Employment and Wages Covered by Unemployment Insurance

The U.S. Bureau of Labor Statistics (BLS) publishes several sets of employment statistics, including the Quarterly Census of Wages (ES-202), the monthly Current Employment Statistics (CES, BLS 790, establishment survey), and monthly Current Population Survey (CPS, households).  The U.S. Census Bureau (BoC) produces County Business Patterns, and the Office of Personnel Management (OPM) measures federal employment. 

How do these five sets of data compare?

The following is quoted from the BLS's documentation at ( http://www.bls.gov/opub/hom/homch5_c.htm ).

Comparison of the Quarterly Census of Wages (ES-202) Program with Other Series

A number of other statistical data series are comparable in some respects to those provided in the ES-202 program. These series all have certain applications, strengths, and shortcomings. Because of its broad universe coverage, continuity, and currency, the ES-202 program is one of the most useful.

Current Employment Statistics

The Current Employment Statistics (CES) program uses a sample of about 390,000 establishments to provide current estimates of monthly nonfarm employment, average weekly hours, and average hourly and weekly earnings. Employment estimates derived from the CES are benchmarked primarily to ES-202 records, which cover about 98 percent of all nonfarm employees and 98 percent of those in the private nonfarm sector. For the remaining industries, the CES program uses other sources to estimate employment not covered by State UI laws.

In addition to being both sample-based rather than a universe count and more current, the CES program differs from the ES-202 program in that the former provides paid hours and weighted weekly earnings estimates for production workers in manufacturing and nonsupervisory workers in nonmanufacturing. The ES-202 program provides total quarterly payroll data for all employees, unrelated to hours paid. Also, the CES program data are available monthly, whereas ES-202 data are available quarterly.

Current Population Survey

The Current Population Survey (CPS) is a sample survey of about 50,000 households selected to represent the entire civilian noninstitutional population and designed to measure overall employment, unemployment, and those not in the labor force. In terms of employment, the sample thus includes categories of workers which are entirely or partly excluded from the ES-202 program: Certain farm and domestic workers, the self-employed, persons working 15 hours or more in the survey reference week as unpaid workers in a family-operated enterprise, employees of certain nonprofit organizations, and railroad workers. The CPS also counts employees uncompensated because of temporary absence but excludes workers under 16 years old.

Because the CPS is a sample and surveys households rather than establishments, it cannot present employment and wage data in the industrial and geographical detail available under the ES-202 program, but it does provide demographic characteristics. As a household survey, its focus is on individuals, whereas establishment-based surveys such as the ES-202 program focus on jobs. When providing geographic information, the CPS program tabulates data by the location of the residence. On the other hand, the ES-202 program provides its State and county data by the location of the job. Both CPS and CES data are released within 1 month of the reference period; ES-202 data become available several months after the reference quarter.

Differences between the CPS and the CES

The CES survey is an employer-based survey that provides data on the number of jobs within industries, while the CPS is a survey of households that provides data by demographic characteristics, occupation, and industry.

Employment estimates from the CPS pertain to persons in any type of work arrangement - including the self-employed and unpaid workers in family businesses - whereas those from the CES survey refer only to persons on nonfarm payrolls. As a result, the count of employment from the CPS is larger than that from the CES survey. Partially offsetting the higher estimate from the CPS is the fact that it is a count of persons and, as such, includes individuals only once, regardless of the number of jobs they hold. In contrast, the CES survey is a count of jobs and includes each job for persons who work in more than one establishment.

A partial reconciliation of the two series is shown in the table below. It should be noted that the data shown are not necessarily representative of the difference between the two surveys at all points in the business cycle. Also, there are some differences in concepts and definitions that are not possible to quantify. For instance, the CPS provides information on persons aged 16 and over. By contrast, in the CES survey, any person who appears on a payroll record, regardless of age, is counted as employed.

Reconciliation of employment estimates from the CPS and the CES surveys (NSA):

Other differences in the surveys' methodology and coverage also preclude a prefect reconciliation of the two employment series. For example, the reference period for the CPS is the week that includes the 12th day of the month, while, for the CES survey, it is the pay period that includes the 12th of the month. It is therefore possible for the CES survey estimate of employment to reflect a longer reference period than that used for the CPS.

Finally, coverage in the CPS includes household members who are part of the civilian noninstitutional population. Persons who are inmates of institutions, such as those confined in penal or mental facilities, residents of homes for the aged and those who are on active duty in the Armed Forces, are excluded from the CPS. Also, only households that are in the US are eligible to be sampled in the CPS. In this regard, the coverage of the CES survey is broader: uniformed military personnel who hold civilian jobs are counted in the CES survey because of their civilian employment, and persons who commute into the US from Mexico or Canada and are employed by companies within the US are counted as employed in the CES survey estimate.

County Business Patterns

Quarterly census of employment and wages data from the ES-202 program differ from employment data published in County Business Patterns (CBP) of the Bureau of the Census in the following major areas: (1) CBP data exclude administrative and auxiliary units from "operating" unit data at the 4-digit level and include these data at the industry division level only. ES-202 covered employment, on the other hand, includes data for these units at the 4-digit SIC level. (2) CBP excludes agricultural production workers and household workers, some of whom are included in ES-202 covered employment data. CBP also excludes government units, all of which are included in the ES-202 program. (3) Every 5 years, data are collected for all multi-units within the scope of business and economic censuses and included in the CBP for that year. Annual updates for the larger multi-units are obtained from the sample selected for the Report of Organization Survey, and data for nonsample multi-units are estimated. Annual updates for single units come from the Internal Revenue Service and the Social Security Administration. ES-202 quarterly census of employment and wages data, on the other hand, include data collected from all active units each quarter.

Office of Personnel Management

The Office of Personnel Management (OPM) maintains and publishes a statistical series on Federal employment and payrolls with information by agency, type of position and appointment, and characteristics of employees. Both the OPM and ES-202 series exclude the Central Intelligence Agency and the National Security Agency, the Armed Forces, temporary emergency workers employed to cope with catastrophes, and officers and crews of certain American vessels. OPM data, but not ES-202 data, include employees working in foreign countries, workers paid on a fee or commission basis, and paid patients, inmates, and certain employees of Federal institutions. Conversely, ES-202 data, but not OPM data, include Department of Defense employees paid from nonappropriated funds as well as employees with Federal appointments to the Agricultural Extension Service, County Agricultural Stabilization and Conservation Committees, and State and Area Marketing Committees.

In comparison with the OPM data, ES-202 program data provide more industry, local employment, and wage detail and more frequently updated detail on employment by State. The OPM data series, of course, have certain statistics that have no parallel in the ES-202 program.