|
Activity Number:
|
207
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #308189 |
|
Title:
|
Modeling Rotation Group Bias and Survey Errors in the Labor Force Survey of Statistics Netherlands
|
|
Author(s):
|
Jan Van den Brakel*+ and Sabine Krieg
|
|
Companies:
|
Statistics Netherlands and Statistics Netherlands
|
|
Address:
|
Post Office Box 4481, Heerlen, 6401 CZ, Netherlands
|
|
Keywords:
|
rotating panels ; structural time series models ; small area estimation ; Kalman filter
|
|
Abstract:
|
The Dutch Labor Force Survey (LFS) is a monthly survey, based on a rotating panel design. A major problem is that systematic differences occur between the subsequent waves due to panel effects, which is known as rotation group bias (RGB). Furthermore the monthly sample size is too small to produce reliable monthly estimates with the generalized regression estimator. Therefore, a multivariate structural time series model that accounts for the RGB and the autocorrelation between the different panels is applied to the monthly data of the LFS. With this model, estimates are obtained that are corrected for the RGB. Furthermore the model borrows strength from data observed in preceding periods, resulting in a remarkable reduction of the standard error compared with the generalized regression estimates. This results in a substantial increase of the accuracy of the monthly unemployment figures.
|
