Skip Main Navigation Skip To Navigation Content
Link to Census 2000 Gateway  

Press Briefing -- September 19, 2000
Director Prewitt

MR. JOST: Good morning. Thank you for joining us this morning. My name is Steve Jost. I am with the U.S. Census Bureau, and we are today making an announcement regarding the response rate through the mail, over the Internet, over telephones for Census 2000.

Most of you have covered us regularly through our operational press briefings. We have a standard routine, which is being adjusted a little bit today, and because we have a couple of special guests with us I would like to, if I can, introduce to the members of the media those folks who will be speaking today. With us is Secretary Norman Mineta of the U.S. Department of Commerce; Under Secretary Robert Shapiro, to my far right, who has presided over the census throughout his term at the Commerce Department; and you all know Director Prewitt of the U.S. Census Bureau. And all three will be making remarks.

Secretary Mineta joins us today on this occasion because in many ways today's story is a local story. It's about how the census occurred within local communities all across the country. And as a former city councilman, a former mayor of the city of San José, California, and as a former member of Congress, he brings a unique perspective to that point of view. And, with that, I introduce to you Secretary Norman Mineta.

SEC. MINETA: Thank you very, very much, Steve, and I want to thank all of you for being here this morning.

I am proud to be here with our Under Secretary Rob Shapiro of the Economics and Statistics Administration, and with Dr. Ken Prewitt, who is the director of the U.S. Census Bureau.

As all of you are very well aware, the census and its accuracy are fundamental to our system of government. The census determines political representation, it is used to distribute government funding at the federal, state and local levels, and it is used by private industry to guide investment in communities across the country. Now, the very heart of the census is our national commitment that everyone in America should be fairly represented in all of these decisions. But a successful census does more than just deliver the number of people in the country. It also gives us a very strong measure of civic engagement.

Dr. Prewitt has rightly called the census a "national civic ceremony." And the degree to which the American people participate in this ceremony can tell us how connected they feel to their communities, to their neighborhoods and to their government. Since the 1970 census, that participation has been falling. In 1970, 78 percent of American households returned their census form. By 1980, that rate had fallen to 75 percent. And by 1990, it had fallen to 65 percent. There were very real concerns that Census 2000 would continue in this decline. Some would worry that the Census Bureau was overly optimistic in predicting a response rate of 61 percent. And I am very proud to say today that those worries have been proven wrong. The final mail-back response rate has been calculated, and it shows that 67 percent of American households returned their census form by mail and the Internet this year. That result halts the 30-year slide in census participation and actually begins to reverse it.

For much of this year, the American people have seen the census in operation, which is America's largest mobilization in peacetime. Forms have been mailed or delivered to well over 100 million households. Hundreds of thousands of census employees have visited the homes of our neighbors across the nation in an effort to get the most accurate picture of who we are as a people. The first-ever paid national census advertising program has blanketed the country with the message that every person counts.

More than 140,000 partner organizations across the country have worked tirelessly to build awareness in their own communities and in their own neighborhoods. The message they brought to American families across the country was very simple, and that is: This is your future, don't leave it blank. And the American people responded. Thirteen of the nation's largest cities and 14 of the largest counties improved on their 1990 response rate. Nearly 9,300 governmental units across the country, including five state, 8,837 municipal governments, 42 counties and 32 tribal governments met the challenge to lift their response rate by 5 percentage points over 1990. And I am very proud to say that my home state of California and my hometown of San José were among them. Now, that is an accomplishment of which all of us can be proud, and one that I hope is an awakening civic re-engagement in the nation.

In a moment I will turn the microphone over to our Under Secretary, Rob Shapiro, who will talk about the unprecedented partnership effort that has been undertaken as part of Census 2000, and then to Dr. Prewitt, our census director, who will give you some additional details on the final response rates and how they were determined. But first I would like to take this opportunity to say something about the men and women of the Census Bureau who have worked so hard to make Census 2000 a success story. Earlier I said that the census is fundamental to our system of government, and that is not an exaggeration. As a member of the city council in San José, California, and as a former mayor of San Jose, and then as a former member of Congress representing that Silicon Valley area, I can tell you how exactly true that is. It is also true that the census is fundamental to the operation of our economy. Now, quite simply, neither government nor business could function well without the information that the Census Bureau provides.

This census, as all of you know, has seen more than its share of controversy. The men and women at the Census Bureau have seen their work come under more scrutiny than, perhaps, ever before. And the job that they have done, in the face of that scrutiny, should surprise no one who is familiar with them and the work that they do. The men and women at the Census Bureau are on the national radar screen for a few years out of every 10. But they work -- but the work that they do every day, even when the rest of the country forgets to pay attention, is critical to all of us. They don't do their work expecting public acclaim. They don't expect to see their names in the newspapers or hear them on television. The statisticians, the geographers, the demographers, public affairs specialists, programmers, technicians and support staff at the Census Bureau do their jobs simply because this country needs that work to be done, and because they have the skills and dedication to do them. Now, they have done a tremendous service for this great country, and for that, they deserve our thanks and our admiration. They have mine.

