Prosecutorial discretion for domestic violence cases: an examination of the effects of offender race, ethnicity, gender, and age

Abstract

Much of the existing literature on courts and sentencing has focused on judicial decision-making. Prior research on prosecutorial decision-making is more limited, with even less attention paid to the prosecution of domestic violence cases. The research that has been conducted has produced inconsistent results regarding the effects of legal and extralegal variables. The current study focuses on the effects of extralegal suspect characteristics on the decision to dismiss domestic violence cases in a large Midwestern county from June 2009 to December 2009. The findings demonstrate that gender and race have a strong influence on prosecutors’ decisions to dismiss charges in domestic violence cases. Contrary to the focal concerns perspective, however, the results indicate that males and Black and Hispanic offenders are more likely to have their cases dismissed. Implications for future research are discussed.

Keywords:

• domestic violence
• prosecution
• race
• gender
• focal concerns

Notes

1. Although some scholars have criticized the focal concerns perspective for its dual treatment of prior record as a measure of blameworthiness and community protection, and the lack of attention to measurement of system constraints (see Hartley, Maddan, & Spohn, 2007, for discussion), Steffensmeier and colleagues (1998) argue that the focal concerns are guided by extralegal factors.

2. Information on the victim and offender relationship was not available. Therefore, analysis could not be conducted examining possible differences in the processing of cases dependent on this relationship.

3. The model also was examined with a hazard rate included for the pretrial decisions using the Heckman two-step method. In the model, significant values were not changed. Because of the similarity in the results and concerns with utilizing Heckman’s sample selection correction with a dichotomous dependent variable (see Bushway, Johnson, & Slocum, 2007), the results without the Heckman correction are presented.

4. .