Statistical Profile of the JOLTS Program

• Activity Number: 342
• Type: Contributed
• Date/Time: Wednesday, August 11, 2004 : 10:30 AM to 12:20 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Charlotte D. Mueller*+ and Mark Crankshaw
• Companies: Bureau of Labor Statistics and Bureau of Labor Statistics
• Keywords: sample survey ; labor statistics ; response bias ; data collection ; nonresponse adjustment

Abstract

The Job Openings and Labor Turnover Survey (JOLTS) at the Bureau of Labor Statistics produces monthly estimates for job openings, hires, and separations. The goal of the JOLTS data is to reflect the current status of the labor market as well as show trends over time. JOLTS sampling procedures such as careful frame construction and frame stratification guarantee a sample that is statistically representative of the U.S. economy. Further, the JOLTS procedures for unit and item nonresponse adjustment were selected to translate the statistical validity of the sample into valid estimates. But methodology is not the last step in guaranteeing statistically valid estimates. Unit and item response rates indicate differences in how likely different types of establishments are to report, as well as differences in which data elements the establishments can or will provide. Predictably, examination of the reporting patterns reveals differences by industry and by establishment size class. As a sample survey, we strive to collect as much data from as many respondents as possible, but we also must aim for balance among reporters at the establishment and item level in order to reduce response bias. Maintaining high-quality estimates depends on monitoring reporting patterns, taking measures to reduce imbalances in unit and item response rates, and applying methodology in the estimation process to counteract large response bias if it occurs despite our efforts during the data collection process.