[Congressional Record Volume 144, Number 29 (Tuesday, March 17, 1998)]
[House]
[Page H1171]
From the Congressional Record Online through the Government Publishing Office [www.gpo.gov]




                              {time}  1245
                              2000 CENSUS

  The SPEAKER pro tempore (Mr. Hobson). 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 talk about the 
2000 Census. I realize there are not many people in Washington focused 
on that subject today or this week. While the country remains fixated 
on the problems engulfing the White House, the business of government 
must go on. The 2000 Census will be the largest peacetime mobilization 
ever undertaken by the Federal Government, and the planning must 
continue.
  I want to begin by complimenting and thanking Acting Director James 
Holmes. Last week we were headed towards a confrontation over the issue 
of congressional access. Last night I received word from Mr. Holmes and 
we have resolved the issue. I think Mr. Holmes understands how 
seriously Congress takes its oversight responsibilities in regard to 
the census. Given all the controversy surrounding the methodology of 
the 2000 Census, the best way to proceed is to have an open 
relationship in the process of information gathering. Frankly, until 
Mr. Holmes arrived, the administration had a different view.
  Mr. Speaker, we need cooperation between Congress and the 
administration because at the moment the 2000 Census is in serious 
trouble. I have said I believe we are headed towards a failed census. 
The Clinton administration, without the approval of the Congress, has 
designed the largest statistical experiment in U.S. history. The plan 
is multifaceted and complicated. If one element of the plan goes wrong, 
it can destroy the accuracy of the entire census. The plan depends on 
an unrealistic time line and if they do not meet the deadlines at each 
step, the plan could easily fall apart.
  The Commerce Department's own Inspector General has called the plan 
risky. The Inspector General said in December, ``We conclude that 
although the 2000 Census design is risky, the bureau's fundamental 
problem is that it simply may not have enough time to plan and 
implement a design that achieves its dual goals of containing cost and 
increasing accuracy.'' The Inspector General goes on to state, 
``Because this process is long, complex, and operating under a tight 
schedule, there will be many opportunities for operational and 
statistical errors.''
  I have a Ph.D. in statistics and marketing, so I understand clearly 
the operational risk of this plan. As a statistician, the 
administration plan raises too many red flags to move forward and spend 
$4 billion of taxpayers' money.
  Let me try and give my colleagues a basic outline of this grand 
experiment. There are 60,000 census tracks in the United States. Each 
contains about 4,000 people. Under this new, untested theory, the 
administration wants to count only 90 percent of the people in each 
census track. That is unprecedented. For the first time in American 
history we will not attempt to count all Americans. First, they collect 
all the census forms returned by mail for each of the 60,000 census 
tracks. They hope to average about 67 percent response rate in each 
track. Then in each of these 60,000 tracks, they will randomly remove 
enough remaining addresses to add up to 10 percent of the total census 
track and then put them aside. Then they will do what is called a 
nonresponse follow-up with the homes not removed so they have actually 
counted 90 percent of the people in each track. Then they will conduct 
60,000 simultaneous polls to estimate the other 10 percent in each 
census track.
  This has never been tried before. The scope of this experiment is 
simply breathtaking. When you see a poll in the New York Times or CNN 
or USA Today the pollsters typically do one poll and survey 1,000 or so 
Americans. I saw a poll this morning that shows the President's 
approval ratings just went up again, which really has to make one 
question the accuracy of polling. But what this administration is 
talking about doing is 60,000 separate simultaneous polls at the same 
time. It has never been tried before and the potential for mistakes and 
errors is quite large.
  That is just the beginning. After all this has been completed, they 
will conduct an extensive nationwide poll of 750,000 American 
households. This is done to adjust the figures in all 60,000 census 
tracks. Some tracks will be added to, some subtracted from, based on 
this poll of 750,000 households. This 750,000 survey is called the 
Integrated Coverage Measurement or ICM. The administration claims the 
ICM will increase accuracy. That is a huge theoretical leap of faith. 
The Commerce Inspector General says, ``Because of its complexity, the 
ICM is highly vulnerable. In particular, the survey's magnitude, 
quality demands, and tight schedule all present serious challenges.'' 
He added, ``Estimation associated with the ICM survey in particular 
faces lingering methodological questions.'' In other words, it is not 
at all clear that the experiment will increase accuracy at all. We need 
to work together and get the most accurate, best census we can for the 
year 2000, not test or try experiments.

                          ____________________