On undercounting Native Americans (and other rural folks) in the 2020 Census

Isolated home in  San Juan County, Utah, near four-corners area
(c) 2008 by Lisa R. Pruitt

Kurtis Lee and Ben Welsh report for the Los Angeles Times from Crownpoint, New Mexico, (population 2,278) part of the Navajo nation, on the challenge of accurately counting Native Americans in the U.S. Census, which is coming up in less than a year.  The story touches in particular on the challenges of remoteness and lack of trust.  It begins with an illustration of remoteness that features Leonard Jones, a Navajo man who doesn't recall being reached by Census workers in 2010:

Only family and close friends make the dusty 10-mile trek from the paved road, down dirt switchbacks lined by sandstone mesas, to his secluded home in northwestern New Mexico. There is no electricity, no running water, in the single-level sandstone structure. 

Lee and Welsch quotes Jones:

Few people know we’re out here.  We live in nature. 

The thought of people coming out here and making us a part of any official count seems like a stretch, you know?

* * *

Who’s going to make it here? No one knows we’re here. I just hope we’re not forgotten.

Here are some other excerpts:

Last month, Census Bureau officials visited New Mexico to meet with state and local officials and tribal leaders. The group traveled to homes near Albuquerque and heard firsthand testimony about the challenges of counting individuals on the Navajo Nation and other rural locales.

* * *

While [ ] reservations [other than the Navajo] are smaller, most are also remote. And all are home to a longstanding distrust of the U.S. government. Those factors help make Native American reservations among the most difficult places to canvass during the census, the once-per-decade federal effort to find and tally every resident of the U.S.

An audit of the 2010 Census showed that 1 in 7 Native Americans living on a reservation did not get counted. That's a total of about 82,000 people, which is equal to the entire city of Santa Fe.
Here is the current U.S. Census Bureau page on Crownpoint.  Crownpoint is in McKinley County, population 71,492.  The poverty rate of McKinley County is an astonishing 37.5%, which is actually consistent with poverty rates in Indian Country, including in Arizona and Utah, where Navajo territory also stretches.  Some 175,000 Navajo live in an area larger than West Virginia. 

Other key data points in the story illustrate what is at stake for New Mexico in getting an accurate count:

New Mexico’s budget coffers depend on nearly $7.8 billion a year from Washington for programs like Medicaid, food stamps and road repairs tied to the census. The New Mexico Legislature recently endowed a new state commission with $3.5 million to spend on ensuring an accurate count.

Montezuma Creek, Utah
(c) 2008 Lisa R. Pruitt

Seth Damon, speaking for the Navajo Tribal Counsel, is quoted:

For the Navajo Nation and Indian Country, the census determines whether your dirt roads get graveled or paved, or whether your people move from dirt floors to a solid foundation.

That's a powerful illustration (at least to someone who grew up on a road that got paved when I was in elementary school; many of my high school friends continued to live on dirt roads).

A map accompanying the story shows in orange all of the places (counties or, in many cases, parts of counties) that are at high risk for undercounting.  Those who know where Native American land is in the west and southwest can see readily the concern for those populations, but the map also shows a lot of other "at risk" places that are rural without also being Native American.  Note for instance my home county in Arkansas, Newton County, a persistent poverty county in the Ozarks.  The population of the county is 8,330, but the bubble that pops up when the cursor hovers over the map accompanying this story indicates that about half of that population is at risk of being undercounted.

At least I take some heart in knowing that some powerful institutions, e.g., Urban Institute, individual states, have an eye on the process.  I guess this is an example (sorta') of what Derrick Bell called interest convergence.  That is, if a state is interested enough in getting the federal funds (however it ultimately spends them), that it invests money (in the case of New Mexico, $3.5 million) to get an accurate count, we can hope that many of these funds will ultimately serve rural interests.