The American Community Survey: Quality of Response by Mode of Data Collection in the Bronx Test Site

• Activity Number: 338
• Type: Invited
• Date/Time: Wednesday, August 14, 2002 : 2:00 PM to 3:50 PM
• Sponsor: Section on Government Statistics*
• Author(s): Joseph Salvo*+ and A. Lobo
• Affiliation(s): NYC Department of City Planning and NYC Department of City Planning
• Keywords: census response ; data quality ; imputation ; 2010 Census ; American Community Survey ; mode of data collection

Abstract

The ultimate success of the American Community Survey (ACS) will rest on its capacity to provide users with reliable data for small areas more than once a decade. Given the complex demography of its population and its abundance of hard-to-enumerate groups, the Bronx poses formidable challenges for the ACS. Previous research on the 1996 ACS in Rockland County, NY showed that, despite lower mailback rates, the ACS had a lower reliance on imputation of key long-form characteristics than the 1990 decennial census. This study extends this line of research by examining the level and quality of response for 10 Bronx subareas, using the 1990 Census, the 2000 ACS, and selected administrative data. Differences in levels of response and item imputation by mode of data collection (mail versus enumerator response) highlight major differences between the ACS and the decennial census. An understanding of these differences is essential as data users attempt to assess the ramifications of eliminating the 2010 Census long-form and replacing it with a continuous ACS. Further, this research attempts to inform future assessments of the ACS as operational data from the 2000 Census become available.