And so now I'd like to turn the microphone over to Under Secretary Rob Shapiro and to Dr. Prewitt. Again, Dr. Prewitt, if you would extend to people at the Census Bureau my deep appreciation and gratitude for their dedication. Rob.

MR. ROB SHAPIRO: Thank you, Mr. Secretary. Census 2000 has achieved a milestone before we even have the first population count -- a 67 percent response rate that has reversed a three-decade-long downward spiral in response rates. So, the first number from the census is very good. Roughly 10 percent greater initial response than anyone had predicted before the census.

How did it happen? Well, many factors played a part. First, leadership, from the splendid census directors over the course of this cycle -- from Ken Prewitt and his immediate predecessors Martha Ritche and Jim Holmes. From superb commerce secretaries -- Ron Brown, Bill Daley and now Norm Mineta. From the great team at Suitland -- now, most of these people are usually anonymous, but this is a day to say their names -- from Bill Barron and John Thompson, from Jay Waite, Marvin Raines, Paula Schneider, Nancy Potok and Steve Jost. And from the president and the vice president of the United States, who stood up and put their prestige and authority on the line for a sound and scientific census time and time again.

This was not an easy or a typical census. The census was under, really, unprecedented attack for its methodologies, even though those methodologies were supported by the entire national scientific community. It had to deal with rising distrust of government and of surveys. It had to deal with the fact that virtually all families now have both parents working with very little time to attend to things like answering census surveys. And yet, despite that, we had this remarkable achievement of reversing a three-decade-long decline in the initial response rate.

An operation like this doesn't ultimately succeed just because, or even mainly, because of the people at the top. Secretary Mineta has mentioned the partnerships that the Census Bureau formed with diverse groups across the country, with more than 140,000 civic groups and businesses, churches, synagogues and mosques, Indian tribal councils and nonprofit organizations, schools and government agencies -- all representing groups across America that appreciated the vital important of a complete and an accurate census count.

I'm particularly proud of the thousands of partnerships with ethnic groups reaching out to new immigrants, and to people who don't speak English -- Hispanic, Chinese, Korean, Japanese, Tagalog, Nigerian, Indian, Pakistani, Ethiopian, Tunisian, Thai, Russian, Belorussian, Latvian, Lebanese partners, and many, many more. The clear message we sent, and which struck a powerful chord with our 140,000 partners, was that the census also can be a declaration of pride in every person's heritage in America.

I also want to mention the nearly 12,000 complete count committees at every level of American public life that also helped to promote the census and produce this great response rate -- governors, mayors, county executives, legislators, assembly people, school districts, and other political leaders used these committees to transmit and amplify the message of Census 2000. These groups became a microcosm of America, including responsible men and women from government, business, social services, nonprofits, minority organizations and religious organizations, all willing to spend time and money to bring up that response rate.

In Orlando, Florida, the complete count committee painted the outside of buses with bright portraits of Americans of every kind and with the census message. These buses circulated in hard-to-count areas, were seen by more than a million people.

In Chicago, where plaster cows became a rallying point for the people of that city, the complete count committee created the census cow. California spent more than $24 million for the state complete count committee to use promoting the census. That's one reason that California and most of its major counties and cities produced response rates among the highest in the country.

Other complete count committees sponsored floats in parades, sound trucks, information booths, leaflets and countless other activities to get the message out that the lifeblood of every community depends on a good census count. The complete count committees, the 140,000 partnerships, the Census in the Schools programs, our $167 million ad campaign -- they all played a part, an integrated and orchestrated part in the civic event that is the census. And for a decade to come, the American people will all be better off for it.

Thank you. I'd now turn to Ken Prewitt.

DR. PREWITT: Well, I'd like to certainly echo what Rob Shapiro just said about the partnership effort, the number of groups, organizations around the country who really did make this census a success. And I particularly want to acknowledge this morning the ones who particularly worked with us in this response rate initiative that we called 'How America Knows What America Needs: Can We Meet Our Plus Five Challenge?'

The National Association of Counties, the National Association of Secretaries of State, the National Association of Towns and Townships, the International City Council Management Association, the National Conference of State Legislatures, United States Conference of Mayors, National Congress of American Indians, and the National League of Cities, the Council of State Governments; indeed, those organizations do have representatives with us this morning. I'm going to ask them just simply to stand, if they will, everyone on the front row, with the exception of Herb Sample. (Applause.) And thank you very much from all of us.

Indeed, when we announced on January 11th our '90 Plus Five challenge, all of our partners understood that we were setting the bar very, very high, indeed. It was a bold challenge to state, tribal, local governments to improve their census response rate. This effort would mean that we had done nothing less than to reverse three decades of rapidly declining rates of public participation in the census, a decline which, by itself, is merely one index among many others measuring broader patterns of long-term civic disengagement in our society.

