|
Activity Number:
|
514
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 2, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Survey Research Methods
|
| Abstract - #308239 |
|
Title:
|
A Multivariate Time Series Model for Small Areas in the Dutch Labor Force Survey
|
|
Author(s):
|
Sabine Krieg*+ and Jan Van den Brakel
|
|
Companies:
|
Statistics Netherlands and Statistics Netherlands
|
|
Address:
|
, Heerlen, 6401CZ, Netherlands
|
|
Keywords:
|
structural time series models ; Kalman filter ; small area estimation ; timeliness ; unemployment
|
|
Abstract:
|
The Dutch Labor Force Survey (LFS) is based on a rotating panel design. The monthly sample size is too small to produce reliable monthly estimates for the unemployment at both the national level and domeins with the generalized regression estimator. Therefore the monthly unemployment figures are based on the data observed in the preceding three months. In this paper a multivariate structural time series model for six domains is applied to the data of the LFS using monthly data obtained in the first wave of the panel. The model borrows strength from data observed in preceding periods and from other domains, resulting in a remarkable reduction of the standard error compared with the generalized regression estimates. This enables us to produces monthly figures instead of an average of the preceding three months with the advantage of improved timeliness and more realistic monthly figures.
|
