Husbands who harm: Predicting spousal violence by men

Previous research on spouse abuse has frequently focused on bivariate relationships between theoretically derived variables and marital violence. This study utilizes a multivariate approach in order to explore the independent and combined effects of several variables derived from the social learning and the frustration/strain perspectives on self-reported violence by husbands against wives. Data for married and/or cohabiting males are derived from a national stratified random sample of couples in the United States. Loglinear analysis is used to identify the main and interactive effects of age, occupational status, employment status, subjective economic strain, and observation of parental violence on reports of violence toward one's wife. Results indicate that age, occupational status, parental modeling and employment status affect the likelihood of violence. Employment status more strongly increases the likelihood of violence for younger men, as opposed to older men, which supports the strain perspective. The independent effect of the observation of parental violence lends support to the social learning approach. Social policy implications are discussed.