Shared Race/Ethnicity, Court Procedural Justice, and Self-Regulating Beliefs: A Study of Female Offenders

Abstract

Using survey data from a sample of white, black, and Hispanic incarcerated females (N = 554), we examine if the theoretically hypothesized and empirically demonstrated relationship between procedural justice and obligation to obey the law is substantiated among a sample of offenders and explore the impact that sharing the race/ethnicity of the defense attorney and prosecutor in their most recent conviction has on female inmates' perceptions of court procedural justice and their perceived obligation to obey the law. The findings reveal that female offenders who perceive the courts as more procedurally just report a significantly greater obligation to obey the law. In addition, white female inmates who had a white prosecutor were significantly more likely to perceive the courts as procedurally just. Non-whites, though, perceive the courts as more fair if they encountered a minority prosecutor regardless of whether the prosecutor was black or Hispanic.

Prior research demonstrates the effectiveness of procedural justice for securing the public's perceived obligation to obey the law (Jackson et al. 2012; Mazerolle et al. 2013; Reisig et al. 2012; Sunshine and Tyler 2003; Tyler 2006; Tyler and Huo 2002; Wolfe 2011). This finding is important because it indicates that criminal justice actors' authority and ability to secure public compliance may be strongly tied to the fairness of the procedures they use—a factor authorities have some control over (Mazerolle et al. 2013; Sunshine and Tyler 2003). There remain, however, several areas for additional exploration in the extant literature on process-based models of regulation.

First, previous studies assessing the impact of procedural justice on people's obligation to obey the law have focused almost exclusively on perceptions of police, devoting far less attention to perceptions of the courts (Mazerolle et al. 2013; Sunshine and Tyler 2003). This is surprising because, as Garland (1990: 71–72) has explained, the contemporary courtroom is “the forum where justice is done” and “a ritual in which society as a whole is deemed to participate.” As such, the perceived fairness of court procedures may play a particularly important role in legitimizing or delegitimizing the authority of the law in the eyes of both the public and offenders. In addition, there are reasons to expect that prior findings concerning the police may not generalize to the courts. Prior research, for example, suggests that people's perceptions of police are consistently and substantially more positive than their perceptions of the courts, which likely reflects citizens' greater familiarity and contact with police (Roberts and Hough 2005). Indeed, Tyler (2006) found that individuals rely more heavily on procedural justice perceptions when evaluating perceptions of the courts versus perceptions of the police. There are also factors unique to court processing—such as the type of attorney (private versus public) representing the defendant and plea-bargaining negotiations—that may influence court procedural justice perceptions. Thus, we explore the antecedents of people's perceptions of the courts and how they may uniquely impact people's obligation to obey the law.

Second, existing theoretical scholarship suggests that racial or ethnic identification with criminal justice actors—that is, sharing the racial or ethnic characteristics of legal authorities—may result in more positive assessments of those authorities (Tyler and Huo 2002; Weitzer and Tuch 2006). Much of this work extends from the theoretical propositions of the group-value model which suggest that being able to identify with an authority may increase perceptions of status which in turn increases perceptions of procedural justice (Lind and Tyler 1988). Yet empirical evidence about this hypothesized relationship is scarce, and the findings from the handful of studies that have explored the issue have been mixed (Kelleher and Wolak 2007; Scherer and Curry 2010; Sharp and Johnson 2009; Tyler and Huo 2002). Thus, there is a question as to whether the diversification of criminal justice organizations, done in part to enhance the legitimacy of those organizations among minorities, is actually effective. For example, an alternative perspective extending from critical race theory (Haney-López 2003) suggests that any support for institutions such as the criminal justice system which historically have protected the rights of whites over minorities is tantamount to supporting white hegemony. As such, interacting with non-white legal authorities may do little to enhance minorities' perceptions of the criminal justice system. Thus, this study further investigates the impact of shared racial identity within the criminal justice system by examining how sharing one's race/ethnicity with court actors affects individuals' perceptions of the courts and their obligation to obey the law.

Third, this study builds on the small but growing body of literature that examines what is known about both the racial correlates and potential consequences of procedural justice perceptions among persons who are actually embedded in a criminal lifestyle and at a high risk for future offending, such as inmates (for example, see Casper, Tyler, and Fisher 1988; Fagan and Piquero 2007; Papachristos et al. 2012; Piquero et al. 2005). Likely because of the difficulties in gaining access to and surveying inmates (Pickett et al. 2014), previous investigations have focused almost exclusively on either the general public or on minor offenders, such as persons stopped for traffic violations or tax offenders (Murphy 2004 2005; Tyler 2006). In addition, the existing studies that explore offenders' procedural justice perceptions have largely concentrated on male inmates' views about the staff and processes of correctional facilities (Beijersbergen et al. 2014; Henderson et al. 2010; Reisig & Meško 2009) with fewer studies examining female offenders (Tatar et al. 2012). We further develop this literature by examining female inmates' views about court procedural justice.

This study addresses these gaps in the literature using data from a survey of female inmates incarcerated at the largest female correctional facility in Florida. The survey included measures of the racial/ethnic characteristics of the court actors with whom the inmates interacted, and the inmates' perceptions of court procedural justice and obligation to obey the law. Our focus on female offenders is important because they have received less attention in research than male offenders (Belknap 2007), yet the proportion of serious crime committed by female offenders has increased in recent decades (Lauritsen et al. 2009). In the sections that follow, we review the theoretical work that informs our analyses as well as the relevant prior research, after which we describe our methodology and discuss the results of our study.

Theoretical Background and Prior Research

Procedural Justice and the Role of Voice

A key assumption on which theories of procedural justice are based posits that regardless of the outcome, individuals may perceive the process of an interaction as just if certain procedural criteria are met (Lind and Tyler 1988; Thibaut and Walker 1975; Tyler 1994). In other words, how a decision is made may be as important to individuals as the actual decision. Just procedures are important because in disputes positive outcomes are rarely achieved for all participants. Research suggests even when negative outcomes occur the individual may be more cooperative and satisfied with the results and the decision makers if the process is viewed as equitable (Cohen-Charash and Spector 2001; Tyler 2006; Tyler and Huo 2002; Tyler et al. 1985).

Within this framework Thibaut and Walker (1975) found that the ability of individuals to take part in the process was critical for promoting the perception that a fair process occurred. They described this action as process control which has come to be referred to as voice in the procedural justice literature (Folger 1977). Tyler (1987) found that value-expressiveness or having one's voice and opinion heard irrespective of the impact was paramount in predicting procedural justice. Several meta-analyses describe voice as a strong consistent predictor of procedural justice (Cohen-Charash and Spector 2001; Colquitt et al. 2001).

