The Accuracy of the 2000 Census
For the 2000 census, the coverage of all groups improved over the 1990 census, when the differential undercount was the highest ever recorded. However, the large number of duplicates (people counted twice or included by mistake) and people “imputed” into the count using statistical methods helped reduce the net undercount for all groups by offsetting the number of people missed, estimated at almost 16 million.
Scientific evaluations also showed that the differential undercount between historically undercounted groups (minorities, renters, and children) and better-counted groups (whites and homeowners) persisted. The Black-White (non-Hispanic) differential undercount, while reduced from 1990, was the highest of all racial/ethnic groups: an estimated 3.0 percent differential in 2000, compared to 3.9 percent in 1990. All other racial groups and Hispanics were also undercounted, while whites were overcounted by about one percent. The owner-rental differential undercount also continued in 2000 with an estimated 2.4 percent in 2000, compared to 4.5 percent in 1990.
Reports & Resources
- Census 2000: Undercount and Accuracy (doc) - Funders' Committee for Civic Participation - February 2009
- The 2000 Census: Counting Under Adversity - National Research Council of the National Academies - 2004
- Census Missed Poor and Minorities Most - LCCR/EF - December 18, 2002
- Census Undercount Has Harsh Impact on Nation's Disadvantaged - LCCR/EF - August 20, 2001
- Effect of Census 2000 Undercount on Federal Funding to States and Selected Counties, 2002-2012 (pdf) - PricewaterhouseCoopers - August 2001
