
Government and Politics Issues | |
Is Census Sampling Suspect? |
The Clinton administration wants to use statistical sampling rather than
an actual count in compiling the 2000 census. But House Republicans are
suing to stop sampling -- contending it would violate the Constitution's
requirement of an "actual enumeration" of the U.S. population. Experts generally agree the census has been undercounting the population
in the past six decades -- with minorities particularly vulnerable to noninclusion.
On the other hand, when it attempted to adjust for the undercount in 1990,
the Census Bureau overestimated those not counted by one million and did
not discover the error for a year. (The Secretary of Commerce kept the estimated
figures from being used in 1990.)
For the 2000 census, once the Census Bureau believes it has reached 90
percent of the population, it plans to estimate the remaining 10 percent
through statistical sampling. The bureau will take another sample of 750,000
households and use those figures to try to correct for both doublecounting
and undercounting. One problem, even the bureau agrees, regards data at the individual census-block
level -- which is extensively used by government and business. In a letter
to Investor's Business Daily, a bureau official wrote, "It is impossible
to tell beforehand whether the adjusted or unadjusted count will be more
accurate." Source: John Berlau, "Will You Be Counted in 2000?" Investor's
Business Daily, July 8, 1998. |
Home | Support Us | All Issues | Social Security | Debate Central | Contact Us
Dallas Headquarters: 12770 Coit Rd., Suite 800 - Dallas, TX 75251-1339 - 972/386-6272 - Fax 972/386-0924
Washington Office: 601 Pennsylvania Ave. NW, Suite 900 South Building - Washington, DC 20004 - 202/220-3082 - Fax 202/220-3096
© 2001 NCPA