Paul Bloom’s essay centers on a major error, arguing that we should use a rational, cost-benefit, “cold cognition” approach in place of an empathy-based approach to decision making. But these are not mutually exclusive, and there are serious dangers in leaving empathy out of decision making.

Bloom argues, “If you want to be good or do good, empathy is a poor guide.” I strongly disagree. Consider Israeli Prime Minister Benjamin Netanyahu’s decision: Should I command the Israeli Defense Force to bomb a rocket launcher that Hamas is firing from within a UN school, and in the process risk killing innocent Palestinian children?

Using the unempathic, rational cost-benefit calculation, Netanyahu might consider the costs to be children’s lives, international political unpopularity, and the risk of angering more Palestinians. Potential benefits are weakening Hamas, reducing the likelihood of rocket attacks against Israelis, and increasing Israeli support for his military campaign. Simply looking at the benefits, his decision is to bomb the Hamas rocket launcher.

Logic has its place, but empathy is a valuable and essential additional resource. And it is free.

Now imagine Netanyahu uses empathy to make his decision. Suppose he says to himself “What would it be like if I were the father of a Palestinian child killed by an Israeli bomb? What would it be like if that Palestinian child were my child, terrified by the bombs raining down?” Using empathy the answer would likely be to find a different way to render the whole region safe.

The same applies to the decision by the leaders of Hamas to fire a rocket at Israel, despite Israel’s possession of the new Iron Dome defense system. If Hamas use the unempathic, cost-benefit calculation, the costs are the risk of Israeli retaliation, the loss of innocent Palestinian lives from Israeli retaliation, and the potential loss of Palestinian support for military tactics. The benefits include sending a protest message to Israel and the world; appearing courageous in the face of a powerful enemy; and inviting Israel to retaliate disproportionately, causing her to lose international support. The cost-benefit analysis leads to firing the rocket at Israel.

But suppose the leaders of Hamas say to themselves, “What would it be like if I were the Israeli child trying to go to sleep at night, when the sirens go off?” Or, “What would it be like if that elderly Israeli woman running for the shelter were my mother?” The answer would be to find a different way to protest against injustice and inequality.

Empathy involves asking yourself “What would it be like if I were the other person?” It is because this elementary step is absent from political decisions that our leaders are capable of hurting others. Bloom’s proposal to leave empathy out of decision making can have disastrous consequences.

President Obama noted how important it is “to see the world through the eyes of those who are different from us.” “It’s the lack of empathy that makes it easy for us to plunge into wars,” he said. I agree. Bloom thinks Obama is wrong and instead says, “Our decisions will be clearer, fairer, and more moral once we put empathy aside,” but I think this is the wrong way around.

If we leave empathy out of our decision making we are in danger of doing what the Nazis did: designing a perfectly rational system such as the Final Solution, with trains taking Jews from all over Europe to the concentration camps and their perfectly designed system of gas chambers and ovens. It all made sense from a Nazi perspective, if the aim was to eradicate anyone with impure blood. All that was missing was empathy for the Jewish victims.

Or consider how the Nazis designed a euthanasia program to systematically eradicate people with learning difficulties. The cost-benefit argument was irrefutable: euthanasia removes “diseased genes” from the population and saves money, since the cost of supporting a person with lifelong learning difficulties was high. What enabled these legal decisions—what allowed lawmakers to believe they were being moral—was the absence of empathy for people with learning difficulties.

Bloom’s family and mine were both persecuted less than seventy years ago, when our relatives were killed in concentration camps that could only be designed and implemented because empathy was absent from politics. We need to make space for empathy, rather than aim for a decision making system without it.

Finally, Bloom says that those who believe in the value of empathy consider it the only way to motivate kindness. But as someone who argues for the value of empathy, I can accept that kindness can result from several sources, including logic. An example of the latter is the golden rule: treat others as you would have them treat you. Logic has its place, but empathy is a valuable and essential additional resource. And it is free.

Below, I offer extra, point-by-point (online-only) commentary to respond more specifically to some of Bloom’s points.

1. Bloom claims that I have claimed that people with Asperger Syndrome have difficulties in both cognitive and affective empathy. In fact, my book specifically argues against this: that people with Asperger Syndrome have impaired cognitive empathy but intact affective empathy. People with Asperger Syndrome frequently stand up for the underdog, against injustice. They may struggle to pick up social nuances, but they do care that others shouldn’t suffer.

