People prefer generous partners over wealthy ones, unless wealth is highly unequal

Two studies using a hypothetical money distribution task found that people are generally more interested in the generosity than in the wealth of a potential partner. However, when potential partners differed more in their generosity, the preference for generosity was strengthened; conversely, when they varied more in wealth, the preference for generosity was weakened. The research was published in Evolution and Human Behavior.

Choosing the right partner—whether in friendship, business, or romance—has played a major role in the evolution of human cooperation. When people can decide whom to interact with, they tend to prefer those who are helpful and reliable, which encourages cooperative behavior overall. In this kind of “social marketplace,” individuals generally try to show that they are generous or trustworthy in order to attract good partners.

A good partner is someone who brings benefits while causing few problems or costs. In everyday terms, people usually judge this based on how kind someone is (their warmth), how capable or successful they are (their competence), and whether they are actually available to help. Both kindness and ability matter, but which one matters more depends on the situation and the environment.

Crucially, people pay more attention to traits that differ noticeably between individuals, because there is little point in evaluating qualities that everyone has in equal measure. Research using computer simulations has shown that when kindness varies a lot in a group, people evolve stronger preferences for kindness, and the same pattern holds for ability.

Study authors Yuta Kawamura and Pat Barclay wanted to explore whether people place more value on a partner’s generosity when partners vary in their willingness to give, and place more value on a partner’s wealth when partners vary in their ability to give. They conducted two online studies in which participants were presented with a pool of hypothetical partners for a money-sharing task.

The first study included 350 U.S. residents recruited from Amazon Mechanical Turk via CloudResearch. Their average age was 40 years, and 163 of the participants were women.

Participants were randomly divided into two groups. To establish the “norm” of the environment, both groups were shown information about how 20 potential partners had shared money on a previous occasion.

In the “Unequal Generosity” group, all 20 potential partners had roughly the same amount of total money (149–149–151), but they differed in the percentage of that sum they gave away (ranging from 3% to 56%). In the “Unequal Wealth” group, potential partners all gave away roughly the same percentage of their money (29%–31%), but the total amount of money they had varied (from $16 to $280). Importantly, the actual absolute dollar amounts distributed were identical across both conditions; only the variance in wealth or generosity changed.

Participants were then asked to choose a new partner for themselves. But before making their choice, they had to decide which single piece of information they wanted to see about their prospective partner: their wealth (total monetary endowment) or their generosity (percentage shared).

The second study included 600 U.S. individuals, also recruited via CloudResearch. The procedure was identical to the first study, with one significant addition: the authors introduced a third, “Control” group. In this group, participants were not shown how potential partners had behaved in the past, meaning they had no information about how much wealth or generosity varied in the population.

Results of the first study showed that participants were generally more interested in the generosity of potential partners than in their wealth. However, the environment mattered heavily. In the group where potential partners varied in generosity, participants were highly likely to seek information about generosity (82%). But in the group where potential partners varied in wealth, the desire to know about generosity dropped significantly (down to 55%), as a large portion of participants shifted their focus to finding out the partner’s wealth.

The second study confirmed these findings and revealed a human “default.” The percentage of people interested in generosity did not differ significantly between the Unequal Generosity group (90%) and the Control group (87%). However, when participants knew that partners differed wildly in wealth, a much smaller percentage (61%) chose to see information about generosity.

“Participants had a default preference to know about others’ generosity rather than their wealth; this preference was strengthened when others varied more in generosity and weakened when others varied more in wealth,” the study authors concluded. “Thus, our study shows that people are sensitive to the amount of population variance on a trait, and flexibly adjust their partner preferences to focus on traits which vary more among others.”

The study contributes to the scientific knowledge about the psychological factors affecting partner choice. However, the authors noted a few limitations: the study asked hypothetical questions rather than having participants distribute real money, and the research was conducted within a single culture (the U.S.).

Additionally, it should be noted that “wealth” in this experimental context was operationalized as the size of a momentary monetary endowment in an economic game (up to $280), rather than real-world accumulated wealth, status, or net worth. Future studies may be needed to see how these flexible partner preferences play out when differences in wealth are on a real-world, macroeconomic scale.

The paper, “Wealth or generosity? People choose partners based on whichever is more variable,” was authored by Yuta Kawamura and Pat Barclay.