by Brian Tomasik
First written: 15 Dec. 2012; last update: 9 Nov. 2014


Is empathy effective at producing altruistic outcomes, or is it mainly a feeling in people's heads? David Brooks urges caution about promoting empathy alone without requiring deeper principles of moral duty. I agree with Brooks, though I'm more optimistic about the prospects for empathy to change hearts and minds. At the very least, empathy is important to inform what deeper principles we should hold in the first place. With less empathy, we might instead attach our sense of life's meaning to religion or conservative moral principles rather than actually helping suffering organisms.


Sometimes it's suggested that promoting compassion could be a way to achieve better futures. But presumably some forms of compassion are more helpful than others, and it's not clear how cost-effective promoting compassion in general would be compared against alternatives.

This article examines an argument by Jesse Prinz and David Brooks that moral principles may matter more than compassion.

An example pro-empathy charity

There are some organizations whose work focuses on promoting empathy. One of the more interesting is the Center for Compassion and Altruism Research and Education (CCARE), which "has collaborated with a number of prominent neuroscientists, behavioral scientists, geneticists and biomedical researchers to closely examine the physiological and psychological correlates of compassion and altruism." Here are some example research projects:

  • Neural Correlates of Compassion in Buddhist Adepts and Novices
  • Investigating the Behavioral and Neural Mechanisms of Compassion Training
  • Neural Networks of Social Compassion and Nurturing: Optical Deconstruction of Altruistic Behavior
  • A Multimodal Study of the Neural Correlates of Experiencing Admiration and Compassion
  • Compassion in the Political Arena.

Note that some of these projects have some downside risk, such as making it easier to reduce/eliminate empathy in humans. On the other hand, greater understanding of the neuroscience of empathy may help with engineering empathy in artificial minds, which would by default likely lack pure (non-strategic) empathy.

A 2012 episode of All in the Mind featured James Doty, Founder and Director of CCARE. I thought the interview was moving and liked many parts of it. However, I had a few points of disagreement:

  • Doty gave the (perhaps unintentional) impression that liberal policies must be better just because they're more compassionate, which ignores the possibility that seemingly non-compassionate institutions could potentially yield more positive outcomes. Compare with the quote often misattributed to Winston Churchill: "If you're not a liberal when you're 25, you have no heart. If you're not a conservative by the time you're 35, you have no brain." I personally have moved somewhat away from the far left and toward more moderate-liberal stances in my political views due to learning more about how the world works. (This was mainly only with respect to economic and international-relations issues rather than social/moral issues.)
  • I also worry that compassion without rationality may lead mainly to more short-sighted charity -- e.g., volunteer work, which Doty mentioned in the interview, or "random acts of kindness," which are endorsed by his Project Compassion -- to the detriment of long-term, indirect thinking.

David Brooks article

A friend pointed me to an excellent article by David Brooks: "The Limits of Empathy." Brooks writes:

People who are empathetic are more sensitive to the perspectives and sufferings of others. They are more likely to make compassionate moral judgments.

The problem comes when we try to turn feeling into action. Empathy makes you more aware of other people's suffering, but it's not clear it actually motivates you to take moral action or prevents you from taking immoral action.

In the early days of the Holocaust, Nazi prison guards sometimes wept as they mowed down Jewish women and children, but they still did it. Subjects in the famous Milgram experiments felt anguish as they appeared to administer electric shocks to other research subjects, but they pressed on because some guy in a lab coat told them to.

Empathy orients you toward moral action, but it doesn't seem to help much when that action comes at a personal cost.

Another example could be the "Good Samaritan" study, where the amount of hurriedness by the subjects had a bigger effect than the moral salience of what they were thinking about.

These claims are generally true. I don't deny that other factors -- including, as Brooks mentions, finding a dime in a phone booth -- can have bigger short-term influences. However, there are a few things to say about this.

People vary in how much they're affected by empathy

Brooks says:

You may feel a pang for the homeless guy on the other side of the street, but the odds are that you are not going to cross the street to give him a dollar.

I have walked across the street several times to help a poor person buy food. In fact, I did so somewhat against my better judgment, because I knew theoretically that the money was better spent helping animals, but I didn't want to become cold-hearted.

So, just as I can generalize from one example by seeing a close link between empathy and action, so too Brooks may be doing the same in his case when not seeing such a link. I think the truth is that people vary.

Brooks goes on to say:

There have been piles of studies investigating the link between empathy and moral action. Different scholars come to different conclusions, but, in a recent paper, Jesse Prinz, a philosopher at City University of New York, summarized the research this way: "These studies suggest that empathy is not a major player when it comes to moral motivation. Its contribution is negligible in children, modest in adults, and nonexistent when costs are significant." Other scholars have called empathy a "fragile flower," easily crushed by self-concern.

So as Brooks admits in the second sentence, some studies have reached the conclusion that empathy matters. And even though empathy is easily crushed, that doesn't mean it always is. Sometimes people aren't in a hurry. Sometimes they aren't pressured by an experimenter to shock a victim. Sometimes they're in a sufficiently calm state that their empathy can lead them to donate to charity, or give up meat, or vote for more humane policies.

