Revisiting the Minority Threat Perspective: Examining the Main and Interactive Effects of Segregation on Sentencing Severity

Abstract

Drawing on the minority threat perspective, this study assesses the main and interactive effects of racial/ethnic segregation on the incarceration and sentence length decisions. Using the State Court Processing Statistics in combination with other data, we employ multilevel models to examine (1) whether racial/ethnic segregation affects sentencing severity, (2) whether racial/ethnic segregation moderates the effects of racial/ethnic composition, and (3) whether racial/ethnic segregation reduces racial/ethnic disparity in punitive sentencing. Both racial and ethnic segregation displayed mitigating effects on punishment severity. In particular, racial segregation reduced the effect of racial composition on probabilities of receiving a prison term, and ethnic segregation reduced the effect of ethnic composition on probabilities of receiving a jail sentence. Our results suggest that the minority threat perspective and its theoretical accounts of segregation can be partially applied to sentencing outcomes in state courts.

Keywords:

• Sentencing
• minority threat
• racial segregation
• ethnic segregation

Acknowledgments

We want to thank Travis Pratt for his very insightful and detailed feedback on earlierd drafts of this paper and Danielle Wallace for her suggestions. We also want to thank Dr. Michael Leiber, Dr. Bryanna Fox, and anonymous reviewers for their constructive feedback and insights.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Although conceptually distinct concepts, the possibility that segregation levels increase perceived levels of threat introduces possible issues with multicollinearity. Accordingly, we conducted bivariate correlations and variance inflation factors. The bivariate correlation between Hispanic population size and ethnic segregation was .368 and the bivariate correlation between Black population size and racial segregation was .660. The variance inflation factors were all below 4, indicating no harmful levels of multicollinearity.

2 The most recent SCPS data that are available are for 2009. As such, it is possible to examine the effects of racial and ethnic segregation on sentencing using the 2009 SCPS data and 2010 US Census data. Despite that, we instead focused on the effects of racial and ethnic segregation in 2000 for at least two reasons. First, the number of immigrants hit the peak levels in 1999 and 2000 and declined after 2001 as the US economy took a downturn after September 11 (Passel & Suro, 2005). Second, the peak levels of immigration in 1999–2000 have implications for racial and ethnic segregation because, as Massey (2001, p. 402) eloquently argued, segregation levels are inclined to increase during periods of rapid immigration. Given this unique social and historical context in 1999–2000, it is especially important to assess the effects of racial and ethnic segregation in this period. It is important to note that ethnic segregation has remained relatively stable since the early 2000s (Tienda & Fuentes, 2014), and racial segregation has also remained entrenched at the macro-level (Lichter et al., 2015). Therefore, our data and analyses may still provide useful insights into the effects of racial and ethnic segregation on the court processing of convicted felons.

3 Our analyses focus on convicted felons. Misdemeanors were removed because previous scholars have noted that different procedures preside over the sentencing of misdemeanors compared to felonies (e.g., Tonry, 1996).

4 Baumer and Martin (2013) suggest that there exist statutory differences between states in the type and length of sentence for each offence. Therefore, our results must be interpreted with caution as these statutory differences may partially contribute to variation in sentencing severity across jurisdictions. Ideally, we could employ a three-level fixed effects model; however, the small number of states (23) and counties nested within states (some states only have one county) in the SCPS data does not allow sufficient degrees of freedom to analyze decisions to sentence convicted felons to prison or jail rather than to noncustodial sanctions.

5 We use the maximum length available in the data (1133 months) for defendants receiving life imprisonment or death penalty. Thus, cases where the defendant received life imprisonment (N = 71) or death penalty (N = 1) are coded as the maximum prison length of 1133 months.

6 The index of dissimilarity is a predominant measure of segregation and captures the broader dimension of separateness (Xie, 2010). We use this measure for two additional reasons. First, after analyzing 19 different segregation indices, Xie (2010) suggested that other measures of separateness did not offer any empirically unique information beyond the index of dissimilarity. Second, Massey et al. (2009, p. 76) have argued that compared to the index of dissimilarity, other measures of segregation are less useful for aggregated units of analysis beyond the neighborhood level.

7 Counties are an ample unit of analysis for the current study for at least two reasons. First, recent empirical work has shown that segregation is best aggregated to the county, metropolitan, or state level in order to include residents who move beyond the urban core to escape increasingly diverse inner-city neighborhoods (Lichter et al., 2015). Second, many courts serve and operate at the county level. Moreover, judicial elections are one prominent avenue through which the surrounding context can influence decision making (Eisenstein et al., 1999), and judicial elections are predominately based on county lines.

8 Percent Black/Hispanic differs from the index of dissimilarity because a county may contain a small Black/Hispanic population while also being 100% segregated if the entire Black/Hispanic population lives in census tracts completely separated from Whites (see Xie, 2010).

9 These offenses include violent offenses (i.e., murder, rape, robbery, assault, and other violent offense), property offenses (i.e., burglary, larceny-theft, motor vehicle theft, forgery, fraud, other property offenses), drug offenses (i.e., drug sales and other drug offenses), public order offenses (i.e., weapons, driving-related, other public-order offenses), and unknown felony. Notably, despite the weaknesses of measuring offense severity in the SCPS data, it seems unlikely that the confounding of the effects of offense severity and race/ethnicity would vary systematically with segregation or minority composition. We thank an anonymous reviewer for pointing it out.

10 We reran all the analyses with listwise deletion and found that the findings concerning the effects of racial and ethnic segregation were almost the same as when using multiple imputation. In fact, we estimated and compared the marginal effects with the complete data because STATA’s margins command with Bonferroni adjustments for multiple comparisons could not be performed with multiply imputed data.

11 We included the squared terms of county-level Black/Hispanic population sizes and racial/ethnic segregation in the incarceration and sentencing length models, but removed them when they were not statistically significant.

12 We also assessed if the interaction between racial/ethnic segregation and racial/ethnic composition is stronger for Black/Hispanic felons by including the three-way interaction terms between race/ethnicity, racial/ethnic segregation, and racial/ethnic composition, as well as all the two-way interactions. None of these three-way interactions, as well as marginal effects in the incarceration model, was statistically significant (results are available upon request).

13 We conducted ancillary analysis to examine if there is a differential set of effects for ethnic segregation in border states as opposed to other states. Our analysis indicated that ethnic segregation had similar effects in border states versus other states.

Additional information

Notes on contributors

Laura Beckman

Laura Beckman, Ph.D., is an Assistant Professor in the Department of Criminal Justice at Shippensburg University. Her primary research interests include race/ethnicity, immigration, and justice with a focus in juvenile justice and sentencing. Her work has appeared in Criminal Justice and Behavior and Journal of Interpersonal Violence.

Xia Wang

Xia Wang, Ph.D., is an Associate Professor at Arizona State University’s School of Criminology and Criminal Justice. She is involved in studies of race and ethnicity and their effects on crime and criminal justice, and the use of various analyses to test and extend criminological theories. Her work has appeared in Criminology, Journal of Research in Crime and Delinquency, Journal of Quantitative Criminology, Justice Quarterly, Law & Society Review, and other journals.