Carter Udel 0060D 13410.Pdf

BLINDED BARS:

RACE AND SOCIAL CONTROL AMONG CORRECTIONS EMPLOYEES

TaLisa J. Carter

A dissertation submitted to the Faculty of the University of Delaware in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Criminology

Summer 2018

BLINDED BARS:

RACE AND SOCIAL CONTROL AMONG CORRECTIONS EMPLOYEES

TaLisa J. Carter

Approved: ______Karen F. Parker, Ph.D. Chair of the Department of Sociology & Criminal Justice

Approved: ______George Watson, Ph.D. Dean of the College of Arts and Sciences

Approved: ______Ann L. Ardis, Ph.D. Senior Vice Provost for Graduate and Professional Education

I certify that I have read this dissertation and that in my opinion it meets the academic and professional standard required by the University as a dissertation for the degree of Doctor of Philosophy.

Signed: ______Karen F. Parker, Ph.D. Professor in charge of dissertation

Signed: ______Aaron Kupchik, Ph.D. Member of dissertation committee

Signed: ______Christy Visher, Ph.D. Member of dissertation committee

Signed: ______Jeff Fagan, Ph.D. Member of dissertation committee

ACKNOWLEDGMENTS

“Every good and perfect gift is from above, coming down from the Father of the heavenly lights” (James 1:17). I think this is a pretty good gift. And I am grateful to Him. To my original clan: The Carters – Jacqueline (Ma), Bruce (Daddy), Nancyia (Chick) and Music (Mu). I love you beyond an acknowledgement. Thank you for loving me back, for grounding me and reminding me who I am. Thank you for talking to and laughing with me every day, for helping me feel less alone, and for checking me when I needed it. It is my hope to give you what you’ve given me – a special kind of love that violates the laws of time and space. Each of you are pieces of me. Ma, I am proud to act and look like you. Daddy, you’ve built in me resilience and confidence. Chick, there is no pact in the world stronger than the one we share. I will never betray you. Mu, you’re too young to read this but one day know that you are TT’s first child and no matter where I am or what I am doing – I will never be too busy to keep your secrets and be in your corner.

Thank you to all my family – blood, adopted, by marriage, via christening, and in spirit. I can’t name you all but Grandma Fannie Mae, Aunt Peppie, Aunt Diane, the Bells, the Thompsons, the Perrys, my God families – love you much. Thank you to my friends –Blue Hens, Quakers, Rev. & Mrs. Curlee Windham & the Liberty Baptist Church family, First Mount Bethel and Bishop & the late/beloved Dr. Joyce Hunter & the Prayer Temple family. Again, I can’t name you all but much love to Shawntai,

Judah Praise, Destination Hip Hop, LBC’s Youth Ministry Choir, LBC’s Usher board,

iv and anyone I’ve ever taught a dance to! It was destiny that you all would love me, and I would love you. Thank you to my committee! Dr. Karen F. Parker, my chair, mentor and the person who has pushed me beyond mediocrity since my first day at UD. Dr. Aaron C. Kupchik, my mentor and the ultimate crises manager. Dr. Christy A. Visher who always has time for a quick random chat, gives the best advice and thinks of things I never would’ve based on your years of experience. And Dr. Jeff Fagan (Columbia Law) for your willingness to serve, responsiveness and guidance in this process (air hug)! My time at UD has been incredible. Thank you to every person on the faculty and staff – for your guidance and assistance over 2 degrees and five years. Shouting out an incredible network of supportive Drs.: Ronet, Santhi, Maria, Cresean, Ellen,

Anne (my academic Aunt), Ann, Barret, Asia, Eric, Tammy, Lana, & Susan, as well as incredible staff support Ms. Judi (my CDHS mom), Grant, Ms. Diane, Linda, Deanna, & Vicky! Cheers to each and every graduate student who has come and will come through the Department of Sociology and Criminal Justice especially Ashley, Aneesa, and those I love in the “Estrogen Eleven” cohort. Love and light to Dr. Erin Kerrison whose laugh, love and faith in me as a scholar and a person has helped me navigate life in countless ways.

Cheers to the Black Graduate Student Association – especially Dr-to-be Jehnae Linkins xoxo. To every person on UD’s campus that wages war against institutional racism, discrimination and inequity with class, grace and skill – including Dr. Carol Henderson, Dr. Keeley Powell, and Dr-to-be Jen Daniels. Thank you for being living examples of “Black Girl Magic” or better put “Black Women Wonders”. My

v experiences wouldn’t have been the same without any of you. I will miss seeing you frequently, but this is far from goodbye. Stay in touch – I promise I will. Much appreciation to a Mid-Atlantic Department of Correction for their continued support and willingness to work with me in research towards improving corrections statewide. Thank you to each person I interacted with. Special cheers to the members of the CEIT class who trained on the edge – special mention to one of the fastest and youngest members of the class: sleep in peace. Much love to my former coworkers in a southern jail. You all inspire me every day and I think of you – what you’ve taught me, your families, your safety – throughout the research process. Special salute to Unit 2 – Big B, Hood, Crpl. C., Rice, Mr. J and the rest of the lot. Love you always – Carter.

