Improving Census Accuracy
GGD-94-116R: Published: Aug 2, 1994. Publicly Released: Aug 2, 1994.
- Full Report:
Pursuant to a congressional request, GAO reviewed the effects of eliminating: (1) the long form used during the 1990 decennial census; and (2) some of the questions on the 1990 census short form. GAO noted that: (1) the Bureau of the Census concluded that eliminating the long form would have minimal effects on its mail return rate and the accuracy of census data; (2) eliminating two-thirds of the questions on the short form would increase the Bureau's mail return rate by about 6 percent and reduce the overall omission rate by about 10 percent; and (3) the Bureau does not believe that its statistical model is suitable for assessing the effects of eliminating questions on the short form.