Murphy · Proceedings of the National Academy of Sciences of the United States of America 2026 · preregistered randomized behavioral experiment · n=602

Partner preferences for resources adapt to income and gender economic inequality.

Cited 0 times in the scientific literature.

Level 2 - randomized trial

Oxford Level 2 by design analogy (randomized controlled experimental trial in behavioral science)

PubMed 41838902 · doi:10.1073/pnas.2527295123 · record verified 2026-08-31

What was done

A preregistered experiment randomized 602 participants across five countries into one of 45 simulated virtual ecologies that manipulated personal income levels and local gender economic inequality. The authors evaluated the impact of these economic conditions on three measures of stated partner preferences for resources (ranked trait preferences, rated trait preferences, and the importance of dating or marrying up financially), as well as preferences regarding age gaps and physical attractiveness.

What was found

Participants exhibited stronger stated preferences for partner resources when they were assigned to lower personal income conditions and when their sex was economically disadvantaged in the virtual ecology, consistent across all three resource preference measures. Preferences for partner age gaps and sex differences in preferences for physical attractiveness were largely unaffected by the economic manipulations. The abstract reports directional patterns but provides no numerical effect sizes, test statistics, or confidence intervals.

Why it matters

These findings suggest that mate preferences for wealth and resources are not fixed evolutionary traits or static cultural norms, but flexible strategic adaptations that respond directly to personal and demographic economic conditions.

Limits

The abstract provides no quantitative data or effect sizes. Mate preferences were measured as stated ratings within hypothetical virtual ecologies rather than observed real-world mating choices or relationship outcomes. The specific five countries and participant demographic characteristics are not detailed in the abstract.

Cited by