TO ALL OF YOU! To anyone who has spoken a word of encouragement, corrected me when I needed it or just saw my potential when my eyesight was cloudy. To you, beautiful people, who have loved, listened to, advised, partied with, cried with, fed, and just grown with me – you are phenomenal. You are my village! I couldn’t imagine life without your consistent support. To Dr. TaLisa J. Carter. Thank you for keeping this promise to yourself.

There’s more work to be done. Do it with Joye!

To the person taking the time to read this dissertation – for whatever reason that may be :-) – thank you for your time. I hope this work moves you in some way. Now, go be great! With respect, admiration and joy, - Dr. TaLisa J. Carter

vi TABLE OF CONTENTS

LIST OF TABLES ...... x LIST OF FIGURES ...... xi ABSTRACT ...... xii

Chapter

1 INTRODUCTION ...... 1

1.1 The Issue: Racial Disparities ...... 1 1.2 Exploring Racial Disparities across the Criminal Justice System ...... 3

1.2.1 Racial Disparities in Policing ...... 4 1.2.2 Racial Disparities in Courts ...... 5 1.2.3 Racial Disparities in Corrections ...... 7 1.2.4 Challenges in Racial Disparity Research ...... 10

1.3 Purpose & Significance of Study ...... 17

2 CRIMINAL JUSTICE ACTOR MISCONDUCT ACROSS THE SYSTEM .. 22

2.1 Who’s to Blame? ...... 22

2.1.1 Police Officer Misconduct ...... 25

2.1.1.1 Empirical Approaches to Police Misconduct ...... 26 2.1.1.2 Theoretical Approaches to Police Misconduct ...... 29 2.1.1.3 Racial Disparities & Police Misconduct ...... 31

2.1.2 Prosecutorial Misconduct ...... 32

2.1.2.1 Approaches to Studying Prosecutorial Misconduct ..... 34

2.1.3 Correctional Officer Misconduct ...... 37

2.1.3.1 Approaches to Studying Correctional Officer Misconduct ...... 39

vii 2.2 Filling the Gap ...... 43

3 THE INSTITUTIONAL RESPONSE MODEL OF SOCIAL CONTROL ..... 46

3.1 Theoretical Underpinnings ...... 46

3.1.1 Colorblind Ideology ...... 48 3.1.2 Black’s Theory of Law & Social Control ...... 51

3.2 The Institutional Response Model of Social Control ...... 53

3.2.1 Visibility of the Action ...... 54 3.2.2 Institutional Context ...... 58 3.2.3 Status of the Actor ...... 61 3.2.4 Research Questions ...... 63 3.2.5 Hypotheses ...... 63 3.2.6 IRM in Correctional Facilities ...... 64

3.3 Research Setting ...... 66

4 DATA AND METHODOLOGY ...... 70

4.1 Data Access ...... 70 4.2 Data & Sampling ...... 71 4.3 Methodology ...... 75

4.3.1 Dependent Variables: Institutional Responses ...... 75 4.3.2 Visibility of the Action ...... 77 4.3.3 Institutional Context ...... 79 4.3.4 Status of the Actor ...... 80 4.3.5 Qualitative ...... 81

4.4 Reflexive Note ...... 82

5 QUANTITATIVE RESULTS ...... 85

5.1 Univariate Statistics ...... 85 5.2 Quantitative Results ...... 88

5.2.1 Logistic Regression ...... 88 5.2.2 Negative Binomial Regression ...... 95 5.2.3 Quantitative Summary ...... 97

6 QUALITATIVE RESULTS ...... 100

viii 6.1.1 Institutional Responses ...... 101 6.1.2 Visibility of the Action: Seeing Silence ...... 106 6.1.3 Institutional Context: “us versus them” ...... 110 6.1.4 Status of the Actor: Before MADOC ...... 113

6.2 Results by Hypothesis ...... 117

7 DISCUSSION AND CONCLUSION ...... 126

7.1 Limitations ...... 128 7.2 Challenges as a Researcher ...... 133 7.3 Further Considerations ...... 135 7.4 Policy Implications ...... 136 7.5 Conclusion ...... 138

REFERENCES ...... 141

Appendix

A TIME IN THE FIELD ...... 160 B MADOC POLICY IMPLICATIONS ...... 161

B.1 Transparency in Response Processes ...... 161 B.2 Visibility as an Objective Tool ...... 162 B.3 Break the Code of Silence and Bridge the Gap ...... 164 B.4 Tenure-Long, Intensive Training ...... 165