Basically, we are asking the country to roll against the tide, to make headway against strong social and demographic headwinds. As a social scientist who has studied these trends for most of my life, let me say that a decade-long trend line in social behavior is hard to change, hard to reverse. Indeed, trend lines are really very stubborn things. And to change a trend line, especially one that's gone across three decades, takes an unusual and enormous effort. And I want to emphasize that what the country has done in Census 2000 is, in this sense, an extraordinary achievement with an historical dimension to it, and it should not really go understated.

On April 19th, we were able to report on the initial response rate at a news briefing much like this one. I searched then for a way to define our challenge to the American people. For me, the best image I could find provided more than a century ago by the English poet Robert Browning, who said, "A man's reach should exceed his grasp." So, too, should a nation's. And, indeed, this challenge was important operationally, because every percentage point increase in the mail-back response rate reduced by 1.2 million households those we had to visit in the Non-Response Follow-Up period. But it also indicated 1.2 million more households who had really stepped forward and done their civic duty.

I do want to remind us all, though most of you know the census story very, very well, but I do want to remind us that the initial response rate is, of course, the response rate of those who mailed in their census form. We then followed up with all of the nonrespondents with our massive field effort. And we have now visited, of course, all 120 million households for which we have an address.

When I reported the initial results in April, I had to acknowledge that while numerous localities reached the 5 percentage point benchmark, no state had done so. But I also pointed out that the rates were preliminary. More forms were still arriving. In fact, we received more forms in Census 2000 after April than in any prior census.

I said within a few months we would report the final results. Well, today is that day. First of all, as has already been suggested, five states met their Plus Five goal, starting with the most populous state in the nation, California, at 70 percent. These five states are specially indicated on the map: Commonwealth of Massachusetts at 69 percent; the smallest state in geography, Rhode Island, at 67 percent; and then two western states, Nevada and Wyoming, at 66 percent. All five of those states improved their response from 1990 to 2000 by 5 percentage points or more, and, as I said led by the largest state in the nation, California. It's fair to say that the Golden State had a golden census, as you will see from some of the other patterns at the city and county level.

Only a short distance behind the four is an honor roll where four more states reached at least 4 percent improvement from 1990: Connecticut, New Hampshire, District of Columbia and Alaska; followed by another four states at 3 percent -- Colorado, New Jersey, Texas and Maine, all of which is, of course, displayed here on the map.

Among the nation's 100 largest cities, 14 improved their performance by 7 percentage points or higher; Santa Ana, California doing so by 14 percentage points; two others showing an 11 percentage point improvement, Anaheim, California and Boston, Massachusetts. Among the others, they range from Anchorage, Alaska; Denver, Colorado; Nashville, Tennessee; Jackson, Mississippi; Baton Rouge to Chesapeake City in nearby Virginia.

Meanwhile, 13 of our largest cities, led by San Diego and Jacksonville, Florida, met or exceeded their 1990 results. They each improved by 6 percent. Just behind them: Houston, Dallas, Detroit and, as the secretary mentioned, San Jose, which improved by 5 percent. In fact, all but one of the 15 most populous counties, starting with the largest, Los Angeles, exceeded their 1990 rate. And 82 of the top 100 equaled or beat their '90 rate. Among the 100 most populous counties, Suffolk County, Massachusetts, scored the highest performance improvement at a whopping 11 percent, followed by Davidson County in Tennessee, 10; New York County at 8 percent.

So right from the beginning, we had a dedicated effort to sort of boost, basically beat 1990. And the country clearly beat 1990, and it did it by having some of its largest cities, large states, out front on that effort. There also, of course, were very high performers who didn't necessarily improve that much over '90, because they were already starting from a very high base. Fourteen states had a response rate of 70 percent or more, starting with five states in the Midwest, continuing the historic pattern of high rates of participation: Iowa, 76 percent; Minnesota, Nebraska, Wisconsin, 75 percent; South Dakota, 74; and two other Midwest states, North Dakota and Kansas, 72 and 71 percent, still above the 70 percent level.

In addition, the four large population states -- California, Ohio, Michigan, Pennsylvania -- are all on this list, as are Virginia, Colorado and Connecticut. Meanwhile, among the 100 most populous counties, Macomb County, Michigan, holds the pride of first place at 81 percent. Fairfax County in Virginia, 80 percent, follows close behind. Fairfax had a 4 percentage point increase from 1990 to 2000. Orange County, California, is also noteworthy at 76 percent because of its impressive 7 percentage point improvement over 1990. Among the 100 largest cities, Chesapeake City, Virginia scored the highest overall at 78 percent, as well as beating its Plus Five goal by a full 2 percentage points. So, well done, Chesapeake City.

Obviously, we did not do this alone. The Census, as has been emphasized, is as much local as it is national. But, at the Census Bureau, we did try to reinvent the census wheel. We redesigned the look and feel of the form. We used new technology, including putting in place a massive data- capture system capable of scanning tens of millions of incoming census forms, the equivalent of 1.5 billion sheets of paper. And indeed, as we reported, all of that happened at over 99 percent accurate optical-character-recognition level.

