Panel Analysis of Household Nonresponse and Person Coverage in the Current Population Survey

Edwin L. Robison and Christopher Grieve

Abstract

Nonresponse and coverage problems affect every household survey, but it can be difficult to determine the impact on estimates. The Current Population Survey is a monthly panel survey, and each panel is included in the survey 8 months. The panel aspect of the survey expands the possibilities for analyzing nonresponse and coverage. For example, panel data on relationships of household members to head-of-household can be examined to determine if relationship structures systematically change during the eight months. Another possibility is to link panel microdata across 8 months. The linked microdata can be used, for example, to measure differential response (and presumably nonignorable nonresponse) among households with different relationship structures. Results from analyses of panel data are presented and suggestions are made for further research.