While much of the literature on voice and procedural justice has been examined in organizational and business contexts, the importance of voice has also been demonstrated in procedural justice judgments of the criminal justice system. For example, prior studies have found direct and significant relationships for the effect of voice on procedural justice judgments of the police and courts (Tyler 2006) and on procedural justice judgments in felony cases (Casper, Tyler, and Fisher 1988). Additional summaries on five separate works are detailed by Lind and Tyler (1988); each study found a positive link between voice and assessments of procedural justice. However, virtually all the prior literature on voice (alternatively, see Casper, Tyler, and Fisher 1988) examines it within lower stakes encounters, especially when compared to the seriousness of a criminal court proceeding.

Court Procedural Justice and Obligation to Obey the Law

The extant research on the impact of procedural justice on people's obligations to obey extends from the theoretical foundation of the process-based model of regulation (Tyler 1989). The model's key theoretical propositions are rooted in the argument that sanction-based models of social control (i.e., outcome-based or deterrence-based models) are limited in affecting individual behavior over the long term. Instead, the process-based approach is built on two alternative assumptions. First, people's perceptions of procedural justice and of authorities' motives affect their felt obligation to obey the law, even if they do not receive the outcome they desire (Sunshine and Tyler 2003). Second, authorities have the ability to affect whether people perceive them to be procedurally just (Tyler and Huo 2002).

Among the general United States population, a positive link between police procedural justice and obligation to obey has been empirically demonstrated in a number of studies (Sunshine and Tyler 2003; Tyler 2006). Indeed the link between procedural justice and obligation to obey is generalizable cross-nationally and in various contexts (Mazerolle et al. 2013). For example, Murphy (2004, 2005) found that Australians were more likely to pay their taxes if they perceived tax authorities' behavior as procedurally just. Levi and Sacks (2009) found that even in developing nations individuals' perceptions of procedural justice predicted obligations to obey the law. However, far less research has focused on the courts.

Within the court setting, Thibaut and Walker (1975) produced the foundational work espousing the importance of procedural justice. They showed that individuals involved in simulated adversarial trials perceived the trial as fairer than those in an inquisitorial setting. Casper, Tyler, and Fisher (1988) extended this line of research by testing the effects of procedural justice among a sample of individuals actually engaged in the criminal justice system—felony defendants. They found that the procedures used influenced defendants' perceptions of court procedural justice. For example, respondents' perceptions about how they were treated by the police affected their views about the courts. However, factors such as defendants' criminal record, and whether they had a public defender or private attorney did not impact views of the courts (Casper, Tyler, and Fisher 1988). Unfortunately, neither of these studies evaluated whether court procedural justice was correlated with respondents' obligation to obey the law, or whether the court actors' demographics affected defendants' views about the courts or the law.

A more recent study of female inmates in Florida found that inmates' age, education, and ethnicity were associated with perceptions of court procedural justice, but not with perceptions of police procedural justice (Baker et al. 2014). Other correlates including perceived judge honesty and voice also emerged as significant predictors of court procedural justice. Importantly, using a factor analytic approach Baker et al. (2014) found that perceptions of police procedural justice were distinct from perceptions of court procedural justice. In addition, similar to Casper, Tyler, and Fisher (1988), the study's most important finding showed that perceived police procedural justice was an antecedent to perceived court procedural justice. This finding suggests that of the two types of procedural justice perceptions—police and courts—perceived court procedural justice may constitute the more proximate determinant of offenders' obligation to obey the law. However, like Casper, Tyler, and Fisher (1988), Baker et al. (2014) did not evaluate whether offenders' procedural justice perceptions were associated with their obligation to obey the law, nor did they investigate whether sharing the race/ethnicity of legal actors impacts inmates' procedural justice perceptions.

While several studies have linked perceptions of court procedural justice to satisfaction with and confidence in the courts (Benesh 2006; Buckler et al. 2007; Olson and Huth 1998; Tyler and Huo 2002), studies examining the effect of court procedural justice on either the public's or offenders' obligation to obey the law remain rare. The few investigations that have examined the impact of perceptions of the courts on people's obligation to obey the law have conflated perceptions of the courts with perceptions of the police by combining the two types of perceptions into a single item—perceptions of legal authorities (Tyler 2006; Tyler and Huo 2002). This conflation of perceptions of police and courts overlooks the nuanced nature of these two types of perceptions, and potentially obscures the unique correlates and effects of each (Baker et al. 2014; Sprott and Doob 2009). For example, research by Papachristos et al. (2012) demonstrated that while prior offenders' procedural justice perceptions of the police were significantly associated with perceived legitimacy, no such relationship emerged when the focus was on perceived procedural justice of prosecutors. It remains unclear, then, whether court procedural justice can impact self-regulating beliefs among offenders, and whether the racial characteristics of legal authorities are important for understanding why offenders' perceive them as being just or unjust.

Importance of Shared Race/Ethnicity with Authorities

The importance of shared race/ethnicity with authorities is rooted in social identity theory which highlights the importance of people's in-group and out-group social categorizations (Abrams and Hogg 1990; Hogg 2006). Such classifications are not necessarily complex but instead “draw on readily accessible social categorizations (e.g., gender, race, profession)—ones that are valued, important, and frequently used aspects of the self-concept” (Hogg 2006: 119). “Groups and their members strive for positive intergroup distinctiveness because in salient group contexts the ‘self,’ as social identity, is defined and evaluated in group terms, and therefore, status, prestige, and social valence of the group attaches to oneself” (Hogg 2006: 120). Shared social identity derives its impact from individuals' ambiguity about either their position/status, or the position/status of their in-group, in the public sphere. As a result, they rely on in-group members, especially those in positions of authority, as valid information sources about their status, and they adjust their views accordingly (Abrams and Hogg 1990).

The socio-psychological significance of shared race/ethnicity with authorities is further articulated within the procedural justice literature through Lind and Tyler's (1988) group-value model. Lind and Tyler (1988) posit three antecedents of procedural justice: status, neutrality, and trust. Lind and Tyler's (1988) status component of the group-value model emphasizes the importance of being a member of a respected and authoritative group. In one of the first tests of this model, Tyler (1989) found that all three factors of the group-value model, in addition to perceived voice, significantly influenced perceptions of procedural justice. However, an important limitation of Tyler's (1989, 1994) research on the group-value model is that it did not directly consider the potential effects of authorities' demographic attributes, such as their race or ethnicity. This is notable because the group-value model suggests that such identification may shape status perceptions and, in turn, influence views about authorities.

