INSTEAD OF providing a definitive ruling on the use of statistical sampling to correct the serious undercounting expected from the 2000 Census, the Supreme Court essentially decided that the hard choices about sampling should be left to politicians.
The justices' ruling this week wasn't a surprise. Several of them made the same point during oral arguments last fall, indicating they weren't interested in settling this controversy for the president and lawmakers.
The court ruled 5-4 against using sampling for the traditional head count of the population that is used to apportion congressional seats among states. That is bad news for states with large numbers of immigrants and low-income residents who have been underrepresented in previous census counts.
However, the justices avoided the constitutional question of whether sampling -- a method favored by statistical experts -- violates the "actual enumeration" stipulated in the Constitution. The court's narrow decision allows for the possibility of using statistical sampling when collecting census data that will be used for other governmental purposes, such as the allocation of federal funds.
Before that can happen, though, Congress and the president must reach an agreement. And the current climate in Washington does not bode well for quick and sensible resolution.
The simplest permanent solution to the undercounting problems that have beset the census would be to create a nonpartisan, independent agency to run the count, a body similar to the Congressional Budget Office and the Federal Reserve Board.
But the most important thing is to make sure that every American is counted.
Pub Date: 1/30/99