Identity non-overlap has weaker and less general consequences for well-being than existing theories imply. A comparison of incompatibility, coherence and spill-over hypotheses.

Most people hold multiple social identities whose trait descriptions overlap to varying degrees. There are several theoretical accounts that suggest that non-overlap of social identities affects well-being, but existing theories make conflicting predictions about the nature of this effect. Cognitive consistency and acculturation models suggest that non-overlap between identities undermines well-being (incompatibility hypothesis), developmental models suggest that non-overlap may foster growth when differences are integrated at a higher-order level (coherence hypothesis), and spill-over models suggest that overlap amplifies the effects of positive or negative identity-based experiences across identities (spill-over hypothesis). The present research directly compared these predictions using frequentist and Bayesian analyses across a total of eleven studies in different dual-identity contexts, including gender, profession, and religious identities. Identity-based experience valence was measured in Study Package 1 (total N = 610) and manipulated in Study Package 2 (total N = 778). Across studies, the effects of identity overlap on well-being were weaker and less consistent than existing theories imply. Among the three accounts, the incompatibility hypothesis received the strongest, though only partial, support: non-overlap showed a moderate negative association with well-being, whereas the coherence and spill-over hypotheses received little support. These findings suggest that current theories may overstate the direct consequences of identity overlap for well-being and point to the need for a more conditional, process-based and time-sensitive accounts of when and how multiple identities shape well-being.