In a seminal study conducted with a sample of Oakland and Los Angeles residents, Tyler and Huo (2002) directly evaluated the importance of citizens' shared ethnicity with legal authorities, broadly defined, with whom they had recently interacted (e.g., the citizens' had called police for help, gone to court as a plaintiff, etc.). However, their study yielded mixed evidence about the effects of such interactions. Although black and Hispanic citizens' willingness to accept decisions made by authorities did not vary depending on whether the authorities were white or non-white, white citizens were significantly more willing to accept decisions made by white decision makers, even after controlling for process issues. To our knowledge, however, there is no existing evidence that specifically evaluates whether serious offenders who interact with racially similar legal authorities, be it judges, lawyers, or law enforcement, are more likely to perceive greater procedural justice or feel a greater obligation to obey the law.

The Role of Race in Developing Perceptions of the Courts

In addition to the socio-psychological aspects of social identity theory and procedural justice theory, critical race theory such as articulated by Haney-López (2000, 2003, 2006) also provides guidance for understanding the potential importance of race in developing individuals' perceptions of the criminal justice system. For example, white and non-white perceptions of obedience, fairness, and legitimacy of legal authorities may greatly differ (Tyler and Huo 2002; Weitzer and Tuch 2006). Haney-López (2006) posits that racial differences may be the result of different approaches to how whites and non-whites approach lawful obedience. According to his theory obedience to the law may be highly rational, that is, instrumental, which can be juxtaposed against acquiescence to the law which involves accepting the norms and assumptions underlying the law and legal system (Haney-López 2006). This may explain the potential for differences among whites and non-whites in how their judgments about obedience to the law are evaluated. The theory postulates that acquiescence to the law creates complicity on the part of individuals for the legal constructions and assumptions of race that have become part of the law—that is, a norm of white superiority. As Haney-López (2006: 21) explicates “Whatever the language used, it is clear that white identity is tied inextricably to non-white identity as its positive mirror, its superior opposite.”

For whites, acquiescence to the law that protects their superior racial identity may indicate that their judgments will be more normative—based on aspects such as perceived fairness—and interacting with white authorities would further reinforce their underlying normative ideals about the law and white superiority. Alternatively, non-whites may be obliged to obey the law as opposed to being acquiescent to the law and as such their obedience may be more instrumental and based on aspects such as having the ability to control the outcome to some degree, for example, through participating in a negotiated plea as opposed to being subjected to a trial or experiencing “voice” in their encounters with the courts. Extending these theoretical principles to shared versus nonshared racial interactions, non-whites interacting with white legal authorities may further reinforce the value of instrumental as opposed to normative aspects of the criminal justice system. This in turn may negatively impact their judgments about normative aspects of the law such as the fairness of the system. Non-whites may be more attuned to the racism, albeit unconscious or “script” racism (Haney-López 2000), present in the criminal justice system particularly the courts.

This unconscious institutional racism may be additionally emphasized when non-whites interact with white court actors who may be unable to fully empathize with the “structural forms of racism” faced by many minorities in the United States (Lynch 2014: 122). Even white prosecutors who may be able to empathize with non-white defendants “run the risk of reprimand within their offices, in addition to marking themselves as soft touches in plea negotiations” if they allow such empathetic judgments to affect their decisionmaking (Lynch 2014: 124). This may be critical for how non-whites view the courts as prosecutors have near total control over charging decisions and penalty enhancements giving them a great deal of power in plea negotiations. As such, white prosecutors who have a great deal of control over plea negotiations but who are unable to empathize with non-white defendants may be less likely to offer favorable pleas, thus, diminishing non-whites perceived fairness of the courts and perceived obligation to the obey the law. Findings from one recent study seem to bear this out as black and Hispanic defendants were more likely to receive custodial plea offers—those involving jail or prison time—over noncustodial plea offers—those involving community service, probation, and so forth—than were whites (Kutateladze et al. 2014).

The Importance of a Focus on Female Offenders

Females provide a special group of individuals to examine due to their shifts in criminal offending over the past several decades. For example, arrests of females more than doubled from 1980 (∼1.2 million arrests) to 2012 (over 2.8 million arrests) (Snyder and Mulako-Wangota 2014). During roughly that same time period, the number of females incarcerated in state and federal prisons grew by a factor of nearly 9 (from 12,746 in 1978 to 111,287 in 2013) (BJS 2014). Examining why females offend or, alternatively, why they obey the law has been widely ignored among contemporary theories (Belknap 2007). But, females now make up a considerable proportion of criminal offenders and are responsible for more crimes than at any other time in United States history. Official statistics suggest that they account for approximately 26 percent of offenses (Federal Bureau of Investigation 2013). This partly reflects the fact that for reasons that remain elusive, the steep decline in criminal offending among males during the past few decades did not replicate among females (Lauritsen et al. 2009). Indeed, the 10-year arrest trends from 2003 to 2012 indicate that while offenses attributed to adult males dropped 12.7 percent, offenses attributed to adult females increased 2.9 percent (FBI 2013). Understanding the dynamics of this increase is critical to devising ways to reduce the growth in female offending.

The nature of and pathways to offending may also be different for females than for males (Jones et al. 2013; Steffensmeier and Allan 1996). For example, responding to sexual and other physical abuse seems to be a key reason females turn to crime (Chesney-Lind and Rodriquez 1983; Hubbard and Pratt 2002; McCartan and Gunnison 2010; Salisbury and Van Voorhis 2009). However, legitimate outlets for females to escape abusive situations could alleviate some of the growth in female offending. One possible outlet is the ability of women to turn to the criminal justice system. But, their perceptions of the fairness and legitimacy of legal procedures and processes could affect their willingness to engage the system. For example, positive views of the fairness and legitimacy of the criminal justice system could provide women with a pro-social pathway to escape abuse. Alternatively, negative attitudes could potentially result in law-breaking as a means to solve problems or escape negative situations.

The Current Study

Based on theoretical expectations and the extant literature on voice, racial identification, procedural justice and obligation to obey the law, and critical race theory, we aim to answer four research questions. First, while the importance of perceived voice for promoting perceptions of procedural justice has been established in low stake encounters, what affect will perceived voice have on perceptions of procedural justice among convicted offenders in criminal court proceedings? Second, does the link between procedural justice and people's obligation to obey the law, which has been observed for perceptions of the police and overall measures of legal authorities, hold true when perceptions of the courts are examined? Third, is the hypothesized and empirically demonstrated relationship between procedural justice and obligation to obey the law substantiated among a sample of serious offenders? Fourth, what impact will shared racial identification with court actors have on female inmates' perceptions of court procedural justice and their obligation to obey the law?