C Search Engine Results as of 4.26.16 ...... 167

ix LIST OF TABLES

Table 1. Variable List ...... 76

Table 2. Univariate Statistics ...... 88

Table 3. Correlation Matrix. Positive Commendations ...... 90

Table 4. Correlation Matrix. Disciplinary Sanctions ...... 91

Table 5. Logistic Regression Reporting Odds Ratios for Presence of Institutional Response (N=246)...... 92

Table 6. Negative Binomial Regression reporting IRR coefficients (standard errors) for Institutional Response (N=239) ...... 99

x LIST OF FIGURES

Figure 1. Word Cloud. Positive Commendations...... 103

Figure 2. Word Cloud. Disciplinary Sanctions ...... 106

xi ABSTRACT

Criminal justice scholarship has long recognized the pervasiveness of racial/ethnic disparities in the system. However, few have explored how criminal justice actor behavior contributes to systemic bias. Even fewer have examined how institutional responses to correctional employee behavior can perpetuate racial/ethnic disparities. A review of the criminal justice literature reveals that while racial/ethnic disparities are present across the criminal justice system, they are researched unequally. This is particularly true in terms of corrections scholarship, where the link between institutions, criminal justice actors and race is not well-developed.

Additionally, literature points to three main challenges in exploring racial/ethnic disparities including issues with data access, inconsistencies in the way data are collected and variations in theoretical models applied. To overcome these challenges, this study proposes a single theoretical framework that considers race as central to the understanding of institutional outcomes. Additionally, this dissertation uses both qualitative and quantitative methodology to strengthen findings.

This dissertation explores how racial/ethnic inequality is perpetuated within criminal justice organizations, specifically corrections, by developing and applying a theoretical framework using mixed methodology. Informed by Bonilla-Silva’s colorblind ideology and Black’s theory of law and social control, the institutional response model of social control prioritizes race and the role of criminal justice actors in contributing to unequal outcomes across the system. To test the model, I use qualitative and qualitative data in a Mid-Atlantic Department of Correction. Results

xii show overall support for the institutional response model’s three key elements: visibility of behavior, institutional context and status of the actor. Each element influences the presence and level of institutional response a correctional employee receives, when controlling for other factors. Noteworthy, race surfaces as a central factor in determining how employees are sanctioned in both the quantitative and qualitative analysis. Quantitative analyses find employees who are Black are likely to receive more sanctions than their White counterparts, controlling for other factors. Qualitative analysis shows how narratives around institutional responses are racialized, perpetuating an institutional atmosphere that considers race when disciplining employees. Limitations, policy recommendations, and suggestions for future research are also discussed.

xiii

Chapter 1

INTRODUCTION

“Welcome to the Thin Blue Line” – Training Instructor The criminal justice system in the United States is flawed; the thin blue line between criminal justice actors and society is tainted. Practitioners across the justice system acknowledge that errors committed by criminal justice employees harm defendants during the justice process (Ramsey & Frank, 2007). At all stages of the criminal justice system a range of procedural errors connected to occupational misconduct are possible, according to the American Civil Liberties Union (2014). While this is concerning, professional deviance and institutional responses to wrongdoing are not researched equally across the criminal justice system. This chapter examines how the behavior of criminal justice professionals is studied by highlighting a problem that has plagued the system for decades: racial disparities.

1.1 The Issue: Racial Disparities

The United States criminal justice system is riddled with racially disparate outcomes at every phase (Mears, Cochran, & Lindsey, 2016). In 2009, Blacks and Whites made up approximately 13% and 67% of the United States general population, respectively (ACLU, 2014; Hartney & Vuong, 2009). Statistics reveal that Blacks are significantly overrepresented in terms of arrests (28%), the incarcerated population (40%), and inmates on death row (42%) (ACLU, 2014; Hartney & Vuong, 2009). In contrast, Whites are either proportionately or significantly underrepresented in these

areas comprising 70% of arrests, 40% of individuals held in prisons and jails and 56% of the death row population (ACLU, 2014; Hartney & Vuong, 2009). While less data are collected for other racial/ethnic groups, according to the Sentencing Project (2008). Hispanics and Native Americans are reported to be overrepresented in the criminal justice system, confirming the overrepresentation of people of color (Hartney

