![Sample our Law journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5942/TOC-Subject-Sample-2021-Law.jpg"})
ABSTRACT
Research has yet to fully investigate whether the effects of citizenship and legal status are moderated by age, gender, and race/ethnicity. This is a significant oversight because negative stereotypes of young, male, black, and Hispanic offenders may intensify overall negative perceptions for noncitizen and undocumented immigrant offenders. Using data from the United States Sentencing Commission for fiscal years 2006–2008, we investigate whether the effects of citizenship and legal status on sentencing outcomes are moderated by age, gender, and race/ethnicity. The results indicate that the effects of citizenship and legal status on incarceration decision, but not sentence length decision, are moderated by age, gender, and race/ethnicity. Further, young Hispanic male noncitizen and undocumented immigrant offenders are among the groups that have the greatest odds of being sentenced to prison. We discuss the implications of these findings and conclude with a discussion of future research.
Acknowledgments
We would like to thank Cassia Spohn for her feedback on the previous draft of this article.
Notes
1. More recently, Donald Trump, then the Republican presidential candidate commented, “But you have people coming in and I’m not just saying Mexicans, I’m talking about people that are from all over that are killers and rapists and they’re coming into this country” (as cited in Scott Citation2015:1).
2. In a recent New Jersey case, the prosecutor moved to increase the defendant’s bail amount after ICE issued a detainer arguing that the detainer increased the risk that the defendant would not appear at his criminal trial. The trial court agreed and set a new (higher) bail; the New Jersey Supreme Court upheld the trial court’s decision on appeal.
3. There were 153,075 cases excluded from the analysis. The majority of cases were omitted because they are not drug offenses. The number of cases with missing information on citizenship and legal status was 13,219 or about 11 percent of the total cases removed from the analysis (n = 153,075). We removed four districts (Arkansas West, Washington East, North Mariana Island, and Oklahoma East) from the analysis because all drug offenders in those districts were sentenced to prison, and dummies representing these districts were automatically removed from the analysis.
4. Although 96 percent of the sample received a prison sentence and this binary outcome of incarceration decision is skewed dramatically toward 1, we estimated the incarceration model for several reasons. First, incarceration is an important sentencing decision, and it has profound and pernicious collateral consequences for offenders who are sentenced to prison (Mauer and Chesney-Lind Citation2002). Second, when analyzing federal sentencing data, prior studies have often examined incarceration decision (e.g., Demuth Citation2002; Johnson and Betsinger Citation2009; Light et al. Citation2014; Steffensmeier and Demuth Citation2000). Following this research practice allows us to compare our results to prior research. Third, whereas 96 percent of the sample received a prison sentence, this percentage varies by several factors, such as citizenship and gender. For example, whereas 97 percent of male defendants received a prison sentence, 89 percent of female defendants did. Thus, it is important to conduct multivariate models to assess whether these factors are indeed related to incarceration decision, net of controls. Fourth, given the unbalanced data, logistic regression can sharply overestimate the probability of prison sentencing by affecting the estimate of the model intercept (King and Zeng Citation2001). Thus, we caution that scholars may want to place more emphasis on the model estimates of the coefficients of variables of interest than predicted probabilities.
5. Sentence length was capped at 470 months. This was done to remove outliers such as life sentences that would affect the mean and distribution.
6. In analyses of two-way interactions between age and citizenship/legal status, we found that age cutoffs at 24/25, 34/35, and 39/40 produced similar findings compared to the 29/30 cutoff. In analyses of four-way interactions among age, gender, race/ethnicity, and citizenship/legal status, we found that using other age cutoffs resulted in multiple groups being omitted from the analysis due to their small group size. It was especially the case when the interaction effect involving legal status was examined. That said, using different age cutoffs in analyses of four-way interactions produced largely similar findings. The results, available on request, indicate the need to examine varying age cutoffs in interactional analyses and the validity of using a 29/30 cutoff.
7. Downward departures for substantial assistance (5K1.1) are initiated through a motion of the U.S. Attorney. This departure is based on the premise that the offender has provided the prosecutor with information that has aided in the furtherance of an investigation.
8. Downward departures under Federal Rule 5K2 are initiated by judges. Judges retain the discretionary power to sentence outside of the guideline recommendations.
