(June 2001) Kenneth Prewitt, former director of the U.S. Census Bureau, recently spoke with PRB about the 2000 Census and its implications for social policy in the United States. Prewitt oversaw preparations for and the execution of Census 2000 as the bureau's director from October 1998 to January 2001. Prior to taking the helm at the Census Bureau, he served as president of the Social Science Research Council, director of the National Opinion Research Center, senior vice president of The Rockefeller Foundation, and as professor of political science at the University of Chicago. He is the author of Recruitment of Political Leaders: A Study of Citizen-Politicians, and co-author of Social Structure and Political Participation: Development Relationships and Labyrinths of Democracy: Adaptations, Linkages, Representation, and Policies in Urban Politics. Prewitt is currently dean of graduate faculty of political and social science at the New School for Social Research in New York.

PRB: In 1999, you noted that the Census Bureau was working hard to make sure that there was no drop-off in the quality of the census in 2000, even though enumerating the population looked to be even more of a challenge in 2000 than in 1990. As it turned out, the 2000 Census may have been even better in some respects. What improvements impressed you the most, and what changes do you think played the biggest role in making that happen?

Prewitt: Obviously, the most important improvement is coverage. The thing that matters most in the census is including everyone in the country in the count, and not including anyone twice. It now appears that the 2000 Census was an improvement on accuracy in the sense that the total number of errors was less than we've had in the past…

Accounting for that success, there are two separate things that I would draw attention to. First, this census was well-funded. It does make a big difference when you're in the middle of an operation and you know you have the resources to solve problems as they happen, rather than having to go back and tease the money out of Congressional appropriations committees. It's hugely important to fund a census at a level that will allow it to meet the expectations that a country has for it.

The other thing is the community involvement in the census, especially in the harder-to-count areas and among harder-to-count population groups. I run into people all the time who hardly remember that there was a census — they filled out the form and forgot about it the next day. That's just because of the social circles that they inhabit. But when you go off into black churches or migrant worker communities, or the southwest, or some boroughs of New York City, you run into people who remember the census experience very fully because they heard about it in school, or church, or in the media. So the effort both by the Census Bureau and by local communities to target hard-to-reach population groups must have improved overall coverage.

One of the indicators [of community involvement] to me is that everyone had feared that we would have labor problems. After all, how are you going to find 920,000 people at a time of very high employment? I'm convinced that putting the census on the public radar screen even as we were building toward it with this community promotional effort and the advertising effort is part of what helped us be successful with recruitment. I think an awful lot of people did not take this job only for the money. They took this job because they thought they were helping their community. It's a hidden indicator, but a very important one. We really did not run into major employment problems. Obviously we pay good wages, but I attribute it in part to the overall promotion campaign.

PRB: Does community involvement account for the turnaround in the mail-back rate? It had declined from 1980 to 1990, but bounced back in 2000.

Prewitt: Let's say you have a mail-back rate of 95 percent, but you never find that remaining 5 percent. That would be a less successful census. So overall coverage is what really matters.

Of course, one of the things that gets you to better overall coverage is to start from a strong platform. If we had started with a mail-back rate of 55 or 60 percent, we would have been straining for resources. That's a huge number of households we would have had to visit — every one of those percentage points is over a million households. So the fact that we started the follow-up period on the heels of a higher than anticipated mail-back response rate enormously helped the operation.

The mail-back response rate itself is also very important in civic terms, but it's the operational importance of it that is key to the overall success of the census. So certainly the promotional effort enormously helped with the response rate. I think other little things helped as well. We tried to make the census more user friendly, something that started under my predecessor, Martie Richie, of course. We also more prominently stated that it was mandatory, an issue that had dropped off the burner for a while. Little things like that do boost the response rate.

After the 1990 Census, they were expecting a mail-back response rate of 55 percent [for the 2000 Census]. They were expecting another 10 point decline, like we had between 1980 and 1990. A lot of work after that got [the projection up to] 61 percent — making the questionnaire shorter, making it more user friendly, making it clear that it was mandatory, the three mailings (the alert letter, the form, and then the follow-up letter). What really surprised everybody, of course, was coming in with an actual rate of 67 percent.

PRB: The privacy issue came to the fore in the run-up to this census. You've talked about the difference between confidentiality and privacy when it comes to census information. What do you mean by that?

Prewitt: …Census information is not about the individual, it's about the statistical aggregate. And it's easier to hold that kind of information confidential because the individual identifiers are simply not important to us. We're fairly confident we can keep the census data confidential.

But that does not take care of the issue of privacy, which is about who has a right to know about me as a private person. That is a big, big issue in the commercial sector — who has the right to market to me, who has the right to know my bank balance — that spills over into the government sector, even in the statistical agencies. Take the health record problem. That is an individual problem — the health system wants to know your individual medical history, your drug-taking history, and so on. They want to know about you. If you're simply doing a survey about medical care, and the survey doesn't really care about individuals, then it's not about confidentiality, but it could still be about privacy.

