|
Activity Number:
|
405
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Business and Economics Statistics Section
|
| Abstract - #306359 |
|
Title:
|
Modeling CPS Labor Force Time Series in Selected Metropolitan Areas
|
|
Author(s):
|
Jennifer Oh*+ and Richard Tiller
|
|
Companies:
|
Bureau of Labor Statistics and Bureau of Labor Statistics
|
|
Address:
|
2 Massachusettes Ave., NE, Washington, DC, 20212-0001,
|
|
Keywords:
|
small area estimation ; state-space models ; seasonal adjustment
|
|
Abstract:
|
Time series modeling of survey estimates is an effective approach for producing small-area estimates when a long time series is available. Beginning in 2005, the Bureau of Labor Statistics began producing estimates for five metropolitan statistical areas (MSAs) using time series models. These models represent the idiosyncratic features of the CPS survey by explicitly modeling the survey error process with conventional modeling of the underlying true time series. An important benefit of this approach is that it produces seasonally adjusted series where the disturbing effects of survey errors have been suppressed. These MSA models are integrated into a nationwide system of state models. Prior to 2005, labor force estimates for five major MSAs were produced using a nonstatistical approach known as the "handbook method."
|