9. Downward departures based on the “fast-track” program (5K3.1) can be initiated by a government motion or by a judge. This program is based on the stipulation that individuals must plead guilty to the charges and waive their right to an appeal (Gorman Citation2009; Office of the Attorney General Citation2003). This is done to expedite the court process and can be viewed as a type of reward for those who participate in the program. It is often the case that participants in the fast-track program are offenders who face a high probability of being deported (Bolla Citation2006; Office of the Attorney General Citation2003).
10. Similar to the issues confronted regarding the sentence length variable, presumptive sentence is used in its natural log form. The natural log of the presumptive sentence was taken to adjust for skewness.
11. The U.S. Sentencing Commission scale rating for prior criminal history ranges from one to six (see Johansson and Betsinger Citation2009). The guidelines expressly state that “The total criminal history points from §4A1.1 determine the criminal history category (I–VI) in the Sentencing Table in Chapter Five, Part A. The definitions and instructions in §4A1.2 govern the computation of the criminal history points. Therefore, §§4A1.1 and 4A1.2 must be read together” (USSC 2012, p. 369).
12. Because sentence length models include only those cases that received an incarceration sentence, we created a hazard rate to control for potential selection bias when predicting sentence length (see Berk Citation1983; Bushway, Johnson, and Slocum Citation2007). Based on prior research, we account for predicted probability of offender exclusion from the sentence length sample (Spohn and Holleran Citation2002). To this point, sentence length is “a function not just of the usual linear combination of regressors (which suffices in the original population), but [of] a hazard rate capturing the impact of the selection equation” (Berk and Ray Citation1982:369). In –, we presented model results with the hazard rate included. We also examined additional models that do not include a hazard rate for selection bias, and found that findings regarding citizenship and legal status were almost identical (results available on request).
13. The results of bivariate correlations (see the Appendix), variance inflation factors (VIF), and condition indexes indicated that multicollinearity was not a problem.
14. One anonymous reviewer suggested that, given the association of blacks with crack cocaine and the harsher sentences imposed on offenders convicted of offenses involving crack cocaine, one might predict that young black male noncitizens would be sentenced more harshly than other offenders. To assess this possibility, we reran our analysis using young black male noncitizens as the reference category (results available on request). We found that young black male noncitizens were not sentenced more harshly than other groups and that using young Hispanic male noncitizens was more likely to reveal statistically significant interactions among citizenship/legal status and age, gender, and race/ethnicity.
15. One anonymous reviewer suggested that drug type and citizenship/legal status may interact. To assess this possibility, we included the interaction terms between citizenship/legal status and drug type and found significant interaction effects between citizenship/legal status and some drug types (results available on request). For example, being a noncitizen amplified the effect of cocaine on incarceration decision such that noncitizens who were involved in cocaine crime were sentenced more harshly than noncitizens who were involved in marijuana crime.
16. Prior to 2010, defense attorneys were not obligated to inform their noncitizen clients that they could face deportation as a result of a guilty plea (Padilla v. Kentucky Citation2010). In some cases, counsel was misinformed or unware of such consequences. As a result, some defendants may have been advised to plead guilty and get what they believed to be a short prison sentence. Deportation may be even more severe than prison for noncitizens because deported defendants are removed entirely from the United States and are often unable to enter the United States through legal channels.
17. We should note that it is likely that prosecutors use fast-track as a tool to get noncitizens to plead guilty with the threat of longer prison sentences. Since only about 5 percent of the sample received a fast-track departure, we are limited in our speculation of the role of fast-track departures. Further, fast-track departures were not widespread across districts between 2006 and 2008, so some defendants would not have had the opportunity to receive this type of departure.
Additional information
Notes on contributors
Mercedes Valadez
Mercedes Valadez, Ph.D., is an Assistant Professor of Criminal Justice at California State University at Sacramento (CSUS). Her research focuses on courts and sentencing, immigration and Latinos. Her current work examines the effects of race/ethnicity and citizenship/legal status on sentencing decisions.
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 she is also interested in testing and extending criminological theories. Her work has appeared in Criminology, Law & Society Review, the Journal of Research in Crime and Delinquency, and the Journal of Quantitative Criminology.