That work, as Rob has correctly pointed out -- importantly pointed out, I should say -- all of that work was initiated long before I got to the Census Bureau. It goes all the way back; we started planning the 2000 census, of course, in 1988-1989. It goes back to Barbara Bryant and then Marty Riche and Jim Holmes, but especially to the full-time permanent professionals at the Census Bureau. They started doing the kind of work, reinventing the census wheel, in hopes that we would have the kind of successful census that we've now experienced in 2000.

According to the surveys that were conducted both before, during and after the most intense period of partnership and promotional activities, by the end of the census period, 99 percent of the population knew about the census and 88 percent took a high degree of interest in it.

You know, I made a number of speeches around the country in which I tried to describe the census in terms of what were called the three R's: Representation, resources and recognition. We went out and asked the American people, "Why did you respond to the census?" What's really interesting is the pattern of their own response, their own motivation for engaging in the census.

The top reason was the benefits, getting equitable shares from federal tax dollars. That, of course, was the theme of our major paid advertising campaign. However, shortly or just below that answer, in second place, was the answer, "Civic responsibility." It consistently outperformed other sources of motivation, including the fact that the census is required by law, the fact that it does give us recognition, and that it provides representation. That is despite the fact that we put on the envelope, "This is mandated by law," despite the fact that there's been much, much talk in this city about the implications of the census for representation, reapportionment and redistricting; and despite the fact that many, many, many organizations stressed the importance of recognition.

All three of those reasons were less serious or less central to the response of the American people than their own acknowledged sense of civic obligation and civic responsibility. And I want to emphasize that, because I do think it's important. It is a civic ceremony or can be or should be. It is the record of our common existence as inhabitants of the nation we call America. It is the meaning within the great Woody Guthrie anthem, "This land is your land, this land is my land." It is what we honor every July 4th, our connections with those who came before us and with whom we sort of share the responsibility for this nation.

And I think the fact that spontaneously so many people did express civic responsibility and civic obligation as their motivating factor is a fact that we should take acknowledgment of. I also want to point out that of the several thousands of people who came in and ran our local offices, our senior management people in the local offices, the more I went around and talked to them about why they did it, the more of them said, "We just believe in it. We just want to do it. We just want to be here. We care about this." An awful lot of civic sense of obligation and responsibility to their communities.

When we launched this, of course, we crossed our fingers. Most of Washington, D.C. did. Certainly some of us out in Suitland did. That is, we did not know back in April what kind of response rate we would get. Indeed, as you know, our own predicted response rate was closer to 61 percent. So I want to join the Secretary in saying really serious thanks to the career professionals at the Census Bureau, and I really appreciate the fact that Secretary Mineta singled them out, because they do deserve that -- that attention.

My own experience -- I'm not a career professional at the Census Bureau, I've only been here two years, but certainly my esteem for the dedicated career scientists and public servants at the Census Bureau has really been enhanced by the opportunity to work very closely with them over this last two-year period.

I will stop there. We will take your questions, as we always do, of course. And Steve Jost will organize that.

MR. JOST: For those of you who are new to us, we have reporters with us by phone. We'll alternate between questions in the room and questions on the phone. We do have folks with mikes so that your questions can be picked up on audio. And if you would identify yourself -- your name and your affiliation -- with your question. With that, we'll take our first question in the room right here.

Q Steve Piasecki with the Charleston, South Carolina, Post and Courier. You have a lot of good news today. But I wonder what your analysis is of South Carolina, which has the worst response rate in the continental United States.

DR. PREWITT: One of the important things about running this program as publicly as we did, and by definition there are going to be states at the top and states at -- not so high -- in areas not so high -- in any situation like this. What it allowed us to do was to then try to re-energize the state with respect to our Non-Response Follow-Up work. The fact that we started from a lower base simply meant we had to put more resources into those states and, indeed, try to get the local leadership to help us in that phase. And actually that worked out just well. In an odd way, we were able to take advantage in some instances where states or cities hadn't done very well, and it really did re-energize the leadership -- the partners, and the mayors, and the governors and so forth.

I went to a number of cities, for example, which had low response rates. And I went there right as we hit the Non-Response Follow-Up period, and it was very impressive. It seemed like they sort of looked at the numbers and said, "My gosh, we've got to really get down to work to make sure we get our Non-Response Follow-Up rates."

So, it's hard for me to explain any given state's place on the ranking. I could explain -- it's easier to explain some of the higher states because I went and saw all the efforts that they put into it. But the good news is that by the time we finished the Non-Response Follow-Up, we had gotten to all 100 percent of our households. It put a bit more work on us in places like South Carolina, but fortunately we had less work to do other places so we were able to devote the resources to it.

MR. JOST: Butch John.

Q Dr. Prewitt, just basically a two-part question. How many households of the 120 million that you've reached do you feel like you got responses from? And (b), compare what the static is as to the undercount in 1990, if you can extrapolate?

