Jeremi sits down with Dr. Teresa A. Sullivan to discuss the U.S. Census, it’s history and how it affects our society today.
Zachary sets the scene with his poem, “Dear Governor”.
Teresa A. Sullivan, university professor and president emerita of the University of Virginia, is currently serving as the interim provost at Michigan State University, her alma mater.
As president of UVA, Sullivan led a team that stimulated the revitalization of the UVA Health System, raised faculty salaries, launched an ambitious program of faculty hiring, raised both the numbers and quality of applications, reached new fundraising records, and launched the university’s bicentennial celebration.
Earlier, she was the executive vice president and provost at the University of Michigan, executive vice chancellor for academic affairs for the University of Texas System, and vice president and graduate dean at the University of Texas at Austin.
In her academic career as a demographer, Sullivan developed analytic techniques for the use of U.S. Census Public Use Sample. She was an investigator on a large international sample survey, and with law colleagues Elizabeth Warren and Jay Lawrence Westbrook, she led several original large-scale data collections of consumer bankruptcy records. The first book-length analysis of the bankruptcy records, As We Forgive Our Debtors, received the Silver Gavel Award of the American Bar Association. The second book, The Fragile Middle Class, received the Writing Award of the American College of Financial Services Lawyers.
Sullivan has held faculty positions at the Universities of Chicago, Texas, Michigan, and Virginia and has received five major teaching awards. She is a fellow of the American Academy of Arts and Sciences and the American Association for the Advancement of Science.
- Teresa A. SullivanProfessor at and President Emerita of the University of Virginia
- Jeremi SuriProfessor of History at the University of Texas at Austin
Speaker 1: This —
Speaker 2: is Democracy.
Speaker 1: A podcast that explores the interracial, intergenerational, and intersectional unheard voices living in the world’s —
Speaker 3: most influential democracy.
Dr. Jeremi Suri: Welcome to our new episode of This is Democracy. This week we’re going to talk about the census and the historical and contemporary role of the census in American democracy. Why do we do a census? Why is it in the Constitution? Why is it so crucial for our democracy? We have with us, I think, the foremost academic and policy authority on this topic, Terry Sullivan, who’s a former professor and leader of the University of Texas as well as the University of Virginia. She’s now a university professor and President Emerita of the University of Virginia and currently serving as the interim provost at Michigan State University, her Alma mater. Terry Sullivan began her academic career as a demographer, developing analytical techniques that the census uses today, including the census public use sample. She was an investigator on a large international sample survey with colleagues at the University of Texas, including one you might have heard of, Elizabeth Warren. They led an original large-scale data collection of consumer bankruptcy records among many other topics. So we have someone who’s an expert on the study of the census and an expert on the doing of the census. Terry, thanks for joining us.
Terry Sullivan: Thank you. I’m really glad to be here.
Dr. Jeremi Suri: So before we turn to our discussion with Terry Sullivan, we have of course, Zachary’s scene setting poem for this week. What’s your poem titled, Zachary?
Zachary Suri: Dear Governor.
Dr. Jeremi Suri: Dear Governor, let’s hear it.
Zachary Suri: I know you mean well, I know you don’t mean to divide us heaven and hell. I know you sleep uncertain every night knowing deep down we are right. I know you are trying to save some remnant of a nation new think has become far too feminine. I know you believe in some American mythology of white christian exclusionary methodology. I know you can count, I know you can see that two and two does not equal three. I know, you know just as well as I, that all of us triumph when none of us cry. I know you understand we all deserve to be heard. None of us can live with dreams deferred. I know you believe in this great democracy, so don’t hide away in your empty hypocrisy. There’s something more important than party, something more important than who came here earlier tardy. There is something more righteous even than patriotism, or which side you fall on of the partisan schism. There is something more mighty than a racist nostalgic, something more powerful than propaganda magic. There is something far more powerful than you or I, and I hope you can see it. We need you to try. It’s the truth and everyone knows it is the divine ever righteous rose. So when it comes time for this 10 year count, remember, we won’t forget, we’ll hold you to account. Don’t let us see our values of fronted, come April, don’t let us see America stunted. Don’t let our voices be discounted and see to it, please make sure we’re all counted. Sincerely, America’s under-counted.
Dr. Jeremi Suri: I love it, Zachary. What is your poem really saying?
Zachary Suri: My poem is really about how important it is that state leaders like the Governor of Texas get involved in the census and make sure that everyone is counted, because there are a lot of threats to it right now and we need to make sure that it’s as accurate as possible.
Dr. Jeremi Suri: That’s great. Terry, why do we do a census in our democracy?
