Why does a population number matter?
Take a place like Sterling in Logan County. Nestled into the northeast corner of Colorado, the small city serves as a hub for the agricultural communities surrounding it. It's got a Walmart, a few blocks worth of downtown, Northeastern Junior College and the state's largest prison, Sterling Correctional Center.
Now what if Sterling was given the wrong population number?
"It's millions of dollars," said Logan County Commissioner Jane Bauder. "And those millions of dollars fix roads, build senior centers, build childcare centers and mental health facilities. All of it! So it's really, really important that we are counted correctly."
It's not just government projects on the line. The city's nonprofits, businesses large and small, schools, housing and more rely on the population number to get money and make important decisions.
The population number Sterling got from the census in 2010 was 14,777 — closer to 12,000 if you don't include the prison inmates.
The Census Bureau counts inmates as living in the municipality where the prison is located. But last month, Gov. Polis signed a bill that will have the state gather data about inmates' demographics and last known addresses, using that information to draw Congressional and state legislature districts.
This means inmates living in the Sterling Correctional Facility will be included in Sterling's census counts, but for redistricting purposes, they will be counted based on their last known address.
In the years since the 2010 count, the Census Bureau estimates the city has lost about 1,000 people — the most of any municipality in the state.
Commissioner Bauder thinks those numbers aren't representative of the city's population.
"I personally don't think that the census got it wrong. I think that our community members did not respond and did not fill out a 10-minute survey," she said. "Which makes me disappointed in my fellow citizens."
She points to the percentage of people in Sterling who did not respond to the 2010 census as evidence of an undercount in the city and, by extension, the county. And while 32.7% seems like a large number of non-responses, that isn't too far off from the non-response rate in most states overall and the country as a whole. The census has ways of making up for those who leave them hanging.
But she's not alone in suspecting an undercount. The idea that Sterling's population is that low and bleeding residents baffles City Manager Donald Saling.
"If we lost a thousand I don't know where they're at. You know, my sales tax indicates that we're strong and healthy and viable," Saling said. "So I wouldn't say that we've lost a thousand."
He points to housing data, which shows that the rental vacancy rates for Sterling are extremely low and the number of occupied households has actually increased in the last decade, according to a housing report for the city and county from Williford, LLC., a land use and affordable housing consultant.
He also points to the number of actively running water taps in the city. But as certain as he is, the manager knows none of that is enough to definitively prove an undercount occured.
He estimates that the city's population is actually around 15,000 to 16,000 — including the prison.
People on Sterling's streets are a bit more scattered in their opinions. Some totally rejected the idea that the city has lost population, others thought it seemed reasonable. And their population guesses ranged from 11,000 to 40,000.
An undercount in Sterling is possible. By one estimate, 289,000 Coloradans were omitted from the 2010 census. Though, at the end of the day, the Census Bureau says the state was actually overcounted.
Both of these things can be true at the same time.
This happens through some complicated math, but in simpler terms: the number of people who are counted more than once is added to the total number of "imputations," or people who didn't fill out the census but are assumed to exist because of some other evidence like an occupied housing unit.
And then that total is subtracted by the estimated number of "omissions," or people who were totally missed. The result is the net undercount or overcount for a state, or the country.
"There were about 16 million people omitted from the 2010 census," said William O'Hare, a demographer who has been studying the census for decades and is part of the Count All Kids Committee. "There are about 10 million who were double counted and about 6 million who were imputed into the census. So the net undercount is near zero, but the omissions are about 5% or 16 million people."
This system keeps national and, to some extent, state counts rather accurate. But there's a problem: there's no guarantee the specific places and demographic groups (age, sex, race, Hispanic origin) with omissions are the same places and populations that get double counts and imputations.
For example, the Census Bureau concluded that Denver County was undercounted in the last census by more than 2,000 people. But an overcount of nearly 3,000 people in El Paso County more than made up for it.
This level of detailed information doesn't exist for Sterling or Logan County because the census only calculated overcount and undercount numbers for places or counties with at least 500,000 people.
So there is only specific information about four Colorado counties and one city — Denver. The rest are balled up into one big calculation, making it impossible to tell where overcounts and undercounts happened within the state's other 60 counties.
That means there's really only one option for Bauder and Saling in Sterling: get it right this time around.
"At least we can prove (it) this census and say, 'look, this is the information that came into the government,'" Saling said. "And of course we all know you gotta trust the government."
They are working alongside community organizations and a Complete Count Committee to give people opportunities to use tablets and computers to fill out the census at various community gathering places, like Walmart and the Sterling Public Library.
The city's first census information event was held in the library a few weeks ago. At the back of the building, past the bookcases and quilt-covered walls, a handful of people inside a fairly large meeting room listened to a presentation by a census representative for the area.
After it was over, Bauder stood in the back of the room, looking at the largely untouched piles of cookies and census fliers.
"It is boring. But at the same time it's so important," she said.
Bauder was clearly disappointed as the other members of the Complete Count Committee put away the 40 chairs nobody sat in. But this just was the first event. She believed the next one would be better attended by the general public and that the count will be more accurate this year.
"I'm confident that it's gonna be good," Bauder said.
What no one knew at the time was that this meeting would be the last census event the city would have before Gov. Polis declared a statewide stay-at-home order due to COVID-19. Now most community-based census efforts are on hold until the crisis is over.
But not all is lost, at least not for Dora Reyes, English Language Learner coordinator and teacher for the RE-1 Valley School District in Logan County.
"We have a lot of access to the families, to the parents and we have the good relationships we've built through the kids, through the students," she said. "So I know for sure that we'll be a part of it to make sure everybody is counted."
Even though in-person classes are not in session, Reyes is confident she can still encourage her students' parents to fill out the census remotely.
She's been at the district for about six years, and in that time she's seen the number of English second language students double. She's also seen the city's Hispanic community grow dramatically.
"ELL population in Colorado is pretty big and pretty varied with all the different dialects and all the different languages that are in Colorado," Reyes said. "But hopefully after this information we'll have more data to prove that we have more need for more support."
Investment in an accurate count of Hispanic people in Sterling runs in Dora's family. Her sister, Maria Ruiz, was working with the Family Resource Center, a local nonprofit, to help Spanish-speaking people fill out the census at the local Walmart. But, due to COVID-19, that isn't going to happen for a while.
"That's limiting our amount of time to actually be able to get face-to-face contact with somebody," Ruiz said. She believes she wouldn't be able to do much to encourage participation online, but is still trying with Facebook posts.
Both Commissioner Bauder and City Manager Saling believe a higher census count would help attract more businesses to Sterling, by giving them a better idea of the demographics they serve and the workforce they could tap into.
"So (for) economic development it's fairly major," Saling said.
Bauder worries the county has been getting skipped over by national food service and department store chains because it just narrowly missed population thresholds.
Census counts impact almost every facet of the city. Whether they believe Sterling was undercounted in 2010 or not, many agree it is really important to get the count right this time.
"I've been watching the census since 1970 and I am convinced beyond a shadow of a doubt that this is the most difficult census in my lifetime," said demographer William O'Hare.
He, along with many experts and the federal Government Accountability Office, have been sounding the alarm on issues with the 2020 census count for years.
And they all felt that way before COVID-19 started shutting everything down. The Census Bureau has extended the count by two weeks to make up for it, but O'Hare is still more worried than before.
41% of residents in Sterling have completed their surveys, according to the latest Census Bureau data. With a few months to go and no signs of the epidemic slowing, all the city can do is try to work around the virus to encourage participation. Then wait for the Census Bureau to give Sterling the population number that will define it for the next decade.
