Online Program

Modeling nonresponse effects for labor turnover
*John Dixon, BLS 

Keywords: nonresponse, bias, administrative data, labor

The Job Opening and Labor Turnover Survey (JOLTS) estimates the number of hires, separations, and job openings that exist for a given month in the US labor market. The current study examines adjusting for nonresponse in the JOLTS survey. Administrative data from the Quarterly Census of Employment and Wages (QCEW) are used to model some of the labor dynamics by matching the establishment survey respondents and nonrespondents to their employment estimates from the same period. The change in employment between months measured by the administrative data is related to the change of employment within months as measured by the JOLTS survey. The estimates from the models are used to study potential nonresponse effects for the survey.