Over the past few weeks, the United States Census Bureau released its counts of city and county population from the 2000 census (as of April 1). Similar to situation with states, as reported in a previous Demographic Brief the Census Bureau estimates of large city population were substantially inaccurate when compared to the census counts.

Based upon the rate of growth estimated by the Census Bureau from 1990 to 1999, the 1990 to 2000 estimates for the 57 cities of more than 300,000 population averaged 105.9 percent lower than the counted increase. Overall, the Census Bureau estimated rate of increase would have resulted in a 1.867 million increase in the 57 cities. In fact, the gain was 3.835 million The largest numeric error occured in New York, where the Census Bureau rate would have resulted in an increase of 117,331 persons from 1990 to 2000. In fact, the increase was 685,714, for an under-estimate of 485 percent. Overall, the

The issue of the Census Bureau's erroneous estimates is important in light of the controversy over proposals to use statistical analysis and sampling to adjust the census counts. The Bush Administration has ruled against such an approach, which appears eminently justified in light of these gross errors.