Why Do We Punish? Proximate Motivations Reveal the Evolutionary Functions of Punishment

by Paul Deutchman


Imagine that you are at an academic conference with a friend. As you walk into the convention hall one morning, you see a table covered in Danish pastries. You grab a couple pastries and put them on a plate. Before you can even take a bite, you remember that you left your poster in your hotel room upstairs. The poster session is about to start so you leave your pastries with your friend and run to your room. When you come back down, you see that your friend has eaten one of your pastries. You would probably be less than thrilled with their friend, maybe even a bit disgruntled. Now imagine a slightly different version of that scenario. This time, before you can grab any pastries you realize you forgot your poster and run upstairs to get it, but not before asking your friend to grab a few for you. When you come back to the convention hall, you see that your friend has a plate full of Danishes for herself, but only grabbed one for you. This might also make you kind of mad, right? However, it is less clear why that is the case in this scenario; your friend didn’t steal your pastries, but it still feels like they did something wrong.

This example captures two of the main, proximate motivators of punishment: revenge, the desire to reciprocate harm, and inequity aversion, the dislike of unfairness. What the latter example highlights, is that unfairness alone can be upsetting enough to motivate punishment. Understanding the motivations behind punishment is an important topic because understanding what motivates punishment can inform our understanding of the functions of punishment and why it evolved in the first place. Revenge-based punishment serves an important deterrence function by discouraging those that harm us from doing so again in the future. On the other hand, inequity-based punishment may serve an important leveling function by ensuring that we’re not worse off than others and potentially giving us a competitive edge, or at least preventing others from gaining too much of a competitive advantage.

Previous research on the relative importance of inequity aversion and revenge has been mixed – some work has found that punishment is motivated primarily by revenge, while other work has found that it is motivated primarily by inequity aversion. Recent work has added more nuance to this debate, finding that punishment is mainly motivated by revenge, with inequity aversion only factoring in as a secondary motivation when people experience losses. However, it remains unclear whether inequity aversion matters in the absence of losses. In other words, do people ever punish others simply because of unfair outcomes?

In our study, we addressed this question by pairing up two people who have never met and having them play an online economic game for real money. The game had two conditions, a Take condition and an Augment condition. In the Take condition, as in the pastry example, one player could steal money from the participant or do nothing. In the Augment condition, one player could give money to the participant at no cost to themselves or do nothing. By examining cases where a player chose to not give this money to the other player, this condition allowed us to investigate whether inequity aversion motivates punishment in the absence loss. Specifically, because the participant was not harmed by the other player in this condition, the only motivation for punishment is to reduce inequity.

Importantly, we also varied the level of inequity participants experienced by manipulating the starting endowments they received at the start of the task: some participants had more (advantageous inequity), the same amount (equality), or less than the other player (disadvantageous inequity). This resulted in participants experiencing different degrees of inequity, which was either worsened, improved, or unchanged depending on whether their partner decided to steal (Take condition), give (Augment condition), or do nothing, respectively. After learning about the other players decision and seeing their relative endowments, participants could punish by paying a cost to reduce the other player’s earnings.

In support of past work, we found that people mainly engaged in punishment when the other player stole from them, suggesting that punishment was motivated primarily by revenge. However, unlike previous work, we also found that, regardless of whether the other player stole from them, participants were more likely to punish when they had less money than the other player as compared to when they had the same amount or more. This finding—that people punished when they experienced inequity but not losses—offers evidence in line with the idea that punishment is motivated by fairness concerns.

Furthermore, our data provide some support for the theory that punishment serves a competitive function by increasing relative payoffs and status. Because punishment in our study used a 1:3 fee-to-fine ratio (i.e., the ratio of punishment’s cost to damage), it is difficult to disentangle any punishment from a competitive motivation to increase relative payoffs. Additionally, a small minority (5-10%) of participants in our study punished the other player for choosing the prosocial option (not stealing or augmenting), even when they had equal or slightly larger endowments. Together, these results potentially suggest that some participants were motivated to punish antisocially, perhaps out of spite or a competitive motivation.

Findings from our new study are exciting because they offer strong evidence that there are different motivations for punishment. That punishment is motivated by both revenge and inequity aversion suggests that punishment likely evolved for multiple functions—both a deterrence function and a fitness-leveling function. That punishment can serve both functions highlights how evolution can design one behavior to serve different goals.

Future work should examine cultural differences in the motivations underlying punishment. We know there are meaningful differences in punishment across societies, and it will be important to explore the prevalence of revenge and inequity aversion motivations in these countries. For example, inequity aversion might motivate punishment more in societies that lack strong formal enforcement mechanisms and where there is more intense local competition for resources, environments where fitness-leveling might be an especially adaptive strategy.

In sum, our work suggests that punishment is motivated by different factors—people are primarily motivated by the desire to seek revenge, but also to some extent by the desire to reduce inequity. These findings underscore the multifaceted functions punishment evolved to serve and highlight the need for greater research on punishment across different societies, especially in non-WEIRD societies.


Read the paper: Punishment is strongly motivated by revenge and weakly motivated by inequity aversion