Data and Methods

The data were collected on a single day in 2010 at a medium security, privately run, female correctional facility in Florida. Inmates are assigned to facilities based on a number of factors but are eligible for assignment at any facility for which they are appropriately scored. Though some literature and debate exist on the efficacy of public and private facilities, most findings suggest little definitive difference between the two types of institutions. For example, a meta-analysis of studies on various aspects of prison administration conducted by Perrone and Pratt (2003) found no conclusive evidence of differences in prison condition, prison management, activities available to inmates, quality of care, security, safety, or order between private and public facilities. Additionally, research specifically conducted on Florida inmates, found no significant difference in recidivism rates for female inmates who were housed at private versus public facilities (Bales et al. 2005). As such, we are cautiously confident that despite the fact that this survey comes from a sample of inmates housed at a private facility there is little to suspect that they are qualitatively different from female inmates housed at public facilities in Florida. Prior to data collection this project was reviewed by the appropriate University's Internal Review Board, the Florida Department of Management Services, and the warden of the correctional institution involved. Data collection was conducted by graduate students trained in survey administration with the assistance of prison educational staff. No incentives were provided to inmates for taking the survey, and they were allowed up to 45 minutes to complete the questionnaire.

A total of 1,256 surveys were administered to inmates housed in the prison on the date of administration. Respondents were limited to any inmate housed in one of the five housing units throughout the facility with the exception of two pods within one of the housing units. Researchers were unable to survey inmates housed in these two pods as a result of limited research personnel. There is little to suggest that the inmates in the two pods fundamentally differed from the other inmates surveyed; inmates at the facility were assigned to nonprogram-based dorms so any error potentially introduced by not being able to survey these pods would be random. In total, 83 percent of the inmates housed at the facility were surveyed. Inmates granted informed consent and filled out self-administered questionnaires containing 82 questions on a variety of topics from attitudes about marriage to perceptions of the criminal justice system. Of the surveys administered, 814 were completed or partially completed. The response rate was 54 percent, and the cooperation rate was 65 percent.¹

The demographic breakdown of all survey respondents, compared with that of the population of female inmates in Florida (in parentheses), is as follows: white, 57 percent (59 percent); black, 32 percent (37 percent); mean age, 35 years (37 years) (FLDOC 2010).² Following listwise deletion for missing data, the final sample for this study was 554.³ The final demographic breakdown of those included in this study is 65 percent white, 31 percent black, and 4 percent Hispanic with a mean age of 35. According to FLDOC (2010), 31.31 percent of female inmates were incarcerated in Florida for a violent offense, approximately 30.52 percent for a property offense, 29.47 percent for a drug related offense, 1.41 percent for a sex offense, and 7.26 percent for a crime other than violent, property, drug, or sex. According to respondents' self-reported reason for incarceration, the final analytic sample contained fewer violent offenders (15.34 percent), slightly less property offenders (27.80 percent), more drug offenders (37.36 percent), roughly the same number of sex offenders (1.62 percent), and a higher proportion of “other” offenders (17.87 percent). Some of the discrepancy could be do to the nature of self-report versus official statistics; however, the lower number of violent offenders is likely due to the medium security nature of the prison used for this study. Indeed, the underrepresentation of blacks in our sample is also likely a function of the smaller number of violent offenders incarcerated in medium security facilities. To be clear, FLDOC (2010) reported that 37.22 percent of black females were incarcerated for violent offenses compared to 27.41 percent of white females. Thus, a larger proportion of black females were likely incarcerated at higher security facilities than the one surveyed.

Measures

Procedural Justice of the Courts

We focus on inmates' generalized perceptions of court procedural justice, rather than their perceptions of their most recent encounter for two reasons. First, many offenders have multiple experiences with the courts, and so the key theoretical concern is whether any single experience can influence broader perceptions of the courts. Second, prior research suggests that generalized perceptions of procedural justice may be the most important source of felt obligation to obey (Mazerolle et al. 2013). Female inmates' Perceived Procedural Justice of the Courts is measured using five items combined into a single standardized index. The wording of all five questions is presented in Table 1. PJ1 is a five-point measure with response categories ranging from very dissatisfied = 1 to very satisfied = 5. PJ2 is a four-point measure (seldom = 1, sometimes = 2, usually = 3, and always = 4). PJ3 and PJ4 are four-point measures with response categories ranging from strongly disagree = 1 to strongly agree = 4. PJ5 is a dichotomous measure (favor some people over others = 0, treat everyone equally = 1). The items combined into a single measure with Cronbach's alpha = 0.76.

Table 1. Promax Rotated Factor Analysis of Perceptions of Procedural Justice and Obligation to Obey Items

Obligation to Obey the Law

Obligation to Obey the Law is measured using three items combined into a standardized index. All three items (shown in Table 1) had four-point response categories ranging from strongly disagree = 1 to strongly agree = 4. The items combined into a single item with Cronbach's alpha = 0.64. While this reliability coefficient is lower than the traditional 0.70 threshold it is considerably higher than obligation to obey alphas reported in prior studies. For example, Tyler and Huo's (2002) measure had an alpha = 0.47. In Tyler's (2006), seminal Why People Obey the Law his obligation to obey measure had an alpha = 0.57. Finally, Wolfe (2011) reported an alpha = 0.55. In addition, all three of the Obligation to Obey the Law items loaded on a single factor (see below). It is also important to note that Cronbach's alpha is heavily influenced by the number of items in the index (Cortina 1993)—a likely explanation for the low alpha associated with our three-item index—and alpha provides a conservative estimate of a measure's reliability (Carmines and Zeller 1979: 45).

Before proceeding, it is important to note that there has been recent concern regarding the discriminant and construct validity of procedural justice and legitimacy measures (Baker et al. forthcoming; Gau 2011; Reisig et al. 2007). To ensure discriminant validity⁴ of Perceived Procedural Justice of the Courts and Obligation to Obey the Law, we conducted an exploratory factor analysis of all eight items included in the two measures. The results of the promax rotated factor analysis are presented in Table 1; they indicate two separate and distinct measures as operationalized above. While the exploratory factor loading for Ob3 (0.353) is low (<0.40), we chose to retain the item in an effort to be consistent with previous studies that have included it as a measure of obligation to obey (see, for example, Sunshine and Tyler 2003; Tyler 2006; Tyler and Huo 2002; Tyler and Wakslak 2004).

Voice

Given the importance of voice in promoting perceptions of procedural justice (Baker et al. 2014; Tyler 2006; Tyler and Huo 2002), we explore the significance of perceived Voice in court proceedings. Voice was a single-item measure in which respondents were asked “How much of a chance or opportunity did the members of the courts give you to describe your problem to them before making any decisions about how to handle it?” Responses to this item were a great deal of opportunity = 4, some opportunity = 3, a little opportunity = 2, and not much opportunity = 1. As evidenced by the descriptive statistics, it appears that on average respondents felt like they had “not much” to “a little” opportunity to express their side of the criminal encounter.

