This report compares the addresses listed, housing identified and the persons enumerated in sample areas for which we have three sources: the 1990 Decennial Census, the Post Enumeration Survey (PES) and an Alternative Enumeration (AE). The Census source was keyed from the census forms geographically attributed to each sample area. The PES source was data processed from a survey conducted by the Census Bureau after and independent from the census. Alternative Enumerations were conducted by ethnographers associated with private, non-profit organizations in part supported by Joint Statistical Agreements with the Census Bureau. (See Brownrigg and de la Puente 1992a.)

We will present findings from a sample area on one side of a city block in the Harlem neighborhood of New York City where the Alternative Enumeration and the resolution of the triple match were personally conducted by one of the authors, A. Hamid (see Hamid 1992, 1993). Dr. Hamid provides detailed explanations for each specific housing unit or person missed or erroneously enumerated to account for discrepancies among the three enumerations. Next, we report briefly on findings from three other sites where enumerations by the Census, PES and an AE overlapped. Finally, a number of generalizations are offered to explain the discrepancies found between the three enumerations. Recommendations for improving the census are offered.