I think the distinction between confidentiality and privacy — confidentiality being information about me being protected and not shared, and privacy being "I don't care if it isn't shared, I still don't want anyone to know that" — unfortunately in the 2000 Census simply got blurred. And once it's blurred, it's very hard to unblur. So an agency that has a practically impeccable record on confidentiality can still be vulnerable to charges of "you're invading my privacy." There's no easy solution to that one.

PRB: What are the implications for social policy of the opportunity to register more than one race on the census? What kinds of policy changes might result from this? What about the implications for how we think about race in America?

Prewitt: I think that's the most important thing that happened in the 2000 Census. I think when the historians write about Census 2000 in 70 or 80 years, the sampling debate will be a footnote, the improved coverage will be noticed, but the books will be about the multiple race item.

From 1790 to 1990, the Census Bureau collected race data in terms of a small number of discrete categories. Throughout that entire 200-year history, a lot of social policy was based upon those discrete categories. From 1790 to roughly 1960, the social policies were essentially detrimental to racial minorities, that is they were first about slavery and sustaining the slave system, and then in the late 19th century they became about trying to establish the principles of racial superiority and inferiority. They got involved in the relocation of Native American Indian populations. They got involved in very hostile immigration controls designed to retain the purity of the American stock, whatever that is supposed to mean, and in the internment of Japanese Americans in World War II.

Then in 1960, the directionality of those policies shifted 180 degrees. The classification system was still essential — the Voting Rights Act and the Civil Rights Act depended upon that classification system to undo the preceding detrimental policies in a beneficial way. So you get equal opportunity programs, quota programs, affirmative action programs, the Voting Rights Act — a whole series of public policies that are trying to right some wrongs that have been in place for the entire nation's history. The classification system was still important because you did this with the basic principle of statistical proportionality. If there is 12.5 percent of the population that is African American, but they make up only 2 percent of the college population, then something is wrong. So you use the classification system as a marker of discrimination, and then you set out to devise social policies designed to correct for that.

That's not going to work anymore. First, once you go to 63 categories (or 126 depending on how you think about it), there is no upper limit. There is no natural ceiling now to how many racial groups there are in the country. Once you're at this level, then why aren't the Arab Americans their own race? And if they become a race, then you enormously expand the total number of categories. You cannot design public policy that depends upon a racial classification system when the number of categories is just exploding like that.

The other thing that is going to happen, is that the scientific community — which already knows it's hard to measure race, and that it doesn't mean anything biologically or anthropologically — is going to get more restless with even the attempt to measure race. And I think the public will get more restless.

I think the multiple race issue is a tremor in 2000 for what is going to become a political earthquake that is coming down the road. It will take a while, but if this works its way through the system, the number of categories will continue to expand. I think proliferation breeds proliferation. And there is no way that the Census Bureau or OMB can say "enough is enough." The pressure to expand will be intense, and the statistical agencies will accommodate that pressure.

PRB: What do you see as the upside and the downside to this development?

The downside is that there is still discrimination in this country. We have now weakened — and I think will eventually eliminate — one of the tools that our political process has to correct for discrimination.

The upside is that we might have to find a more reasonable way to find discrimination and correct for it. Statistical proportionality is actually a crude measure because it lumps everybody together. Oprah Winfrey is part of the same category as the poorest African American, for example. I'm now dean of a university program dealing with affirmative action issues all the time — is the son of a Brazilian diplomat going to count as an Hispanic hire? It doesn't make sense if what you're trying to do is correct for the injustices suffered by the poor Mexican worker.

Since we are now in the process of rendering the old system useless, the consequences can be that we will be forced to come up with more targeted and effective tools. I don't have the magic answer for that, though.

The upside is that it is very healthy for a country to be beyond placing people into discrete categories. The fact that we're going to have to dismantle and construct other ways to deal with discrimination and lack of opportunity for certain individuals is no easy task, but it was not easy to do the affirmative action and civil rights work in the first place. We're going to have to do it again, but without a certain kind of tool of policy.

In the long run, if we quit thinking in racial categories, that is nothing but good, nothing but a positive development for our country. We got multiple race responses from 2.5 percent of the population this time, which is a reasonable level at this point in time. But I think it will grow enormously. First, it's related to age-structure: as older people who never [identify themselves as multi-racial] die off and younger people who are more comfortable with it take their places, it will grow just through cohort replacement. It will also grow because it has now been legitimated. The first time out, people did not quite understand it, but after a decade of experience with it, and when it starts filtering through all kinds of surveys and so on, it becomes legitimate. I don't know where it will level out, but in 15-20 years I would expect to see about a quarter of the population identifying themselves as multi-racial.