Shared Race/Ethnicity

Prior research often operationalizes social identity as a sense of attachment, belonging, or pride in belonging to a particular group (Murphy 2013; Murphy and Cherney 2011). In this study, however, we use shared race/ethnicity with court actors—defined as the correspondence (or lack thereof) between a female inmates' self-reported race or ethnicity and her perceptions of the race or ethnicity of the court actors involved in her most recent conviction—as a proxy for social identity. This approach is similar to the approach of shared group identity used by Tyler and Huo (2002). Respondents were prompted with the statement: “For the next set of questions I would like for you to think about the conviction that sent you here.” They were then asked to identify the race/ethnicity of the attorney representing them, the prosecutor, and the judge. Additionally, participants were asked to self-identify their race/ethnicity. Using these measures, we created three dummy variables: Shared Race with Defense Attorney, Shared Race with Prosecutor, and Shared Race with Judge. For each variable, respondents were coded “1” if their self-identified race/ethnicity matched that of the court actor, and were coded “0” if their self-identified race/ethnicity did not match that of the court actor. The percentages of whites, blacks, and Hispanics reporting they shared the race/ethnicity of the defense attorney, prosecutor, and judge in their most recent conviction are reported in Table 2. As would be expected, whites were more likely than either minority group to have shared the race/ethnicity of a court actor. Most strikingly, very few non-whites, less than 1 percent of blacks and approximately 4 percent of Hispanics, shared the race/ethnicity with the judge in their case. Even in combining black and Hispanic respondents into a single category, there are still too few non-white respondents that interacted with a non-white judge to reliably estimate the effect of judge race/ethnicity on respondents' perceptions. For this reason, the analyses focus only on shared race/ethnicity with defense attorneys and prosecutors.

Table 2. Shared Race/Ethnicity with Court Actor by Racial/Ethnic Group

Theoretically, it is plausible that among non-whites it may be less important that a court actor is of the same race/ethnicity as oneself than that the court actor is a member of any minority group. That is, the most salient consideration for non-white offenders may simply be whether legal authorities are white or not. For this reason, in their study of Oakland and Los Angeles residents, Tyler and Huo (2002) dichotomized legal authorities into two groups: whites and non-whites. In this study, we assess this possibility with the female offenders in our sample. To do so, in a separate model estimated for non-white inmates, we incorporate binary variables contrasting interactions with non-white versus white court actors (Non-white defense attorney, Non-white prosecutor). These variables are coded “1” if the court actor was black or Hispanic, and “0” if the court actor was white.

Control Variables

Three different types of controls are included to minimize the potential for omitted variable bias. The first set involved respondents' court experiences and criminal background. Individuals were asked about the type of attorney they had during the proceedings (Public Defender = 1, other = 0). Respondents were also asked if they had been previously incarcerated (Prior Incarceration = 1). In addition, respondents were asked to indicate the type of crime for which they were currently incarcerated (Violent Offense = 1, all others = 0). The final measure indicated whether the respondent accepted a plea bargain (Plea Bargain = 1).

The next set of controls examined respondents' lifestyle. They were asked about their marital status (Married = 1) and parental status (has Children = 1). Finally, we controlled for respondents' Age in years; individuals self-identified their race/ethnicity which was converted into two dummy variables (black = 1) and (Hispanic = 1) leaving white as a reference category; and level of Education (1 = grade school only, 2 = some high school, 3 = high school graduate, 4 = some college, and 5 = college graduate or more). The descriptive statistics for all the variables included in the models are presented in Table 3.

Table 3. Descriptive Statistics of Sample

^a Items are mean standardized indexes.

Analytic Strategy

To address our research questions, we conduct eight Ordinary Least Squares Regression (OLS) models. In the first two models, we examine the impact of voice and shared race/ethnicity with each court actor on perceptions of court procedural justice and then examine the subsequent effect of voice, shared race/ethnicity, and procedural justice on respondents' obligation to obey the law among the full sample. Because some prior research suggests that the importance of the race of legal authorities may be more salient for whites than non-whites (Scherer and Curry 2010; Tyler and Huo 2002), the next four models examine the same relationships as above separately for whites and non-whites. In a final set of models, among non-white inmates, we analyze the impact of respondents' interactions with non-white (but not necessarily a direct racial/ethnic match) court actors on procedural justice perceptions and obligation to obey the law.

Results

We begin by examining the results of the analyses on the full sample (Table 4). The first model examines procedural justice as the dependent variable. Consistent with prior literature voice has a strong positive significant effect on procedural justice perceptions representing the strongest predictor. The findings provide no evidence that shared race/ethnicity with the defense attorney is associated with procedural justice perceptions. By contrast, a strong effect emerges in relation to shared race/ethnicity with the prosecutor. On average, female inmates who shared the race/ethnicity of the prosecutor in their case report significantly higher perceptions of court procedural justice. In fact, sharing the race/ethnicity with the prosecutor is the second strongest predictor of court procedural justice among the inmates after voice. Of the court experience and lifestyle related controls, only having plea-bargained is positively and significantly associated with court procedural justice. Inmates who committed a violent crime, had children, and had higher educational attainment, conversely, tended to report lower levels of perceived procedural justice.

Table 4. OLS Models Examining Shared Racial/Ethnic Identity, Perceptions of Court Procedural Justice, and Obligation to Obey the Law Among the Full Sample (N = 554)

^* p < 0.05; **p < 0.01; ***p < 0.001 (two-tailed test).

Abbreviations: b = unstandardized coefficients; s.e. = robust standard errors; B = standardized coefficient.

A key question is whether the link between procedural justice and obligations to obey, which has been supported in examinations of public perceptions of police procedural justice (Mazerolle et al. 2013; Sunshine and Tyler 2003), is generalizable to serious offenders' perceptions of court procedural justice. Consistent with this framework, the results (presented in Table 4 Model 2) show that female inmates' perceptions of court procedural justice are the strongest predictor of their obligation to obey the law. Specifically, inmates' who perceive that the courts are fair are more likely to report that they feel obligated to abide by legal regulations.

