Analyses on unit-non-response in the IAB establishment panel

Abstract

"Within the framework of the IAB establishment panel more than 4000 establishments have been interviewed annually in western Germany since 1993 and the same number in eastern Germany since 1996. Although the response rate is high - also in comparison with other establishment surveys - the unit-non-response can cause distortions in the results if the non-response processes are systematic. In addition the distortions can increase over the course of time in the case of panel surveys in which the same study units are interviewed each year. Against this background the aim of this paper is to examine whether the occurrence of unit-non-response in the IAB establishment panel is connected with certain characteristics. The analyses refer to the longitudinal analysis of 1994 to 1997 for western Germany. On the whole the results do not indicate a selection of the participating establishments which is relevant as regards contents and which increases over the course of time. In addition to the reported analyses which should be continued, other further studies appear necessary: for instance relevant analyses should also be carried out for eastern Germany, it would make sense to do similar studies for the individual cross-sections, and also the item-non-response needs to be examined more closely." (Author's abstract, IAB-Doku) ((en))

Cite article

Hartmann, J. & Kohaut, S. (2000): Analysen zu Ausfällen (Unit-Nonresponse) im IAB-Betriebspanel. In: Mitteilungen aus der Arbeitsmarkt- und Berufsforschung, Vol. 33, No. 4, p. 609-618.