Assessing the bias due to non-coverage of residential movers in the German Microcensus panel
Rendtel, Ulrich ;  Basic, Edin ;  ;  Universität <Berlin, Freie Universität> / Fachbereich Wirtschaftswissenschaft

Main titleAssessing the bias due to non-coverage of residential movers in the German Microcensus panel
Subtitlean evaluation using data from the socio-economic panel
AuthorRendtel, Ulrich
AuthorBasic, Edin
EditorUniversität <Berlin, Freie Universität> / Fachbereich Wirtschaftswissenschaft
No. of Pages23 S.
Series 2007,6 : Volkswirtschaftliche Reihe
Keywordspanel survey, labour market analysis, residential mobility, non-coverage bias, log-linear modelling, inverse probability weighting
Classification (DDC)330 Economics
Also published in
AbstractThe German Microcensus (MC) is a large scale rotating panel survey over three
years. The MC is attractive for longitudinal analysis over the entire participation
duration because of the mandatory participation and the very high case numbers
(about 200 thousand respondents). However, as a consequence of the area sampling
that is used for the MC , residential mobility is not covered and consequently
statistical information at the new residence is lacking in theMCsample. This raises
the question whether longitudinal analyses, like transitions between labour market
states, are biased and how different methods perform that promise to reduce such
a bias.
Based on data of the German Socio-Economic Panel (SOEP), which covers
residential mobility, we analysed the effects of missing data of residential movers
by the estimation of labour force flows. By comparing the results from the complete
SOEP sample and the results from the SOEP, restricted to the non-movers,
we concluded that the non-coverage of the residential movers can not be ignored
in Rubin’s sense.
With respect to correction methods we analysed weighting by inverse mobility
scores and loglinear models for partially observed contingency tables. Our results
indicate that weighting by inverse mobility scores reduces the bias to about 60
percent whereas the official longitudinal weights obtained by calibration result in
a bias reduction of about 80 percent. The estimation of loglinear models for nonignorable
nonresponse leads to very unstable results.
If your browser can't open the file, please download the file first and then open it
FU DepartmentDepartment Business and Economics
Year of publication2007
Type of documentBook
Terms of use/Rights Nutzungsbedingungen
Created at2008-06-10 : 09:25:38
Last changed2016-01-05 : 02:38:10
Static URLhttp://edocs.fu-berlin.de/docs/receive/FUDOCS_document_000000000181