Office of Survey Methods Research

Abstract

Anne E. Polivka (1994) "Comparisons Of Labor Force Estimates From The Parallel Survey And The CPS During 1993: Major Labor Force Estimates," CPS Overlap Analysis Team Technical Report 1 .

The Current Population Survey (CPS), a national survey of 60,000 households, provides a monthly picture of the Nation's labor force including measurements of total employment, unemployment, and characteristics of those not in the labor force. Conceived in the 1940's, the CPS has remained virtually unchanged since 1967. However, since then, there have been many societal changes, including growth in service sector employment; an increase in the role of women in the labor force, particularly mothers; and shifts in the nature of employment, such as more part-time work and less permanent attachment of employees to their employers. In addition, there have been many advances in survey design techniques and data collection procedures which facilitate the development of more accurate measures. To reflect these changes, in 1986, the Bureau of Labor Statistics (BLS) and the Bureau of the Census began a collaborative effort to redesign the CPS. The primary goal of the redesign was to improve the quality of the data derived from the survey by modifying the questionnaire to elicit more accurate data from respondents and converting all data collection to computer-assisted interviewing in order to reduce the potential for interviewer error.

To assess the effects of the redesign, a parallel survey of 12,000 households was conducted using the new collection procedures and questionnaire for the period July 1992 through December 1993. In this report, we compare labor force estimates obtained from the Parallel Survey with estimates obtained from the CPS . These comparisons provide our best estimate of the effects of changing from the questionnaire and data collection procedures used prior to January 1994 to those used from January 1994 forward.

This document contains two sections. The first section highlights survey design features relevant to interpreting the estimates presented in this report. The second summarizes differences in the estimated unemployment rates, employment-to-population ratios, and labor force participation rates from the two surveys (the major labor force estimates.) Salient findings are highlighted at the start of each section that contains estimates. Brief discussions of the reasons for differences between the surveys are provided in each section. More detailed discussions of the effects of questionnaire changes and computer-assisted interviewing can be found in the reports of the Questionnaire Evaluation Work Group and the Mode Effects Work Group The questions used in both surveys to classify an individual's labor force status appear side-by-side in appendix A. The questions are presented by series. A second technical report will discuss estimates of the characteristics of the unemployed, employed, and those not in the labor force (e.g., the duration of unemployment, the estimated percentage of the employed who are part time for economic reasons, and the estimated percentage of those who are not in the labor force who are discouraged workers.) Throughout this report, the estimates generated from the data collected with the "old" questionnaire and the State-based 60,000 household sample will be referred to as the CPS estimates or the CPS. Estimates generated from the data collected with the "new" revised questionnaire using the 12,000 household national parallel survey will be referred to as the Parallel Survey estimates or the Parallel Survey.

This report is preliminary. More detailed discussion and analysis of the data will be presented in a final version of this report. In addition, the final version will include an examination of data related to the consistency of the estimates, such as gross flow and month-in-sample estimates.

Last Modified Date: July 19, 2008

Recommend this page using: