The intersection of race and gender: an examination of sentencing outcomes in North Carolina

Abstract

This study examines the intersection of offenders’ race and gender in the sentencing process using data on felony cases sentenced in North Carolina. Analyses examine the likelihood that charges were reduced in severity between initial filing and conviction, the likelihood of imprisonment, and the length of sentence imposed, and test whether race affects punishment similarly for men and women. Results indicate that status characteristics predict both reductions in charge severity and the severity of the final sentence, and that racial disparity is conditional on gender. However, the results are not entirely consistent with predictions derived from the extant literature. Gender significantly predicts case outcomes at each stage, but black men were not uniformly disadvantaged, and black women received the least severe treatment in two out of four analyses. Theoretical implications for the intersection of race and gender in sentencing theories are discussed.

Keywords:

• sentencing
• charge reductions
• race
• gender
• incarceration

Notes

1. Although imputing missing values in many of these cases might be possible, there is little to be gained. The small number of cases dropped solely due to missing data are not likely to bias our results and do not affect the statistical power of the analyses. We limit the sample to white and black defendants because there were too few women in any other group to include race x gender interactions. Asian, Hispanic, Native American, and Other races are 4% of the entire sample. It is still possible that some Hispanics are coded as white, which could result in underestimation of disparity between blacks and whites. According to the 2000 Census, 4.7% of North Carolina’s population was ‘Hispanic or Latino’.

2. We do not measure reductions in the amount of charges because multiple sentences are predominantly served concurrently in North Carolina. Thus, reductions in the number of charges have little effect on sentence length.

3. Holleran and Spohn (2004) and Harrington (2008) demonstrate the benefits of differentiating between jail and prison sentences. However, in North Carolina, sentences over 90 days must be served in state prison. The shortest possible active sentence for a felony is three months.

4. We examined a more common measure of age with four categories: 17–20, 21–29, 30–29, and 40 and older. Since there was no evidence of a curvilinear effect, we present the simpler two-category measure. Additionally, we ran the models with age as a continuous variable, which did not change the significance or direction of the results for age and all other variables in the equation.

5. This measures the plea that was in effect when the defendant was convicted. A separate variable in the data-set ‘mode of disposition’ (judge or jury) indicates that all ‘not guilty’ pleas were convicted by juries. Nearly all pleas of guilty or no contest were convicted by judges, although 20 cases are coded as pleading guilty and convicted by a jury. These may be cases in which the defendant plead guilty after a trial hadegun.