Census change sparks undercount fears, with dollars and representation on line

Past U.S. Census Bureau surveys likely missed hundreds, perhaps thousands of minority Northwest Arkansas residents, according to the bureau and local research. Some researchers said an added question on citizenship to the 2020 census could push the missing number higher.

Researchers and local officials emphasized the importance of taking part in the census and reaching everyone. Census results each decade steer government functions as varied as setting school districts' board zones, giving Arkansas seats in Congress and distributing billions of federal and state tax dollars each year for roads, school lunches and rent and health coverage assistance for low-income families.

"This is big bucks, you know?" said Pam Willrodt, demographer for the Arkansas Census State Data Center at the University of Arkansas at Little Rock. "The census weights everything."

President Donald Trump's administration last month decided to include a question asking census respondents if they're citizens. If the question outlasts several legal challenges, it'll be the first time in decades it's included in the full census questionnaire.

The U.S. Justice Department said it requested the addition to protect citizens' voting rights against racial discrimination -- to see if the proportion of eligible people voting suddenly drops, for example. Supporters noted the American Community Survey, a smaller bureau survey that goes out to a sample of households each year, asks for citizenship information.

Arkansas Republicans in Congress said they supported the change to get a full view of the U.S. population.

"Counting the number of U.S. citizens in the country should be a high priority of the census, and the only way to get an accurate count is to add a question about citizenship to the census itself," Sen. Tom Cotton said in a statement last month. Last week, Cotton and Sen. John Boozman both noted the Census Bureau keeps responses private.

"Filling out these few questions is an important civic duty that is easy, important and safe," Boozman wrote in an email.

New fears

Several population researchers and immigrant advocates around the country took a different view, saying the question's inclusion could push immigrants to avoid the count even if they're in the country legally. California and other states and organizations have sued to remove the question.

Immigrants who overstayed their visas or came into the country illegally might avoid census takers and forms to avoid law enforcement, but so might their children or other family members in the country legally, said Juan Jose Bustamante, an assistant professor of sociology at the University of Arkansas, Fayetteville.

"We don't talk about individuals per se, we talk about entire families who are going to fear for their significant others," he said. "That is the challenge."

The Pew Research Center estimated the Fayetteville-Springdale-Rogers metropolitan area was home to 30,000 immigrants present illegally in 2014.

Willrodt said fear in response to anti-immigrant policies and rhetoric might also amplify immigrants' concerns. Trump has expanded deportation cases, slowed admissions for refugees fleeing conflicts and supported cutting all legal immigration, saying the moves protect Americans.

"Whenever you have that anti-anything sentiment, people are less liable to raise their hand and say, 'Yep, I belong to that group,'" Willrodt said.

Erick Sanchez is a Northwest Arkansas Community College student and a "Dreamer," taking part in a federal program that gives temporary permits for several hundred thousand people who were brought illegally into the country as children. His family includes older relatives who don't have valid visas and younger siblings who are citizens. They filled out the census in 2010. Sanchez said they probably won't this time.

"With me, personally, I feel like history could repeat itself," Sanchez said, pointing to the forced removal of Japanese-Americans to camps during World War II. Researchers at the University of Wisconsin at Milwaukee and Fordham University in New York have found the Census Bureau provided locations of Japanese residents for the effort in multiple states, including Arkansas.

The bureau apologized in 2000 and has stressed that by law it can't share individual responses with any agency. Sanchez said he isn't convinced, and the data could help plan mass deportations.

"We just don't want to take that chance," he said. His status in the deferred action for childhood arrivals program doesn't allow him to apply for citizenship.

Old patterns

Concerns about undercounting are nothing new, regardless of the citizenship question, Willrodt said. The Census Bureau estimated it missed about 2.1 percent of black people, 5 percent of Native Americans on reservations and 1.5 percent of Hispanic people in the 2010 count, which would amount to about 1,000 people in Northwest Arkansas. The bureau said non-Hispanic white people and homeowners were slightly overcounted.

People living in the Arkansas Delta region along the state's eastern border, an area that skews toward black and low-income residents, were among the least likely to respond to the mailed census surveys in 2010, according to data compiled by the City University of New York.

When people don't return the forms or leave questions blank, the Census Bureau sends workers to knock on doors and talk to neighbors.

As with Sanchez, Willrodt said history can play a role in this response pattern. Earlier in her career, Willrodt worked with a hospital in South Carolina trying to survey low-income people. Several of them brought up the federal government's 40-year Tuskegee Study, which deliberately withheld treatment for syphilis from hundreds of black men to study the disease's effects until the study was ended in 1972.

"I got to hear firsthand, 'Why should we help you? Why should we trust you?'" Willrodt said.

Northwest Arkansas' Marshall Islander population is heavily undercounted, based on research from the University of Arkansas for Medical Sciences and community leaders. Recent census surveys counted about 6,000 Pacific Islanders while the university pegs the number at roughly twice that, based on school enrollment and other data.

Many Marshallese adults don't speak English, but can travel and work in the U.S. without the need for visas under a 1986 treaty.

Benetick Maddison and his family left the islands in 2001, when he was 6. He said his family didn't even get the census form in 2010, and several extended family members also didn't get or didn't know how to fill it out. This time around, the Arkansas Coalition of Marshallese has begun training members to explain and help, and Maddison hopes to join them.

"I think we need to do a better job of reaching out to the community," said Maddison, who works as a community health worker at UAMS and is part of a project to gather data on local islanders from several nations. "We don't want to feel left out."

Willrodt and Bustamante agreed the best way to encourage participation is to involve local communities, including minority groups, in the census-gathering process. Kevin Fitzpatrick, a University of Arkansas sociology professor who has surveyed the local homeless population, said his work depended on building trust.

"We have to do a better job of developing and implementing a positive campaign to alleviate the fear that some folks have -- who see any count as threatening," he wrote in an email. "I think that the more engaged the local community is in these public campaigns to raise awareness and alleviate fear, the more likely we will see both a higher count and less under-representation among minority populations."

Arkansas's members of Congress recently introduced resolutions to honor the Marshallese community and call for better census data. A spokeswoman for Rep. Steve Womack of Rogers said the effort didn't conflict with his support of the citizenship question.

"He believes that it will better serve our communities and our country as a whole if we can accurately account for all populations -- including the Marshallese," spokeswoman Hannah Shea wrote in an email.

Dollars and sense

Local governments have millions of incentives to make the census as complete as possible.

More than $7 billion in federal money goes to Arkansas each year in programs that use census data, including Medicaid, highway programs, adoption and foster care, and vouchers that help low-income households afford housing, according to the George Washington University Institute of Public Policy.

At the local level, turnback money from Arkansas sales tax revenue also flows back to counties and cities based on population.

Cities often participate in the census in several ways. In recent decades, Bentonville, Fayetteville, Rogers and Springdale have held special censuses, which happen in between the decennial counts and can cost cities hundreds of thousands of dollars. The cost can be worth it if a city's population is growing so fast that a more up-to-date count brings in more money.

Fayetteville often provides mapping data to help the Census Bureau's work, chief financial officer Paul Becker said. The city will also likely reach out to students on the university campus to take part.

Springdale spokeswoman Melissa Reeves said officials plan to meet with the bureau in May and to help raise awareness when the count begins in 2020.

Willrodt said she's working with the governor's office for a state "complete count committee," which would form partnerships with groups like the Marshallese coalition, trusted organizations within undercounted communities.

"I have been traipsing around the state with my few facts and figures to get people aware of it," she said, emphasizing again individual census information is protected.

NW News on 04/22/2018