2. Bloom’s attack on empathy uses a recent meta-analytic review that found measures of empathy do not predict more than 1 percent of the variation in aggression. I find this very hard to believe so would want to look more closely at how this review was done and how this conclusion was reached. Regrettably Bloom provides no citation to follow this up. If it is about the variation in type of aggression then perhaps that is because all aggression involves low empathy, so that empathy wouldn’t predict which type of aggression is involved. But without being able to read the review and see how empathy was measured, or how the analysis was done, this claim is hard to interpret.

3. Bloom suggests that it may be no coincidence that those who are super-high in empathy tend to be female, and that females are more prone to depression. However, just because women have higher rates of depression doesn’t mean this is due to having high levels of empathy. Higher rates of depression in women could be for a number of reasons. For example, it has been argued that men experience the same rates of depression but that it is manifested differently between the two sexes. (In males, there may be more drinking, for example). Bloom cites Oakley’s concept of “pathological altruism” but there is still very little data to support this.

4. Bloom spends some time discussing the finer differences between “sentimental compassion” and “great compassion” in Buddhism, but this part of his argument seems to depend on a single case study of a Buddhist monk in an fMRI scanner, given either empathy training or compassion training. Scientifically the results are hard to interpret, given the single case design.

5. Bloom says “empathy is biased; we are more prone to feel empathy for attractive people and for those who look like us or share our ethnic or national background,” but I am not convinced any lab studies correspond to real-world behavior. When the tsunami hit Southeast Asia, for example, charitable donations flooded in from countries from around the world, fuelled by empathy for victims, not based on how attractive they were, and not just for those from the same ethnic or national groups.

6. Bloom writes that “Laboratory studies find we really do care more about one than about the mass,” the implication being we tend to empathize with one victim, not many victims. Again, the public philanthropic outpouring of help after large-scale natural or man-made disasters suggests this is incorrect. And such philanthropy is not based on cost-benefit calculations, because when we send $100 to a charity in Southeast Asia, we have no expectation or desire that it will ever be reciprocated. We just want to help, and this help is motivated by sympathy.

7. Bloom also makes the suggestion that empathy prevents us from appreciating that “a hundred deaths are worse than one.” I am sure I am not the only reader who will find this claim odd, since empathy is not incompatible with such appreciation. He also suggests that “without empathy” we are better able to grasp public health issues such as the importance of vaccination programs, but this implies that when people use their empathy, their ability to appreciate public health statistics goes out of the window, which is again untrue. He says that to support public health programs requires “overriding our empathetic responses.” Again, this doesn’t follow: being intelligent and being empathic are not mutually exclusive.

8. Bloom quotes Jamison who says, “Empathy is always perched precariously between gift and invasion.” He cites this as a criticism of empathy, that it intrudes into another’s privacy, but I agree with Jamison and would add that someone with good empathy would know when to enquire as to how someone was feeling, and when to hold off from doing so.

9. Bloom discusses empathy in a medical setting, suggesting that a doctor who is empathic would become upset at a patient’s pain, which would be unhelpful. I agree with him that what a patient needs is a doctor who remains calm and confident in the medical setting. Bloom’s error here is to assume that for a doctor to show empathy he or she must mirror the patient’s feelings, which ignores a widely quoted definition of empathy: experiencing an appropriate emotion triggered by another person’s emotion. It would not be appropriate for a doctor to become upset by their patient’s pain, just as it would be inappropriate for a mother to burst into tears when her child falls over and hurts his knee. Her personal distress could make her child’s distress even greater, so her job as a sensitive, empathic parent is to offer soothing comforting reactions, not to cry when he cries. That would arguably be selfish, not empathic.

10. Bloom argues that empathy and anger have a lot in common. He says they are both social emotions, but this ignores how much they differ. For example, empathy is a pro-social emotion, whilst anger is an anti-social emotion. He says that both empathy and anger are both moral emotions, but anger is really just a way for a person to communicate, “Don’t disrespect, hurt or unfairly treat me or my family or my friends.” Anger is only a moral emotion in a very limited sense, by impelling one to fight an attacker or seek redress for a perceived grievance. Empathy is much more of a moral emotion because it can lead one to help people, alleviate their distress, care for them, include them, and avoid hurting them or their feelings. Bloom also overlooks a big difference between empathy and anger: that when you are angry, blinded by the desire for revenge and feelings of hate, you cannot empathize. Indeed, it is often only by over-riding anger that one can begin to empathize.