My Two Census

Formerly the non-partisan watchdog of the 2010 US Census, and currently an opinion blog that covers all things political, media, foreign policy, globalization, and culture…but sometimes returning to its census/demographics roots.

Transcript from Tuesday’s Press conference…

Note my questions near the bottom of the first page, and further note how they weren’t clearly answered…(My questions discussed allegations made by Census Bureau employees about QACs…)

STEPHEN BUCKNER: Good morning, everyone. I’m with the Public Information Office at the U.S. Census Bureau. I’d like to welcome everybody joining us on the telephone today, and also here at the National Press Club. Today, Dr. Groves, the Director of the Census Bureau, is going to be talking about mailing back your form. We have five days left to get your form back in the mail. He’s also going to be providing a brief overview of operations since our last operational press briefing on the 22nd of March where we launched the Take Ten Program, challenging local areas to mail back your form and beat your participation rate from the 2000 census. We’re off to a good start, and Dr. Groves will touch base on that.

In your press kit, you’ll see a variety of materials on the topics of today’s operational press briefing. Online, they’re also available under the news conference page for those listening and online. We will have a Q&A session following the Director’s comments, and we’ll try to alternate between the telephone and those here in the room. With that, I will give you Dr. Robert Groves. Thank you.

DR. ROBERT GROVES: Great, thank you, Stephen. Thanks a lot for coming today. This is an operational update, but really the headline of today’s briefing is that we have five days left for the over 120 million households around the country to mail back their form. And by that we mean if you get your form in the mail by April 16th, this Friday, there’s a much, much, much lower probability that anyone will come to your door to do follow-up work in the later phases of the census.

Today, this week, we begin a week of transitions. The transition is moving from this massive phase that we call the mail out/mail back phase where people fill out their census by mail, to one where we begin to hire a large number of people who will go out throughout the country, knocking on doors of houses and taking the questionnaire information in a personal interview.

I want to end with remarks on that, but I want to begin, really, by doing a quick update of recent operations. And in a word, things are going quite well. I guess that’s two words, quite well. Let me run through the things we’ve done. We have finished three operations that are notable that are complicated and are now in the hopper. The update leave operation, by that we mean in areas where postal delivery is not reliable, where many people have postal boxes, and also in those areas where in the, say for example the gulf coast where the housing unit stock is actually changing quite rapidly, we drop off questionnaires. We completed that on April 2nd, the dropped off questionnaires are being mailed back now at great rates. This operation is complete, it was on schedule and under budget.

We also finished, today we will finish, a count of people living in transitory locations. By that we mean RV parks, campgrounds, hotels, motels, marinas, circuses, carnivals throughout the country. We’ve completed that work on schedule, we’re happy to say. And then in a complicated operation called service based enumeration. We reached out and counted people who are affected by various types of homelessness. We counted people in shelters, soup kitchens, at regular stops of mobile food vans, outdoor locations and a variety of other places, about 65,000 locations throughout the country. This, as you might imagine, is a complicated one and we’re happy that we did that on time and with safety, relative safety, of all our enumerators and the people counted.

We had a problem in the New York/Boston area. You might remember there were torrential rains around the time we were doing this. We had to postpone one day to finish that work there.

Then I want to note two operations that we’re right in the middle of because they’re relevant to some folks who haven’t received forms because we do the enumeration in different ways. We don’t mail out forms to different areas. Ongoing right now is an operation called Update Enumerate. By that, we mean we go out with census takers and house by house do interviews with people in those areas. These are areas like the remote parts of Maine, certainly parts of Alaska, American Indian reservations, disproportionately, a lot of areas with seasonal housing. And the settlements called colonias on the border of Texas and Mexico are handled this way. It’s about 1.4 million households that we’ll visit in this way. We’re about 38 percent through with that operation. We’re way ahead of schedule on that, so that’s going well.

And then finally group quarters, another category of folks who don’t receive forms in the mail. These are areas that are like nursing homes, assisted living facilities, prisons, dormitories, barracks, and so on. We’re in the middle of that operation. We’re about 22 percent complete, that’s on schedule. That’s going to go through May 21st. That’s ongoing now. So for people in those situations, they should not have received the mail questionnaire, and indeed we’re reaching out to them and doing the measurement in different ways.

