Native Americans are tired of being lumped in as ‘something else’ in polls, other surveys

On election night, Jodi Owings and her family watched the results reported live on television in their Oklahoma home.

She noticed a CNN graphic that displayed returns by race as white, Latino, Black, “something else” and Asian.

Is “something else” us, Owings, a citizen of the Seminole Nation of Oklahoma, asked her family.

Shortly after CNN aired its graphic, Native Americans took to social media and displayed family photos denoting their heritage and shared T-shirts and face masks with the hashtag #somethingelse. Others found humor in the phrase, replacing the word “Indigenous” with “something else” in memes.

Stephanie Fryberg, a University of Michigan psychology professor who has surveyed Native American populations and is a citizen of the Tulalip Tribes of Washington, and Crystal Echo Hawk, who is Pawnee and the founder of one of IllumiNative, helped push out a “Something Else Survey” to gather data on how people voted, get-the-vote-out efforts and expectations under Biden’s administration. More than 3,600 people responded within a week. Results showed they overwhelmingly supported Joe Biden, Fryberg said.

The survey also asked whether people felt “angry,” “invisible” or “humored” by the network TV exit poll, which also was used by other networks in addition to CNN.

“I think, for non-Native people who have not had to face the things that we’ve had [to], they did not see the graphic like we did,” Owings said.

Hope Craig, a citizen of the Muscogee (Creek) Nation of Oklahoma who lives in Valencia, California, said she was offended and “tired of being excluded, of being invisible, not being seen.”

It’s unclear whether Native Americans were surveyed in the exit poll used by CNN and other networks and who else was included in the “something else” category. The research manager for Edison Research, the company that conducts the poll, did not provide details about the survey.

CNN later removed the category from the graphic.

“Our exit poll results included a poor choice of words and in no way did we intend to minimize the importance of Indigenous communities and the Native American vote,” CNN spokeswoman Alison Rudnick said.

The issues the exit poll raises reflect broader questions about how Native Americans — who make up less than 2% of the U.S. population — are reflected in survey data, which often list Native Americans in datasets as “other” or with an asterisk.

Even when surveyed, the results can be considered statistically insignificant because the sample size isn’t large enough or the margin of error is too great to accurately reflect the population.

The National Congress of American Indians has said there’s a critical need for data that is accurate, meaningful and timely within tribal communities. The data limitations have an impact on everything from community planning and resource allocation to policy development.

Heather Shotton, co-editor of the book “Beyond the Asterisk,” said missing data and incomplete representations of Native Americans have been a part of U.S. history. Shotton, a citizen of Wichita and Affiliated Tribes in Oklahoma, said the “something else” label is a lazy way to represent groups that weren’t included in the graphic.

“It speaks to how Indigenous people remain outside of the consciousness of everyday Americans and exist either in an elsewhere or in the past,” she said.

In contrast to what the “Something Else Survey” found, The Associated Press’s nationwide “AP VoteCast” survey of more than 133,000 voters and non-voters showed 52% of Native Americans supported Donald Trump and 45% favored Biden — a statistic that was largely discounted in Indian Country because Native Americans long have been considered the Democratic Party’s constituency. Emily Swanson, the AP’s director of public opinion research, said the margin of error was too great to say which way the population swung.

The survey allowed participants to self-identify as Native Americans and was conducted primarily by mail and online — methods that present challenges on tribal reservations that lack broadband access and don’t have home delivery for mail.

“There’s a real concern that not the entirety of that population is represented,” Swanson said.

University of Michigan

Fryberg said better survey methods are needed to accurately measure views in Indian Country and to capture the diversity within it. Random sampling doesn’t work well, she said, particularly in reaching tribal members who live on reservations or rural areas.

She and others have had success getting people to respond to surveys by forming partnerships with Native organizations, tribes, tribal colleges and Native news outlets.

“I think we are showing there are ways to get really good information from our population,” said Echo Hawk, whose IllumiNative is among the partners. “It’s not impossible to do.”

Scholars also have pushed to get tribes more involved in how data are collected and applied. Some tribes require researchers to get permission from a review board or tribal council before gathering data on their land, said United Tribes Technical College President Leander R. McDonald, a citizen of the Spirit Lake Nation in North Dakota.

In urban areas, McDonald said there might not be enough people who identify as Native American where a survey is conducted, so researchers “might end up with five out of 100 people” — a number too small to report to draw conclusions regarding an entire population.

The coronavirus pandemic has dealt a setback to efforts to count people on Native American reservations. The once-a-decade U.S. Census determines how much federal money is allocated for schools, health care and other needs.

And not all states have kept data on how COVID-19 has been affecting Native Americans and other communities of color.

Some tribes have argued in court that they didn’t get what they deserved in coronavirus relief when the Treasury Department turned to federal housing data to distribute a portion of the $8 billion due to tribes based on population. The federal data conflicted with tribes’ own enrollment figures.

Crime statistics, though not collected by polling or surveys, are another area where limited data hurts Native Americans.

For example, no one knows how many Native American women and children are missing or have been killed in the United States.

Some tribes over the past few years have worked to collect information before state and federal governments assembled task forces aimed at finding the missing. Some states are engaged in pilot projects to better coordinate state and tribal efforts.

Elaine Thompson / AP