For six days a week throughout August and September, Cindy Hoedel went door-to-door across mostly rural Kansas counties — sometimes covering over 200 miles in a day — to find out how many people lived in a given address on April 1.
But as a Census worker, Hoedel’s biggest challenge wasn’t in the rural areas she covered — it was student apartments in Emporia, Kansas.
“It became a lot of, kind of, detective work in looking for people who might know something about four months ago, which is like a dynamic that this Census is not set up for,” she said. “It’s never been that way before.”
The obstacles Hoedel faced reflects just one of the difficulties faced by the Census Bureau in gaining an accurate count of every man, woman and child in the U.S. — all in the middle of an ever-worsening pandemic. In addition to COVID-19 impacting traditional complete count efforts in Kansas City and surrounding areas, abruptly changing deadlines caused by legal battles to complete the Census and a new online format has caused worry among local groups over incomplete data that may not accurately reflect diverse communities.
But when combining self-response rates with the work of Census workers, the estimated percentage of households counted by the Census was 99.9%, both nationally and in Kansas and Missouri, according to government data.
That 99% is split into two categories: Those who voluntarily responded to the Census before the deadline — either by phone, mail or online — and the households counted by Census workers like Hoedel. Kansas’ final Census response rate was 69.8%, slightly less than the 2010 final self-response rate of 70%, but still higher than the national self-response rate of 67%.
Missouri fared a little worse, with a final self-response rate of 65.9%, which is lower than the 2010 self-response rate of 67.5%. Kansas City, Missouri has an even lower self-response rate of 60.7%.
“That leaves us in a worse place than we were 10 years ago. Because 10 years ago, we finished up at a 65% self-response rate,” said Jenny Garmon, legal and government information specialist at the Kansas City Public Library.
Getting a complete count
As a Census worker, Hoedel was primarily sent to addresses in Lyon, Chase, Greenwood, Marion, Butler and Sedgwick counties to gather demographic data on households who did not already self-respond to the Census.
Part of the difficulty Hoedel faced in student apartments was figuring out how many students lived in a given apartment on April 1. As the pandemic caused universities to end on-campus classes, many students went back to their hometowns.
“Now fast forward, you’ve got the pandemic. The school quits having classes, everybody leaves, everyone goes home, and then they try to open up again for the Fall, and you have completely different people,” Hoedel said.
Anytime a household or apartment refused to answer her questions or weren’t present, Hoedel had to find other ways to find answers.
Getting an accurate count was easier in rural areas because there’s traditionally less turnover in who lives at an address, she said. In most cases, a neighbor who is familiar with the community could often provide information on a household if no one was there or refused to answer questions.
“You’d be surprised how much your neighbor will tell me,” she said.
In lieu of neighbors to speak with, Marlene Nagel, director of community development at the Mid-America Regional Council, said the Census Bureau also uses national commercial databases to count people. But Garmon said those records only help if people have things like Social Security numbers and file income taxes.
The Regional Complete Count Committee in the Kansas City metro, which has been working to spread awareness about the importance of the census, included government groups, faith-based organizations, emergency assistance agencies, business groups, neighborhood groups and more.
In spite of the education and these organizations’ ability to reach far and deep into communities, self-reporting didn’t go as planned in many neighborhoods.
“It’s been a difficult process all along, and COVID made it that much harder,” Nagel said. “The constantly moving target of the end date made matters worse. There were many organizations that were working toward that (Oct. 31) deadline, then it got changed and got changed again. It made the whole process both difficult and confusing.”
Erin Royals, neighborhood outreach and research coordinator for Center for Neighborhoods KC, a program with the University of Missouri-Kansas City focusing on neighborhood revitalization, said the changing deadlines to complete the Census largely impacted the work of Census workers.
“They do a lot of important work of going door to door, talking to people, getting a sense of community challenges,” Royals said. “You can’t do that work the way it needs to be done when the deadlines keep changing.”
Additionally, in 2020, the Census was almost entirely online for the first time rather than through the mail. That added even more groups to the populations at risk of being undercounted, like those with limited or no computer or internet access and those who are suspicious of recording personal information online. When COVID-19 forced library closures in the spring, computer availability became even more limited.
Garmon said she and others who participated in the Regional Complete Count Committee supported by the Mid-America Regional Council anticipated hiccups that the online format might bring, and the Greater Kansas City Community Foundation gave the group funding for laptops and hotspots.
What’s at stake
The U.S. Constitution mandates that the federal government count every person in the nation every 10 years — starting in 1790. The count this year will be used to help allocate $1.5 trillion in federal funds sent to communities for services like Medicaid and school lunch programs, as well as to determine representation in Congress and state legislatures. Additionally, public and private investments for locating and funding community centers and even grocery stores rely on information collected by the census.
“That would make the difference in the quality of life for our whole city, but really for the people who are at the most risk for being undercounted,” Garmon said.
She said those who are at risk include renters, children ages 5 and under, those with low incomes, and those with limited English proficiency. These are the same people who are at risk in other Census years, so this concern isn’t new.
During the 2010 census, Katina Jones worked for the Missouri Secretary of State’s office as the state data coordinator. She knew children 5 and under had not been counted accurately, and says those children are now 10 to 15 years old and don’t have the quality of life they could have had if the count had been more accurate — for instance, more funding for programs at their schools.
This census, Jones worked at the Mid-Continent Public Library as a statistical research analyst.
“If we don’t count everyone in the neighborhood, then that neighborhood doesn’t get the benefit of funding for schools or the fire department or roads,” she explained.
Census self-response rates also vary by neighborhood. In areas east of Troost Avenue — often considered an economic and racial dividing line in Kansas City — self-response rates hovered around 50%. The historic northeast saw even lower self-response rates ranging from about 30% to less than 50%.
But the Census is still not over — the Census Bureau will now spend the next two months to check the data of about 330 million people nationwide. In previous years, the bureau had six months to verify and ensure the accuracy of the data.
“If you don’t have an accurate count, then that data is not very accurate,” Royals said.
“It’s not white people who are being undercounted, it’s black and brown folks. And so when they’re not counted, when you go to map that data, they’re literally not appearing on a map. … If you’re not appearing on the map, it’s like you’re not there.”