Terry Sullivan: Great question. It’s actually on the sixth sentence of the Constitution. So the other question is, why did they put it in the Constitution? In fact, it was a bold thing to do in 1789 because Europeans had a deep superstition about conducting censuses based on an instant recounted in Book of Chronicles about King David doing an unauthorized census and then being punished by God. So they were afraid that if a census had not been explicitly divinely authorized, it was sacrilegious. The founder’s decided to go against that. The reason was they wanted the House of Representatives to be base proportionate to the numbers of people in the states. To do that they had to know how many people there were. So the first census that was conducted in 1790 and there’s been one every 10 years since. This would be the 24th census, this year.
Dr. Jeremi Suri: Chase West, one of our students asks, how has the census been politically used? How has it been politicized historically since this first census of 1790?
Chase West: My name is Chase West, and I’m a third year Bio Chemistry major. My question is how has this census been politicized historically, and how does that relate to the politicization of the census today?
Terry Sullivan: Thomas Jefferson was in charge of the first census and he complained to George Washington that he would certain there had been an under-count. That wasn’t so much political, that was just a reflection of the difficulties of the operations then. But there’s been political concerns of various kinds. It shows itself in terms of, first of all, budget. Congress can restrict the budget the Census Bureau has. That happened after the 2010 census. The Census Bureau has never gotten what it’s requested or even close to it. As a result, they cut back dramatically on a lot of the testing they would normally have done. So the GAO has put this census on its top 10, highest risk endeavors by the federal government because of the concern that inadequately tested technology may lead the census to fail. One of the cost-cutting measures was to move as many people as possible from responding in paper to responding online. In addition, enumerators and others are going to be using tablets and other electronic devices. It’s more efficient, but it’s only been through one test in the providence pretest which happened 2018. So there’s some concern that things could go wrong. I should also note that the Australian Census Bureau got hacked last year. There’s been a lot of concern that somebody will try to either hack the census or hijack information from individuals before or as it’s being transmitted to the census. There’s lots of things people are worried about. Some of them are political and some of them I’d call quasi political. Maybe the biggest issue is under-count as always. The under-count is not randomly distributed across the population. The hard to count people are the most likely not be counted and the big hard to count group, and now think about Texas when I say this, are children, minority groups, rural residents, and renters. Texas has a big share of all of those. But Texas actually is a state very likely to have a large under-count.
Dr. Jeremi Suri: This is an old story, before we talk about the contemporary issues with technology and hacking, which I do want us to talk about. The problem of an under-count obviously goes back to Thomas Jefferson’s concerns. Over time and prior censuses, how has the government dealt with these concerns of an under-count?
Terry Sullivan: Well, the principal way has been by going back to a household over and over again to try and get a completed return. That’s expensive because you’re sending a human being on a payroll to go knock on doors. But the Census Bureau has been very diligent about doing that, and one way to save money is to cut back on how much of that you do. That will also be done in this census, instead, they will turn to what are called administrative records. So if you live at 121 Pine Avenue, and there’s no census return for 121 Pine Avenue, they’ll go to some other administrative record like social security records to see if they can find out who’s in that household. That’s different from what they’ve done in the past. But the Census Bureau works very hard to get a complete count. It’s just that America is hard to count. We’re a mobile population. We’ve got people who live in unusual or perhaps illegal housing. We’ve got people for one reason or another don’t want to be counted, worried that they’re engaged in something illegal. So there are lots genuine problems to counting. The solution to it has principally been money. Now, after the 1980 round of censuses, big city mayors were very concerned that their cities had been under-counted and launched a series of lawsuits, I think there were something like 180 lawsuits against the census that year. The issue came up of whether you could adjust the census with sampling, with post enumeration survey taken after the census. Ultimately the issue was, could you adjust all the way down to a block level? While the Census Bureau was cleared by the Supreme Court to do this for some purposes, not for apportioning the House of Representatives, but for other purposes, they decided that for technical reasons, they didn’t think they could do it. They didn’t think it was robust enough to adjust all the way down to very fine geographic detail. So we’re back to the problem of how do we get people counted who are hard to count.
Dr. Jeremi Suri: Another one of our students, Abby Hammer, asks about the ways in which certain groups have intentionally been under-counted, particularly due to racial categorization. Can we hear that question?
Abby Hammer: Hi, my name is Abby Hammer. I’m a second-year International Relations and Spanish double major from Dallas, Texas. My question is, the first census ever taken in 1790 and the following seven were taken in an era where slavery had yet to be abolished. Through that, how has the census in regards to racial categorization changed, and do you think it’s fully inclusive to all races and racial identities?
