|
Activity Number:
|
106
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Government Statistics
|
| Abstract - #304335 |
|
Title:
|
Redesign of SOI's Individual Income Tax Return Edited Panel Sample
|
|
Author(s):
|
Yan K. Liu and Gerald Auten and Michael Strudler and Valerie Testa*+
|
|
Companies:
|
IRS and U.S. Department of the Treasury and IRS and IRS
|
|
Address:
|
Statistics of Income, Washington, DC, 20013-2608,
|
|
Keywords:
|
Cross-sectional estimation ; Longitudinal analysis ; Panel sample ; Sample size allocation ; Stratified sample
|
|
Abstract:
|
The Statistics of Income (SOI) of IRS started the Individual Income Tax Return Edited Panel Sample in 1999. It has been used for both longitudinal and cross-sectional purposes. Because of the panel drift over time, the efficiency of the panel is reduced in terms of the longitudinal study of taxpayers' behavioral changes. Further, the supplemental sample is unable to capture enough new high-income returns to support the cross-sectional estimations. Therefore, it was decided to start a new panel in 2007. The new panel follows the same stratified random sample design as the old panel. The key is the sample size reduction and allocation. Using the information from the current panel, the stratum sample sizes are determined by balancing the cost and precision and factoring in the needs of panel users. In addition, a refreshment sample is designed to better represent the out-year populations.
|
