Share this article

print logo


Barring an unwelcome power play by the Bush administration, a detailed statistical picture of America at the dawn of the 21st century should begin to emerge in stages throughout March.

The lawsuits will take a lot longer, even though they already have been filed.

Census 2000 has turned into a political battleground. Democrats and Republicans have been arguing for months over which set of numbers should be released - the actual count that has always been used, or an adjusted count that uses the best available methods to correct the raw figures with estimates of how many people were missed and how many may have been counted twice.

The adjusted figures should be used. They will be, unless the new administration intervenes. Neither set of numbers will be completely accurate, but the adjusted ones won't just be the most accurate available - they will be the most fair.

Republicans have opposed adjusted counts because census "undercounts" always have been largest in the minority communities that tend to favor Democrats. Democrats are pushing hard for the adjusted figures, for the same reason.

The actual count of 281.4 million Americans reported late last year will be used to apportion congressional seats among the states. The Supreme Court, in a 5-4 vote in 1999, bypassed the constitutional issue but ruled that a Census Act provision allows only an actual count for that purpose.

But a Census Bureau rule debated last year and finalized last fall calls for adjusted numbers to be used, for the first time, for other purposes. That includes redistricting, and the division of federal aid.

In the 1990 census, Buffalo had an actual-count population of 328,123, including 212,449 whites, 100,579 blacks and 16,129 Hispanics. If adjusted figures had been used, the city population would have gained 6,155 people to 334,278, including 211,645 whites, 106,777 blacks and 17,034 Hispanics.

There are no definitive numbers, but each person generally is estimated to equal an additional $1,500 in population-based federal aid. That could have made a $9 million difference for the city. For New York City, by all accounts the most undercounted district in America, analysts estimated lost aid at $2.5 billion.

Census count adjustments aren't guesses. The Census Bureau used an Accuracy and Coverage Evaluation survey of 314,000 households, taken after the census, to examine the actual count's results. By examining undercounts and overcounts in the areas it covered, statisticians can estimate how many people were missed or double-counted in similar neighborhoods elsewhere.

There has been no effort so far to interrupt the painstaking calculations of the adjustment, according to census officials. But even if the Bush administration doesn't intervene, courts may. Virginia and several other states already have passed laws banning the use of statistically corrected counts, and Los Angeles and a group of other cities and counties have sued to make sure they are used.

Ironically, even the opponents of adjusted counts are comfortable enough with statistical science to rely heavily on polling data. Science can be applied to provide even more reliable population data, with safeguards to make sure decisions are left to professionals and not partisans.

There are no comments - be the first to comment