Given the strong relationship observed between perceived court procedural justice and obligation to obey, we examine whether shared race with the prosecutor, which significantly impacted procedural justice perceptions, exerts an indirect effect on obligation to obey through procedural justice perceptions. The appropriate test for indirect effects is the product of the coefficients approach (Hayes 2009, 2013), which involves multiplying the coefficient for the effect of the independent variable (i.e., shared race with prosecutor) on the mediator (i.e., court procedural justice), with the coefficient for the effect of the mediator on the outcome variable (i.e., obligation to obey) in the full model. Traditionally, the Sobel z-test has been used to determine the significance of the indirect effect. However, scholars have recently emphasized that Sobel's z should be abandoned in favor of bootstrapping because it inappropriately assumes that the indirect effect has a normal sampling distribution (Hayes 2013; Zhao et al. 2010). For this reason, we test for the indirect effect of shared race with prosecutor using the product of the coefficients approach, and we estimate a percentile-based bootstrap (k = 5,000) confidence interval. We find that shared race with prosecutor has a significant positive indirect effect (b = 0.048, 95 percent confidence interval = 0.013–0.092) on obligation to obey the law, through procedural justice perceptions. Stated differently, among female inmates, shared race/ethnicity with prosecutors is associated with a greater obligation to abide by the law because it is correlated with an increased likelihood of perceiving the courts to be procedurally just.

To further unpack the relationship between voice, shared race/ethnicity with court actors, perceived court procedural justice, and female offenders' obligation to obey the law, we reestimated the above model separately for whites and non-whites. The results of the disaggregated analyses are presented in Tables 5 and 6. We first focus on the results for white inmates (Table 5). As was the case for the full sample, among white inmates, we observe strong positive significant relationships between both voice and perceptions of procedural justice and having shared the race with the prosecutor and perceived procedural justice. The strong significant positive relationship between procedural justice and obligation to obey the law seen in the full model is also present in the whites only model. The results reveal that, among white inmates, shared race with the prosecutor has a significant positive indirect effect (b = 0.050, 95 percent confidence interval = 0.009–0.103) on obligation to obey the law through procedural justice perceptions.

Table 5. OLS Models Examining Shared Racial/Ethnic Identity, Perceived Court Procedural Justice, and Obligation to Obey the Law Among Whites (N = 358)

^* p < 0.05; **p < 0.01; ***p < 0.001 (two-tailed test).

Abbreviations: b = unstandardized coefficients; s.e. = robust standard errors; B = standardized coefficient.

The story is different among non-whites inmates (Table 6). Voice remains a strong positive predictor of perceived procedural justice. None of the measures of shared race/ethnicity have significant effects on either perceived procedural justice or obligation to obey the law, though. Stated differently, among non-whites, shared race/ethnicity with court actors does not appear to be an important concern influencing views about the courts and law. However, court procedural justice is significantly and positively associated with non-white inmates' obligation to obey the law. As noted previously, it is possible that among non-whites, it may be more important that a legal authority is a member of any minority group than that he or she is a member of a specific minority group.

In our final set of models presented in Table 7, we explore whether interactions with non-white court actors, regardless of whether those court actors were black or Hispanic, influence non-white inmates' views about the courts and law. The findings show that unlike in the previous model for non-whites (Table 6), which focused on shared race/ethnicity, contact with non-white prosecutors has a significant effect. Specifically, among non-white female inmates, having a non-white prosecutor is associated with greater endorsement of the view that the courts are procedurally just. As such, we again tested for an indirect effect of having a non-white prosecutor on obligation to obey the law and found a positive but nonsignificant effect (b = 0.064, 95 percent confidence interval = −0.005–0.170). The remaining results of both models are the same as those found for non-whites in Table 6.

Table 6. OLS Models Examining Shared Racial/Ethnic Identity, Perceived Court Procedural Justice, and Obligation to Obey the Law Among Non-Whites (N = 196)

^* p < 0.05; **p < 0.01; ***p < 0.001 (two-tailed test).

Abbreviations: b = unstandardized coefficients; s.e. = robust standard errors; B = standardized coefficient.

Table 7. OLS Models Examining Non-White Court Actors, Perceived Court Procedural Justice, and Obligation to Obey the Law Among Non-Whites (N = 196)

^* p < 0.05; **p < 0.01; ***p < 0.001 (two-tailed test).

Abbreviations: b = unstandardized coefficients; s.e. = robust standard errors; B = standardized coefficient.

Discussion and Conclusion

The purpose of this study was to test theoretical predictions about the relationships between voice, shared race/ethnicity, court procedural justice, and perceived obligation to obey the law among female inmates. Several important findings emerged from our analyses. First, the results reveal a significant positive relationship between voice and perceptions of procedural justice. Being able to participate in the process and give one's side of the story plays a critical role in promoting perceptions of procedural justice. In fact, voice was the strongest single predictor of procedural justice across all four models in which perceptions of procedural justice was the dependent variable. The realization by court actors, and more broadly criminal justice personnel, that they could enhance perceptions of procedural justice (which in turn enhances legitimacy) by simply providing accused offenders the opportunity to express their side of the story is encouraging. It indicates that there may be factors within the control of legal authorities that they could use to enhance and promote procedural justice perceptions. And, given the low perceived voice among this sample, it may be an avenue through which a great deal of growth in procedural justice is available.

The findings from our analyses also strongly support the link between procedural justice and obligation to obey the law. That is, they show that in the context of courts, procedural justice perceptions are central for understanding obligations to obey the law. The findings also demonstrate the generalizability of the procedural justice framework for individuals at high-risk for future criminal behavior—prior offenders. It bears emphasizing that this study is the first to present findings showing that female offenders' obligation to obey the law is directly and significantly affected by their perceptions of court procedural justice. The implication is that factors that influence female offenders' perceptions of court fairness have considerable potential to affect their views about the law and also potentially to shape their behavior, by increasing self-regulation (see Tyler 1989). Accordingly, additional theoretical and empirical work directed at identifying the full range of such factors is warranted. Such scholarship is important because, as Tyler and Huo (2002: 12) have explained, when self-regulation occurs, “the result is a law-abiding society in which the need to use force and coercion to achieve compliance with the law is minimized.”

Our study also provides the first evidence that a correspondence between the race/ethnicity of defendants and court actors may shape the formers' perceptions of the courts and law. Specifically, we found that female offenders' who shared the race/ethnicity of their prosecutor perceived the courts as more procedurally just and, in turn, felt a greater obligation to obey the law. Indeed, our analyses revealed that shared race/ethnicity with prosecutors was the second strongest predictor of female inmates' perceptions about court procedural justice. However, when we further explored this relationship we found that the effect of shared race/ethnicity with prosecutors was only significant among white inmates. For non-white inmates, sharing minority status more generally with a prosecutor, rather than sharing the court actor's specific race or ethnicity, increased perceived court procedural justice. From the perspective of social identity theory and the group-value model, this may indicate that non-white offenders interpret encounters with legal authorities, regardless of those authorities' specific racial or ethnic backgrounds, differently than do whites. It may also demonstrate, as the group-position thesis asserts (see Weitzer and Tuch 2004), that for whites seeing non-whites in a position of authority reduces their affinity for the system and their perceptions that it can be counted on to protect their interests. In addition, minorities may believe it is more likely that they will receive fairer treatment and improved justice from authorities of any minority group regardless of race or ethnicity.

