4,130
Views
65
CrossRef citations to date
0
Altmetric
Articles

Disproportional Imprisonment of Black and Hispanic Males: Sentencing Discretion, Processing Outcomes, and Policy Structures

Pages 642-681 | Published online: 25 Sep 2014
 

Abstract

Disproportional incarceration of black and Hispanic men has been the subject of much critical commentary and empirical inquiry. Such disproportionality may be due to greater involvement of minority men in serious crime, to discretionary decisions by local justice officials, or to the differential impact of sentencing policies, such as mandatory minimums or sentencing guidelines, that differentially impact minority men. This study investigated the extent to which the disproportional punishment of black and Hispanic men, and local variation in such disproportionality, can be attributed to unexplained disparities in local sentencing decisions, as opposed to the extent to which such differences are mediated by sentencing policies, or case-processing and extralegal factors. We use 2005–2009 federal court and Pennsylvania state court data. Our findings suggest, particularly in Federal courts, that most disproportionality is determined by processes prior to sentencing, especially sentencing policies that differentially impact minority males.

Notes

1 United States v. Booker 543 US 220 (2005) and United States v. Fanfan 542 US 296 (2004). We analyze the federal data following these decisions because they rendered the US Sentencing Guidelines advisory, and granted greater sentencing discretion to district courts.

2 The formula for converting b-coefficients of logged sentence length to % difference is ((Exp(b)) − 1) × 100 for positive coefficients and −1((1/Exp(b))−1) × 100 for negative coefficients (see Kurlychek & Johnson, Citation2004). The formula for negative coefficients is designed to remove the floor effects of odds ratios, since odds can never be less than zero (note: odds ratio = exp(b)). This yields, for example, the percent difference between whites and blacks in sentence length.

3 We do not examine incarceration type for federal sentences, because there is no analogy to jail sentences in federal courts.

4 A constant of 0.1 is added to all zero values for the presumptive sentence variable but not the sentence length dependent variable, because we want to retain those cases where an offender’s presumptive sentence was 0 months but s/he still received a prison sentence.

5 Prior studies in PA have used a PCS variable indicating whether a mandatory was applied (Kramer & Ulmer, Citation2009). However, a recent analysis by PCS (see Bergstrom, Kempinen, Ruback, & Tinik, Citation2009) found that this variable undercounts whether a mandatory has been applied. We therefore used an alternative method proposed by PCS that used a combination of information about mandatory eligibility and sentence length to estimate actual application. We examined the influence of both mandatory eligibility and actual application. Substantive results were the same.

6 The inclusion of criminal history did not result in problematic collinearity with the presumptive sentence measure in either the federal or PA analysis (the correlation between logged Guideline minimum and criminal history is .35 in the federal analysis and .51 in the PA analysis).

7 The percent reduction in the imprisonment effect is based on the difference between logit coefficients, or log-odds, across models.

8 Supplemental models show that most of the decline in the black and Hispanic male imprisonment effects in Model 7 is due to presentence detention.

9 This may indicate that these men, especially Hispanic men, are more likely detained before sentencing (bivariate correlations suggest this). Regarding mediation from the trial variable, it may also be that black and Hispanic males receive less favorable guilty plea agreement terms (see Ulmer et al., Citation2010).

10 Doerner and Demuth (Citation2010), using earlier federal data, differentiated race/ethnicity/gender groups with refined age categories found that larger punishment disadvantages characterized young black and Hispanic males, while the effects for older minority males were more similar to ours here.

11 In supplemental models for PA, we entered certain guideline sentencing enhancements into the model. Though enhancements had strong positive effects on sentence severity they did not meaningfully mediate sentencing outcomes for the race-gender groups beyond other legal variables.

12 The observed increase in between-county variation in effects for the incarceration-type decision from model 1 to model 9 may result from controlling for individual level factors that magnify differences in extreme cases across counties that are highly different in their likelihoods of different incarceration sentences. As a robustness check, we ran models using jail as the reference category and found substantively similar results.

13 In the federal sentence length models, we found statistically significant positive interactions for black and Hispanic male status * presumptive sentence. However, the relatively small size of these interaction effects (black: b = .03; Hispanic: b = .01), along with the substantive size of the main effects for black male (b = .01) and Hispanic male (b = .01) in our final length models, mean that these interactions resulted in differences that were not substantively meaningful. There were no other statistically significant interactions in the Federal or PA models.

Additional information

Notes on contributors

Jeffery Ulmer

Jeffery T. Ulmer is a professor and an associate head of the Department of Sociology and Criminology at the Pennsylvania State University. His published research spans topics such as courts and sentencing, criminological theory, symbolic interactionism, religion and crime, criminal enterprise and criminal careers, and the integration of ethnographic and quantitative methods. He was awarded the 2001 Distinguished New Scholar Award and the 2012 Distinguished Scholar Award from the American Society of Criminology’s Division on Corrections and Sentencing. Ulmer and coauthors Darrell Steffensmeier, Casey Harris, and Ben Feldmeyer won the American Society of Criminology’s 2012 Outstanding Article Award. He is the author of Social Worlds of Sentencing: Court Communities Under Sentencing Guidelines (1997, State University of New York Press), and coauthor (with Darrell Steffensmeier) of Confessions of a Dying Thief: Understanding Criminal Careers and Illegal Enterprise (2005, Aldine-Transaction) which won the 2006 Hindelang Award from the American Society of Criminology. His most recent book (with John Kramer), Sentencing Guidelines: Lessons from Pennsylvania was published in 2009 by Lynne Rienner Publishers.

Noah Painter-Davis

Noah Painter-Davis is an assistant professor of Sociology at the University of New Mexico. His main research interests center on how criminal behavior and punishment are stratified across groups (e.g., race/ethnicity, gender, age) and how these differences are shaped by social change and spatial contexts. He has published in Homicide Studies, Criminology, and Social Science Research. His current research projects include studies on immigration and crime, the sources of race-age-gender differences in criminal sentencing, and the effect of the economic downturn on the nature of crime.

Leigh Tinik

Leigh Tinik is a research analyst at the Pennsylvania Commission on Sentencing. Her major focus is on public policy and evaluation research. She has completed numerous legislative reports on alternative sentencing programs in Pennsylvania including the State Motivational Boot Camp Program and the State Intermediate Punishment Program and is published in Crime & Delinquency.

Log in via your institution

Log in to Taylor & Francis Online

PDF download + Online access

  • 48 hours access to article PDF & online version
  • Article PDF can be downloaded
  • Article PDF can be printed
USD 53.00 Add to cart

Issue Purchase

  • 30 days online access to complete issue
  • Article PDFs can be downloaded
  • Article PDFs can be printed
USD 386.00 Add to cart

* Local tax will be added as applicable

Related Research

People also read lists articles that other readers of this article have read.

Recommended articles lists articles that we recommend and is powered by our AI driven recommendation engine.

Cited by lists all citing articles based on Crossref citations.
Articles with the Crossref icon will open in a new tab.