Terry Sullivan: Well, I would say that it is an issue, but it’s a function of these other things I’ve mentioned. Particularly, low-income members of minority groups are more likely to be renters, they’re more likely to move often, they live in crowded housing conditions. Very often they’re subject to eviction and other such things, so there’s housing instability. And they often have — Because they’re often younger and have more children, they’ve also got more children who may be undercounted. Some children are undercounted because they’re living with grandparents or in other kinds of irregular situations, and whoever is filling out the census form doesn’t actually know if this child should be counted here or not. But some children aren’t counted because the housing unit doesn’t permit children, or the parents who try to shield the child from being known by the government. There’re all kinds of reasons why this could happen, but an undercount has really serious consequences down the road. Let me just tell you about one of them.
Dr. Jeremi Suri: Please, please.
Terry Sullivan: About one and half trillion dollars of federal revenue sharing has something in the formula that is based on population or a particular population. And so, a single poor child who is missed can cost a school district $1,700 a year in Title 1 funds. So, if you have a school district that has a lot of poor children who get undercounted, they’re going to lose out on a lot of federal money. Of course, that federal money will be spent. It’ll just be spent somewhere where the count was better.
Dr. Jeremi Suri: Right. And I guess underlying Abby’s question, right, is have the race categories that are defined within the census, and the way they’re counted been used intentionally to affect or prejudice the ways in which money is spent?
Terry Sullivan: You know, that’s a hard question to ask. I think it’s unlikely the Bureau itself did it. But sometimes the Bureau is reliant on other federal agencies.
Dr. Jeremi Suri: Right.
Terry Sullivan: Until in 2017, after a lot of expert analysis, the Bureau recommended adding a category of Middle Eastern and North African to the count, a separate category that you could check. The Office of Management Budget sat on that recommendation for so long that it could not be enacted for this census. But that’s an example of something where there was certainly political opposition mounted against the category. There were interest groups who wrote to demand that it not be included, and simply the names you use. How do you define Mexican-Americans? What the Bureau tries to do is to come up with a variety of terms that people might call themselves by. So both Latino and Chicano will be included in that census.
Dr. Jeremi Suri: Great. Zachary?
Zachary Suri: Why is it that when a lot of us realize how important this census is, and how vital it is to our democracy, to people and to governments, why is it that it’s something we don’t talk about very often when it comes to discussing government? It seems to be something that we talk about when it’s about to happen and then completely forget about for the next nine years.
Terry Sullivan: Well, because it does happen just once every 10 years, we tend to focus a lot of attention on it then, and also the Census Bureau spends a good bit of time and effort mobilizing the population once every 10 years. That’s going on right now. You should be seeing public service ads. Schools and pediatricians, and other trusted individuals are being encouraged to encourage their clients to participate in the census. It’s probably the public relations effort. But in fact, the census data are used heavily by lots of people. Sometimes they don’t even realize they’re using census data. But it’s used by marketers, I mentioned school districts that choose to plan highways, businesses when they’re thinking about where to put a new location, they’re using census data to figure that out. There’s just lots and lots of uses of the census data all throughout the economy. The other thing that’s technically important for those of us who are social scientists is, the census is used as the sampling frame for all future samples that are drawn, both in the private sector and the public sector, because it’s the only database that maps every individual into a unique geographic location.
Dr. Jeremi Suri: I was thinking about that, Terry, that we use it so often as social scientists, as citizens, as policy analysts. How is it, coming back to your earlier points, that we can put ourselves in a situation now where there seems so many risks to the veracity of this data due to hacking, due to insufficient preparation this year, how did we come to this point?
Terry Sullivan: Well, any demographer who’s not worried about it, hasn’t been paying attention. I think a lot of us are very concerned that this might not be a good census. I will say in the political realm, there is a belief, I want to underscore, this is a belief and not necessarily a proven fact. But there’s a belief that a poor census benefits Republicans. I think that the flap over the citizenship question, which got decided by the Supreme Court last year, is an example of this. The citizenship question, and there was some experimental data to support this, suppressed the responses by Latinos, whether they were legally present in the United States or not. They saw that question and they simply didn’t answer. If you know that’s going to be the case and you put the question on anyway, is that an effort to suppress the count of Latinos? Latinos are believed, and I think this is demonstrated more often, to vote Democratic, so is this an effort to cut them out of the count? There’s lot of supposition about this. What the Supreme Court ended up saying was that the question could not be added in 2020, because the Secretary of Commerce gave a pretextual reason for adding it.
