Humans might not be altruistic ‘avengers’ after all, study finds
CU Boulder research contradicts the long-held belief that humans interfere when they see the abuse of strangers
Picture this: You’re walking down the street when you overhear someone spewing nasty insults at a stranger. Would you intervene, even if it meant putting yourself in harm’s way?
While most of us would like to think we’d step in, even going so far as to punish the bad actor, new research led by the University of babyֱapp Boulder suggests that might not be the case.
The study, , adds a new, contradictory perspective to the established wisdom that humans have evolved to punish people who mistreat strangers, even when intervening could lead to social costs.
In their paper “The Unresponsive Avenger: More Evidence That Disinterested Third Parties Do Not Punish Altruistically,” a team of researchers tested this scientifically accepted theory using a novel set of experiments. Their findings suggest that, on average, people aren’t inclined to selflessly punish people who abuse strangers, except for when they are faced with a very specific set of conditions in a lab.
“We’ve drastically overestimated the extent to which we think that third parties are willing to intervene on behalf of strangers,” said Eric Pedersen, assistant professor of social psychology at CU Boulder and the study’s lead author. “Our findings highlight how important it is for researchers to make sure they’re testing ideas in a multitude of different ways to make sure we get converging evidence from different methodologies.”
Using babyֱapp games
Since at least the early 2000s, psychologists have been studying whether humans evolved to altruistically punish people who act badly toward strangers. The most common way researchers have tried to answer this question is with laboratory-based babyֱapp games in which participants can pay to “punish” another participant who had purportedly harmed someone.
In those experiments, participants tend to punish others for treating strangers unfairly. The working theory is that humans have evolved this way in order to encourage cooperation in society.
“There has been lots of research suggestions that people are very, very willing to engage in third-party punishment when they don’t have a personal stake in the conflict,” Pedersen said.
However, Pedersen and his colleagues weren’t convinced that these babyֱapp games were the best way to test this concept. In essence, the researchers guessed that people were being influenced by the setup of the experiments, a concept known as experimental demand.
“If you ask people to focus their attention on the idea that a third party has been unfair toward another person, and then invite that person to invest a little of their own income in punishing that unfair third party, they will do it,” said Michael McCullough, a psychology professor at the University of Miami and one of the study’s co-authors.
“But, if any of those pieces are missing, you don’t get punishment. What we’ve found is that if there is such a thing as altruistic third-party punishment, it really unfolds under a very specific, very restricted set of circumstances. The horizon of real-life events to which that original finding is applicable is really narrow.”
Designing robust tests
The team decided to test the theory of altruistic third-party punishment in a new way, using five experiments that did not involve babyֱapp games.
The experiments varied, however, in that a person either insulted the participant directly, insulted a stranger or insulted a friend of the participant. Then, researchers gave the participant the opportunity to blast an annoying sound at the bad actor, letting them choose the duration and volume. In another experiment, participants simply imagined that a person had insulted them directly or had insulted a stranger.
What we’ve found is that if there is such a thing as altruistic third-party punishment, it really unfolds under a very specific, very restricted set of circumstances. The horizon of real-life events to which that original finding is applicable is really narrow.”
Taken together, the findings contradicted past research, suggesting that people will punish others who have harmed them directly or who have harmed a friend, but will not punish someone who has harmed a stranger.
“The fact that people did not take advantage of an opportunity to punish on behalf of strangers, despite the lack of barriers to doing so, was striking,” said William McAuliffe, a psychology graduate student at the University of Miami and one of the paper’s co-authors.
“Participants had no reason to fear retaliation from the insulter and did not have to pay a price to punish, as they do in the babyֱapp game method. This suggests that most participants were truly quite apathetic that some other stranger was insulted in an unwarranted way.”
The researchers note that their study has limitations and caution that their findings assess average human behavior, noting that there will always be exceptional “heroes” who intervene on behalf of strangers.
“There are always individual differences, and our experiments are no exception in that regard,” McAuliffe said.“Experiments are not well-equipped to make all-or-nothing statements about human nature anyhow. They are better equipped to demonstrate what is typical under various circumstances. Exceptional behavior is better documented by studying real-world heroes.”
Training bystanders effectively
Though the researchers say that more work needs to be done to better understand third-party punishment, their recent findings have important scientific and practical implications.
In the psychology community, the paper challenges a widely held theory that scientists use as a starting point for other research and arguments, which points to the acute need for further study in this area.
“(The paper) casts serious doubts on these assumptions, and going forward, it’s important to use a variety of methods in addition to what we have used in this paper, such as going out into the field and using real-world data so we can get convergence on what might the best approximate answer,” said Pedersen, who started this research as a graduate student at the University of Miami.
In the real world, the findings reinforce the usefulness of a criminal justice system that doles out punishment on behalf of strangers. For people who create policies or programs intended to improve human cooperation and behavior—programs to prevent bullying in schools, for example—the findings suggest that people need help intervening.
Since the research suggests that stepping in on behalf of strangers is not a natural tendency, people need to be made aware of the innate psychological barriers they’re up against and given strategies for overcoming them. In addition, the findings highlight the importance of tools that don’t require direct confrontation, such as anonymous tip lines and ombuds offices, Pedersen said.
“We can really try to prompt people to be on the lookout for their own reluctance to intervene and find a way to do something,” Pedersen said. “We can incentivize people to step up in various ways by making it clear that there are benefits to standing up for others or highlight ways in which people can intervene without putting themselves in harm’s way.”