SC Johnson Chair in Global Marketing; Professor of Marketing; Professor of Psychology, Weinberg College of Arts & Sciences (Courtesy)
Most of us have moments in our lives when we wonder what would have happened if we’d done something differently. A manager might kick themselves for botching a critical presentation: Would they have gotten a promotion if they’d prepared better? Or someone might have missed an investment opportunity that could have brought about a windfall, if only they’d made a different choice.
Researchers call these what-if scenarios “counterfactual thinking.” But counterfactual thinking is not limited to our personal or professional lives. In a recent study, Neal Roese, a professor of marketing at Kellogg, and his colleagues investigate how it plays out in the political realm: What do we think the world would be like if politicians had acted differently?
Given that the American left and right are becoming increasingly polarized over how to interpret facts, the team wondered if similar partisanship would color counterfactual thinking.
They suspected it might. After all, Roese says, facts are somewhat constrained by observation, but counterfactuals are “more open to the imagination.”
The researchers found that Democrats and Republicans were more likely to agree that a counterfactual scenario was plausible if it aligned with their political views. Partisan differences also emerged when they were asked to write their own what-if scenarios—for instance, what would have happened if Republicans hadn’t blocked Merrick Garland’s Supreme Court nomination or if the U.S. hadn’t rejoined the Paris climate agreement.
“Partisans differ not only in how they interpret facts, but also in how they construct scenarios of what might have happened,” Roese says.
Counterfactual thinking can go in either of two directions. The first is “upward”: how the outcome could have been better, if only something different had happened. For instance, if a student gets a C on a test at school, they might think that if they’d started studying earlier, they would have scored higher.
The second is “downward”: how things could have been worse. In that case, the student might think that if they hadn’t crammed the night before, they would have failed the test.
In general, research has shown that people’s counterfactual thinking tends to lean upward. They’re more likely to ruminate about how they could have achieved more of their goals and dreams if they’d made different choices, or if obstacles hadn’t gotten in the way. A key reason behind this pattern is that counterfactual thinking is usually connected to goal pursuit. Counterfactual thoughts, when they focus on one’s own actions, aim toward an understanding of how best to avoid failure and move toward better states of being. The more optimistic a person is, the more they focus on actions they might have taken to have achieved a better outcome, which connects to the expectation of better outcomes down the road.
To find out if this general tendency toward upward counterfactual thinking holds true as people ponder political issues, Roese and his colleagues, Kai Epstude
at the University of Groningen and Daniel Effron at London Business School, surveyed both Democrats and Republicans through the online research platform Prolific. Each person was presented with descriptions of six political topics of recent interest, such as taxes or immigration.
This research shows that . . . partisanship plays an important role in how people use counterfactuals to bolster their blame for political foes.
For each topic, participants read a counterfactual scenario—sometimes upward, sometimes downward—and rated each scenario’s plausibility on a scale of 1 to 5. The upward scenarios all favored Democrats’ views, while the downward scenarios favored Republicans’ views.
For example, on the topic of the COVID-19 pandemic, the upward Democrat-leaning counterfactual was “If the Trump administration had worked directly with more doctors and scientists, then the current COVID situation would be a whole lot better.” The downward Republican-leaning counterfactual, on the other hand, was “If the Trump administration had not launched Operation Warp Speed, then the COVID situation would be a whole lot worse.”
The team then repeated the experiment with another set of participants but reversed the political alignment of each scenario. For instance, this time the Republican-leaning counterfactual was upward: “If Biden were not president, then the current COVID situation would be a whole lot better.” And the Democrat-leaning counterfactual was downward: “If Biden were not president, then the current COVID situation would be a whole lot worse.”
Participants tended to rate scenarios that aligned with their views as more plausible, regardless of whether they were upward or downward. For instance, in the first half of the study, Democrats rated the left-leaning upward scenarios as more plausible than the right-leaning downward scenarios. In the second half of the study, they rated the right-leaning upward scenarios lower than the left-leaning downward scenarios.
While this may not feel particularly surprising in today’s hyper-partisan atmosphere, the results did contradict the long-held finding in previous research that counterfactual thinking tends to be upward. “In the political sphere, that general pattern goes out the window,” Roese says. The research shows that the realm of the hypothetical is fair game for partisan politics.
The researchers saw similar results when they asked participants to write their own counterfactual scenarios, as opposed to evaluating counterfactuals provided to them. That is, participants responded to prompts such as “If Biden had not adopted a harsher tone with Russia, then …” or “If Republicans had been able to pass the tax cuts earlier than 2017, then …” Participants could use their own words, and their own imagination, to describe how the situation might have been better or worse.
Again, participants didn’t automatically gravitate toward upward thinking. The direction of their written scenarios tended to align with their political views.
The team then investigated a subtler question. They wanted to examine reactions to potentially disastrous scenarios that people might believe had come close to happening, such as war with North Korea or a massive cyberattack on the United States.
In their personal lives, people tend to cast more blame when a bad outcome is only narrowly avoided. For instance, imagine that your spouse delays you with a question on your way out the door to the airport, and you nearly miss your flight. If you make it to the gate with only 30 seconds to spare, you’d likely blame your spouse more than if you arrive a full ten minutes before the doors close.
The researchers wanted to explore two related questions: Did the perceived “closeness” of a hypothetical catastrophe affect how much people blamed the political leader in charge? And was the amount of blame influenced by partisanship?
They presented a new set of participants with several political scenarios. For example, on the topic of cybersecurity, they explained that the military had increased defenses against cyberattacks under President Biden, but that some critics argued that those efforts were insufficient, and the U.S. could have been debilitated by an attack on millions of people’s electricity and bank accounts.
Then the researchers asked people to rate how close they thought the disaster came to happening and how much they blamed the president for nearly letting it happen.
The team found that the closer people thought a catastrophe came to happening, the more they blamed the president. This was true regardless of whether the president was from their favored political party. But partisanship also affected how much blame participants placed on the president, with participants tending to put more blame on leaders from a non-preferred political party.
This research shows that counterfactuals and blame go hand in hand, Roese says, but also that partisanship plays an important role in how people use counterfactuals to bolster their blame for political foes.
Some observers have claimed that those on the left are more biased than those on the right; others claim the reverse. An intriguing question, then, is whether Democrats or Republicans are more likely to use counterfactuals to back up their partisan views.
In all of the studies, the researchers were careful to sample equivalent numbers of participants from either side of the partisan spectrum to try to answer this question. Interestingly, no evidence of unique bias on the left or right was found. Overall, Roese explains, Democrats and Republicans were similar in their use of counterfactual thinking to support their preferred political viewpoint.
Roese says that while the worst part of the polarized political atmosphere in the U.S. is that “we can’t always agree on the basic facts,” this research suggests that the divide goes further. “We can’t agree on the counter-facts either,” he says.
Roberta Kwok is a freelance science writer in Kirkland, Washington.
Epstude, Kai, Daniel A. Effron, and Neal J. Roese. 2022. “Polarized Imagination: Partisanship Influences the Direction and Consequences of Counterfactual Thinking.” Philosophical Transactions of the Royal Society B. 377: 20210342.