DR. PREWITT: Certainly. It is very, very important to stress that you can have a quite good census -- both in terms of the mail-back response rate, and in terms of the level of cooperation once we're in the Non-Response Follow-Up -- and still not solve the undercount problem. The undercount issue is something that happens in the last 2 percent of a count, and its differential, as we know, which means it's not the last 2 percent of every geographic area and every demographic group equally. For some groups it may be 5, 7, 8, 10 percent. For other groups it will be 0. We cannot extrapolate on the basis of what we currently know whether we will have a smaller undercount in 2000 than we had in 1990. We have said from the very beginning that with respect to the last 2 to 3 percent of the population, it's extremely difficult to get them. But we will not be able to give the country, obviously, an answer to that question until we've completed our Accuracy and Coverage Evaluation. This is by far our most ambitious and extensive quality control operation. We've already fielded that survey -- it's now out of the field -- and we're beginning to do the data analysis. But we will certainly be reporting back to the country, but it won't be until about February/March when we've finished all of that work, whether we were able to make a dent in the undercount. And that we simply -- the jury is still out on that one.

Q Thank you.

MR. JOST: Right over here in the front row.

Q Pat Joplin with Washington Radio. Your map basically shows that the upper Midwest had the highest response rates; the southern part of the country the lowest. Does that mean there's a differential in civic responsibility according to the area of the country? Or what do you attribute that to?

DR. PREWITT: Well, this is an historic pattern. In fact, one of the things that did happen in Census 2000 is we were able to close the gap a little bit. But I think a lot of things account for that. In the southern-tie cities, for example -- especially the larger cities -- you have a very, very large immigrant population, obviously, in Texas and in Florida. The new populations coming to this country are less familiar with the census. Sometimes they're afraid of the census. They don't understand its purposes. That accounts for a lower response rate. We've also got extremely complicated cities with a lot of rental units. People who rent their homes respond at a lower rate than people who own their homes. So, you do have some demographic characteristics that separate the upper Midwest from some of the Southern states. You also have higher levels of poverty and lack of education in some of the Southern states. That we know correlates with the response to the census. So, there are demographic, as well as attitudinal correlates, that I would point out.

The upper Midwest, sometimes called the Lake Wobegon effect, where everybody's above average. And, lo and behold, they're all above average.

MR. JOST: With that, we'll go back to the telephones. Adam Bell of the Charlotte Observer.

Q Actually, my question was just answered. Thank you.

MR. JOST: Okay, we'll go to another on the telephone. Clarissa Walker with the Augusta Chronicle.

Q I wanted to know how much the 2000 census cost to date?

DR. PREWITT: Surely. The 2000 budget itself is $4.2 billion. The overall cost of the census -- as I say, we started planning this census more than 10 years ago -- what we call the cycle cost, that is the total cost of the census, is closer to $6.5 billion. For example, it was almost a billion dollars that went into address work, and that was back in '98 and '99. We, of course, don't yet have the final accounting. We're closing offices right now. We closed -- this weekend we will close about 120 offices. No, sorry -- last weekend we closed 120 offices. And, in about another 10 days we'll close another 350 offices. Obviously, as we close offices, we collect all the remaining bills and we pay them. And so it's not until we can really finish that task, which will take us through all of this month of September, can we give a final accounting of the cost of the census. But it will be in the neighborhood of that which was budgeted.

Q Thank you.

DR. PREWITT: I should -- just one other thing -- and I do want to say, for the record, as I've said before, Census 2000 will be in the black. We simply don't know by what magnitude, but Census 2000 will be in the black. And part of what is going to explain that, as we begin to get all the final numbers, will be this response rate. The American people stepped forward and sent in those forms, and that made it a less expensive census.

MR. JOST: Over here. Steve Holmes, New York Times.

Q Two questions. Preliminary data indicated that the improvement was mainly fueled by improvement amongst minority respondents, primarily Hispanics. Does your final data confirm that? And second question, any -- would you like to make any projections on your return rate, your final return rate?

DR. PREWITT: Both reasonably complicated questions. We tried to do, as Steve Holmes knows, and you can read it in the New York Times, some analysis of the response rate by demographic groups. Now, the only way you can do that is go back and look at the areas in terms of the demography in 1990, because we don't have the 2000 census data yet. We're working with 1990 census data, of course, when we're mapping the demography of the population. Based upon that analysis, it suggests that the minority areas, the Hispanic areas, African-American areas, Asian areas, improved more substantially than some of the White areas. And yes, that pattern sustained itself, Steve, through this most recent initial response rate analysis. It's preliminary analysis. We'll be glad when we've got definitive data. But it does appear -- indeed, as you listen to me read off the cities -- Santa Ana, Anaheim, Houston, Miami, Hialeah -- these are cities with very large Hispanic populations, other cities -- Dallas-Fort Worth, New York County, Boston -- other cities with large minority populations, and so we really think it made a difference, that enormous effort that was put in by many civil rights groups and many other organizations working to improve the count in the hard-to-reach, hard-to-count areas.

The second part of the question has to do with what we call our return rate. And it is important for the press to understand the difference between the response rate and the return rate. The response rate is based upon our 120 million addresses. Now, some of those addresses have no one living there, so by definition they could not have returned a form. And they're vacant. They're on the market. They've been demolished. We still have them on the address file, but until we actually send someone there, we do not know the status of the housing unit. We are now processing the data, calculating the percentage of the houses which got a form, which were either vacant or second homes, seasonal homes and so forth.

