Issues Concerning Imputation of Hispanic Origin Due to Potential Administrative Record Enumeration for the 2020 Census

• Activity Number: 40
• Type: Contributed
• Date/Time: Sunday, August 3, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Survey Research Methods Section
• Author(s): Richard Griffin*+
• Companies: U.S. Census Bureau
• Keywords: Non-response follow-up ; Non-ignorable nonresponse ; Administrative Records

Abstract

For the 2020 Decennial Census, we are looking into reducing the cost of Nonresponse Follow-up by using administrative records as a substitute for field follow-up for housing units that do not respond themselves by, for example, mail or the internet. Enumeration using administrative records is referred to as ADREC Enumeration. In many cases, with the sources we currently have available to us, administrative records have no information on Hispanic Origin or the available information may not be accurate. With these existing sources, the imputation rate for Hispanic Origin for ADREC Enumeration would be very high. Missing data on Hispanic Origin (binary variable, Hispanic or Non-Hispanic) is ignorable if true origin status is independent of whether the origin status of a person could be resolved without imputation. It may be that those for whom data is not available are more likely to be Hispanic. This paper describes a potential non-ignorable nonresponse procedure and an ignorable nonresponse procedure using methodology adapted from Little and Rubin (1987). Both the ignorable and non-ignorable procedures are applied for two simulations using ADREC Enumeration in place of NRFU.