Identifying the causal effect of alcohol abuse on the perpetration of intimate partner violence by men using a natural experiment

Intimate partner violence (IPV) is widespread and has substantial negative consequences. Researchers have documented a strong positive correlation between alcohol abuse and IPV. However, alcohol abuse is potentially endogenous to IPV. We deal with this problem by exploring a unique instrumental variable—the September 11 terrorist attack—in Wave III of the National Longitudinal Study of Adolescent Health. Ordinary least squares results confirm a strong positive correlation between alcohol abuse and IPV. However, our two‐stage least squares results are statistically insignificant. These results indicate that alcohol abuse might not have a causal effect on IPV and, therefore, have important policy implications.