Selective empathy is the tendency to feel empathy for some people but not others, based on factors like group identity, moral judgments, or personal closeness. It’s not a clinical diagnosis or a rare quirk. It’s the default way human empathy works. People consistently show greater empathic responses toward those they see as part of their own group and reduced empathy, sometimes none at all, toward people they perceive as outsiders.
This selectivity shows up everywhere: in how you react to news stories, which victims move you to tears, and whose suffering you scroll past. Understanding it helps explain not just personal blind spots but some of the largest patterns in political division, social conflict, and moral inconsistency.
Why Empathy Is Selective by Default
The idea that empathy is a universal, evenly distributed feeling is a popular misconception. Research consistently shows that people display more empathy toward ingroup members and are more sensitive to perceived harmful behaviors committed by outgroup members. “Ingroup” and “outgroup” can be drawn along nearly any line: race, nationality, political party, sports fandom, or even something as abstract as perceived moral character. People who are seen as morally good get treated as potential ingroup members, while those judged as immoral get mentally sorted into the outgroup, and empathy drops accordingly.
This filtering happens fast and often without conscious awareness. Brain imaging studies show that empathy-related neural responses are stronger when people observe someone from their own group in pain compared to someone from an outside group. The bias isn’t just emotional preference. It’s built into how the brain processes social information.
The Biology Behind Biased Empathy
Two biological players are central to understanding selective empathy: oxytocin and the amygdala.
Oxytocin is often called the “bonding hormone,” but its effects are more nuanced than that label suggests. It enhances emotional empathy, the visceral feeling of sharing someone else’s experience, but it does so selectively. Oxytocin strengthens empathic responses within trusted social circles while doing little to extend empathy outward. It deepens the bonds you already have rather than building new ones with strangers. This means the same hormone that makes you cry at your friend’s grief can coexist comfortably with indifference toward a stranger’s.
The amygdala, a brain region involved in processing emotions and social signals, plays an essential role in both social learning and emotional empathy. Research published in The Journal of Neuroscience found that patients with bilateral amygdala damage were significantly impaired in emotional empathy tasks but performed normally on cognitive empathy, the intellectual ability to understand someone else’s perspective. This suggests the amygdala is part of a core circuit for feeling what others feel, and oxytocin modulates that circuit. When the system works normally, it prioritizes emotionally relevant social information, which in practice means it prioritizes people who already matter to you.
Evolutionary Roots of Ingroup Favoritism
Selective empathy isn’t a modern problem. It likely evolved because caring selectively was more useful for survival than caring universally. The theory of parochial altruism proposes that two traits coevolved in humans: cooperativeness toward the ingroup and aggressiveness toward outgroups. Groups whose members were willing to sacrifice for each other, while remaining wary of outsiders, had a survival advantage when competing for resources. Over time, natural selection favored populations where ingroup loyalty and outgroup suspicion reinforced each other.
A related framework, the group heuristic model, offers a slightly different angle. It suggests ingroup bias isn’t an unconditional strategy of discrimination but a mental shortcut. Your brain treats fellow group members as people you’re more likely to encounter again and who are more likely to reciprocate. Empathy flows toward them because, evolutionarily speaking, they’re the people whose cooperation you need. Strangers and outsiders represent unknown quantities, so the brain conserves empathic energy.
Even among other primates, something similar plays out. Chimpanzees form coalitions and engage in coordinated aggression against rival groups, with individual members joining attacks when the risks are low and the potential benefits high. The pattern of bonding inward and competing outward appears to be deeply rooted in primate social life, not something humans invented with politics or religion.
How Selective Empathy Fuels Political Division
One of the most striking findings in recent research is that higher levels of empathy don’t reduce political polarization. They increase it. A study published in the American Political Science Review found that people with higher dispositional empathic concern showed greater affective polarization, meaning they felt warmer toward their own political party and colder toward the opposing one. Moving from the lowest to the highest level of empathic concern increased the probability of expressing the least favorable view of the opposing party by about 27 percentage points.
The mechanism is straightforward: highly empathic people feel deeply for those they identify with, and that emotional investment makes perceived threats to their group feel more urgent and more personal. The study found that empathic concern increased the desire to censor opposing political viewpoints. People one standard deviation above average in empathy rated their desire to silence the other side at 3.77 on a five-point scale, compared to 3.22 for those at the mean. Empathic people were also more likely to experience pleasure at the misfortune of political opponents. The probability of feeling some degree of satisfaction when someone from the opposing party was harmed was 68%, compared to 54% when the person was from their own party.
This paints a counterintuitive picture. The most empathic people in a society may also be the most tribally motivated, not because empathy is bad, but because it operates selectively. When you feel someone’s pain deeply, you also feel more anger toward whoever you believe caused it. And when “your side” and “their side” are clearly defined, empathy becomes fuel for division rather than a bridge across it.
Selective Empathy vs. Empathy Deficits
It’s worth distinguishing normal selective empathy from the empathy patterns seen in certain personality conditions. Everyone is selective about empathy to some degree. But in people with traits associated with psychopathy, narcissism, or Machiavellianism (collectively known as the Dark Triad), the pattern looks different.
These personality profiles are consistently linked to deficits in affective empathy, the ability to feel what someone else feels. But cognitive empathy, the ability to understand what someone is thinking or feeling without sharing the emotion, often remains intact. This combination is what enables manipulation: understanding someone’s emotional state well enough to exploit it, without being moved by it. Research on narcissistic personality disorder found that individuals scored significantly lower than controls on affective empathy but showed no difference in cognitive empathy.
Among the three Dark Triad traits, narcissism shows the smallest deficit in affective empathy. This makes sense when you consider that narcissists depend on social recognition and admiration, which requires some capacity to read and respond to others emotionally. Psychopathy and Machiavellianism, by contrast, show steeper drops in emotional resonance. The key difference from ordinary selective empathy is that most people’s empathy narrows based on group identity or moral judgments but remains emotionally real within those boundaries. In Dark Triad profiles, affective empathy is broadly diminished regardless of group membership.
Compassion as a Corrective
Psychologist Paul Bloom has argued that empathy, defined as literally feeling another person’s pain, is inherently biased and innumerate. You feel the suffering of one identifiable child more than the suffering of thousands of statistical victims. You feel the grief of someone who looks like you more than someone who doesn’t. Empathy, by this view, is a spotlight: vivid but narrow, and easily aimed by the same tribal instincts that produce selective empathy in the first place.
Bloom’s proposed alternative is what he calls “rational compassion,” which others have described as “intelligent kindness.” Where empathy means feeling someone’s pain, compassion means caring about their well-being and acting to help, guided by reason rather than emotional contagion. Compassion can account for scale, consider long-term consequences, and extend to people you’ve never met and can’t easily imagine. It doesn’t require you to feel devastated in order to act helpfully.
This distinction matters practically. If you notice your empathy activating strongly for some people and going quiet for others, that’s not a character flaw. It’s the system working as evolution designed it. The corrective isn’t to force yourself to feel more but to recognize the bias and make deliberate choices about who deserves your concern and action, even when the emotional pull isn’t there.