Alternative interpretations to the lack of a significant shared race finding for non-whites also exist. For example, it may not be that non-whites necessarily perceive non-white prosecutors as fairer but instead that non-whites who encountered a white prosecutor perceived the courts to be significantly less fair. That is, it is not that their perceptions of procedural justice were enhanced by having a non-white prosecutor but rather that having a white prosecutor was a significant detriment to their perceptions of court procedural justice. This alternative interpretation is supportive of the propositions espoused by critical race theories (Haney-López 2006; Lynch 2014). As Haney-López (2006) suggests, non-whites negative perceptions of encountering a white prosecutor may be indicative of their inability to identify generally with a predominantly white court system. Non-whites may perceive the courts as part of a criminal justice system unable to empathize with their life circumstance and their current situation (Lynch 2014), a court system that has played a critical role in controlling minorities for the sake of white dominance (Haney-López 2006).

Theoretically, these findings are suggestive of the importance of the status component of the group-value model for predicting procedural justice. The results also suggest that social identity theory may be generalizable to aspects of the criminal justice system that have been widely ignored, especially in terms of the relationship between offenders and the key authority figures in the courts. As the group-value model would suggest, the mechanisms connecting shared race/ethnicity with views about the courts and law may largely be psychological as opposed to the objective legal characteristics of fairness. Specifically, shared identification may increase offenders' procedural justice perceptions by fostering more positive judgments of personal status as well as the belief that one is a member of a respected group. For whites, who may have an expectation of conferred status, encountering white authorities may reinforce and enhance their perceptions or perhaps even restore (as they may have come to view themselves not as a member of a respected group but instead as a criminal “other”) their perceptions as members of a respected group. Importantly, if this latter theoretical perspective is correct, then the effects of shared identification would, for the most part, be independent of the objective characteristics (e.g., fairness) of the procedures used and treatment of defendants by court actors. However, for non-whites shared identification may be less important than encountering court actors who empathize with their current situation (Lynch 2014). Future research should consider the importance of perceived empathy in assessing minorities' perceptions of the courts.

Respondents expressing a perceived lack of procedural justice and reduced obligation to obey the law may also be engaging a particular “script” of condemnation, externalizing the blame of their current incarceration (Maruna 2001, 2004). Blaming an unfair court system for their current situation may allow offenders unable or unready to reform from internalizing responsibility for their incarceration instead believing that had they not been so “unlucky” to get an unfair judge or prosecutor or public defender they would not be in prison. Alternatively, inmates expressing that the process of the courts was fair and in turn a high felt obligation to obey the law may be internalizing the causes of their situation (Maruna 2001) and providing signals as to their desire for rehabilitation and desistance from crime (Bushway and Apel 2012). Maruna (2001) has suggested the development of “rituals of redemption” is a critical part of the process of transitioning from a persisting to desisting offender. These rituals involve the normalization of offenders from criminal “others” back to “normal” members of society and one aspect of this process is having respected members of the communities, with whom the offender can identify, engage in a formal redemption process—often in the courts (Maruna 2001). To the extent that, as Lynch (2014) suggests, court actors of the same race are better able to empathize with offenders, especially non-white court actors with non-white offenders, and confer to offenders that they are members of a valued and respected group, a court process further perceived to be fair by an offender may actually begin the redemption ritual necessary to move offenders from persisting to desisting.

Another important avenue of inquiry for future studies is to investigate potential explanations for why shared race/ethnicity matters more in relation to some types of court actors (e.g., prosecutors) than others (e.g., defense attorneys). One possibility is that the effects of shared race/ethnicity may be conditional on the nature—adversarial versus supportive or cooperative—of offenders' interactions with specific authority figures. For example, from the perspective of the group-value model, we might expect that the importance of shared race/ethnicity with court actors for defendants' judgments about their group-based status would be greatest when those actors have an adversarial role in the proceedings. Thus, shared race/ethnicity may matter little in the case of defense attorneys because their interaction with defendants is generally not adversarial. Researchers might also consider exploring whether in jury trials, shared race and ethnicity with the jury—operationalized, perhaps, as the correspondence between defendants' race and their perceptions of the racial composition of the jury—influences defendants' perceptions about the courts and law. It may be that individuals with a jury composed of individuals of similar race or ethnicity would perceive that jury to be more representative of a jury of their peers, able to empathize with their current situation and past life experiences. Shared race and ethnicity of probation officers may also be important for offenders under community supervision. Again, as Lynch (2014) posits the ability to empathize with offenders may play a critical role in authority-subordinate relationships. As such, same race authorities may be better capable of conferring legitimacy through the promotion of procedural justice than different race authorities.

Future studies are also needed to examine the effect of shared race/ethnicity of judges on people's perceptions of the courts. Given the lack of variability in the race/ethnicity of judges encountered by the offenders in our sample, we were unable to reliably examine this relationship. The lack of variability may also speak to a greater issue of a lack of diversity in the race/ethnicity of judges in Florida. For example, a study conducted by the Florida Supreme Court Standing Committee on Fairness and Diversity (2008), found that although 15.7 percent and 19.5 percent of Floridians are black and Hispanic, respectively, only 6.7 percent of judges in Florida are black and only 7.2 percent are Hispanic.

This limited diversity speaks to another limitation of the current study, the small number of individuals who interacted with non-white court actors. For example, four cases were dropped for being outliers, three of which were non-whites. The inclusion of these cases has a great deal of undue influence on the results.⁵ It is important to note that regardless of the inclusion or exclusion of these cases the prosecutor effect on procedural justice and the effect of procedural justice on obligation to obey the law among whites are stable. Still, future studies should oversample non-whites who interacted with non-white court actors.

It is also important to note that this study relied on female inmates' obligation to obey the law and not actual obedience to the law. This is an important distinction that future studies should address. That is, future work needs to examine the extent that felt obligation to obey the law translates into actual compliance. Almost every single one of the inmates involved in this study (and the vast majority of individuals incarcerated in the United States for that matter) will eventually reenter society, and their views of the criminal justice system and their felt obligation to obey the law could have a profound impact on their recidivism risk. Future studies should seek to follow offenders reentering society to evaluate how their perceptions affect their actual compliance with law.

