Fixing the Numbers

Although advertising campaigns and public-relations efforts are already underway, the U.S. Census for the year 2000 will officially begin April 1. April Fool's Day will be an appropriate date for a counting process that the last time it was conducted missed several million people, especially African-Americans, Hispanics, urban dwellers, immigrants and the poor. Although later studies estimated the extent of the undercounting, the official figures were not adjusted, and undercounted states suffered corresponding cutbacks in their representation in Congress and shares of federal funds.

Getting an accurate assessment of the U.S. population, as well as demographic data vital to school districts and businesses alike, would be next to impossible with a door-to-door enumeration. Only 65 percent of Americans returned their census forms in the last census in 1990; in some areas, rates were as low as 39 percent. Undercounted states like California lost as much as $2 billion in federal aid.

The Census Bureau had suggested a way to avoid this undercount, using statistical techniques to determine a more accurate estimate of the total population. Unfortunately, these techniques (known as "statistical sampling") were banned last year by a GOP Congress concerned that the minorities and inner-city voters who comprise a disproportionate number of the undercounted would elect Democrats and that more accurate numbers would threaten Republican seats. (The same Congress later attempted to include the census as an "emergency" appropriation, presumably because no one could have predicted that the year 2000 would come exactly ten years after 1990.) The Supreme Court ruled last year that current law forbids sampling, and without a change from Congress, the undercounts are likely to continue indefinitely.

Democratic principles require a compensatory accounting for those "missing" from the census, especially since many of them are members of marginalized groups. To settle for an inaccurate enumeration of the population is akin to relegating the uncounted to second-class citizenship. Without an effort to compensate for the difference, inevitable and expected statistical error becomes a concerted effort to ignore minorities and the poor when conducting the business of government.

The Census Bureau and undercounted states have already begun multi-million-dollar public education programs to encourage participation in the census. We wish them success and hope that the count will be as accurate as is possible under the circumstances. We look forward to the day when political factions will no longer attempt to maintain power by fixing the numbers. A sensible census is definitely within our reach.