That number will also grow as the number of categories grows. Right now, if an Arab American marries someone whose family is originally from Finland, then that's a white-white couple, and their children would be white. But if Arab Americans become their own category, then those children would be multi-racial.

One very important thing is that the question could have been asked in a different way. We could have simply had a question asking: Are you multi-racial? There would have been a new group in society of multi-racial people, we would have reified that population. We would just be reproducing the same 200-year history but now we add another category to it. The fact that we allowed people to check more than one box, instead of a single discrete category, is what makes it a tremor leading up to a political earthquake that is going to happen.

PRB: Because the implications are so big, do you think some people with a stake in the old system will push to go back to the old system or push to make multi-racial a discrete category?

Prewitt: Absolutely. I think there will be a push in that direction, and I think it will be unsuccessful. I just don't think you can put this genie back in the bottle. It will now be thought of as racist. Trying to put all multi-racial people into one category will cause people to say things like I'm Indian-Black and this person is native Hawaiian-Japanese and we belong in different groups. I cannot believe that [the creation of a discrete multi-racial category] will happen.

There are social and political interests who do want to get away from race-based social policy, and since this hastens that, there will be both liberal and conservative resistance to going back [to discrete categories].

PRB: Are there other respects in which this census may change the way we think of ourselves as a country?

Prewitt: There is another dimension to this. You don't need a census to tell you that we're 10 percent foreign-born — you can get that out of ACS data. You don't need a census to tell you that we're more diverse than ever before. What the census did was to get on the public radar screen. More people knew about the census and thought about the census than has been true in earlier censuses as a result of all this mobilization and advertising and so on. The press coverage on the census was extensive. Therefore, the public was more positioned to pay attention to census results. The very first census results were those concerning the diversity of the country — how many Latinos there were, how many Asians there were, what the growth rates were, the fact that three states are now majority-minority (one of them being the largest in the country), the fact that Texas is now only a few percentage points from being majority-minority. The American public probably registered that in a more profound way than they otherwise would have. Newspapers put lots of money into their graphics, and the census became the story. The story became not numbers, but diversity. So I think that the census came along at a very interesting time to remind the American public that this is a very different world and we have to deal with it. I think that's very healthy.

And we have not even gotten that much data out yet. When you start getting the long form data coming out, and you start doing cross tabs on income and education by all the different groups, and you get ancestry data as well as race data out, we'll start the story all over again.

PRB: What new uses do you see for census data, especially given the increasing ease of accessing it and using it?

Prewitt: The key issue, it seems to me, is to get groups who otherwise don't have access to census data today to start using it. That's classrooms, civic groups, advocacy groups, local planning groups. We need to do some real technical assistance for lots and lots of groups who otherwise would never think about census data. You take black churches who worked so hard on the census. It would be wonderful if they had someone, maybe five hours a week, who understood census data and was looking at their local community. Once the long form data are out, they will really have something to look at. Then you start to develop a sophistication about data that the country has never had. We talk endlessly about literacy, but we talk very little about numeracy. It would be very useful if Census 2000 could result in an increase in an understanding of data and information. I'm not predicting this will happen, but I think it's within reach. Money would have to be put into it, but not big money.

PRB: What do you think should be done differently next time? What are some possible innovations in how the census will be conducted in the future?

Prewitt: You have to start with the politics of it all. I would hope that the contending political interests about how to conduct the census and how it should be managed during its operational phase would reach agreements early and then live up to them. That is to say, whatever the basic methodology is going to be, settle it by 2005 and don't still be debating it a year before like we were this time.

Then once the agreement is struck, let the Census Bureau do it. The oversight apparatus' attempt at micromanagement is not an operationally simple way to do this. You would never run a military campaign that way. I testified [before Congress] 17 times in less than two years. Schwartzkopf, when he was running Desert Storm, wasn't back and forth to Capitol Hill to get permission for every step. It's just not an intelligent way to conduct a major effort like this. I understand the politics, and I understand why they are intense. But resolve it — if at all possible — by some deadline.

The other thing — and this is very hard for Congress to understand — is early planning money. Already there should be planning and evaluation money [for the 2010 Census]. Very small investments now that help us figure out what worked and what didn't work can really pay dividends. We hired 920,000 people to work on the 2000 Census — do we really know how that happened? No. But can we learn that? Sure, we have enormous amounts of data on that.

The big thing, of course, is whether there is going to be an American Community Survey [ACS] or not. If there is going to be an ACS, then you do Census 2010 very differently. It could practically be a "postcard census" because you have all of the basic data already there. That kind of census, along with much greater use of the Internet, could perhaps really improve coverage and eat away at the differential undercount problem. But if you don't have the ACS, and you're still carrying the whole burden of the long form in the census environment, it's a very different kind of operation. You have to train people, there's data capture, everything about the census is different because you have that one-sixth survey going out at the same time.