There's an unspoken rule most people absorb early in life: if someone does something nice for you, return the favor. A friend covers your coffee tab, you get it next time. A coworker brings donuts to the office, you bring bagels next week. For decades, this back-and-forth has been treated as a cornerstone of social psychology. But a new study suggests that rule only holds when the two people involved are on equal footing. When there's a power difference between them, the whole playbook changes.
Researchers at MIT ran six experiments testing what people actually expect after someone acts generously. When two people are equals, observers expect the favor to be returned. But when there's a rank difference, say between an uncle and a nephew, people don't expect the recipient to pay it back. Instead, they expect the original giver to simply give again. That pattern held whether the giver outranked the recipient or the recipient outranked the giver, and it showed up consistently whether the relationship was described in abstract terms or with concrete, familiar labels.
Published in the journal Open Mind, the study was led by Alicia M. Chen and Rebecca Saxe of the Department of Brain and Cognitive Sciences at MIT and involved 599 U.S. adults recruited through an online platform across six experiments.
In the first four experiments, participants acted as outside observers. They read short scenarios, one person picking up the coffee tab, preparing a meal, or deferring to the other's restaurant preference, then answered: who will be generous next time? Each scenario was paired with a description of the relationship, ranging from abstract labels like "symmetric" or "asymmetric" to familiar pairings like cousins versus uncle-and-nephew.
When the relationship was equal, participants consistently expected the other person to return the generous act. When even a small power difference was introduced, expectations flipped. Participants predicted the original giver would give again, regardless of whether that person ranked higher or lower.
To test whether these results reflect how people actually behave, rather than just what they think should happen, the final two experiments introduced real monetary rewards.
Participants took part in coordination tasks with simulated partners described as having more, equal, or less power than themselves. Their choices about who would take the generous role affected their earnings, and in one version, acting generously paid less, meaning repeated givers gave up real earnings.
Even so, the same pattern emerged. After a successful first interaction with an unequal simulated partner, participants acted as though whoever was generous the first time would be generous again. With equal partners, they were more likely to switch roles the second time, even at the risk of a less efficient outcome. The willingness to take turns, and accept some personal cost to do it, was specific to relationships between equals.
Anthropological research has long documented this pattern in real communities worldwide. In many cultures, gifts flow consistently in one direction within hierarchical relationships, not as part of a trading system but as a way of sustaining the relationship itself. Psychology experiments, the researchers suggest, may have focused too narrowly on strangers and equals, producing a skewed picture of how generosity actually works.
Researchers also found that expectations weren't simply based on guessing which role fit a given type of person. Even when a generous act broke what observers initially anticipated, they still predicted that same person would do it again. A single observation was enough to reset expectations, as if the brain files away what just happened and treats it as the new normal for that relationship.
Reciprocity may be a real social force, but this research suggests it operates in a narrower lane than previously assumed. It shows up reliably between equals, the kinds of interactions that happen to dominate psychology experiments. In the broader landscape of daily life, where most relationships carry some degree of rank or role difference, a different rule appears to take hold: whoever gave last time gives next time too.
Many everyday relationships, from mentors and mentees to uncles and nephews, may run on that logic. If the pattern holds outside the lab, the uncle who always picks up the tab and the mentor who always makes the introduction may not be especially generous individuals. They may simply be following a rule that set itself the first time they gave.