Using Administrative Record Persons in the 1996 Community Census

I. Introduction Information from administrative records are Medicare Enrollment Database from Health Care a possible source of people and housing units missed in Financing Administration (HCFA), and the the Census. This paper documents the evaluation of Registration File from Selective Service. the administrative records person level data used in the-In addition to the files above, for Chicago only, a 1996 Community Census. A coverage measurement 1996 Medicaid file for Illinois was obtained. A personal interview was conducted in all housing units Student Information System file from Chicago Public in the three test sites (seven tracts in Chicago, IL, the Schools was obtained. Characteristics Systems (MTCS) were obtained from interview was conducted via a Computer Assisted Housing and Urban Development. Personal Interview (CAPI), and will be thus forth-For the Ft. Hall and Acoma sites, we also obtained referred to as the Integrated Coverage Measurement an Indian Health Service file from the Department of ICM CAPI. Prior to the ICM CAPI interview, Health and Human Service, Public Health Service. administrative records people and people enumerated Unfortunately the files did not include all ZIP Code during the initial phase of the census were areas of the sites. Leggieri, 1996 discusses these unduplicated and loaded into the CAPI instrument for limitations. the appropriate housing unit. During the ICM CAPI,-The Social Security Administrations NUMIDENT it was possible for an interviewer to ask questions file was only used to validate the social security about people thought to be living in that housing unit, number and to gain missing demographic which could include asking about administrative characteristics for people found on other files. records people. 2. For each individual file, name and address Documented in this paper are: (1) the amount information were standardized. Census geography was of duplication between the administrative records and added to each file where applicable. people enumerated during the initial phase of the 3. Files were merged to form one file of person census; (2) the quality of the administrative records records. Records were dropped if the address was people by source; and (3) available information on the located outside of the site boundary or if the record sensitivity surrounding use of administrative records lacked a name. person information in the 1996 Community Census. 4. Based on a predefined order of selection, duplicate