Patriarchy and punitiveness: Spouse abuse and Protestant Fundamentalism

Given the scant research on spousal violence and religion, the present study focuses on denominational affiliation—specifically Protestant Fundamentalism—and violence in the marital relationship. To illuminate potential violence within Protestant Fundamentalism, special attention was paid to the economic and political inequalities, both in the family and in the church, as well as Fundamentalists\u27 attitudes toward public and private punitiveness. With violence toward the marital partner as the dependent variable, it was hypothesized that Protestant Fundamentalism would be associated with significantly higher odds even after applying control factors and secondly that there would be a significant interaction of Fundamentalism and gender. Traditional gender role attitudes based on patriarchy was believed to be the intervening variable linking Fundamentalism to spousal violence. The data were derived from the third wave of the Marital Stability Across the Life Course study (N = 1737) (Booth et al., 1988) and the 1988 National Survey of Families and Households (N = 6020). Logistic regression, in which the dependent variable is the natural log odds of the response category (violence versus no violence) was used for the analysis. Data analysis of the MILC provides support for the first hypothesis, although traditional gender role attitudes did not provide an explanation. Those who specified a Protestant Fundamentalist denomination were at significantly higher odds of engaging in spousal violence during the current marriage even after controlling for a variety of factors. Using the NSFH data, all five measures of Fundamentalism (i.e. denomination, attitudes, self-identification, convert, lifetime Fundamentalist) were negatively associated with marital violence. The interaction of gender and Fundamentalism upon violence was not in the direction which would support a patriarchy thesis, as Fundamentalism predicts lower odds of violence for men. A variety of explanations are offered for the seeming inconsistent findings using the two data sets