I don't have the number yet, Steve. I can tell you that there's reason to believe that the 1990 number will be predictive, which was about 9.3 percent of the housing units turned out to be, one way or the other, vacant or not the kind of housing you can send in a response for, which means that the overall return rate -- we don't calculate that return rate until we've really scrubbed the data a great deal and really worked it over and gotten rid of all the inconsistencies and looked for all these vacancies, gone back out and double-checked them and so forth.

One of our big operations this summer was to send an enumerator out to all of the houses which we thought were vacant to make absolutely certain -- it was a very big, expensive operation. As many of you know, we revisited, even after we had gotten some kind of information from 120 million households, we revisited 12 million households as part of our quality-control work.

Long answer, but I would suspect, then, if the 1990 pattern holds, that what we're easily looking at is a return rate that will exceed 75 percent, and it could reach the high 70s, if not 80 percent. But we will not know that until next February.

MR. JOST: Returning to the telephones, Mark Skertic of the Chicago Sun-Times.

Q Thank you. Dr. Prewitt. I wonder if you could discuss the response rates, long form and short form. I'm wondering how the responses shook out for getting long-form data.

DR. PREWITT: Yes. The differential between the long form and the short form in the mail- response rate was about an 11 percent higher return rate on the short than the long form. However, we were able to make that up in the Non-Response Follow-Up; that is, it put a burden on us. We had to go to more homes that needed the long-form attention, and that takes longer to collect those data and so forth. But we were able to close that gap. So by the time we had finished the census with Non-Response Follow-Up, we are not talking about any differential between long and short form.

We are still analyzing the quality of the data on the long form. That will take some time. But right now we are convinced that the work that we did in the field period managed to overcome some of the concerns that we'd had about the long form last March during the public discussion of it.

MR. JOST: Herb Sample, Sacramento Bee.

Q Hi. What can one extrapolate from, say, the California number, 70 percent, in terms of apportionment of housing, if anything?

DR. PREWITT: Nothing yet. What this number means is it simply made it easier to get the census done in California. I mean, we simply started with less of a workload in a very big, complicated state, and that was enormously helpful for us and it's one of the reasons that we were able to move through Non-Response Follow-Up in California so expeditiously.

But the final count is the final count. All of this is preliminary, of course, to the numbers. I think the most important thing I can say about the apportionment count at this stage is that we're really on schedule to deliver it. Going into something as massive and complicated and unpredictable as a census, you always have your fingers crossed. Things can go wrong, seriously go wrong. And they could go wrong in such a way that by the time we got them fixed, we were running out of time.

I can now say with some confidence that the probability of not producing the apportionment numbers well on schedule is minuscule. But actually what it will be is just -- we won't know till we've crunched the numbers and processed the data.

MR. JOST: Back to the telephones. We have Julia Harding with WRO-TV.

Q Hi, Dr. Prewitt. I'm not sure if you can answer this as specifically as I'd like, but I know the census has shown there's been a population decline in Augusta, Georgia, which is Richmond County. Can you tell me what this decline is, what the response rate was, and if Augusta is still Georgia's second-largest city?

DR. PREWITT: No, I'm sorry. We have some of those numbers. We've obviously issued here a press kit, and it's up on the Web. You can look at the response rate specifically for Augusta, but I cannot really tell you yet what it means about the population size of Augusta.

I do have to keep emphasizing that we are talking at this stage still about the operations of the census, how well they went, that they were on schedule, that they were at or under budget, that the American people cooperated and so forth. But we're still talking about the operations of the census and not yet the results of the census. And the first major result will be the apportionment number, and it will be available to the country in December. That apportionment count will only be a state-by-state count, so you will not begin to get city-level counts, place-level counts, for another two months.

Q Okay, thank you.

MR. JOST: We have time for a couple more, but I just want to -- that's a good question to make a footnote. Immediately following this presentation, we'll have more detailed regional information for members of the Regional Press Association; the associate director, Marvin Raines, who oversaw our field operations; Brenda August, the head of our partnership programs; and more detail on some of the local angles to this national story.

And we'll go back to the room, and I think I saw Dee Cohn with her hand up.

Q Dee Cohn, Washington Post. I'm wondering if you could do a little more analysis of any patterns that you discovered in looking at what the response rates were. Specifically, did places that did a lot of outreach all improve their responses? And can you talk a little about the places that did not do so well? Baltimore, for example, had a much lower response. And lastly, did places where there was doubt raised by some political leaders on whether people should respond, did that seem to affect your mail-back?

DR. PREWITT: An important, though complicated, set of questions. And as a social scientist, I'm reluctant to speculate too much about and to explain patterns where we simply haven't had time to collect all the data that we'll collect.

