Colorado Finally Has Some Local Census Data After A Months-Long Delay. Here's What Happens Next
After months of waiting, Colorado finally got the local Census population counts needed to draw congressional and state general assembly electoral districts on Thursday. The massive delay has caused a lot of unexpected problems for the state’s brand new independent redistricting commissions.
This new set of local data, released by the Census Bureau in a press conference Thursday, show the ways Colorado’s population has changed over the last decade.
“On average, smaller counties tended to lose population and more populous counties tended to grow,” census senior demographer Marc Perry said, noting a national trend that Colorado’s state demographer Elizabeth Garner has drawn attention to locally in the past. “Population decline is even more widespread this decade, with 52% of all counties having smaller populations in 2020 than in 2010.”
Highlights from Thursday’s data release include:
- Weld and Broomfield counties have seen the highest rate of population growth since 2010. Greeley is the fourth fastest-growing metro area in the nation.
- Denver is one of only 14 cities in the nation with a population growth of over 100,000 people since 2010.
- Denver, Adams, Morgan and Weld counties all have Hispanic/Latino populations around or above 30%. In general, Colorado’s diversity has increased along with the nation's. The state’s non-Hispanic white population is now at about 65%, down from about 70% in 2010.
- Counties like Logan, Kit Carson and Bent have lost over 5% of their populations since 2010.
The data was made available through several interactive maps and charts. Learn more on the Census Bureau’s website.
But more detailed information won’t be easily accessible in this release because it’s in a format that requires time, specialized experience and software to process.
The Bureau needed more time to double-check its counts due to the pandemic, which would have been fine if states didn’t need the data to draw new districts months ago. Recognizing that, and to get the information out faster, the Bureau put the data in this “legacy format” that states have to process themselves. A more publicly accessible version of these numbers will come out by the end of September.
Can Colorado’s redistricting commissions make up for lost time?
In some ways, they already have: using population estimates to draw preliminary maps allowed both commissions to hit the road to gather public comments in meetings across the state a bit sooner than they would have waiting for this release.
Still, much time was lost and using estimates didn’t make up enough of it. Commission staff managing attorney Jerimiah Barry told congressional commissioners in mid-July that “there’s no way staff will be able to prepare a plan to have the commission approve it by Sept. 1. It’ll be physically impossible” because of the delayed full count release.
Since the commission’s deadlines and rules are enshrined in the state constitution, commissioners voted to ask the state supreme court for help. The court approved extending the deadline to October. This will not only give staff more time to integrate the new data into the maps, it may also allow commissioners to take the new maps they create from this data on a second, much shorter public comment tour.
Counts versus estimates
The census count’s role in district mapmaking is enshrined not only in Colorado’s constitution, but in many other states and federal law. It’s a primary reason the count exists.
Trying to count every individual in a community will yield different (and, in theory, better) results than surveying a portion and doing a bunch of complicated math to extrapolate it to the whole community. And staff constantly recognized this issue while presenting their preliminary maps for the first time earlier this summer.
In the state House map, districts are so small in some areas that even the tiniest differences or errors between the estimates and the count can change district lines.
Observers hope the extra time the Bureau spent reviewing the data will reduce the inaccuracies at the local and demographic levels they’ve been worried about.
“We have conducted one of the most comprehensive reviews in recent census history,” said Census Bureau acting director Ron Jarmin during Thursday’s release. “The data we are releasing today meet our high data quality standards.”
However, there is still a lot of doubt about the count’s accuracy at the local and demographic level for several complex reasons, including the new way the Bureau scrambles data to ensure privacy, a process called “differential privacy.”
“Just as with protections we used in the past, differential privacy protects personal information by adding noise or fuzziness to the data,” Jarmin said. “This noise is carefully calibrated following numerous consultations with stakeholders to protect the data at the most granular level, but ensure accuracy across larger geographies and groups.”
The Bureau likely won’t have a full assessment of how much it over- or undercounted certain areas or demographic groups in this count until 2022, Jarmin said. Whether it is accurate or not, it’s the only data Colorado’s final district maps can use.
Moving incarcerated populations around
Concerns about the Bureau's differential privacy process have bled into the redistricting commissions’ internal back-and-forth about whether they should follow a 2020 state law directing them to count incarcerated people as living at the address of their last known residence when drawing districts.
A state Supreme Court decision on another redistricting bill a few months ago rebuked the legislature for trying to get involved in the process. The commissions argued they were created by voters to almost totally remove any legislative influence on redistricting.
That decision emboldened commissioners to vote to potentially ignore the prison population reallocation law, HB20-1010.
Proponents of adjusting prison populations say it’s fairer because it avoids giving small communities with big prisons unfair political power through a population that can’t vote and would likely return to their previous communities if released.
“People in prison are paying a debt to society,” 10 state civil rights groups wrote in a letter to the commissions in support of reallocation, noting that the state’s prison populations are disproportionately non-white. “However, a collateral consequence of that sentence should not be to provide an unearned political benefit to more rural communities at the expense of more urban communities just because the state decided to build prisons in one particular part of the state rather than another. That’s an arbitrary rationale for designation of residency for purposes of redistricting.”
Some commissioners and staff were initially opposed or concerned about the reallocation law for a few reasons: it came from the legislature, it would take extra time and worries about how accurately they can reallocate those incarcerated individuals.
A legislative redistricting commission memo lays out the pros and cons of reallocation ahead of the commissioners’ vote on Friday. In its list of points against, the document notes that incarcerated individuals' former addresses may be inaccurate if not properly collected by correctional officials and “the introduction of noise in the 2020 census can cause distortions and additional complications in small populations such as prison populations.”
Both commissions will plan to vote on the issue by the end of the week.