As of Friday, I’m happy to note, if you’ve been following this on the website, 65 percent of American households we’d estimate have returned the forms. This is over 77 million households in the country who have completed the short form and mailed it back. We’re going to post an update today at 4:00 on our website. If you’ve been following that, that will make that number be higher, no doubt. It is notable, if you look at those data, that there are states that are above 70 percent at this point already, 10 states have that status. It includes large states like Michigan, Ohio, Pennsylvania. When we started this operation, the states in the upper middle west dominated the early returns. That’s mainly because they received those questionnaires by hand using our census takers to drop off questionnaires at their houses. Now you see states, other states, catching up that received them by mail.

It is interesting to note that on April 8th, South Carolina as a state, surpassed its entire 2000 census performance. And today, both North Carolina and South Carolina have beaten their own 2000 participation rate. So that’s a notable and noteworthy event, I think. Kentucky is really close, this could happen today for Kentucky. We have hundreds of jurisdictions around the country who have beaten their 2000 participation rate already, and to all of them we salute you for your civic participation and we hope you’ll be joined by hundreds of others in a matter of days.

We remain focused on return rates, participation rates in large cities, in some rural areas. The large cities pose interesting problems to sample surveys and censuses. They are low this decade, as they were last decade. Notable is New York, Chicago, Detroit, Philadelphia and a set of others. You can go to our website and see this. In fact, if you look at this map here, the colors we love to see on this map over here are those that are yellow, orange and red. And the colors we don’t particularly like to see are those blues. You can see the Texas/Mexico border as an area that is challenging for us. You can see the middle of the state, or the middle of the country with those higher than average performances.

We are focused on all of these areas. We are looking at this daily through a variety of statistical models and intense scrutiny of partnership and other activities. We’ve made changes in our advertising to have much more targeted advertising into the areas and into the subgroups that look like they’re responding at a lower rate.

I want to turn to an issue that you in the media could help us get the word out on, and that is those people who didn’t receive a form, what should they do at this point? And we have a variety of reasons that that might have happened. You might be living in one of those areas where census takers are coming to your door right now over the next few days and weeks. You might have a post office box where you receive your mail, rather than having your mail delivered to your home. You might be in a house newly built that was added just recently and we tried to get about two million of those kind of cases into the mail stream. You may have just received a form, but you may not have gotten it yet.

For all of those people who have not received a form, we have a very simple thing you can do, two alternatives. First, you can call our telephone assistance center, and these numbers are posted right here. The English number is 1-866-872-6868. And starting today, you can call between eight a.m. and nine p.m. local time every day, every day of the week. And there are numbers for different languages and for the hearing impaired.

And starting today, you can take the interview right on the phone. You can answer your census questions right on the phone. It’s the simplest thing you can do. You supply the address where you’re living, and you answer the questionnaire over the phone. Alternatively if you wish, you can go to over 40,000 sites around the country that are questionnaire assistance centers or labeled Be Counted sites. How do you know where they are? You can call these same numbers to find out where they are. You can go to our website,, and find where the questionnaire assistance centers are. And there, you can pick up what we call a Be Counted form. You supply your address information and fill out the form that looks very similar to the mail out form.

We want to make it as easy as possible for those who haven’t received the form to get it. This is a massive operation, going to over 134 million households. Missing a few households is something that happens every census, and we want to make sure that you have an easy way to get a hold of the form.

But in closing, I want to turn to the most important message. We are counting down the days, the clock is ticking, and we are asking you, if you have a form sitting in your home that you haven’t filled out yet, to take a few minutes to fill it out and mail it back. If you can mail this form back by Friday, April 16th, the odds that someone will come to your house to follow up and ask census questions are much lower. Each passing day makes that likelihood higher. We would love to avoid sending census takers to large numbers of households around the country. And the easiest way for those of you who have a form sitting in your house to avoid that is to fill out the form and mail it back.

I remind us that for every one percent of the households that do that, we save us taxpayers $85 million nationally, a very large sum of money, for a very small act on your part. Taking it down to the personal level, if you fill this form out and mail it back, it costs us taxpayers 42 cents. If you don’t, it costs us taxpayers about $60 to send someone out and take the information in person.