The Census Bureau conducts several hundred evaluations of the census operation. We are collecting systematic information about the partnership program, about the advertising program, all of our outreach effort and so forth. Where we are also trying to get information, to speak more specifically to these questions, is the effort that was put in by states, communities, townships and so forth, independently of the effort. We don't have that information collected yet.

All I can say is, anecdotally, it seems fairly clear to me, in terms of the places that I visited, that places that actually made a local commitment to make certain the census was successful were rewarded for it. That doesn't mean it's a one-to-one correlation, but certainly the cities that I visited where there was very, very strong leadership taken by the mayor's office, the local leaders, especially where money was put in -- Houston is a city where the Houston City Council put some serious money into the census, and they created their own major promotional effort. Detroit was another one; enormous effort by the mayor in Detroit. And that also paid off. So was New York; the mayor's office in New York really put a lot of effort, a lot of organizational groups. New York is a tough, tough place to count. Sometimes that effort did not have the boost in the response rate that we had hoped for, but it did make it much easier once we were out in the field.

One of the things that the local effort really helped on was recruitment. Let's not forget that when we went into this census, everyone was concerned. Could we recruit what turned out to be 920,000 people moving through our payroll during the census period. And so some of the payoff of the local effort shows up that way rather than in the response rate.

I can't be more definitive than that. My sense is, as a social scientist, and just what I saw in the country, is that once we have all the data, there will be a fairly clear pattern that emerges between the quality and the effort that was put into it.

Now, you should always be looking at the magnitude of the improvement, which is to say you can have a state which has a very high response rate where not that much was done, but it didn't move much between 1990 and 2000. What's really hard -- to go back to the first point -- what's really hard is to change trend lines. The best predictor of the response rate in 1990 was the response rate in 1980 by region, by state, by locality and so forth. And we expected that to happen again between 1990 and 2000. That is not what's going to show up. We're going to find some very interesting bumps in the response rate that are -- and as I say, cities that improve by 14 percent, 11 percent, 10 percent. So obviously that's going to be attributable to the local effort.

MR. JOST: Back to the telephones. Gordon Trowbridge of the Detroit News.

Q Yes, Dr. Prewitt, a question that you sort of touched on earlier as it relates to Detroit and other cities that did make that substantial effort, local effort, to boost participation. The argument that Detroit and other cities were making is that "There's a lot at stake for our city in terms of making sure that everyone is counted and the federal dollars that are based on population counts," and so forth. Given that you don't know a whole lot, at least right now, about how participation will affect issues of undercount, what, if anything, can we say about whether or not that argument, you know, holds some water? Will the effort that these cities, Detroit and other cities, put into boosting participation have any eventual impact on things like federal funding that's dependent on census figures?

DR. PREWITT: Well, as we all know, federal funding is on a share basis, and so the effort that was made by many of our city leaders, especially some of our larger cities with large minority populations, had they not made it, there's every reason to believe that their undercount would have been higher. That is, we would like to believe -- we won't know this until we get the data -- that they have cut into the undercount.

The real solution to the undercount problem, however, as we said right from the beginning, is not simply trying to count everyone. The census cannot count everyone. And when the situation is -- when we finally get all of the data in, we will find out, as we've been saying now for over a year, that it's very, very unlikely that 100 percent of the American -- I know 100 percent doesn't. I have too many letters from people saying they refuse to cooperate with it. Many of them sent me $100 saying, "I'd rather pay the fine than fill out this thing."

There's a lot of anger out there. There's a lot of mistrust. However, it's really small in number. It made more noise than its numbers, of course. But there are people who are not going to fill this out, who are going to be evasive, who are fearful of the census, all of the reasons that we talked about before. We need the Accuracy and Coverage Evaluation. We need the quality survey in order to be able to measure the magnitude of the undercount and then correct for it. That's what a Census Bureau does.

Just to be repetitious of things I said before, the census is an estimate of the total population. Everything that we can do to get that estimate closer to the truth improves the census. Long answer to your question, but I just have to say that at this stage, I do think that had the effort not been made, it would have -- if we had kept going in the same direction we were going between 1980 and 1990, when the undercount did increase and turnout or participation levels did decrease, that if we kept going in that direction, we are fearful we would have seen a larger undercount in 2000.

So I would like to believe, but I won't know until we have the data, I would like to believe that not only did we arrest the decline in the participation, but that we also arrested the increase in the undercount. But we won't know that till we've finished our Accuracy and Coverage Evaluation work.

MR. JOST: Time for two more. We'll take the front row here.

Q Hi. Mark Heller with the Watertown Times in New York. I wonder if you can tell us, generally speaking, what kind of a response you had from Indian tribes. And I wonder, kind of on that note, whether a great number of tribes actually are not reflected in these numbers because their numbers are actually taken by census-takers, not by mail.

DR. PREWITT: Our partnership program with the Indian tribal leaders we were very, very pleased with it. If you want to afterwards, you can ask the experts in the room. So we actually got some high mail-back response rates in Indian land, as well as across other parts of the community. As you recognized, though, in your question, most of our work in the Indian areas was done by direct enumeration; that is, direct visiting of the households by enumerators. So the partnership there was much more important for creating an atmosphere of cooperation and responsiveness to the enumerators.

