Flexible Morals: A Key Reason American Voters Support Divisive Misinformation

Over the last decade or so, concern has grown over the tendency for Americans to support their favorite politicians even when those politicians share misinformation. A common assumption about the problem is that partisan voters are apt to believe what they should question (and vice versa). And research backs up the idea that voters are “factually flexible,” either due to laziness or bias.

But what if factual flexibility isn't the whole story? What if a key part of the story is that partisan voters are also “morally flexible” — that they hold opposition politicians to strict standards of factuality but allow their favorite politicians to share misinformation — even socially divisive misinformation. For morally flexible voters, such statements are permissible because they articulate a “deeper truth” that captures their grievances.

This is the message of a new study forthcoming in the American Journal of Sociology, “When Truth Trumps Facts: Studies on Partisan Moral Flexibility in American Politics,” written by Minjae Kim, assistant professor of organizational behavior at Rice Business (Jones), and Ezra W. Zuckerman Sivan, professor of strategy and entrepreneurship at MIT Sloan School of Management.

Along with Oliver Hahl from Carnegie Mellon University Tepper School of Business and Ethan Poskanzer from University of Colorado-Boulder Leeds School of Business, Kim and Zuckerman Sivan found that when American voters evaluate statements made by politicians they support, they tend to be less concerned about whether those statements are based on objective evidence. What matters more is the “deeper truth,” or the overarching message, the statements express.

According to Kim, “People insist on strict factuality when it comes to politicians they don’t favor. They don’t give the opposite partisans the same leeway, in part because they don’t like their message. People shift their standards as it suits their partisan interests.”

To arrive at their findings, the researchers conducted a series of six online surveys of American voters: five during the last two and half years of Republican President Donald Trump’s administration, and one in the spring of 2023, during Democratic President Joe Biden’s administration.

Five of the surveys asked respondents to evaluate divisive, fact-flouting statements made by Trump. Two compared responses to similar statements by a Republican politician (either Trump or Florida Gov. Ron DeSantis) and a Democratic politician (either U.S. Rep. Alexandria Ocasio-Cortez, D-N.Y. or Biden). The statements pertained to such hot-button political issues as immigration, the 2020 Black Lives Matter protests, the COVID-19 pandemic, and Trump’s “Big Lie” that the 2020 U.S. presidential election was “rigged” or “stolen.”

“There is no doubt that partisans are factually flexible,” said Zuckerman Sivan. “Supporters of a politician are much more likely to say that one of these statements is based on objective evidence than are opponents of that politician. But we also found consistent evidence of partisan moral flexibility — that voters distinguish between ‘objective evidence’ and ‘truth’ — and that when you consider who is morally flexible, you’re better able to predict who supports divisive misinformation than if you focus on factual flexibility alone.”

The researchers found that voters care more about truth when evaluating favored politicians. But they care more about facts when evaluating disfavored ones. This effect is slightly stronger for Republican voters, but it applies to Democrats, too.

Participants in each study were shown a divisive statement made by a politician, along with a note that clarified that the statement had been verified as non-factual by a third-party fact-checking organization. For each statement, participants indicated whether they thought the statement was based on objective evidence or subjective impressions; whether the statement was “true” and whether it was more important for the statement to be based on objective evidence or “to send the right message about American priorities.”

For example, in a survey called the “Caravan,” the researchers showed participants a real post from Trump that was not factual — that there were criminals and unknown Middle Easterners mixed in “the Caravan heading to the Southern Border of the United States” that “Mexico’s Police and Military are unable to stop.”

Across the five studies that included such statements, Trump supporters were more likely than non-supporters to say that the statement was based on “objective evidence,” giving an average rating of 3.65 based on a 7-point scale. This was 0.9 points higher than the average given by non-supporters. But while this gap was large, the gap grew to three times larger when respondents rated the truth of these statements. Across the five studies, Trump supporters gave an average rating of 5.2 for truth. Non-supporters averaged 2.2.

A similar gap was found in responses to a statement from Biden that stated erroneously that the COVID vaccines were effective in stopping “the spread of disease to anyone else.” In a follow-on question, one Biden voter explained their rationale for affirming Biden’s false statement:

In a public health crisis such as the COVID-19 pandemic, it was more important for President Biden to appeal to American values of patriotism and the willingness to step up for others. Most people are aware that while vaccinations greatly reduce the spread of disease, there is no vaccine that can completely, utterly stop it. However, Biden was using strong, emotional, positive language to encourage Americans to do what was morally correct and patriotic at that moment, and I feel that was entirely appropriate.

This emphasis on truth over facts gave the researchers’ article its name. But the more direct indicator of partisan moral flexibility was the preferred standard for evaluating these statements. “In short, when the statement is by a Republican, Democrats insist the statement should be evaluated on its factuality whereas Republicans say that what matters is whether it conveys an important message. And the reverse is true when the statement is by a Democrat,” Kim said.

For instance, in the “Summer of 2020” survey, participants were asked to respond to a post by Ocasio-Cortez. Like the Trump post, this statement included a line that was highly misleading: “Police brutality is now a leading cause of death for young men across the board in the US.” Democratic respondents were much more likely than Republicans to say that the statement should be evaluated on the basis of whether it sends “the right message about American priorities” than on whether it is based on “objective evidence” (3.53 on a 7-point scale for Republicans; 2.13 for Democrats).

“The study illustrates how both Democrats and Republicans shift their standards to suit their interests,” Zuckerman Sivan said. “It seems that we all do it.”

Kim and Zuckerman Sivan believe it’s important for people to recognize that moral flexibility plays a significant part in our decisions and behaviors, because coming to terms with that could ultimately lead to a different set of interventions — though exactly what those interventions might be is still unclear.

“But let’s at least get the diagnosis right before we start talking about the treatment,” Zuckerman Sivan concluded. “Our hope is that if people come to recognize the extent of partisan flexibility, then we’ll be able to start making some progress from there.”