[Congressional Record Volume 144, Number 39 (Tuesday, March 31, 1998)]
[House]
[Page H1780]
From the Congressional Record Online through the Government Publishing Office [www.gpo.gov]




                 HEADING TOWARD A FAILED CENSUS IN 2000

  The SPEAKER pro tempore. Under the Speaker's announced policy of 
January 21, 1997, the gentleman from Florida (Mr. Miller) is recognized 
during morning hour debates for 5 minutes.
  Mr. MILLER of Florida. Mr. Speaker, I rise today to once again 
express my deep concern that we are headed towards a failed Census in 
2000. Last week, the General Accounting Office released a new report 
stating that the risk of a failed Census had increased since their last 
report in July. Census 2000 was already in their high-risk category, 
and now things have gotten worse. We are just 2 years away from Census 
day, and the risks are increasing.
  Why are we headed towards a failed Census? For one very simple 
reason: The Clinton Administration has unilaterally designed the 
largest statistical experiment in U.S. history. And despite their 
sincerity, the Census Bureau just does not have the technical 
capability to pull it off.
  The plan that they and their statistical experts developed is 
breathtaking in its complexity. I have a Ph.D. in marketing and 
statistics, and I must say, from an academic standpoint, it is an 
interesting theory.
  But the Census is not a theory; it is a massive field operation, and 
the more complex you make it, the more the chance of failure.
  Now, some in the media who have sided with the administration do not 
want to face reality. They have invested so much in this polling theory 
that they want to find some other reason why this Nation is headed 
towards a failed Census. So now they, with the help of my friends in 
the Democratic Party, have come up with a new reason: It is Congress' 
fault.
  Of course, it is the administration's plan that is headed towards 
failure. The majority in Congress has been warning for almost 3 years 
now that the administration's plan cannot work, but that does not 
matter. The defenders of polling theory have to blame someone, so it is 
Congress.
  Now, I am fair-minded, so at the first hearing last week of the new 
Census Subcommittee, we decided to ask the GAO some questions. We asked 
if Congress was responsible for the following problems that are leading 
towards a failed Census. We asked the following questions:
  We asked if the Commerce Department's Inspector General finding that 
the decennial census software is not being developed in accordance with 
any well-defined process; and the answer was, Congress has nothing to 
do with it.
  The Commerce Inspector General's finding that estimates of software 
development schedules and resources are not realistically for the dress 
rehearsal or the Census; the answer was Congress has nothing to do with 
it.
  The Commerce Inspector General's conclusion that he questions the 
Bureau's ability to develop and implement complete accurate software 
for the Census; no congressional fault.
  The Commerce Inspector General's reporting that the Bureau's matching 
and unduplication programs are so geographically restricted that they 
will virtually guarantee more errors; again, no congressional fault.
  The fact that the ICM sample drawn by the Bureau mistakenly included 
commercial addresses which would have thrown it completely off; again, 
no congressional fault.
  The vague and incomplete guidance provided by the Bureau to local 
governments that, according to GAO, hindered efforts to establish 
complete count committees; no congressional fault.
  The Commerce Inspector General's finding that the Bureau is not 
giving itself enough time to follow up on households that do not 
respond in the first 2 weeks; no congressional fault.
  The fact that the Bureau's plan forces nonresponsive follow-up to be 
completed in just 6 weeks, instead of a more realistic time frame given 
that it took 13 weeks last time we did a decennial Census; this is not 
Congress' fault.
  The fact that the Bureau's plan for the ICM assumes it can contact 
five times as many people as it did in 1990, and do it in half the 
time, 13 weeks versus 28 weeks; that is not Congress' fault.
  The fact that if the response rate in this short 13-week time frame 
for the ICM falls below 98 percent, the Census will become less 
accurate.
  The Commerce Inspector General reporting that experimented field 
managers feel the ICM sampling plan is unrealistic and they are 
assuming a 98 percent response rate; this is not Congress' fault.
  The incompatibility of the Census Bureau's plan to start the ICM 
before nonresponsive follow-up is complete with the findings of the 
Inspector General that ``the integrity of the ICM hinges on the 
assumption that it is fully independent of nonresponsive follow-up;'' 
again, this is not Congress' fault.
  The strategy of hiring moonlighters as Census enumerators, that the 
GAO has described as questionable; this is not Congress' fault.
  The high rate of duplicative or nonexisting households on the address 
lists; that is not Congress' fault.
  The problem with accuracy and completeness of the address list and 
matches provided to the localities by the Census Bureau; it is not 
Congress' fault.
  The lack of information and resources provided by the Bureau to local 
communities that wish to review the address list; again, not Congress' 
fault.
  The Bureau's failure to complete and present a comprehensive design 
review in January 1998, as promised, to the Inspector General; that is 
not Congress' fault.
  The answer to all these questions was the same. Congress has nothing 
to do with the problems. These are specific design flaws in the Clinton 
Administration's unprecedented plan.
  If you want to save the Census, simplify the design and go back to 
what you know works.




                          ____________________