KP: Thank you Al. I'll move right into the issues. It seems to me that in no small measure the fate of the census rests in part with the quality of journalism that will cover and report on the census. I'd like to explain that point.

We'll start with the differential undercount. That's very familiar to everyone in the audience. The undercount itself we've had since 1790. We know that the decennial census does not reach everyone. Not everyone can be found and not everyone will cooperate.

The differential undercount, however, was "discovered" in a major way in 1940. We counted in 1940 as best we could. Then, after Pearl Harbor, more young African American males were conscripted into the U.S. Military then the Census Bureau had counted in 1940. So we realized that we were undercounting at a higher rate in certain population groups.

So starting in 1950 the Census Bureau went to work trying to understand and correct for what we now all recognize as the differential undercount. You also know that the differential undercount persisted in 1990.

It's because of the differential undercount and our increasingly refined understanding of it that we have come by this point in census history. In 2000, the decennial census is widely viewed as a key indicator and reflector of social injustice in American society. Eleanor Holmes Norton, the Congress woman from Washington, D.C. declared the census to be the civil rights issue of this decade. Strong words, but not inappropriate. The differential undercount carries on its back three big burdens.

First the burden of the misallocation of power. That is, because census data are used to allocate Congressional seats and to draw district, legislative, city council, school district boundaries within states, any differential miscount distorts the structures of power in American society. This is to say that the census carries with it the burden of an accurate and adequate representational system.

Also, the census carries with it the burden of the resource flows. The federal funding now is about two hundred billion dollars. That's an annual figure but because the census data lasts for a decade you multiply the funding by ten, which totals two trillion dollars in federal funding that are misallocated if the data are in error.

So we have the burden of the representational system and also the burden of the resource system resting on the back of this differential undercount. Third, and increasingly in this society for exactly the themes of this conference, the census is also about recognition. It's about who's included.

It's about a statistical portrait of this society that leaves no one out. Indeed to oversimplify history just a little bit, the 19th Century interacted with the census primarily around the representational system. The first R of the census.

The Twentieth Century interacted with the census along the additional dimension of resource flows. The second R of the census, resources.

The Twenty-First Century, I am convinced, is going to interact with the census data along a third dimension, the issues of recognition and respect and inclusion. That is, the census is going to be strongly interacting with the racial politics of this society: who's in and who's out. Therefore, the undercount, the differential undercount, carries all of that burden on its back. No wonder we wish we could fix it.

Now let me shift to another point. The census since 1970 has been a mailout/mailback census. In 1970, 85% of the public returned their questionnaire. In 1980, 75%, or a 10% drop in one decade. In 1990, 65%, another 10% drop. The census response rate is an indicator of the declining levels of civic engagement in this society. It didn't create them. It reflects them.

Census form return rates reflect the apathy, the indifference, the anger, the mistrust, the lack of confidence in government. That response rate is a powerful indicator of where this society is today and its lack of civic engagement.

The issue addressed in this unity conference will not be solved unless this country reverses its overall decline of civic involvement and engagement. In this respect, what is Census 2000? It is an opportunity to reverse that decline of civic engagement. We do not have to tolerate a response rate of 65%, as experienced in 1990. We certainly do not have to watch it go any lower although we are currently expecting it might do so.

What we have done and are continuing to do at the Census Bureau are activities designed to re-engage the public in the census itself -- a census in the schools program, a program with tribal leaders involving them in series of meetings across the country, a major paid advertising campaign for the first time in history to try to increase awareness and participation, a partnership program in which we have already signed 25,000 agreements with organizations that will promote the census. All of that has transformed the nature of the census in 2000 from one which is just the activity of this statistical bureau into one that is literally being shared with the American public.

So here is the issue. Can all of that effort -- can the partnerships, the liaison programs, the complete count effort, the census in the schools, the paid advertising -- can it reverse the decline in response rates? If so, we will have used Census 2000 to say to the American society that civic disengagement and civic withdrawal is going to be reversed.

How does that connect to the responsibility of this particular conference? It is the journalists represented in this room who are in contact with those population groups which have the lowest rates of response. They are also the population groups with the highest rates of differential undercounting.

But it will be difficult to solve the problem of differential undercount unless we increase that initial response. In some of the population groups in the areas of the cities that you're most familiar with response rates are as low as forty-five or fifty percent.

Here therefore is what seems to me the really interesting challenge of Census 2000. If people of color in this society can prove to "majoritarian society", though soon to be not a majority anymore, can prove to the white parts of society that they are going to civically engage, that they're going to connect with this country, they're gonna be part of the country, they will have sent a very powerful signal.

The diversity that many want to celebrate in 2000, which by the way the Census Bureau will not only measure but will recognize, especially through its advertising campaign, should to be a diversity that's connecting to the nature of the society itself.

In this manner, Census 2000 can become a civic ceremony which is inclusive and particularly inclusive for those population groups which had low rates of civic engagement and participation in the past. That strikes me as the challenge.

Two big issues are to be addressed -- differential undercount civic disengagement. They're two sides of the same coin. The census bureau has programs to address both, and we hope that American journalists, especially the journalists represented in this room, will help us tell that story.

I believe, it's going to be a big story, one way or the other. If we fail to reverse the decline of participation that itself is a story. It's not a very happy story. It means all of the effort -- the money and investment and partnerships and liaisons and complete count committees and census in the schools could not fix the problem.

That's a very strong statement as to who we are as a society as we enter 2000. I actually think we're going to reverse that decline. If so that will be the big story in 2000. So I invite you to pay attention to us, reporting our flaws when and where appropriate. That's all right with us. But recognize that beyond any particular problem, we are trying to do something that's very big and important: not only do a complete count, not only do an accurate count, but to make a statement to American society that we do not have to tolerate the degrees of apathy, indifference, civic disengagement that we've come to accept and take for granted. Thank you.