Abstract

Sarah Goodale and Darrell Greene "Improving the Job Openings and Labor Turnover Survey’s Sampling Procedure"

The Job Openings and Labor Turnover Survey (JOLTS), conducted by the U.S. Bureau of Labor Statistics, is a panel sample of 16,000 establishments, in which participants remain in the sample for two consecutive years. Every year a sample of approximately 8,000 establishments is drawn from the 1st quarter data from the Longitudinal Database. Each month one new panel is rolled into the sample while the oldest panel in the sample is rolled out. Previously, each annual sample was weighted to the current sample frame while previous samples remained unchanged. There was also no consideration made for younger units, which are thought to be more dynamic and should be included in the sample as early as possible. With the new sampling procedures, all establishments will be updated to reflect the current frame, and quarterly birth samples will be taken to update JOLTS with younger establishments.