Do moral ideas have power? We would like to think so. We all appeal to certain actions being “right” as a way to justify why we want certain things done. And implicitly, we expect those moral appeals to carry force, to disarm opposition, to persuade, and to drive changes in the concrete world of policy, conflict, jobs, and money.
The ubiquitous nature of moral appeals can be puzzling to an economist: We have traditionally assumed that we live in a material world, and that material incentives are the most powerful driver of actions.
That said, behavioral economics and the exploration of non-material motivations have taught us much over the past two decades. Economists now accept the fact that individuals can be altruistic, and that they are sensitive to the intentions and motivations of those with whom they interact. We have learned that we are highly social animals: We care strongly about what others do, and we are reluctant to fall out of line.
Experiments in psychology and economics have shown that it is possible to affect behavior by indicating that a particular action is popular. For example, hotels have found that a good way to save money on laundry is to get customers to re-use their towels, and an effective way of achieving this is to tell customers that the vast majority of other customers re-use their towels.
But manipulating beliefs about what “the majority does” appeals to a herd mentality, which can often lead people to immoral acts. To me and my brother Pedro Dal Bó, an economist at Brown University, it seemed crucial to learn whether strictly moral ideas can be used to affect behavior. Could we prove that pure moral suasion, which is so widely attempted, actually works? And could we do this by invoking unadulterated ethical principles, devoid of any herd mentality or social components?
For me, this was a personal question. Every spring, I teach Ethics to nearly 240 Berkeley-Haas MBA students. If ethical discourse has no impact on behavior, why train people to think and express clear moral ideas?
As strange as it may sound, a rigorous, experimental investigation of the power of strictly moral ideas to affect behavior had never been conducted. My brother and I performed a series of experiments in Berkeley’s XLab, the results of which we recently published in the Journal of Public Economics.
In our experiments, subjects played games where everyone did better when individuals cooperated but each person had a material incentive to withdraw cooperation for the sake of individual profit. Individuals got to interact a total of 20 times. Halfway through their play, some people were randomly “treated,” as in a clinical trial, with moral messages. One message was based on the Golden Rule, that you should treat others as you would want others to treat you. Another was based on the utilitarian principle that an action is moral if it benefits the overall group. Other subjects were treated with different, non-moral messages.
We found something a hard core economist should find surprising: Compared to non-moral messages, the “immaterial” moral messages made people more cooperative… but only for a while. On its own, the moral appeal lost much of its impact as the game went on. But that changed when the game allowed players to punish those who had been uncooperative. Under the punishment option, people could reduce the returns to people who cooperated very little. When that option was available, moral suasion worked permanently. In other words, we found a powerful interactive effect between a very material element of punishment and the very non-material element of moral sensibility.
We then investigated the mechanisms behind the moral suasion effect. It’s possible that exposure to moral ideas makes us more cooperative, regardless of what others do. Another possibility is that moral suasion leads us to expect better behavior from others, and we become more cooperative because we want to be “closer to the pack.”
With two additional experiments, we demonstrated the presence of both effects. Moral suasion does change our preferences, but it is much more powerful when people know that others are getting the same message. This implies the existence of a “moral social multiplier.”
These findings have important implications for the management of organizations: Moral discourse can be an effective way to promote cooperation, especially when everyone knows that shirkers can be punished and that everyone got the memo.
The more general takeaway is that moral ideas can be wielded to affect material outcomes. In a recent book, Steven Pinker claims that what President Lincoln called “the better angels of our nature” have played a gradual role in making the world a less violent place. Our findings suggest that we can talk to those better angels and put them to work for the common good.