This is a moment, these last few days, where our over 225,000 partners throughout the country and social and political leaders need to come together and get the word out that we are ticking away the clock to return these forms. This is the moment where we can all come together. Ask your neighbors if you filled out your form whether they filled out their form. Pass the word that we have just a few more days left, and this will be a much cheaper census if we can do that.

Starting at about the third week of April, we will begin to build the large files that will identify the addresses that our interviewers will have to go out and call on case by case. We will begin that work on May 1. It will continue through July 10th, and we will have other briefings on that large operation that we call non response follow-up. But this moment, at this time, the message is very clear. If you have a form sitting in your home and you haven’t turned it back, you haven’t returned a form at all, now is the time to fill it out and mail it back. Thank you very much, I’m happy to take questions. Carol?

CAROL MORELLO: Can you talk a little bit about what’s going on?

MR. BUCKNER: Just one second. So as we start our Q&A process, we’ll start here in the room. Please state your name, organize, and your question. Wait for the mic, and then we’ll jump to the telephone as well. First up, Carol Morello, question posed?

CAROL MORELLO: Hi, could you talk a little bit about what you think is going on in the blue states primarily, or blue sections, primarily in the south and the west? How does it compare to 2000 at this point, and how much do you think maybe politically motivated by people who resent the questions being asked?

DR. GROVES: Some of those rural areas that are in blue are traditionally what we call hard to enumerate areas. So you might first ask, so why is that? The challenge of rural areas is both that the kind of– first of all, these are based on either what we call update leave– these are the houses– the blue areas you’re looking at are houses where they have a questionnaire, either by mail delivery or dropped off. In some of those areas, we know we’re actually going to do enumeration, direct enumeration. They don’t even have a mail questionnaire yet. But rural areas pose difficulties, both in making sure we get all the areas. And then the other challenge, I think, in rural areas is that the impact of media that we use is more dispersed. If you think about it in an urban area, there are local media that we can use and local partnership activities that we can stimulate to get communities involved in the census. That kind of organization’s a little tougher in rural areas.

At the same time, in some of those areas, there are language impediments. The Texas/Mexico border, a whole lot of those areas got a bilingual questionnaire, but not all of the areas, and that’s an impediment. So, these are not surprises. If you looked at the 2000 census, or the 1990 census, this is a fact that is replicated over censuses.

CAROL MORELLO: So what impact do you think that politically the opposition to the census has had?

DR. GROVES: Actually, this is a topic of press commentary right now. And the City University of New York has just done an analysis that’s kind of interesting. We’ve done our own analyses. We can’t find empirical support for that, but I can tell you no one has the right data. We can look at the characteristics of counties that are returning the questionnaire at different rates. We see no evidence that some counties that may have voted in one way are lower than other counties. But we actually don’t know individual level attributes of people returning the form or not returning the form.

I can say one other thing, and that is the rate of forms that are being delivered with partial completes that force us to actually go back, they’re so incomplete that we have to go back and call on people’s houses, those are within or below the kind of tolerances we expected from our prior studies. So, we don’t see the evidence there for the impact of that on behaviors.

MR. BUCKNER: Carol, I’m going to have to go to the telephone here and then I’ll come back. All right, on the telephone, I believe we have a question? Please state your name and organization?

STEPHEN ROBERT MORSE: This is Stephen Robert Morse at I have a couple of questions about the questionnaire assistance centers operation. Today, I posted an article on my website that was written by an assistance center employee that suggests that not all of the QACs are listed on the census bureau’s website. And it also suggests that all the organizations and corporations that are listed partners are not necessarily cooperating and working with QACs as they– and are not necessarily partners anymore. Please talk about this?

DR. GROVES: I don’t know anything about the second question. The first question, we’re doing the best we can at getting the questionnaire assistance centers posted on the web and updated. This is a big challenge for us because they’re, as you know, these 40,000 units are moving around cities for different days and different hours. But we’re doing our best on that.

MR. BUCKNER: Thank you. We’ll go to the room here.

IRENE MOORE: Irene Moore with the Just have one question. How do you account, or what’s your estimation, for the people that are unaccounted for? How do you count them, illegal immigrants, those that you couldn’t–?

