Empathy Is Tearing Us Apart

Americans are as polarized as they've ever been. Could the problem be that we're caring for each other too much?
donald trump standing in front of an american flag
Photograph: SAUL LOEB/AFP/Getty Images

There are people who believe that the political polarization now afflicting the United States might finally start to subside if Americans of both parties could somehow become more empathetic. If you’re one of these people, the American Political Science Review has sobering news for you.

Last week APSR—one of the alpha journals in political science—published a study which found that “empathic concern does not reduce partisan animosity in the electorate and in some respects even exacerbates it.”

The study had two parts. In the first part, Americans who scored high on an empathy scale showed higher levels of “affective polarization”—defined as the difference between the favorability rating they gave their political party and the rating they gave the opposing party. In the second part, undergraduates were shown a news story about a controversial speaker from the opposing party visiting a college campus. Students who had scored higher on the empathy scale were more likely to applaud efforts to deny the speaker a platform.

It gets worse. These high-empathy students were also more likely to be amused by reports that students protesting the speech had injured a bystander sympathetic to the speaker. That’s right: According to this study, people prone to empathy are prone to schadenfreude.

This study is urgently important—though not because it’s a paradigm shifter, shedding radically new light on our predicament. As the authors note, their findings are in many ways consistent with conclusions reached by other scholars in recent years. But the view of empathy that’s emerging from this growing body of work hasn’t much trickled down to the public. And public understanding of it may be critical to shifting America’s political polarization into reverse somewhere between here and the abyss.

Like many past studies, this one gauges people’s level of “empathic concern” by asking them how strongly they agree or disagree with a series of seven statements such as “I often have tender, concerned feelings for people less fortunate than me.” If it seems strange that people who identify with this statement might find amusement in someone’s being injured at a protest, maybe putting the paradox in a more extreme context will help.

Imagine these avowedly empathetic people hearing about the death of ISIS leader Abu Bakr al-Baghdadi last month. There’s no denying that on the day of his death, Baghdadi was in some sense “less fortunate” than they—but do you expect them to have “tender, concerned feelings” for him? And would you be surprised if they reported that, actually, they got a bit of a lift from his demise?

What seems obviously true in the Baghdadi case—that people don’t deploy empathy indiscriminately—turns out to be true in less extreme cases, too, ones that don’t involve terrorist masterminds. Various scholars have found, in various contexts, an “empathy gap” between “in group” and “out group.” In one study, soccer fans showed more concern over pain felt by fans of their favorite team than over pain felt by fans of a rival team.

Of course, this new study does more than find meager empathy for the out group. It finds that high-empathy people view the out group more unfavorably (relative to their own group) than low-empathy people; and that they may even take more delight in the suffering of some out-group members. Here, too, the Baghdadi case is illuminating.

After all, high-empathy Americans presumably felt more acutely the suffering of the in-group members who were beheaded, on camera, by the out group that Baghdadi led. And this could translate into more antipathy toward the out group and its leader. (In President Trump’s colorful ramblings about the special forces raid, he peppered his fond reminiscences of Baghdadi’s death with vivid references to the beheadings, as if trying to make the death feel more gratifying to his audience. Whether consciously or not, he was harnessing the fact that in-group empathy can elevate ill-will toward the out group.)

The authors of the APSR study—Elizabeth Simas and Scott Clifford of the University of Houston and Justin Kirkland of the University of Virginia—have this kind of dynamic in mind when they write, “Polarization is not a consequence of a lack of empathy among the public, but a product of the biased ways in which we experience empathy.”

Or, in the more general formulation favored by the late American scholar Richard Alexander: the flip side of “within-group amity” is “between-group enmity.”

Alexander was a biologist. He believed that our feelings and their patterns of deployment have been shaped by natural selection in accordance with a simple principle: genetically based tendencies that were conducive to the survival and proliferation of our ancestors’ genes are the tendencies we inherited (whether or not they have that effect in a modern environment). These tendencies constitute human nature.

In this light it seems only natural that our most beautiful emotions—love, compassion, empathy—would be deployed selectively, tactically; and only natural that this tactical deployment can wind up deepening hatred and violence. Helping genes proliferate can be a nasty business.

Alexander, like many Darwinians, also believed that our frequent obliviousness to the tactical logic governing our sentiments is itself a part of human nature; it was favored by natural selection because there are benefits to having a sunny view of your own motivations. That way you can make declarations such as “I believe that only bad people should suffer,” without adding, “plus, sometimes people should suffer because their in group happens to be my out group.” Our genes, Alexander wrote, delude us into thinking that we are “law-abiding, kind, altruistic souls.”

The new APSR study may be, in part, a study of this very delusion. When people who take that seven-item empathy survey are reflecting on their level of empathy, they’re likely to focus on occasions when they do feel empathy. They’re probably not pondering the fact that they feel no empathy for the Baghdadis of the world, or that they feel little if any empathy for the Trumps of the world or, as the case may be, the Nancy Pelosis of the world. It may not even occur to them that Trump or Pelosi (or their followers) might deserve better. So, in rating their own empathy, they don’t take off points for this empathy gap.

The link between within-group amity and between-group enmity works in both directions. Just as the former can elevate the latter, the latter can elevate the former. As usual, this dynamic is most conspicuous in extreme cases: Ask New Yorkers how they felt about other New Yorkers the day after 9-11, compared to the day before. But it’s also evident in day-to-day politics. Trump’s outrage du jour strengthens bonds among his opponents.

And that’s not the end of it. These strengthened bonds can help sustain or even elevate antipathy toward Trump and Trump supporters. And this antipathy can then strengthen bonds among Trump supporters, thus helping to sustain or even elevate their antipathy toward … and so on.

Such feedback cycles are one reason that political polarization, once it’s underway, can be so hard to stop, let alone reverse.

It might help if we all learned to be less blindly obedient to the various feelings—including the beautiful, affiliative ones—that push and pull us through life. In the book Against Empathy, Yale psychologist Paul Bloom, after documenting various ways empathy can lead us astray, recommends “rational compassion”—a thoughtful, reflective deployment of affiliative feelings guided by well-informed skepticism about more instinctive patterns of deployment.

Unfortunately, this is super hard. It’s one thing to absorb all the evidence that human beings are less good than they think. It’s another thing—given the natural penchant for self-delusion that Alexander and others have emphasized—to really reckon with the fact that you’re one of these human beings. In one study, after experimenters informed people of various cognitive biases—like our tendency to claim lots of responsibility for successes and little for failures—the average person said they were less prone to these biases than the average person. Not a promising start.

And even if you concede that you’re probably no better than average, even if you acknowledge the depth of your biases, recognizing them in real time is a challenge, given how subtly they do their work. The feelings that drive intergroup conflict—empathy, righteous indignation, loyalty, honor, pride, vengeance, hatred, and so on—just about always feel right. (That’s part of their job!) So it’s hard to get enough distance from them to reflect on whether they’re leading you in a morally defensible direction. As I’ve written before, I think mindfulness meditation can help, but I don’t claim it’s a miracle cure, or that it’s the best approach for everyone.

In any event, a good first step on the path to national recovery would be for more people to recognize that what Adam Smith called “the moral sentiments” aren’t naturally deployed in a consistently moral manner. One way to cultivate that recognition is to get in touch with the long line of academic literature on empathy, of which this new study is just the latest, unsettling installment.


More Great WIRED Stories