Gender and the Peremptory Challenge: Separating the Effects of Race and Gender in Jury Selection

Abstract

Gender- and race-based discrimination in jury selection is unconstitutional in the United States. Nevertheless, court cases and empirical evidence suggest that discrimination on such characteristics, especially race, continues to occur through peremptory challenges. Statistical evidence of the gender effect, however, is more limited and has not previously incorporated controls in race-specific analyses to address non-gender-based explanations for gender differences. The present study examines gender differences in peremptory challenges using data from criminal jury trials in Mississippi. Race-specific analyses and propensity score matching are used to isolate the impact of gender from other measurable effects. Results indicate weak and non-significant gender differences for peremptory challenges used by the prosecution, and weak-to-moderate and non-significant gender differences for peremptory challenges used by the defense. This suggests that gender differences in the use of peremptory challenges may be the result of racial differences and other factors rather than a true gender-based effect.

Keywords:

• Gender
• gender discrimination
• peremptory challenge
• propensity score matching
• race

ACKNOWLEDGMENTS

The author thanks APM Reports for collecting and publishing the data used in this study.

Notes

1 The body of literature on jury selection discrimination typically refers to “gender,” but has sometimes used the terms gender and sex interchangeably. Likewise, some studies describe venire members as men and women, and others describe them as male and female. To avoid introducing addition inconsistencies, this article uses the same terms as the respective studies being referenced, and uses “gender” with “male” and “female” in describing the present study’s findings, which reflect the terms used in the underlying dataset.

2 These particular summary statistics were not included in the original article and have been calculated here using the raw counts provided (Eisenberg, 2017, p. 332 and p. 334). Significance is (χ² = 8.62, p < .01) and (χ² = 6.10, p < .05), respectively.