Abstract
Intimate partner violence (IPV) perpetrated by police officers has been largely attributed to the unique critical incident stressors encountered in the field. While a large body of studies has documented a link between critical incident stressors and IPV among law enforcement, these studies have not examined whether experiencing child maltreatment and witnessing interparental violence also contribute to IPV perpetration. The current study uses data from the study Police Stress and Domestic Violence in Police Families in Baltimore, Maryland: 1997–1999 to examine the effects child maltreatment, interparental violence, and critical incident stressors have on IPV perpetration within a social learning and general strain framework. Results indicate that personally experiencing child maltreatment was associated with police officers perpetrating IPV later in life. In addition, women were more likely to report perpetrating IPV when compared to men. Studies that have not controlled for child maltreatment may not have provided a comprehensive test of the correlates of IPV in police families.
Notes
1. Agnew (Citation2002) also pointed out the importance of vicarious (real life strain experienced by others around the individual) and anticipated (an individual’s expectation that his or her current strain will continue into the future) strain. However, since the current data does not contain measures of vicarious and anticipated strain, a formal discussion of these two types of strain will not be provided.
2. Information regarding how the study’s aims were explained to police officers, who would have access to their responses, and whether their participation would be kept confidential are found on the final report published by Gershon (Citation1999).
3. This is how the primary investigators of this study operationalized child maltreatment. No further detailed questions were asked regarding this behavior. Nevertheless, this measure has been used by other scholars to examine the effects of child maltreatment and IPV perpetration by police officers (Zavala, Citation2013b).
4. Some readers may notice that the questions corresponding to negative emotions included items examining feelings about work and items examining more general feelings that do not pertain to work. However, this is in concert with Anderson and Lo (Citation2011).
5. Criminologists and other social scientists have long practiced the tradition of “racial lumping,” or creating racial variables in which white respondents are compared with all other races/ethnicities at once. When such events occur, it is unknown how blacks, Hispanics, Asians, or other ethnicities compare to each other. Given the low variation of Hispanic and “other,” (12 Hispanic and 23 “other” respondents) we created the current variable as presented, a coding scheme that is consistent with the work of others (Anderson & Lo, Citation2011; Gibson et al., Citation2001; Swatt, Gibson, & Piquero, Citation2007).
6. We test the unique importance of the “definitions” portion of social learning theory by generating a two-step regression model and excluding the nonviolent values variable in the initial regression model. Including the nonviolent values in the second step increases the Nagelkerke R2 from .21 to .22 and does not change the statistical significance of the other important variable. Additionally, the nonviolent values variable is also statistically significant, as is the model’s chi-square. This indicates the importance of values as a unique variable for social learning theory and for this reason we report only the second model in .
7. A third analysis was conducted to explore the possible interaction effects between gender*child maltreatment and race*gender on IPV perpetration. This analysis (not shown) found these interaction effects to be nonsignificant.