Individual, interpersonal, and socioeconomic predictors of divorce: A 12-year prospective study.

Divorce is among the most common and costly life disruptions adults may face. Maladaptive and ineffective couple communication is routinely implicated in leading models as the cause of divorce, and interventions designed to prevent divorce accordingly prioritize changes in couple interaction patterns. Yet research linking couple communication to divorce has been inconsistent, and failures of communication-based interventions highlight the need for an expanded view of early signs of risk. Alternative perspectives on divorce draw attention to the vulnerabilities individuals bring into marriage and the socioeconomic circumstances couples must navigate once married, yet the predictive power of these perspectives has never been directly compared to the behavioral model. The current analyses address this gap through a study of 368 different-sex newlywed couples who provided data on six clusters of predictors within 6 months of marriage: relationship satisfaction for each spouse, individual vulnerabilities for each spouse, couples' adaptive behavior, and socioeconomic risk. Over the next 12 years, 39% of couples divorced. Preregistered Bayesian structural equation models and survival analyses demonstrated that each baseline risk factor independently predicted divorce in the expected direction. When all predictors were included simultaneously, only wives' individual vulnerabilities predicted divorce. Adaptive communication behaviors attenuated the risk conferred by wives' vulnerabilities. The best performing model explained 34% of variance in divorce, suggesting that at least one third of couples' divorce risk is detectable early in marriage. Exploring vulnerabilities and their effects promises to deepen understanding of how relationships develop and deteriorate and can clarify boundary conditions of behavioral models of divorce. (PsycInfo Database Record (c) 2026 APA, all rights reserved).