• Overview
• Concepts
• Data sources
• Design
  • The sampling frame
  • Reliability
  • Methods for maximizing response rates
  • New BLS initiatives
• Calculation
• Presentation
• History
• More Info

• Methods Home
• About
• Programs
• Glossary
search

Handbook of Methods Quarterly Census of Employment and Wages Design

Quarterly Census of Employment and Wages: Design

This section is based largely on the 2014 Office of Management and Budget (OMB) clearance of the Quarterly Census of Employment and Wages (QCEW) program. The program’s universe of business establishments consists of those establishments which are respondents to the U.S. Bureau of Labor Statistics (BLS) in partnership with the 50 states, the District of Columbia, Puerto Rico, and the U.S. Virgin Islands (a total of 53 entities). The original source of data is the Quarterly Contribution Reports (QCRs) submitted to State Workforce Agencies (SWAs) by employers subject to state and federal unemployment insurance (UI) laws. The QCEW data, which are compiled for each calendar quarter, provide a comprehensive business name-and-address file with employment and wage information by industry at the six-digit North American Industry Classification System (NAICS) level and at the national, state, Metropolitan Statistical Area, and county levels for employers subject to state UI laws. Similar data for federal government employees covered by the Unemployment Compensation for Federal Employees (UCFE) program also are included. The QCEW data serve as the sample frame for establishment surveys conducted by BLS and selected states. In addition, the QCEW has served as a sample frame for other U.S. government surveys. The BLS role in the QCEW program is to establish and enforce uniform methods and processes that yield consistent data quality for the multifaceted uses of the QCEW data. In this role, BLS takes in raw UI administrative data, seeks to understand the error components involved, and uses methods and processes to reduce error and yield a high-quality economic dataset and sample frame.

The sampling frame

The QCEW is a virtual census of nonagricultural establishments’ employees and their wages. It also includes about 53 percent of the workers in agricultural industries. The BLS process for enhancing the quality of the data collected is completed by the state’s workforce agency staff. The process includes, but is not limited to, training staff; collecting data efficiently from large multiple-establishment employers, achieved through Electronic Data Interchange (EDI); applying statistically valid procedures for editing data, estimating missing reports and data elements (imputation), addressing standardization, and linking records; and using standardized processing systems for quality control procedures for data review. (See “Calculation” section for details about estimation procedures and reliability of the data.)

After the data have been extensively reviewed at the state, regional, and national levels, BLS uses the resulting file as a business establishment sample frame. The data also are summarized to produce counts of establishments, employment, and wages for all counties; all Metropolitan Statistical Areas; each of the 50 states, the District of Columbia, Puerto Rico, and the U.S. Virgin Islands; and the nation, for all levels of industrial aggregation.

As shown in table 1, the number of private business establishments (worksites) covered by the QCEW in 2014 was about 9.07 million and the covered employment was about 115.6 million. In addition, about 61,000 federal government, 68,000 state government, and 165,000 local government establishments are covered; private households are excluded, as is all agriculture except forestry. The total number of covered establishments was about 9.36 million and covered employment about 136.61 million. The QCEW series has broad economic significance in measuring labor trends and major industry developments, in time series analyses and industry comparisons, and in special studies such as analyses of establishments, employment, and wages by size of establishment.

Sampling weight. The QCEW is a census of establishments, or worksites; hence, every unit is in the sample and represents itself only. That is, each unit has a sampling weight of 1.00 (absolute certainty).

Reliability

The QCEW is designed in such a way that it can identify industries by geographical location and link establishments by multiple worksites in order to produce firm-level data. The QCEW has a longitudinal database in which it can link data over time and capture business mergers and acquisitions. It can also link by UI number.

An inherent strength of the QCEW is its basis in mandatory UI reporting and the built-in potency of that system. For example, each initial claim for UI benefits includes a check ensuring that the employer exists and is contributing to the UI compensation fund. Based on a weekly survey of over 270,000 initial claims and more than 14 million claims in total, the system ensures complete coverage.

The QCEW data are censuses of establishments and therefore are not subject to sampling error. However, some other types of errors can occur, such as invalid county, industry, or ownership codes; data entry mistakes; and over- or underreporting of employment and wages. To control for these errors, BLS has put in place extensive quality control procedures. Among these procedures are (1) improving data collection methods, including EDI and especially for large multiple-establishment employers; (2) standardizing data-processing systems so that they include edits, imputation, record linkages (with standardization of addresses), and industrial classification coding; and (3) standardizing training of staff at state, regional, and national levels in reviewing data according to the guidelines provided by the QCEW program office and stated in official memorandums.^⁠1 Any reports that are identified as suspect after completing these procedures are individually reviewed. Contact with respondents is frequently used to validate significant movements or to correct the data.

The two most important initiatives undertaken by BLS to enhance the quality of QCEW data are the establishment of the Multiple Worksites Report (MWR) and the establishment of the Annual Refiling Survey (ARS). Two separate OMB clearances are obtained for these surveys. The MWR is sent quarterly to multiple-establishment employers for the purpose of asking them to break out their consolidated reports to the establishment level. For example, the UI system requires some employers to provide data for all of their operations within a state; the MWR asks the employer to provide information on each establishment so that all records on the file can be at the establishment level, which is generally the sampling unit for most BLS surveys. The information improves the quality of local economic data by reporting the location and type of economic activity more accurately.

The ARS is conducted annually on about one-third of the establishments in the sampling frame, for the purpose of updating businesses’ industrial classification, name, and reporting and physical location addresses. The establishments surveyed are selected randomly. State and regional staff are trained extensively in industrial classification coding. In addition, standardized systems are provided to the states and regions to process the data.

Methods for maximizing response rates

Because employers are required to file Quarterly Contributions Reports (QCRs) under state and federal UI laws, response rates for these documents are generally very high. Response rates are about 95 percent for employment (table 2) and about 96 percent (table 3) for wages (like filing of QCRs, reporting of wages is required by UI law). Response rates for total covered employment are about 97 percent (table 3), because nonresponse is mostly from small establishments.

The growth of EDI—the direct transfer of data from the firm to BLS—also provides a high level of response and stability. BLS currently collects more than 432,000 reports from nearly 250 large firms with about 12.2 million employees via EDI. Virtually all of these firms provide data that make up the final estimates.

New BLS initiatives

BLS has undertaken several new research initiatives in the control and measurement of nonsampling error. The 1991 Current Employment Statistics (CES) survey’s benchmark employment estimate transmitted to the QCEW revealed a substantial nonsampling error problem caused by payroll-processing firms. The American Statistical Association formed a committee to review BLS procedures and issued a report in January 1994.^⁠2 BLS has adopted most of the report’s recommendations. BLS has also conducted a Response Analysis Survey (RAS) of payroll-processing firms.^⁠3 The purpose of the survey was twofold: to identify practices that can affect the data collected by the CES survey and the QCEW (the benchmark source data) and to educate payroll processors on proper reporting procedures. In addition, BLS has conducted a RAS of CES and QCEW covered employment reporting.^⁠4 The survey identified factors that affect both CES and QCEW reporting within the same firm. On the basis of these RAS studies, BLS has undertaken an extensive education program with CES respondents. The program includes highlighting correct reporting of problematic items on the CES report form and including special notices on correct reporting in the monthly advance notice fax message. A new RAS was conducted in 2008; a report details new findings of the survey.^⁠5

^⁠1 Official memorandums to the states and regional staff about the QCEW program (U.S. Bureau of Labor Statistics, various dates), available upon request by contacting QCEW, https://www.bls.gov/cew/contact.htm.