DR. GROVES: There are actually two ways to evaluate how good a census is. One is called demographic analysis. Those estimates will be available in late December of this year. They’re based on birth and death registration, and immigration and emigration. And then the second way we do it is do a very large sample survey that will be done, actually, in August and later. The results of that, unfortunately, won’t be ready until 2012. Those will be the data that will have the best kind of drill down capability of any.

MR. BUCKNER: Okay, we’ll go back to the telephone if there’s any questions?

BILL GLAUBER: Dr. Groves, I have the shameless local question. As you know, Wisconsin has a return rate of about 75 percent, I believe most of those were mailed out, not handed out. Do you have any reason why the return rate is high in Wisconsin this year, and historically high?

DR. GROVES: Well, you know, before I had this job, that was my life, to try to answer questions like this. I spent 20 years trying to figure out why some people participate in sample surveys and censuses and others don’t. I can tell you, and it’s an important thing to note, that since 2000 participation in surveys is going down every year. For Census Bureau large sample surveys, that decline has been about 5 percentage points over the ten year period. Yet when, as you’ve just said, when you look at states, it appears that there are– when you look at a map, you’re forced to think of geography as the reason for these differences.

I don’t think it lies at the geographical level, there’s nothing about the water or the trees in Wisconsin that are anything different than other places. It does appear to have to do with the social organization of areas, so it is true worldwide that rural communities respond at higher rates than densely populated urban communities. I’m talking about small towns. It is true that income and education affects likelihood of participation; higher income people, higher education people tend to respond to surveys and censuses much more quickly. Wisconsin and Minnesota, as you know, if you look at those attributes, stand out among the country on that. But there are many, many other things that are influences on our decision to participate.

There’s great variability across types of censuses and surveys and how people respond. It is clear that when the individual people receiving the survey request understand that it’s a benefit to them to respond, they tend to respond. And that’s been the message of our advertising and partnership activity throughout the census.

MR. BUCKNER: Okay, we’re going to go back here in the room. Carol?

CAROL MORELLO: When you talk about the rates of partial completes being within or below the tolerance that you expected, can you give us a percentage? Tell us what it is or what it’s been so far, what you were expecting and what it was in 2000?

DR. GROVES: I don’t have those numbers here, you’re asking for too many. I mean, all those numbers are knowable, we could get you those numbers. These are teeny percents. This is single digit numbers, close to one or two, that we’re talking about. And our percentage centers, we have three processing centers, actually read the forms as they come in immediately and sort of put aside those cases where someone just answered the number of people in the household. Or, if they just answered a few number of personal attributes. And those we have to follow up on, since by law we have to get answers to all the questions. But we’ll get you those numbers.

MR. BUCKNER: Okay, we’ll go back to the telephone?

ANDREA SHALAL-ESA: Hi, Dr. Groves, Andrea Shala-Esa with Reuters. I have two questions for you. One is that the head of the Lockheed Martin project, Jolie Dunlap, told me that Lockheed Martin actually come in under budget on its work. And I was wondering if you had any sense of how much you could save on that contract, potentially? And then I wanted to ask you, too, about what you’re seeing about the effects of the housing crisis and all of the foreclosed homes that are out there around the country, whether that’s having a significant impact on your ability to get these forms back?

DR. GROVES: Okay, two great questions. Let me do a global comment on the budget. You know, ever since summer 2009 where we had an overrun on a big operation, we’ve completed, I think now, four operations on time, under budget. One thing that most people don’t know is the entire cycle budget for the entire decennial budget, we are under budget for the entire cycle. We did have that overrun, and I think we’ve diagnosed what caused that overrun and the operations that are going on now are on time and largely under budget.

Now, why is that happening, you might ask? I don’t have an answer for the Lockheed Martin case, but I do have an answer for those activities that involve human beings out on the field, large field staff. And I’m convinced that the quality of the staff we’ve been able to recruit is so far superior to those that were recruited and worked on the 2000 census that we’re just more efficient. We have a staff that’s doing better, and we’re proud of that, and it will not surprise me that these operations going forward are on time and under budget. That’s what we’re hoping for, that’s what we’re working for.

Now, the second one has to do with– your second question has to do with– foreclosures. And maybe you could repeat it, because I want to make sure I answer your concerns?

