The evaluations suggested that because the A.C.E. did not identify a substantial number of these overcounting errors (mostly duplicates), the use of the original A.C.E. data to adjust the census could lead to overstating the population. Therefore the Census Bureau recommended against adjustment of the 2000 complete-count and long-form-sample data used for fund allocation and other purposes (Executive Steering Committee for A.C.E. Policy, 2001b).
Two principal evaluations of the E-sample identified problems with the classification of erroneous enumerations in the A.C.E. The first was the Evaluation Follow-up Study (EFU), in which field staff, using a more detailed interview than in the original A.C.E., revisited a subsample of the E-sample housing units in one-fifth of the A.C.E. block clusters to determine who should have been counted there. The EFU subsample was about 70,000 people; it was subsequently reduced to 17,500 people for a detailed clerical review that focused on unresolved cases (Adams and Krejsa, 2001; Krejsa and Raglin, 2001). The second evaluation was the Person Duplication Studies, which included computer matching of E-sample records by name and date of birth to census enumerations nationwide. This matching could be conducted because optical character recognition technology used by the Bureau for the first time in 2000 to process the questionnaires made it feasible to include names on all of the computerized sample records. The results of the EFU clerical review and the Person Duplication Studies were combined to develop an estimate of 2.9 million duplicates and other erroneous enumerations in the census that were not measured in the original A.C.E. (Fay, 2001). Such cases included college students who were counted both at their college dormitory and at their parents’ household; prisoners who were counted both at prison and at their family’s residence; children in joint custody who were counted in the homes of both parents; and people with more than one house, such as those who lived part of the year in the South or West and the rest of the year in the North or Midwest.
The results of Fay (2001) were used by Thompson et al. (2001) to construct revised preliminary estimates of the 2000 net undercount.