Also, while studying female inmates is a unique quality of this study it has the potential to limit the generalizability. As a result of historical and cultural subordination, females, especially non-white females, may be particularly attuned to the presence of implicit racial bias in the courts and as such be more likely to value shared race court actors (Haney-López 2006). Future research should evaluate how serious male offenders perceive same race court encounters and examine if they affect perceptions of procedural justice. At the same time, while this sample represents some of the most serious offenders with whom the procedural justice-obligation to obey the law relationship has been demonstrated, the sample was drawn from a medium security prison and as a result the sample contained fewer violent offenders. While there is little empirical evidence to suggest that the findings would not generalize to higher custody inmates, future research should more fully explore this possibility.

Before closing, it is important to situate our findings within the current theoretical debates in the literature concerning justice and legitimacy. Much of the prior work in criminology and socio-legal studies on justice and legitimacy has taken a Tylerian approach to the conceptualizations of justice and legitimacy specifically espousing the distinction between procedural and distributive justice and defining legitimacy as an individuals' obligation to obey the law. While this study takes a Tylerian approach to defining justice and legitimacy, it is necessary to discuss alternative views on these critical concepts.

First, recent work by Lerner and Clayton (2011) is highly critical of Tyler's (1994) and Tyler and Blader's (2003) conceptualization of justice. Lerner and Clayton (2011) suggest that poor methodology and conflating generosity and kindness with fair treatment incorrectly caused Tyler and colleagues to conclude that distributive justice and procedural justice are separate and distinct concepts and to incorrectly conclude that fairness leads to compliance. Accordingly, they suggest that no such distinctions or a relationship actually exists. Rather they posit that “Justice consists of people getting what they deserve, or what they are entitled to by virtue of who they are and what they have done” (Lerner and Clayton 2011: 117). While their differential conceptualization of justice is discordant with Tyler's and the one espoused within this study (specifically the importance of an explicit distinct concept of procedural justice and the measures used to capture it), the empirical findings expressed relating to obligation to obey are not wholly inconsistent with Lerner and Clayton's (2011) theory of justice.

To be clear, in contradiction to our findings, they suggest in part the possibility that in the face of injustice people may actually find their resolve reinforced believing that “good things happen to good people”. That is, especially when given time to reflect on an injustice people are more likely to respond with norm-appropriate behavior. However, they also suggest, that when faced with an injustice that elicits an emotional reaction individuals may react more punitively in an attempt to punish the individuals who unjustly transgressed. While we did not capture the emotional responses of individuals when they were giving their perceptions of court justice, it is possible that those who believed the courts were unjust felt angered by such injustice and in turn retaliated or rebelled with perceptions of disobedience, that is, a lack of obligation to obey the law. While the above discussion is highly speculative, further examination, especially real world examination, may help to ascertain which theory or aspects of these competing theories of justice most accurately reflect individual perceptions. Future studies should seek to set up competing models of justice as well as explore integrating aspects of these theories that can be validated and are logically consistent across both models of justice.

In addition, there is currently a debate about what actually constitutes legitimacy both empirically and conceptually. Bottoms and Tankebe (2012) and Tankebe (2013) have posited that obligation to obey should not be considered a measure of legitimacy because other factors such as fear or feelings of powerlessness, which have little to do with legitimacy, may influence whether people feel obligated to obey authorities. Instead of obligation to obey, Bottoms and Tankebe (2012: 166) “envisage a multidimensional measurement of legitimacy embracing (at least) legality, procedural justice, and effectiveness.” Tankebe (2013) recently tested this conceptualization presenting procedural justice, distributive justice, effectiveness, and lawfulness as measures of legitimacy, not antecedents of it, as has been consistently used by studies espousing Tylerian approaches to procedural justice and legitimacy.

Although this important discussion of the operationalization and conceptualization of key criminological constructs is ongoing, it is beyond the scope of this article to evaluate Bottoms and Tankebe's (2012) and Tankebe's (2013) arguments. However, even if one accepts their position that obligation to obey is a “much wider concept than legitimacy,” it remains true that obligation to obey “can be considered a ‘dependent variable,’ sometimes explained by perceived legitimacy, and sometimes not” (Tankebe 2013: 105). From this perspective, then, our study is the first to examine the impact of one key component of legitimacy—procedural justice—on female offenders' obligation to obey the law, while also taking into account the effect that shared race/ethnicity has on Tankebe's conceptualization of legitimacy and obligation to obey. The findings show that while perceived procedural justice may constitute only one of many factors (e.g., fear, felt powerlessness, etc.) that influence obligation to obey the law, among female inmates, it seems to have a strong impact. Again, we do not intend our results to represent actual tests of these competing theories but instead to highlight different theoretical perspectives and how future examinations of procedural justice and legitimacy should consider alternative interpretations for their findings. As important, and regardless of which theoretical model is correct, it appears that shared race/ethnicity with court actors is an important antecedent of legal subordinates' justice attitudes.

The policy implications of our findings are twofold. First, promoting procedurally fair behavior in the courts through the emphasis of due process may have significant effects on promoting self-regulatory behavior among prior offenders. Specifically, it seems that practices that allow people to take part in the process by having the opportunity to share their side of the story can increase offenders' perceptions of procedural justice. This could be enhanced through initiatives that train court actors to act in a manner that is perceived to be procedurally just and demonstrates areas in which they can give accused offenders the opportunity to be heard. Such training may make it possible to communicate a greater sense of fairness in cross-race encounters, specifically non-white offender—white authority encounters.

Second, the appointment and election of more racially and ethnically diverse prosecutors able to more fully empathize with many of the offenders entering the courts may help to improve minorities perceptions of the fairness of the courts specifically, and potentially the criminal justice system as a whole more generally. To be sure, increases in minority representation in positions of authority have important social policy implications that extend far beyond the criminal justice system. Policy makers should note, though, that to the extent that the findings from our research with female offenders as well as those from previous studies of the public (Scherer and Curry 2010; Tyler and Huo 2002) are generalizable to male offenders and other samples, increases in minority representation among court actors may result in a loss of court legitimacy among whites. So, when considering the added benefits that diversification would have above and beyond training it is important to recognize the value of training which would promote procedural justice when weighed in concert with the generation it will take to improve diversification within the courts.

Thus, while this study adds to an extensive body of literature signifying the importance of increasing racial diversity among those in positions of authority, policy makers may be faced with a precarious balancing act between increasing support among minorities at the risk of losing support among whites. Advanced training in promoting perceptions of fairness generally and cross-race encounters specifically may be of critical importance for maintaining the balance between offenders' perceptions of procedural justice and legitimacy and promoting increased diversity throughout the criminal justice system.