How Is Your Data Used When You Turn In Your Census?
Next year, the United States Census Bureau will send out its 10-year census to collect demographic data on every person in the country. That survey happens once a decade and asks a handful of questions, but the agency also sends out the yearly American Community Survey, or ACS, which is an ongoing survey that collects more detailed data on smaller populations. How is your data used once you turn in your survey?
Demographer Catherine Fitch talks about how the information surveys are used for research and policies, why certain questions appear on the forms, and new ways that the census is trying to survey the country.
Catherine Fitch is the Associate Director of the Institute for Social Research and Data, and Innovation Co-Director at the Minnesota Research Data Center, at the University of Minnesota in Minneapolis, Minnesota.
IRA FLATOW: This is Science Friday. I’m Ira Flatow. Next year the Census Bureau will send out their survey, you know, that big packet that they have to every person in the country, asking information about households, gender, all the basic info stuff. And that census happens once every 10 years. But you know– maybe you didn’t know. I didn’t know.
There is another census called the American Community Survey that is sent out all the time to smaller populations. And both of these surveys are big data collection projects. Add into the mix that next week, the Supreme Court will rule if a citizenship question can be included on the 10 year survey, and we’ve got lots to talk about. So how do certain questions make it into the census? And once you fill out the survey, how do researchers and policy makers make use of that data?
My next guest is here to walk us through those questions and, hopefully, the answers. Catherine Fitch is the Associate Director at the Institute for Social Research and Data Innovation at the University of Minnesota in Minneapolis. And we did reach out to the Census Bureau to supply its spokesperson from the agency. I think that phone is still ringing. We’re waiting for a response.
CATHERINE FITCH: Dr. Fitch, welcome to Science Friday. Good afternoon. Thanks for having me.
IRA FLATOW: You’re welcome. The census that is sent out every decade is updated each time. What is new on the upcoming census?
CATHERINE FITCH: Well, I think the main thing– well, there’s a couple of things, you might say. One thing that’s going to be new this year is how we’ll take the census. And so this year, there will be an internet response option. So that’s one thing that’s new. And the other thing is the thing everyone is talking about, which is the possibility that they would include citizenship on that survey, which hasn’t been done before.
IRA FLATOW: And why are people fearful of that question?
CATHERINE FITCH: Well, I think people fear that– and evidence suggests– that it will depress the response rate on the survey. And that response rate is really important for two reasons. The higher the response rate, the more accurate the census is going to be and the less expensive it will be. When you don’t respond, they need to send someone to track you down and get answers to those questions. And I think, from previous studies the Census Bureau has done, they can see that this is likely to have a significant response rate, particularly on people who have noncitizens in their household.
IRA FLATOW: That’s why we’ve asked our audience will they take the census? And what would they think about having the citizens questions. We have lots of people who are responding. Well, let me go to a tweet.
Meg tweets, I’m a statistics teacher from Muskogee, Oklahoma. And I’ve asked my students how the citizenship question will affect the accuracy of the census? She asks, what’s your take?
You’ve already told us that you think it’s going to suppress the number of people who might answer the question. Let me go to the phones. We have some interesting– let’s go to Jacksonville to Richard in Jacksonville. Hi, Richard.
RICHARD: Hey. How you doing?
IRA FLATOW: Hi there. Go ahead.
RICHARD: The census, I actually think it’s very personal. I mean, we pay taxes. So if they’re talking about getting their money for roads or for different things, or for, I don’t know, different races, creeds, and so on and so forth. I mean–
IRA FLATOW: Are you going to participate in the census?
IRA FLATOW: Will you participate in it?
RICHARD: No, I never have. But one thing I might want to say and add is that, say a hundred years from now, if somebody wanted to track their great grandfather, that would be hard for them, because I was not part of the census. But other than that, I don’t see the reason for a census.
IRA FLATOW: OK. There you have it. He doesn’t see the reason for a census.
CATHERINE FITCH: Well, I think if you go back to the Constitution, it’s required. And that’s how we decide how many representatives are apportioned, allocated to each state. So that’s a really important role the census plays and one of the reasons that all states are working with the Census Bureau to increase turnout and response to the census.
IRA FLATOW: Right. And now you’re a social scientist who studies marriage formation.
CATHERINE FITCH: That’s right.
IRA FLATOW: Can you break down how you might use the data from the census?
