A common assumption in economics and psychology is that a generous act creates a debt that's expected to be repaid. Yet anthropologists and sociologists have long thought of generous acts as not just based on their payoffs, but as ways to create and sustain different kinds of social relationships.
We also asked whether minimal descriptions of relationships shape people's behavior in incentivized economic games. In these cases, people also expected a precedent — even when it meant repeatedly giving at a personal cost — except for the relationship was equal.
These expectations held regardless of whether the lower- or higher-rank person was generous, in both abstract and concrete descriptions of relationships (e.g., "asymmetric" vs. "advisor-student"), and were not based on the costs/benefits of acting generously.
In this paper, we experimentally tested the prediction, articulated most explicitly in Debt by David Graeber, that in hierarchical relationships, people expect generous acts to set a precedent (for the same person to be generous again).
Expectations of following a precedent came from two sources: priors about who "should" be generous in a given situation, and an abstract rule that whatever happened last time should happen again.
New paper with @rebeccasaxe.bsky.social "Expectations of reciprocal generosity are specific to equal relationships," in @openmindjournal.bsky.social. Thread below!
Article: doi.org/10.1162/opmi...
MIT News: news.mit.edu/2026/would-y...
Hi everyone! My paper on cognitive representations of relationships and their developmental origins is finally out! I had a blast responding to the commentary and was honored to be able to engage with people who laid the foundation for my research program. www.cambridge.org/core/journal...
People generally only engage in reciprocal generosity with others of equal or unknown status, MIT researchers experimentally demonstrated. During repeated interactions between people of different soci...
Cognitive representations of social relationships and their developmental origins - Volume 49
Overall, most classic evidence for reciprocity and turn-taking comes from experiments between anonymous strangers, who are implicitly equal. Our results suggest that this might be a special case, and that much of social life seems to run on following a precedent.
We first tested third-party judgments of vignettes in familiar U.S. social situations (e.g., bringing coffee to a meeting). Participants observed one generous act and predicted who would be generous in the next interaction.
Reciprocity was specific to equal relationships, and following a precedent was expected in hierarchical relationships:
