Line of Inquiry: Shereen Chaudhry on Why We Hesitate to Apologize

Generally, I’m interested in how people navigate interpersonal relationships using language or speech acts. One type of interpersonal interaction that’s really important is conflict, right? It happens to all of us, in all domains of life. And apologizing is a central way that people use . . . a central kind of behavior people use to navigate conflict and overcome conflict.

There’s a ton of research on apologizing showing that this simple speech act can really help with reconciliation. In this particular project, I’m interested in mutual-blame conflicts. So that’s situations where there’s not only one person who did something wrong, but both people in some sense are the offender and both people are the victim.

The specific question we’re asking in this project actually was inspired by my own thinking in my own mutual-blame conflict. I realized that I was hesitant to apologize to a friend in a situation where I also felt like she did something wrong. And I realized one thing I was worried about was whether she would return the apology if I apologized. And I was reluctant to apologize if I thought she wouldn’t. And I wondered if this was a more general phenomenon, and if we could learn something about human psychology from this situation.

So what we did was we ran three different surveys across three different public talks of people and had them recall any mutual-blame conflict from their life. And then we asked them to tell us how they would anticipate feeling in several different situations. So imagine that you apologized and they didn’t apologize back. You apologized and they did. They apologized; you didn’t apologize back. And so on. And what we consistently found across these three different populations was that people expected to feel much better if they apologized first and got a return apology than if they apologized and didn’t get a return apology.

And in fact, they expected that apologizing would be a negative experience if they didn’t get that return apology, but that it would be positive if they did get that return apology. But we also asked people: How would you feel if you apologized and they forgave you? And people anticipated not feeling as positively as in the case where they get this return apology. So there’s something special about the other person also taking blame in this mutual-blame conflict that people really value.

So an open question, though, is why does this matter for behavior? How does people’s preferences over getting this return apology affect their willingness to apologize at all? It might not. So if the status quo is that neither person has apologized and you just think regardless of what the other person does, I just wanna apologize first, then the fact that you want a return apology is not going to affect your willingness to apologize. Or let’s say you just don’t . . . you categorically don’t want to apologize first. Then regardless of whether you feel better, getting that return is not going to matter for your willingness to take the first step toward reconciliation. Instead, it might matter, though, if you find the status quo to be better than ending up in a world where you’re the only one to apologize, but not as good as ending up in the world where you apologize and they return the apology. And if that’s the case, you’ll display something we call conditional initiation, where you’ll conditionally initiate the reconciliation process based on your expectations of what the other person is going to do.

And so I think the most compelling way that we tested this was to create a mutual-blame conflict in the lab, where we had pairs of participants interact on a real-world task, a real-effort task. And we constructed it so that people would make decisions in a way that would cause the other person to lose money in one part of the task. And then in the other part, the other person would cause them to lose money. And then they would learn that they’re in this kind of mutual-blame conflict. And one person in the pair was called the first mover, because this is the person we gave the first chance to communicate to, and then they knew the other person would get to respond to their message. OK, so at this point is where we manipulated people’s expectations of what the other person would do—specifically, whether they would get a return apology from their partner or not.

So to do this, we took advantage of something called the 10 days of atonement in the Jewish religion. This is the time between Rosh Hashanah and Yom Kippur where theoretically people who are observant are more likely to apologize to people that they’ve harmed throughout the year. And so what we did was we had half of the first movers paired with somebody that we said was not observing this 10 days of apologizing. And then the other half were paired with somebody who we said was observing this 10 days of apologizing. And this study was conducted during that 10 days. Then we coded the first movers’ messages for the presence of apology. And what we found was that when first movers were paired with somebody who was not observant, not observing this 10 days of apologizing, they only apologized in 17 percent of cases. But when their partner was observant, they apologized 26 percent of the time. And so what we find is that people are sensitive to the expectations of getting a return apology from the other person in the sense that their willingness to go first is affected by that information.

We also ran a study to see if this held on the group level. Is this a dynamic that applies to country-level, group-level apologies? So to examine this, what we did was we surveyed US citizens, and we asked them to imagine that President Biden was gonna take a trip to Japan and he was considering apologizing for the US dropping the atomic bombs in World War II. And we asked them: How willing . . . how supportive of this action are you? But we randomized the information we gave people before they answered that question. Half of the people, we told them, “You read an article about the prime minister “of Japan being interviewed and you learned “that he said that he would be willing to apologize “for the bombing of Pearl Harbor “if the US first apologized for the atomic bombs.” And the other half of people read that the prime minister would not be willing to apologize in return. And what we find is that US citizens are much more supportive of Biden apologizing for bombing and for the bombing in World War II if they expect the Japanese prime minister to respond in kind.

So this is a dynamic that extends beyond interpersonal relationships. We have observations that people care about getting this return apology, and their willingness to go first depends on it. But the question remains. Why? What is the psychology driving people to care about this return apology? I mean, if you think about it, the cost of an apology is seemingly the same whether you get a return apology or not. If you are responsible for making an error in a report, and you say that out loud, “I’m sorry I made this error in the report,” regardless of what the other person does, it’s now clear to people who heard you that you made an error in the report.

But when you apologize for a specific act, you’re sort of taking blame for that specific act. We call that . . . judgments on that act “absolute-blame judgments.” But in this project, we propose people also care about relative-blame judgments. So that is, how responsible you are for a conflict relative to the other person or people involved.

Let’s go back to the workplace example I discussed where, let’s say you’ve made an error on a report and your colleague missed it and didn’t catch it in time. And then following this, your supervisor is furious, and you’re having a meeting with the three of you. Now imagine the supervisor has no idea who was responsible for what part, like how you divided the work. So they actually don’t know who is responsible for the mistake. And in this case, the reality is is that it’s mutual blame. So if you were to say to your colleague, “Hey, I’m sorry that I made the error in the report,” and they don’t apologize back—they say, like, “Hey, it’s OK. Don’t worry about it”—the supervisor is going to assume you are the only one responsible for that. This is a case of unilateral blame. But if your colleague were to respond and say, “Oh no, I’m sorry. I didn’t catch the error “as I said that I would try to do.” Now the supervisor knows that at least it’s not unilateral blame, that to some extent both of you are responsible. It’s a case of mutual blame, and this feels better than ending up as the only person to blame.

What this tells us about apologizing is that it’s not a one-sided decision, especially when blame is mutual, that people’s decision to apologize is affected by their expectations of what the other person is gonna do. How are they gonna react? How are they gonna respond? Will they apologize? And why this is especially interesting is that people are often miscalibrated about other people and about how interactions with other people will go. And so if people underestimate how likely it is that they’ll get a return apology, let’s say both people in a dyad do, then neither person will be willing to take that first step. Even though both of them are willing to apologize, they won’t find out. And we might observe a lot of suboptimal apology stalemates, where neither person decides to take that leap of faith and resolve the conflict.