CATHERINE FITCH: Sure. And do you mind if I take a second, explain a little bit about the American Community Survey first?
IRA FLATOW: I’m putting my feet up. Go right ahead.
CATHERINE FITCH: All right. So the census, which is asked every 10 years, has a small number of questions asked. And in fact, this year they’re not asking marriage on the census. So the census itself would not be a great source for me.
But the American Community Survey, which is, as you mentioned, always in the field, has a rich array of questions, including demographic questions, marital status, as well as changes to marital status in the last year. So one question I have asked is– there’s been a big increase in marriage age since the middle of the 20th century and trying to think about what are the causes of that and/or what things are associated with increased marriage age.
And one thing I wanted to look at is economic opportunity. What role does other opportunities for women play in decreasing or increasing their likelihood they’re going to be married? So having a data set that includes information about demographics and marital status, but also about educational background and occupation, other information, allows that kind of analysis.
IRA FLATOW: Now the American Community Survey, as you talk about it, that does include a citizenship question, does it not?
CATHERINE FITCH: Yes. That’s correct. Which is one of the reasons questions have been raised about the need to have it on what was referred to often as the short form. The reason it’s called the short form is, before there was an American Community Survey, every census year had a short form that everyone got and a long form which asked similar questions to the American Community Survey. And that was asked of maybe around 20% of the population.
IRA FLATOW: We have lots of people who– well, let me read another tweet. He says if the US census includes a citizenship question, I will not respond to it. The last time there was a controversial question I refused. My wife finally relented. So this is– how do you expect the Supreme Court– give me your over under on the Supreme Court?
CATHERINE FITCH: I am not a Supreme Court watcher.
IRA FLATOW: I know it’s an unfair question. You could throw a dart [INAUDIBLE]. We’ll accept–
CATHERINE FITCH: Throwing out the question. I was talking to someone today who threw out an expectation that maybe they would not allow the citizenship question on the census. But–
IRA FLATOW: Wow.
CATHERINE FITCH: –that’s what I’ll say–
IRA FLATOW: OK.
CATHERINE FITCH: –on guessing what the Supreme Court is going to say. But I focus a little bit more on what we know. And that is, you know, the data exists on the American Community Survey and can be used. And this is just likely to decrease the accuracy and increase the cost of the census.
IRA FLATOW: Let’s go to Don in Prescott, Arizona.
IRA FLATOW: Hi, Don.
DON: Hi, Ira. How are you doing?
IRA FLATOW: Hi there. Go ahead.
DON: I took part in the 1980 census. And I can tell you there were some improprieties that happened, such as some of my co-workers putting the census in mailboxes and coming back later to pick up the completed forms. A second thing I’d like to talk about is the gerrymandering that occurs and has been occurring for decades now with blue and red legislatures using this data to put noncompetitive districts together. And I advise all my friends– put down how many people live permanently in your household, but other than that the information is, in my opinion, intrusive.
IRA FLATOW: OK. Catherine?
CATHERINE FITCH: Yeah.
IRA FLATOW: Not everybody fills in every question on the census, right?
CATHERINE FITCH: That’s right and they will have techniques for trying to fill in the missing information. And they will try to do that. And I encourage people to think about the other thing that the census does.
About $80 billion in federal funds will be allocated based on census responses. And that’s important. If you have kids and you’re in a growing school district, they need this information to help plan for coming years and make sure they have the schools and the teachers available to support your family and your neighbors. So there’s a lot of important information to be gathered from the census.
IRA FLATOW: A tweet from Tina actually backs that up. And she says I use census data to help determine how transit access affects minority and low income populations. Without it, equity concerns would be anecdotal at best.
CATHERINE FITCH: That’s exactly right. There are a lot of topics that policymakers and planners would simply be taking on on anecdote. And with the American Community Survey data backed up by census data, they are able to make a lot more precise decisions. And I think that that’s another reason to consider filling it out.
IRA FLATOW: Well, let me ask you the difference between the American Community Survey and the census. To you, which is the more important census or survey?
CATHERINE FITCH: The American Community Survey, by far, I think– it’s a source that most researchers are using. And so we focus on the American Community Survey at the University of Minnesota and making that available to researchers in an easy to use format. But the census is sort of an underlying necessity for, also, a strong American Community Survey. So they can’t really exist separately.
IRA FLATOW: I don’t ever remember seeing a form coming in the mail– is that– from the American Community Survey. Or if it had, I thought it might be a book offer, you know? American something– just throw it out. How do I know it’s even coming in?
