Seeing good people do bad things makes the world feel like a more confusing place

In two initial studies, Kate W. Guan and Steven J. Heine at the University of British Columbia recruited online participants who reported having had this type of experience. After reporting details about the event and the perpetrator, participants answered a series of questions about how they felt before, during and after it happened. These questions probed their perceptions of the perpetrator's character, their general beliefs about the world and other people, and confidence in their ability to judge character.

The results suggest that the experience of a "good" person doing something immoral is so unsettling, it makes us doubt our ability to judge character and also makes the world feel like a more confusing place. The starker the contrast between the participants' impressions of the perpetrator's moral character before versus after their bad behaviour, the bigger the dent in their sense of being able to comprehend the world and their place in it. In the second study (though not the first), Guan and Heine also found that this experience leads people to a dimmer view of the moral character of people in general.

The researchers also report that some people were more affected in this way than others. Those who reported liking closure in life — who find it hard to tolerate uncertain situations — and those who felt that a person's moral character is fundamental and unchanging reported the biggest shifts in their own attitudes.

However, in these studies, the participants had to think back into the past, and try to recall how they felt at various points. This experimental approach is notoriously vulnerable to mis-remembering. So the researchers then attempted a more real-time investigation.

A fresh group of 446 online participants imagined meeting a new person. Some were told that this person seemed "like a kind and likeable person" who was "objectively caring, helpful and upstanding"; others were told that they seemed "like an unkind and dislikable person", who was "objectively selfish, cruel and hateful". The participants then read about this person doing something bad — either engaging in hate speech or committing domestic abuse. Finally, they answered questions about how this information would influence their own beliefs and perceptions, similar to those in the first studies.

The researchers report that those who'd been led to believe that the person was kind and likeable reported a bigger threat to their sense of meaning in the world, as well as to their confidence in judging character — and to their moral impressions of other people, in general. It's notable that a shift from good-to-bad had a bigger negative impact on the participants than a bad person engaging in another example of bad behaviour, the researchers write. But there is an obvious weakness to this study too — the participants only imagined how they'd feel; perhaps they'd feel differently in real life. 

If we set the various limitations of this study aside, why should witnessing someone going from "good" to "bad" have a bigger impact?

The pair's pretty convincing argument is that we feel a strong need to work other people out. To succeed in a social environment, we really need to know if another person is likely to be a trustworthy, helpful ally — or a threat.

Of course countless books, TV series and movies that grip us do so because they on play on this drive. Is a character really a goodie or a baddie? The more we are made unsure, the more we just have to know, and we keep reading or binge-watching. It makes sense, then, that moral character violations should make us reconsider our ability to judge others, and make us more cautious. Guan and Heine conclude: "The present findings advance a novel reason why people may bemoan such experiences. They threaten a core way in which we make sense of our social landscapes."

Further reading

– When Good People Break Bad: Moral Impression Violations in Everyday Life

About the author

Emma Young (@EmmaELYoung) is a staff writer at BPS Research Digest