information technology and service based, and the All Employee Hours and Earnings data will capture this. CES wants its hours and earnings data to better reflect the 21st century economy. Currently there are no estimates of all-employee average weekly hours and all-employee average hourly earnings, only estimates for production/nonsupervisory workers, which by definition are limited in scope and usefulness. There is no estimate for total gross monthly wages in CES, and its addition would be beneficial. The program has also faced increased difficulty in collecting accurate hours and earnings data from CES reporters who have sometimes found the CES definition of production/nonsupervisory workers confusing. Who gets reported and who doesn’t? For example, an administrative assistant working in a manufacturing industry is not a production worker and is not reported, but an administrative assistant working in a financial industry is a nonsupervisory worker and is reported. Many respondents’ payroll systems do not allow them to identify workers according to the CES definition of production/nonsupervisory workers, so hours and earnings data cannot be extracted.

What are the improvements and uses?

The new data series additions will improve the CES hours and earnings program in breadth and depth. The new All Employee Hours and Earnings series, with its broader definition, will provide more comprehensive information for analyzing economic trends than the present production/nonsupervisory workers series. The base average hourly earnings series will continue to provide a measure of underlying wage trends exclusive of irregular payments. The new series will also provide improved inputs for other major economic indicators, including series on nonfarm productivity and personal income. The GME should improve the accuracy of preliminary estimates of personal income in the national income accounts. Further improvements are expected for states and metropolitan statistical areas (MSAs) with expanded publication of hours and earnings data into more private industry sectors. Increased response rates from CES reporters are expected since “all employee” data are readily available from the payroll records of most employers. For the time being there will be two data sources for hours and earnings as BLS has no plans to discontinue the production/nonsupervisory worker data series at this time.

Where can all this be found and when?

The new All Employee Hours and Earnings and GME experimental data became available in April 2007 at the national level for a limited number of industries. See www.bls.gov/ces/cesaepp.htm .

BLS has designated these first released data as experimental because BLS has limited experience in editing and reviewing the sample data as well as the estimates produced. These data are also not seasonally adjusted. BLS should have sufficient historical data by the end of 2009 to permit seasonal adjustments to the All Employee Hours and Earnings payroll and hours series. The data will be published as official CES data in early 2010. Table 1 shows the industry detail published at the national level for this data series.

A comparison of the “all employee” data to the “production worker” data shows:

• Trends are similar for both “all employee” and “production worker.”
• Average Hourly Earnings are higher for “all employee” compared with “production worker” for all industries.
• Average Weekly Hours for “all employee” compared with “production worker” are mixed, depending on the industry.
• Average Weekly Earnings for “all employee” in nonmanufacturing industries are greater than for “production worker.”
• Average Weekly Earnings for “production worker” in manufacturing industries are greater than for “all employee.”

The new all-employee data series for Minnesota

Currently the Labor Marker Information (LMI) office of the Department of Employment and Economic Development (DEED) in Minnesota publishes employment estimates for all employees for all supersectors and an expansive set of industries at the statewide level and for four MSAs: Minneapolis-St. Paul, Duluth-Superior, Rochester, and St. Cloud. LMI publishes average weekly earnings, average weekly hours, and average hourly earnings for production/nonsupervisory workers for various industries at the Minneapolis-St. Paul MSA and statewide level. For more detailed information visit the web site: www.deed.state.mn.us/lmi .

Minnesota will be publishing its All Employee Hours and Earnings experimental data series in March 2008, concurrent with the release of January 2008 preliminary estimates. The experimental data will be published in limited industry detail, mainly at the supersector level for statewide and at the total private level for the MSAs. Total private gross monthly earnings will be published only at the statewide level with no gross monthly earnings data for MSAs.

Table 2 shows the industry detail that Minnesota will be publishing at the statewide level in March 2008. Included is the industry detail for both the experimental All Employee Hours and Earnings data series as well as the production/nonsupervisory data series.

The addition of the new “all employee” data series will greatly increase the breadth of what Minnesota currently publishes. For the production worker data series, Minnesota currently publishes information for eight detailed manufacturing industries plus two supersectors under goods-producing, and 12 detailed service sector industries plus two supersectors under private service- providing. With the addition of the “all employee” data series, Minnesota will be publishing eight additional supersector levels plus three rollup industry levels that include: goods producing, private service providing, and total private, an increase of almost one-third in the total number of published industries. In short, the new data series will improve Minnesota’s published hours and earnings estimates with its expansive industry coverage and representation of a larger portion of employment.

