Objective To assess whether local violence is associated with the timing and type of women's first union formation.

Background Local violence may cause disruptions to marriage markets and psychological and behavioral changes that may affect union formation patterns.

Method The authors exploited the variation in homicide rates caused by a shift in national drug-enforcement policy in Mexico in December 2006. Competing-risks Cox models and union histories from a nationally representative survey of women (N = 33,292) were used to assess whether a recent increase in violence was associated with the timing of the first union transition, which could be either marriage or cohabitation. Analyses were conducted separately by education level.

Results A recent increase in the local homicide rate was associated with delayed first marriage formation for less educated women. Supplementary analyses suggested that a decrease in the number of employed men per women, as well as reduced social interaction due to fear of victimization could be plausible causal mechanisms. No statistically significant associations were found between a recent increase in violence and transitions to first cohabitation for the less educated, or with any first union transition for the moderately and more educated.

Conclusion Among less educated women, a recent increase in violence was associated with a delayed entrance into marriage as a first union transition.

Implications By increasing their barriers to marriage, local violence may contribute to the accumulation of disadvantage among disadvantaged women and families.