Native groups: Exit polls on Native voter preferences were flawed

FILE - Voter information is on display at a Phoenix Indian Center booth during an Indigenous Peoples' Day event in Phoenix, Arizona, Oct. 14, 2024.

In the days following the November 5 election, media outlets widely cited NBC exit polls indicating that 65% of Native American voters cast their ballots for Donald Trump.

While Native American advocacy groups acknowledge a trend of Native voters shifting toward the political right, they argued that the polling in this case was flawed and did not accurately represent their demographic.

NBC did not conduct the polling itself. It is one of four major news networks in the National Election Pool, or NEP, which relies on marketing research company Edison Research to question voters as they leave the polls.

Edison conducted phone, email and text message surveys of absentee and early voters in Georgia, North Carolina, Nevada and Ohio before Election Day and polled exiting voters at more than 300 polling stations across the U.S.

FILE - A woman walks near a polling station on the day of the 2024 U.S. presidential election in Raleigh, North Carolina, Nov. 5, 2024.

Once polling stations shut down on election night, Edison forwarded the voting data to NEP members so that they could analyze answers and project the winners.

NBC was the first to report exit polling data that included Native Americans. ABC News lumped Native Americans into the "all other races" category, as did CBS News and CNN.

The 65% figure prompted skepticism and confusion among some Native American observers. Native News Online, working with Northwestern University's Medill School of Journalism and national survey firm Qualtrics, surveyed 865 Native voters and found that 51% of Native voters voted for Donald Trump.

In contrast, 60% of the nearly 5,000 Native Americans who participated in Illuminative's Indigenous Futures Survey in 2020 identified as liberal, and 51% said they were Democrats.

A sign marks the Turtle Mountain Reservation, home of the Turtle Mountain Band of Chippewa Indians, in Belcourt, North Dakota. Native groups complain that national exit polls were not conducted on any tribal lands on Election Day, Nov. 5, 2024.

A closer look at the numbers

Stephanie Fryberg (Tulalip), who leads Northwestern University's Research for Indigenous Social Action and Equity collaborative, was puzzled by the NBC calculation.

In a November 8 editorial in Native News Online, she said that the Native voter sample size was too small and couldn't accurately reflect Native voter preferences nationwide. She pointed out that 80% of respondents were from urban and suburban areas, while less than 20% came from rural areas.

In a separate editorial on November 18, Native News Online editor Levi Rickert questioned whether the respondents were legitimate members of federally recognized tribes.

“Perhaps the most challenging aspect of research and data collection among Native Americans is self-identification,” he wrote. “For various reasons, many people claim Native American ancestry. Among Native Americans, a common joke is that the largest 'tribe' in Indian Country is the 'Wannabe' tribe.”

VOA reached out to Fryberg to ask how she arrived at the numbers.

"We wanted to better understand the sample of Native voters that major news organizations were using to draw broad conclusions about voting patterns in the 2024 election," she answered by email. "Reports from the NEP show that only about 1% of these respondents identified as Native American, equating to roughly 229 individuals … underscoring the urgent need for more comprehensive and inclusive sampling strategies that genuinely reflect Native voices in electoral data."

FILE - This combination photo shows Republican presidential nominee and former President Donald Trump, Aug. 15, 2024, in Bedminster, New Jersey, and Democratic presidential nominee and Vice President Kamala Harris at a campaign event in Raleigh, North Carolina, Aug. 16, 2024.

Edison Research Executive Vice President Rob Farbman agreed that the sample size was small, but he told VOA that it met the NEP's minimum criteria for reporting subgroups.

"This is a national survey meant to represent the country, but subgroups as small as Native Americans … are difficult to measure," he said. "It's certainly possible that our survey is underrepresenting people that live on reservations."

He also noted that results at the county level are still coming in.

"And the votes that we've been seeing so far are showing that Trump is doing way better among American Indians than last time,” Farbman said.

Allison Neswood, a staff attorney at the Native American Rights Fund, said that exit polls should be viewed "with a lot of skepticism," and that the Edison poll "should really be discarded."

"Real hard conclusions will take a little bit more time," she said. "We're going to have to get more granular data, below the county level, to the precinct level."