We think we got a much, much better count in the Indian areas than we got in 1990, and it's partly because the partnership opened up the areas for the enumerator staff, and it's largely because we hired from the local population groups, from the local tribes, where you obviously then create a higher level of trust and confidence. But you can't learn a whole lot by mail-back response rates about the success of the census in the Indian areas.

MR. JOST: Last question; on the phone, Mae Cheng of Newsday.

Q Hey, Dr. Prewitt. I'm wondering if you could go back and address New York City once again.

DR. PREWITT: I'm sorry, you're going to have to say that again. We had interference here.

Q Okay. I'm wondering if you could go back and address New York City once again. The elected officials in the community have made efforts to get the word out. And you were talking about how, in minority communities, the response rate seemed to improve. But in New York City, barely half the population responded. And in our most diverse county, Queens County, the response rate was even lower than 1990.

DR. PREWITT: No, exactly. I mean, I will go back to that. As we all know, New York has increased its population -- New York City has increased its population of foreign-born in the last decade. I visited New York City -- Queens, Brooklyn, Bronx -- a number of times. I visited schools there. I think I went to one school in Queens where 166 languages were represented among the student group.

So New York is always difficult. And it was certainly made more difficult by its rapidly changing demography over the last 10 years. We actually ended up feeling very good about the count in New York. We maintained our recruitment pool quite successfully and we maintained schedule. New York is always the city that's the hardest to count in the final weeks of the census. And we were able to maintain schedule.

So, as I say, you have to look at the partnership and the promotional effort against a number of tasks. And one of these was to, as I say, recruit the enumerators. Another was to get the cooperation, not just of the mail-back response rate, but to get the cooperation when we knocked on the doors. And so, in the final analysis, we feel very good about New York, even though they started out with -- parts of them started out with a quite low response rate.

New York County, was, however, one of the highest counties in terms of improvement. I think that improvement was 8 percent from 1990 to 2000. So it was mixed. But, look, I don't mean to say -- I obviously want to tell a positive story. And to go back to the earlier question about the impact of the partnership program on the benefits that will come to the communities -- that will be seen as we get out the raw numbers. However, as of today I can tell you, and I happen to think it's important, that the partnership/promotional effort improves civic engagement, improves civic responsibility. That is a dependent phenomenon of this effort. It is also something worth noting and celebrating. So it is not just simply the distribution of benefits; it's whether the country is willing to step forward and perform its civic obligations.

Indeed, the census suggests to me that "bowling alone" may not be the metaphor that we want to use to describe at least the potential of civic responsibility in the country. And I say that based upon hundreds of visits to local communities where I simply saw civic spirit and community commitment -- people turning out early in the morning for coffees, people working late at night. You saw it, those of you who were out in the field. So I do think that the important phenomenon today is not just the final count, but whether the response rate itself, as we would like to suggest, does indicate a willingness to engage this duty.

MR. JOST: We can squeeze one quick one in from Chuck Holmes, who has been propping his arm up in the back, and I've overlooked him several times.

Q Dr. Prewitt, you mentioned that 61 percent was used as a planning benchmark. Does the 6-point differential mean that money was saved? Is there a calculation of these 1.2 million households that you didn't have to visit -- does it translate into dollars? Conversely, does the 11 percent differential between short form and long form cost the Census Bureau that much more money on the other end?

DR. PREWITT: No. And indeed the way that Chuck framed that question was exactly the right way to frame it. When you are doing a complicated field operation like this there are tradeoffs. Some things go down and some things go up. Indeed, having to go out and collect more long-form data was a more expensive operation. But the fact that we had to knock on fewer doors meant a less expensive operation.

What I can say today is basically what I have said; Census 2000 is in the black. The magnitude of that is very, very hard to assess until we have finished paying all of our bills, which we are doing through this month as we really close down all of our field operations. Whatever the magnitude will turn out to be of savings, it will certainly be attributable to this higher response rate. It has simply changed the dynamic of the census enormously. To not have a 61 percent response rate, to have a 67 percent response rate -- changed it dramatically.

Now, as you know from following this story well, the 67 is today's number. Earlier in April it was the 65. And that's when we had to actually cut for Non-Response Follow-Up. So eventually we did have to knock on a lot of doors where people had already sent their form in. Well, that doesn't save us money. In fact, if anything, that created some frustrations in the public. We had to constantly explain why we were coming back anyway -- because a lot of these forms came in after that cutoff date, and that just was a fact of the census process this year.

But the census will be in the black, and there is no doubt, when we do the analysis of that, that one of the largest explaining factors will be the higher response rate.

MR. JOST: Okay, thank you very much. We'll take a couple of minutes for those of you who need to leave, and we'll have Marvin Raines and Brenda August join us up front for members of the regional press who want to stay here for more details about local stories.

[END OF EVENT.]


Source: U.S. Census Bureau
[an error occurred while processing this directive]

Last Revised: June 14, 2010 at 01:38:51 PM