Make Full Use of the Census

January 3, 2001

With the Census 2000 figures out, Americans can be congratulated for reversing a downward trend in census participation. That, combined with what appears to have been a more rigorous count and a more detailed set of questions on the long form than ever before, will do much to provide a richer profile of who the American people are and where they live.

While there are no dramatic surprises in terms of where the country is growing (population and hence political representation continue to shift to the South and West), the census numbers can still generate excitment. The raw population totals released last week will soon be followed by more detailed data about the country's demographic makeup. And here's where things get a bit complicated and politically contentious.

Last year, the Supreme Court ruled that federal law requires the use of raw census data (straight from the national headcount, without any statistical adjustments using sampling techniques) to reapportion the 435 seats in the US House of Representatives. But it also ruled states could decide which numbers to use to redistrict their own legislatures.

So the long-running debate over which set of datais best - the raw or adjusted, - isn't over.

Many experts extol the merits of statistical sampling. This method, which uses a base of hard, headcount numbers to project probable populations in hard-to-count areas, promises to reduce persistent undercounts of African-Americans, Hispanics, and recent immigrants.

These groups tend to have a lower mail-in rate, are relatively mobile, and harder for census enumerators to locate.

On the other hand, the census tends to overcount affluent Americans who have children away at college or own more than one home.

In political terms, the adjusted figures could benefit Democrats, whose constituents are more likely to be in the undercounted groups. Republicans, predictably, prefer the raw numbers. It will be up to the incoming Bush administration whether to release the adjusted figures.

The correct decision would be to make both sets of census data available. The adjusted figures can provide a quality-control check over the raw numbers. Different states have different needs and different legal constraints; some are better able to make the necessary determinations.

Moreover, states still under court orders because of past civil rights violations must prove they have not "harmed" minorities by their redistricting. Having both sets of data should help them meet that standard.

The decision also has far-reaching implications for social programs in states with large minority and immigrant populations. Federal grants and social programs are distributed according to population figures. These states have a stake in the best possible tally of needy residents.

All the more reason to have all the data available, along with an explanation of the methods used. Then, informed wisdom can take a front seat in the challenges that lie ahead this year for the states.

(c) Copyright 2001. The Christian Science Publishing Society