Abstract:
|
The U.S. Census Bureau has been investigating ways administrative records can be used to imrpove data quality and reduce respondent burden in the American Community Survey (ACS). Possible burden-reduction scenarios include the removal of the income section or individual items for all or a subset of ACS respondents. I evaluate various potential uses of administrative records, including: 1) modeling survey responses for the respondents that did not receive the income questions and 2) using administrative data to replace survey income. Currently available administrative records presents challenges for direct replacement in many cases. I conclude that more administrative data is needed than is currently available to either eectively model susurvey responses or improve income estimates through replacement with administrative records. However, there is considerable scope to improve the data, especially to address under-reporting, and reduce respondent burden using administrative records.
|