ANDREA SHALAL-ESA: Well, I understand that there are quite a lot of foreclosed homes out there where a form might have arrived and not come back. You know, is that going to have a significant impact on your need to hire census workers to go out and check those presumably empty homes?

DR. GROVES: Absolutely. So, first you have to understand what we have to do by law. We have to, by law, attempt to count everyone. We don’t know where people live, so we send– we don’t know that every house is occupied or unoccupied. We’ve sent questionnaires, literally, to millions of houses that may be unoccupied now. Some of those came back to us as undeliverables. Others are sitting in the mail boxes of those homes right now. Because by law we’re required to get a determination of the status of every one of these households, our census takers will go out to those households. This is in contrast to your prior question, one source of higher cost of the 2010 census versus the 2000 census. We had fewer vacant homes in 2000 than we do now. We’re going to spend money on those homes, making sure they’re vacant because if there are people living in there, we have to make sure they’re counted.

That operation will take place both through the non response follow-up operation, and then in later operation. We’ve already estimated the increased costs of that operation. We will incur larger costs, but it’s fully within our budget targets given our estimates right now.

We’ve also tried– we’re worried about the people who used to live in those houses, too, and we’ve tried to target advertising in areas with high foreclosure rates to people who may now be in doubled-up houses, to make sure they’re counted where they’re now doubled up. And both of those things are things that we’re going to see how well we’ve done as time goes on.

MR. BUCKNER: Okay, we’re going to get another question on the phone. I believe Reuters has a question?

ANDREA SHALAL-ESA: Yeah, I just wanted to follow up, actually, on both points. One is you said you are under budget on many of the operations that you’ve had. I wonder if you could just put some flesh on those figures. So the entire budget, I think, is $14.7 billion? Do you have a sense of where you might come out under, or how much? And then also if you could sort of go back. You say that you’ve estimated the increased costs linked to foreclosures. Can you share that with us, or at least define it somewhat?

DR. GROVES: I can give you a sense of, I think, the foreclosure side. On the entire budget, let me tell you how we’re thinking about this. We still have a variety of uncertainties in the final cost of the 2010 census. One is the focus of this news conference; how many people are going to return this form? For every one percent, we save $85 million. So, the uncertainties are getting smaller as the operations are completed, and it’s a great feeling for all of us when we see these smaller operations come in on time and under budget. But I don’t have a current estimate for the overall cycle, how much we’re going to come in under. And then I’ve blanked on your follow-up point.

ANDREA SHALAL-ESA: The foreclosure estimate.

DR. GROVES: On the foreclosure, we don’t have final estimates of this, but this is probably a major jump up, maybe a 50 to 75 percent increase in the number of houses that we’re going to have to go out and make sure they’re vacant. So that’s an operation that goes on at the tail end. Those are our current estimates. The non-response follow-up stage will harden those estimates quite well. But that’s what we’re looking at now and we’ve estimated that that’ll cost us more money at that stage, but we’ve got enough money to pay for that.

ANDREA SHALAL-ESA: Is the cost millions, or–?

DR. GROVES: In almost every operation involving the census, the first number is at least an eight digit number. You’re always talking about tens of millions of dollars when you send people out to visit households. And remember, from now on, starting May 1, after this April 16th big targeting date, we’re going to have people on the streets. And that’s where the money is spent at very high rates. But as time goes on, we’ll be able to give you much better numbers on these things. There’s too much uncertainty to give you a number right now.

MR. BUCKNER: I’m sorry, I’m going to have to cut you off right there. Hang on a second, I think we have some other questions on the telephone and I want to get to a couple of more before we have to call the end of the Q&A period. I believe Terri Ann Lowenthal with the Census Project was next?

TERRI ANN LOWENTHAL: How are you? Hello?

MR. BUCKNER: You’re live.

TERRI ANN LOWENTHAL: Okay, I have a very quick question just to clarify. If people telephone in their responses or send in a Be Counted form by this weekend, I know you have to verify the existence of that address, either against the master address file or during your later vacant delete check. But will those people probably avoid a visit by a census taker as well?

DR. GROVES: On the Be Counted forms, I’m pretty sure that the processing of those are such that the likelihood of actually getting a visit is pretty high, to verify that we got the address right and the information right. On the TQA, we still have to verify the addresses, and so that’ll be a process that will go matching the master address file contents. And then the decision will be made.

