US Population Estimate Accuracy: 2010

Intercensal population estimates, while generally reliable, are prone to substantial variation in some cases. This is especially so with municipal population estimates.

Between 2000 and 2010, the average discrepancy between the US Census Bureau 2010 estimates and the 2010 census counts at the county level was 3.1% (absolute value). By comparison, among the 50 largest municipalities and census designated places, the average discrepancy was more than one-half higher, at 4.7 using the 2000 to 2009 estimates (there were no 2010 sub-county population estimates). The variations, however, can be substantial in sub-county population estimates. Between 2000 and 2010, the Census Bureau estimated that New York had added more than 410,000 residents. However, the 2010 census count showed a much smaller gain, at approximately 165,000 (2010 estimates are available for New York because it is composed of whole counties).

There were even more substantial variations. The 2009 population estimates for Atlanta and Detroit were more than 25% higher than the 2010 census count. In the case of Atlanta, the 2000 to 2009 population growth estimate was more than 120,000, more than 100 times the actual increase of approximately 1,000. The discrepancies in Atlanta and Detroit were greater than in all but a three of the nation's more than 3,000 counties and each of the counties with larger discrepancies had populations of less than 1,000 in 2010.