I did a brief web search and didn't find a lot of studies, but "Empathy, Emotional Expressiveness, and Prosocial Behavior" suggested once again that people vary in the degree of connection between empathy and altruism:

Boys' empathy, in turn, was a strong predictor of prosocial behavior, R2 = .55. In contrast, girls' empathy was related to prosocial behaviors with friends, R2 = .13, but not to cooperation with peers.

The argument isn't about whether empathy sometimes has an effect but whether it's the most efficient point of leverage for inducing greater altruism.

Is empathy delicious or difficult?

To some extent I think Brooks and I are talking about different things. Brooks portrays empathy as a fleeting feeling, "a way to experience delicious moral emotions without confronting the weaknesses in our nature that prevent us from actually acting upon them." This is not really what I'm after. For instance, empathogens can produce a feeling of "I love the world and the world loves me," but this in itself doesn't actually make things better. Feeling at one with other people and animals doesn't by itself do anything to help fish that are being eaten alive by predators. What matters is whether people are actually motivated to do something about suffering. Likewise, Buddhist monks are often praised for their empathy, but how many of them go out and lobby on important issues?

Moreover, I would dispute that empathy is a "delicious" emotion. For myself, empathy often makes me feel upset, and my desire to reduce suffering is driven by knowing that I stop feeling upset when I'm doing something to help. Empathy for suffering is not warm bubbles and sunshine. Its unpleasantness is probably why so many people avoid factory-farming videos or articles about torture. Seeing suffering temporarily makes people feel bad, not good. (Of course, there can also be a spirit of meaningfulness in knowing you're doing the right thing by learning about suffering, even if doing so is not necessarily hedonically desirable in the moment.) Sometimes I have to turn down my level of empathy in order to write about painful issues, although I worry about doing so too frivolously lest it become a habit.

Empathy is one ingredient in action

Brooks suggests that moral principles play a bigger role in action than empathy:

People who actually perform pro-social action don't only feel for those who are suffering, they feel compelled to act by a sense of duty. Their lives are structured by sacred codes.

Think of anybody you admire. They probably have some talent for fellow-feeling, but it is overshadowed by their sense of obligation to some religious, military, social or philosophic code. They would feel a sense of shame or guilt if they didn't live up to the code.

I think this might be true. Even for myself, I'm more often motivated by a general feeling of "this is the right thing to do" rather than "I feel sorry for this particular suffering worm," although I often can't cleanly separate the two in my brain.

However, where do moral principles come from in the first place? For me and probably for a decent fraction of altruists, this strong sense of duty comes from past empathy. Watching a factory-farming video can change a person's life. The compassion that we feel in a particular instance doesn't necessarily impel short-term action, but it can redirect your life orientation for the long term. If there weren't suffering in the world, I might still be playing video games to pass away the days. Empathy is what gives me a purpose in life; without it, I wouldn't care about moral principles.

Sure, empathy is often not sufficient to provide follow-through without other personality traits (motivation, persistence, intelligence, etc.), but of course that's the case. No single emotion is going to produce effective altruism on its own; you need the combination of several of them. Brooks would probably agree with this, because he acknowledges that promoting empathy does some good; he just points out that it's not enough. That said, I think the tone of his article may lead people to feel as though promoting empathy is valueless rather than recognizing that empathy is just one ingredient of the recipe.

Empathy guides moral principles

Plenty of people find meaning in life without significant empathy. Where does it come from? Well, some examples are religious devotion and commitment to conservative moral principles (purity, loyalty, respect for authority, etc.). But some of these principles are harmful and others are at best occasionally misguided.

Just following principles is not necessarily enough. As Jonathan Glover notes in a Philosophy Bites conversation, the psychopathic patients that he interviewed didn't lack knowledge of ethics entirely. For instance, they held that swearing, bullying, and damaging royal property were bad. However, they didn't have a sense of why these were bad, nor that bullying was worse than swearing because it hurt others. In general, the patients respected the authority of the Queen, the police, and the military, but they had no particular empathy toward suffering. As John Alt summarizes:

The inmates' sense of right and wrong pertained to doing what parents told a child to do. It was authority-based, not based on empathy or compassion. In short, it was rule-following.

Empathy associates moral-principle brain regions with reducing suffering. This presumably makes people more liberal, because liberal morality tends to focus mainly or even exclusively on care/harm. Maybe that's part of why Brooks is skeptical of empathy?

In the paper that Brooks cites, Jesse Prinz notes this as well:

For conservatives, there is little tolerance for transgression; three strikes and you're out. Lakoff captures the liberal value system by saying that for liberals, morality is empathy. The construct of empathy is essential. Liberals try to empathize with both victims and transgressors, and, instead of dividing the world into good and evil, they try to put themselves in the shoes of people on both sides of every divide.

Prinz goes on to enumerate some dangers with purely empathy-based morality, including the possibility of preferential treatment for those you empathize with more, cuteness bias, greater concern for those who express more emotion, in-group favoritism, caring more about those near you, and so on. These are important points, and they illustrate why rationality is also essential.

Focus on targeted empathetic morality

My guess is that generic "empathy promotion" is less valuable than specifically advancing empathetically grounded moral arguments -- for instance, the ethical significance of wild-animal suffering. Doing so channels empathy toward a concrete target, rather than letting the emotion sit in the brain or lead to random acts of compassion toward whatever is most visible to the person at the moment.