MR. BUCKNER: Terri, you’ve got one quick follow-up?

TERRI ANN LOWENTHAL: Well, it sounded a little uncertain, but I guess that people– it sounds like people might still have to expect that someone will come to their house, even if they answer by phone?

DR. GROVES: Here’s one thing I can say with great certainty. With each passing day, if you wait to return this form or wait to pick up a Be Counted form or do an interview on the telephone, our ability to do these kind of matches with the address file and avoid sending someone out goes down. That is absolutely crystal clear. We will do the best we can, and certainly this week is the best week to do all of these things rather than waiting until later weeks. That is absolutely clear.

MR. BUCKNER: Okay, we have time for one more question. I believe Hope Yen at the AP on the telephone?

HOPE YEN: Two quick questions. One, I just wanted to clarify that in terms of the areas and the reasons for lagging mail response for those areas, I wanted to clarify if there are any real differences from 2000, or is it pretty much a similar pattern? And then the other question was just wondering if you could offer a broader assessment looking ahead for the challenges and preparations for the non response follow-up?

DR. GROVES: We’re looking at these patterns every day. In fact, this is usually my nighttime behavior. I look at all these patterns in the evening very carefully. I don’t see any surprises. We actually have better data in 2010 about the characteristics of areas that are high responders and low responders. So sometimes, I don’t have a comparison back to 2000. But knowing what I do about the American public’s reaction to surveys, there’s no surprise here at all. So that’s number one.

I think there’s variability across the country in areas that have organized their communities much more effectively than others that we see, and that’s a wonderful, wonderful thing. But no surprises. And then I’ve forgotten–

MR. BUCKNER: The challenges of NRFU?

DR. GROVES: The challenges of NRFU, well we actually need a whole review of what happens in non response follow-up. We want to make sure that’s our next press conference. The typical challenges of non response follow-up, if you go back over the decades, is can you recruit census takers with the skills and the devotion to the job that you need? So far, so good. I can tell you, all of the people involved in 2000 recruiting agree that this is a much better recruitment pool, recruit pool, than we ever had in prior decades. They are very optimistic about this. There remain pockets of challenges, and we haven’t yet seen attrition rates once they get in the field. But so far, so good on that.

The other challenge of non response follow-up is really the reaction of the American public. Now, instead of asking them to fill out a form and mail it back, we’re going to ask them when they receive a knock on the door, or a door bell rings and they see someone standing out there with a badge and a census briefcase, that they are there to make it easy for them to participate in the census just by answering a few questions on the doorstep. That will be the focus of our partnership activities and our paid media in the coming weeks. We will try to make sure everyone understands that’s the next phase.

MR. BUCKNER: Okay, with that that concludes our operational press briefing on the 2010 census. If we did not get to one of your questions, or if you have additional questions, please submit them to the Public Information Office. In terms of recruitment and the hiring of those positions for the non response follow-up that begins in May, we had a very large applicant pool that exceeded our goals, over three million people applied for the 600 to 700,000 workers we’ll be needing to go door to door. That hiring will begin in about two weeks. So just to sort of build out the timeline of when to expect perhaps a call back to come in for training for those positions.

If you have any questions about the recruitment, please feel free to call the Public Information Office at 301-763-3691. With that, thank you very much and get those forms back. You’ve got five days left. Thank you.


Tags: , MyTwoCensus, , , , , , Robert M. Groves, , , , ,

2 Responses to “Transcript from Tuesday’s Press conference…”

  1. My Two Census » Blog Archive » Fact-Checking Census Bureau Director Robert M. Groves: Are assisted living facilities group quarters? Says:

    [...] a reader pointed out to me an inconsistency from the transcript of Dr. Robert M. Groves’ most recent press conference. Dr. Groves said, “And then finally group quarters, another category of folks who don’t [...]

  2. HermHollerith Says:

    How can listing 40,000 Questionnaire Assistance Centers on the Census Bureau’s web site be an IT challenge?
    Even if the sites move around and have different hours on different days. Relational Database Design 101.
    Lamest thing I ever heard.
    This kind of remark destroys confidence in the agency’s ability to enumerate the population accurately.