|
Abstract:
|
Linked survey and administrative data files have great potential for policy research, as they can mitigate known problems with the survey data, itself. By linking the survey data with administrative data, two major limitations are overcome. Survey data contain a significant amount of measurement error especially when it comes to measuring public program enrollment. Administrative data do not include data on those eligible, but not enrolled, in the program or those who would potentially be eligible, if the program rules changed. By linking the files, the eligible, but unenrolled, information can be provided by the survey data and the program enrollment data can be provided by the administrative data (correcting the survey measurement error). This work is crucial for program simulation modeling used to score federal legislation and to evaluate particular public programs.
|