Dr. Jeremi Suri: Right.
Terry Sullivan: That is, they didn’t believe the reason he gave. But it has led people to think that there is a political advantage to the undercount. I think the other thing, which is interesting, is that President Trump has ordered the Census Bureau to try to construct a best estimate of who is a citizen and who is not. But not using census data, using other data, and to seek to give state legislatures two data files. One, which is the count by small geographic area, and the other which is the count of voting age citizens, so that presumably legislatures could choose how they wanted to redistrict. This doesn’t affect Congress and the reapportionment. It only affects how districts are drawn inside the state for those members of Congress. Then most states also use it to district, state legislatures as well.
Dr. Jeremi Suri: Just to be clear on that, and you mentioned this before, the Constitution requires that Congress is apportioned its seats based on the census not on other data, but states can use other data for their state seats, correct?
Terry Sullivan: It is certainly the case that Congress must use the raw data as it comes from Census Bureau. That is true. It is believed that legislatures have a good bit of leeway in how they redistrict. I say believed, because I feel certain there will be another round of lawsuits about it, and particularly in states that have a record of seeking to suppress voting rights of minority groups. But I believe right now, I’d say, the general opinion seems to be state legislators have a lot of leeway and the last summer Supreme Court decision about gerrymandering, which said it was a political matter and not a matter for the courts, probably reinforces that.
Dr. Jeremi Suri: In the past, Terry, as a historian myself, I don’t know the answer to this. In the past have legislatures used non-census data in apportioning state assembly and state Senate seats?
Terry Sullivan: Yeah, I understand that Hawaii has used voting rolls in part because there are so many people living in Hawaii who are not really resident for the state. They station with the military. They are tourists, they are short-term visitors, and so on. I do understand that it’s one state that has done it, but there are other states who are certainly looking at it.
Dr. Jeremi Suri: I guess what we have to turn to now, what we always turn to in the latter part of our discussions, is what can we do about this? For those of us who care about an accurate count, first of all, because we’re good social scientists and care about accuracy, but also because we care about fairness and your point about schools is so poignant. The funding for children, for schools, for child health, for so many issues is deeply dependent upon an accurate count. What can we do as citizens and activists to encourage a good census?
Terry Sullivan: Well, each of us has a network. We have friends and neighbors and co-workers and others that we interact with and it’s a great time to encourage the people we know to complete the census form. I think that we can put to rest a lot of people’s concerns about their confidentiality of their personal data. I firmly believe that the Census Bureau does keep that information confidential since its employees take a lifelong oath to keep it confidential and they’re not permitted by law to share it with any other agency. I think that the data themselves are about as secure as we can make it. I think people should feel free to answer it. There are also complete count committees that are seeking to turn out the count. I’m pretty sure that there is one in Travis County or possibly in City of Austin and there are other localities in Texas that have it. Unlike many states, Texas does not have a complete count committee for the whole state, but states like California are putting forward a lot of money to basically do what is almost the parallel census to help ensure that they have the maximum count. So a lot of states are doing a lot of work to make sure people get counted.
Dr. Jeremi Suri: What about going forward beyond this census? What can we do as we hope and hypothesize on our weekly podcast here, that we’re entering a reform period for our government soon? How do we encourage this to be part of a series of what we might call progressive reforms for our government going forward?
Terry Sullivan: Well, I think one thing is to the extent possible to try and insulate the Census Bureau from politicization. Right now, the people in the Census Bureau tend to be very dedicated professionals. Highly skilled by the way. Some of the biggest reforms in social science data-collection have come out of the Census Bureau. I try to insulate them from this. Of course, part of that insulation is don’t starve them of the funds they need to do the research to do a better count. It’s easy to do that in years that don’t end in zero because you figure the census is far away. We can save money on that now. Well, actually you can’t. So you got to keep going because people in the Bureau need to continue to test and refine their techniques as our population changes and as the technology around us changes. I think those are maybe the most important issues.
Dr. Jeremi Suri: As I listen to you, it’s so clear that we need talented people, leading thinkers, as scholars, as demographers to be involved with these issues and to devote some of their own career to them. How did you get involved in this?