CATHERINE FITCH: It would come in the mail. It’s got the Census Bureau stamped on the outside. And I think they’d follow up with a couple more mailings if they were trying to get you. But–
IRA FLATOW: How many Americans? One in what?
CATHERINE FITCH: How many Americans– I think it’s– I’m going to guess it’s, like, 2.2 million respondents. I could be off on that. What they get is a 1% sample in the micro data we use of the population. And they have a broader sample to get that 1%.
IRA FLATOW: That’s about 3 million, 1%.
CATHERINE FITCH: Right.
IRA FLATOW: Roughly. That’s usually what people do, you know, in marketing, whatever. They get a 1% sample. They consider that to be significant.
CATHERINE FITCH: Yeah, well, it’s actually pretty amazing what you can do with it. We sometimes talk about the American Community Survey and the decennial census that preceded it as the Hubble telescope for social science research. It offers a lot of detail. And it does allow you to look at small sub-populations, whether that is an age group or an occupational group, a geographic region.
And you have this great detail, as well as this broad scope of time that really is fundamental to understanding the social demographic and economic transformations of the last few decades.
IRA FLATOW: I’m Ira Flatow. This is Science Friday from WNYC Studios. Being educated by Catherine Fitch, who is telling me all kinds of stuff I didn’t even know existed about the census data. Let me go to the– see if I can get one call in, and then another question.
Yeah. Let’s go to Don in Columbia, Missouri. Hi, Don.
DON: Yeah, hi. The question I’ve got is why can’t government just use existing data– income tax data, social security data, people receiving government benefits, and then maybe some mechanism to account for people that don’t fall into any of those categories, rather than having a formal census?
IRA FLATOW: Isn’t the census in the Constitution? We have to take it.
CATHERINE FITCH: We have to count every one every 10 years. But the strategy that the caller suggests is something the Census Bureau has been implementing. And they’re working on gathering some of the information from administrative records. As he describes, there’s a lot of things we already tell the government at different times during the year. And it would be great to capture that instead of inquiring about it.
I think there’s still some work to be done to get through the complete array of information that’s included on the American Community Survey to completely replace it with outside data. But they are working on that. And that also helps with the accuracy of the survey.
IRA FLATOW: If I want to, having heard about this survey, the American Community survey for the first time, if I want to see the results, is it online? Can I access the data for myself?
CATHERINE FITCH: Sure. You can see descriptions on census.gov. And they have data access tools to help you look up, say, information about your neighborhood, which is one way they disseminate data.
You could also go to IPUMS.org. IPUMS is the database that we run at the University of Minnesota. We also have census data. And there you can download smaller extracts of the data. You can also look at some of those tables that describe places and analyze online.
IRA FLATOW: I’m going to go look.
CATHERINE FITCH: Please take a look.
IRA FLATOW: Tell me what the biggest misconception people have about the census or how the data is used.
CATHERINE FITCH: I mean, I think we’re already kind of tackling this with– when people refer to census data and they’re really referring to this annual survey that’s called the American Community Survey. I think that’s one of the biggest things. I think one of the other things is conceptualizing how important this is to research and the richness of it.
Another factor I talked about how you can get at small populations. Another thing you get is you have individual level records describing people in their families, in their households. And that’s really helpful for a lot of research questions.
One that I was just describing to someone is if you wanted to look at the impact of the new family leave law in California, the ACS will allow you to look at before and after that legislation. And you can look at households with small children. And you can look at the work and income of both– of any parents in the household.
And you don’t want to look at, say, a mother without the context of a partner who might also be working or who might be at home with a child. So that richness is really important.
IRA FLATOW: I have about 30 seconds left. And I want to ask you this.
CATHERINE FITCH: Sure.
IRA FLATOW: You say they take the survey online? Is there any fear it could be hacked to change the data?
CATHERINE FITCH: The census will be working to make sure they have security in place, to make sure responses are not hacked. I’m sure that is a great preoccupation at the Census Bureau.
IRA FLATOW: Well, thank you very much. This is quite informative. Catherine Fitch, Associate Director at the Institute for Social Research and Data Innovation at the University of Minnesota in Minneapolis. Happy census to you.
CATHERINE FITCH: Thank you very much. Thanks for taking the time to showcase social science.
IRA FLATOW: Well, we really depend on it.