Table 3 is a sneak preview of the All Employee Hours and Earnings experimental data series for Minnesota. Included are January 2007 through August 2007 data. For more detailed data see: www.deed.state.mn.us/lmi/tools/ces/default.aspx . The series will be available with the January 2008 data release on February 28, 2008. At this time the new data series is labeled experimental because of our limited experience with editing and reviewing the sample data as well as producing the estimates. For Minnesota the experimental data in Table 3 may change when it is published in February 2008 as a result of additional data added to the series, the end-of-year review, and re-estimates.

A brief analysis of Table 3 shows that for Minnesota statewide, “all employee” in the Financial Activities and in the Professional and Business Services industries receive the highest average hourly earnings, while employees in the Leisure and Hospitality industries receive the lowest. According to the average hourly earnings across sub-state areas, Rochester MSA employees collect the highest average hourly earnings ($28.70) in sharp contrast to Duluth-Superior MSA employees, who receive the lowest ($18.34). A closer look at the two MSAs reveals the reason for this difference.

In the Rochester MSA almost 42 percent of private employees are concentrated in the Education and Health Services industry and receive fairly high earnings, while only 27 percent of the Duluth-Superior total private employees work in this supersector. In addition Duluth-Superior has nearly 8 percent more of its private employees working in Leisure and Hospitality and Retail industries, both known for lower average hourly earnings.

Average weekly hours for “all employee” across the state show that Manufacturing and Construction industry employees work the most hours per week — at least 1.4 hours more per week than employees in any of the other industries — while employees in the Leisure and Hospitality industry work the fewest hours. This finding was also expected as the goods-producing industries have a larger percentage of employees who fall under the CES “production worker” definition, and they tend to work more overtime hours in contrast to the Leisure and Hospitality service industry where a larger percentage of the employees work part- time.

The resulting average weekly earnings show that employees across the state in both the Construction and the Financial Activities industries bring home the largest earnings, $978.88 and $1,023.08 respectively, since both groups receive high average hourly earnings ($25.78 and $27.97 respectively) and work a large number of hours per week (38.0 and 36.6 respectively). The opposite holds true for Leisure and Hospitality workers, who receive the lowest average weekly earnings of $281.60. Employees in this industry receive the lowest average hourly earnings ($11.88) and work the fewest hours (23.7) as well.

This new data series will provide important information about “all employee” that has been missing and is needed by the CES program to make it more comprehensive in its coverage. The data will also provide improved information for analyzing economic trends.

The NAICS 2007 conversion

The conversions to the NAICS 2007 structure are minor for the CES program. The revisions reflect content changes within the Manufacturing supersector as well as the professional, scientific and technical services sector. The telecommunications subsector is restructured, and the real estate investment trust industry is being eliminated from the finance and insurance sector. In addition several industry titles and descriptions have been updated. At the national level this revision will reclassify approximately 1 percent of employment into different industries. For more information on NAICS 2007 go to www.bls.gov/sae/home.htm . More information will be available from the BLS on Jan. 4, 2008.

For Minnesota the major changes in the NAICS 2007 structure affect eight industries at the statewide level and five industries at the Minneapolis-St. Paul MSA level. The largest data conversion is in 517110 — wired telecommunications carriers — at the statewide level with an employment revision of almost 9 percent. With such a large employment revision, the data series for this industry must be reconstructed to maintain a reliable historical time series. The next largest conversion is in the Minneapolis-St. Paul MSA telecommunications industry —NAICS code 517000 — with an employment change of nearly 5 percent. This adjustment, however, did not require historical time series reconstruction. The effects of all the other NAICS 2007 conversions resulted in employment revisions of less than 1 percent.

 ¹The sample design is a stratified, simple random sample of worksites clustered by unemployment insurance account number. The sample strata are defined to yield a state-based design with sampling rates that minimize the overall variance or sampling error on the statewide total nonfarm employment level.

² This includes all types of workers except those whose major responsibility is to supervise, plan, or direct the work of others. Working supervisors and group leaders who may be in charge of a group of employees, but whose supervisory functions are only incidental to their regular work, are included as production workers.

³ Average weekly overtime hours are currently collected and published for manufacturing industries only. The data series is optional for states to report. Minnesota will not be reporting this data series.

Top