Terry Sullivan: Well, I did my dissertation using the 1960 and the 1970 public use sample. These are developed by the Census Bureau, they are extracts of the census data, but they’re actual samples. You don’t have to worry about whether you have a random sample or not, you do. Any identifying information is gone, so you can’t tell who this person is. What you know, however, is the generic information about age, sex, race, level of education, and so on. These are pretty detailed samples, and I use them to estimate the amount of underemployment in the United States. Not just unemployment, which the measurement of that is pretty well developed now, but measures of underemployment. I learned a lot about the census in the course of doing that. Of course, I taught demography courses at Texas for many years using census data in class. But eventually, the Census Bureau asked me to serve on their — They have a professional advisory group made up of statisticians and others who are trained in areas the census is interested in. I eventually chaired that group. We were consulted on a wide variety of things. I served on a couple of national academy panels that also looked at particular issues about the census. I’ve had the opportunity to see the Census Bureau from a number of different perspectives and I’ve always been impressed with how hard they work and how hard they try to do the right thing. The issue is with any kind of technical innovation, what is the right thing is hard to figure out.
Dr. Jeremi Suri: Sure.
Terry Sullivan: Right now that issue is about what’s called differential privacy, which is a statistical technique the Bureau is planning to deploy in the census. A lot of the professional community is unhappy about it. There’ll be a lot of conversation about that. But it’s an effort to protect the privacy of individuals by injecting what’s called noise into the data. The intention is that you’ll never be able to identify a particular individual in the census. But what users are worried about is you will never know if your data are accurate.
Dr. Jeremi Suri: That’s a problem.
Terry Sullivan: Yeah. It is.
Dr. Jeremi Suri: Terry, my final question for you is, it has been just a wonderful in-depth examination of these issues. Do you see other young talented people, the next generation of Terry Sullivans out there, getting involved with these issues, bringing their talent to bear, not just on the count, but on the statistical measures and the uses of technology and demography around these issues?
Terry Sullivan: Yeah. I do think that that is happening. I had the privilege of giving the president’s invited address at the American Statistical Association last summer, and I talked about these very issues. I was really impressed by how many statisticians, including students who were present there, were intrigued by the issues and by the opportunity to really be at the cutting edge of important developments. I think the Census Bureau will be able to continue to recruit very talented people, assuming they got the money to do it.
Dr. Jeremi Suri: Right. Zachary, do you think young people like yourself will be excited and energized to care about these issues and to get involved with doing the work of making sure that they were counted in a democracy? Counted in a way that protects privacy, but also allows for the accurate use of our resources?
Zachary Suri: I think we will, and I think it’s mainly because it’s so important to our future, more so than really any other group in American society. It’s most important to America’s young people. But I think that we really need to think about going forward is how do we bring the census to the forefront of our political discussions even when it’s not year when we’re about to conduct a census.
Dr. Jeremi Suri: To keep our attention on the issue. Terry, really last question now, what should our listeners be looking forward to? What will they hear from the census in the next few weeks? We in Austin have not received anything yet. What should we be looking for?
Terry Sullivan: Well, you won’t get anything probably until starting March 12th. It will be a letter that every household receives. Now, Austin is an area that has a high level of connectivity and Internet usage. So it’s very likely that most of the Austin population is going to get a letter that has a code. If you enter that code online, that will uniquely identify your household address and you can complete the census online. It’ll take about 10 minutes. You could do it with a smartphone too. For people who don’t have access to the Internet, there will also be paper questionnaires, and they’re available in 12 or 13 languages. So, even if English is not your first language, you should be able to get the census form in a language that you understand.
Dr. Jeremi Suri: Will it be sent to people’s homes? How does that work?
Terry Sullivan: Yes. It goes to your household address. There was an effort made starting last October to do what’s called canvassing, that is to identify every residential address in the United States. If for any reason your household has not received this letter in the mail by the end of March, you will certainly have plenty of opportunities and you’ll see lots of public service announcements giving you a phone number you can call to ask for a census questionnaire.
Dr. Jeremi Suri: But the main thing is to be looking for the letter in the postal service delivery to your home.
Terry Sullivan: That’s correct.
Dr. Jeremi Suri: Great. Terry, thank you so much for sharing your expertise and historical perspective with us. I hope many of our listeners take from you, not just the knowledge you share, but also the commitment to this important institution is vital institution in our democracy. Thank you, Terry.
Terry Sullivan: Thank you, Jeremi and Zach. And let me just say, everyone counts.
Dr. Jeremi Suri: Everyone counts. That’s right. Zachary, thank you for your poem and for your questions today and for helping to inspire young people to care about the census in our democracy. Thank you to our listeners for joining us for this week’s episode of This is Democracy.
Speaker 3: This podcast is produced by the Liberal Arts Development Studio and the College of Liberal Arts at the University of Texas at Austin. The music in this episode was written and recorded by Harrison Lemke and you can find his music at harrisonlemke.com.
Speaker 1: Subscribe and stay tuned for a new episode every Thursday featuring new perspectives on democracy.