Skip to main content

UW-Madison demographer gets far-flung questions and far-reaching results

July 21, 1999

When Paul Voss tells you, “The day is punctuated by phone calls and e-mails,” you might respond, “So is mine.”


What makes the census so costly and time-consuming?

What is the American Community Survey?


But a look at Voss’s phone log gives a glimpse into the far-flung questions he gets. The Madison Symphony Orchestra recently asked him how to use demographics in marketing. The Wisconsin Department of Natural Resources wondered what the population was in the area of the Crandon mine. And the House Subcommittee on the Census invited him to testify about methods for the census.

It’s a typical day for the director of the College of Agricultural and Life Sciences’ Applied Population Lab. In his 30 years of work there, Voss has fielded such questions as:

  • Which Wisconsin counties have a high percentage of log homes?
  • How dense is the human population in areas near national forests?
  • How many sewerage plants will a city need in the year 2020?

Who gets that information? Anyone who needs it. “My job is solid outreach for communities, business, government, individual faculty members and the media,” Voss says. “I have a large Extension appointment, and the nature of that appointment is helping people.”

This is a difficult time for providing data, Voss says. “My biggest challenge right now is trying to help people with data in a part of the decade when the census data is very old.” One professor asked about data on marital status, and Voss had to give her the latest census information he had – which was from 1990.

Lately, many of the questions have involved the census, which Voss has been involved with for years. As a member of the Commerce Secretary’s Advisory Committee for the 2000 Census, Voss helped the U.S. Bureau of the Census find ways to make the census less costly and time-consuming.

To get more people to respond to the 2000 Census, the Census Advisory Committee met regularly to see how they could shorten the time it takes to fill out the questionnaires. The advisory committee included 35 representatives, with stakeholders ranging from the National Committee for the Complete Count of Asian and Pacific Islanders to the Business Roundtable. Voss represented the professional demographic community.

Still, the committee faced one major roadblock: figuring out which questions could be cut. Every question is required by at least one federal statute. An item on how people got to work in the previous week, for instance, is required by the Departments of Labor, Transportation, Justice, and Commerce. The Federal Reserve and the Environmental Protection Agency also require the question.

But even when federal statutes require the questions, the Census Bureau has to determine whether the census is the best way of getting that data. For some questions, the Census Bureau could get the data from the national surveys it conducts on a yearly basis. So the 2000 Census forms won’t ask how many children have been born to each woman, for example.

Another new feature of the 2000 Census, created in response to public demand, is that people can respond to more than one racial group as they identify themselves. The census now has five groups to choose from: white, Asian, African American/black, American Indian/Eskimo, and Pacific Islander/Native Hawaiian. In addition, another question asks whether people are Hispanic.

While it’s more realistic for people to say they are from more than one race, the variety of answers creates a new problem for researchers, Voss says. “You take the Tiger Woods of the world, and you have the potential of getting very complex sets of multiple race responses. Then you tabulate those five with Hispanic origin.” Someone looking for a correlation between education, income and race has a tabulating nightmare.

Besides working to improve the census itself, Voss works with Cooperative Extension Service to promote the census. He helps extension agents in Wisconsin support mayors and local agencies in gathering and using census data. Their 1990 promotion efforts paid off: Wisconsin had the highest mail return rate in the nation at 77 percent.

No matter how well the state promotes the 2000 Census, however, previous estimates from the Census Bureau show that Wisconsin may still lose a Congressional seat. Even in 1990, Wisconsin came close to losing a seat. Since then, “Our population has grown, but relative to the Sun Belt, it’s grown more slowly,” Voss says.

The census and other research are only a part of Voss’s job, though. He also teaches a course on American demographics, and expects to create a new course while he’s on sabbatical in Laxenburg, Austria this school year. Besides that, he’ll finish a book manuscript. And when the 2000 Census begins in April, he’ll be out of the country.

Tags: research