& Vuong, 2009). Do these disparate outcomes reflect systemic discrimination in the criminal justice system? Is the criminal justice system “discriminatory” because differences are observed between racial groups? There is an important distinction between disparity and discrimination. Disparity simply represents a difference. Some scholars define criminal justice disparities in terms of proportions. If the proportion of a racial/ethnic group is unequal to their proportion in the general population, a disparity exists (TSP, 2008). A pattern of disparities is suggestive of systemic discrimination (Rhodes, Kling, Luallen, & Dyous, 2015). Walker and colleagues define discrimination as the treatment of individuals according to their membership in a group rather than their personal behavior or traits (Walker, Spohn, & DeLone, 2012). Scholarship highlights the systemic nature of discrimination in the criminal justice system by emphasizing the statistical patterns of disparity in mass incarceration (Alexander, 2012), policing

(Blauner, 1969) and court proceedings (ACLU, 2014). Taking the position that racial/ethnic disparities in an institution reflects systemic discrimination, it is important to understand that capturing these disparities involves more than counting the differences in criminal justice outcomes between groups. Researchers consider a variety of factors that may lead to differences in criminal justice outcomes. This is not a simple task, as researchers consider a wide range of legal (within the scope of the

government/law) and extralegal (outside of the scope of the law) variables that may also account for racial/ethnic disparities (Mitchell, 2005; Petersilia, 1983). This dissertation directly addresses the critical (though often overlooked) role criminal justice actors play in perpetuating racially disparate outcomes in the system by theoretically and empirically exploring institutional responses to employee behavior. As with any social institution, criminal justice employees are responsible for translating institutional policy into practice by their decisions and deeds. Therefore, to truly reach a deeper understanding of the longstanding racially inequitable outcomes across the criminal justice system, it is vital to consider criminal justice actors alongside other legal and extralegal variables. Furthermore, because criminal justice institutions are theoretically purposed with ensuring justice in society, the way these institutions discipline and praise employee behavior can also speak to the ubiquity of racial injustice within the system. This chapter first briefly summarizes criminal justice statistics on the existences of racial inequalities in all three phases of the criminal justice system: police, courts and corrections. Demonstrating the existence of racial disparities in every phase of the system makes the pervasiveness of racial disparities clear.

1.2 Exploring Racial Disparities across the Criminal Justice System Disparities in criminal justice outcomes in the system have been examined by scholars and practitioners for decades. Understanding what causes and perpetuates these disparities is vital to promote justice and equality. The following sections will briefly discuss racial disparities in the three main phases of the criminal justice system: police, courts and corrections. Although these race-based differences will be summarized separately for organizational purposes, it is important to note that this

work is concerned with the systemic nature of racial disparities rather than the nature of these differences at each phase independently. This is especially true because disparities increase in gravity and implications in each subsequent phase the system (TSP, 2008). That is, as an individual gets processed through the system racial disparities widen. Therefore, it is important to consider the three phases of the criminal justice system simultaneously when discussing racial/ethnic differences.

1.2.1 Racial Disparities in Policing Statistics on policing outcomes across the United States illustrate wide gaps in the treatment of people of color. The New York Police Department stopped 4.4 million individuals over an 8.5-year period; 83% of the individuals stopped were African American or Latino (ACLU, 2014). According to the Bureau of Justice,

African Americans were arrested at a rate that was 2.5 times higher than Whites from 2005-2012. In fact, all minority groups except Asian Pacific Islanders were arrested at a rate higher than Whites (Rhodes et al., 2015). Sixty-six percent of African American men, compared to 34% of White men, are likely to be arrested before they reach 30 years of age (Tillman, 1987). Additionally, Hispanics and African Americans are arrested with weaker evidence justifying arrest than Whites (Petersilia, 1983). This results in minorities being detained and released at a higher rate than Whites. In line with these findings, Gelman and colleagues examined data from 125,000 pedestrian stops and found that after controlling for precinct-level differences, African American and Hispanic individuals were stopped more frequently than Whites (Gelman, Fagan, & Kiss 2007). Beyond initial contact, the nature of citizen-police interactions is also raced. Males, Blacks and younger citizens were found to be more likely to experience

force when interacting with the police, making young, Black males most at risk for experiencing police force (Durose, Smith, & Langan, 2007). While it is true that generalities cannot be made about police departments nationwide, consistent patterns of difference suggest institutional discrimination. Going further, scholars have linked the institution of policing and the behavior of officers to the racial injustices that trend in the profession (Legewie, 2016; Voigt et al., 2017). Because police officers usher individuals into the system, scholars have identified them as crucial actors in maintaining the status of Black Americans as a part of a perpetually oppressed population (Blauner, 1969). That is, the actions of police officers – which entail a lot of discretionary decision-making - are directly related to the realities of race and policing. Therefore, by analyzing actions of police officers, greater insight on racial disparities may be gained. Suggestions in the literature that have been made to reduce disparities in policing include: community approaches that integrate the public in policing strategies, alternatives to arrest, cultural competency training for police officers, and continued research and assessment of racially disparate outcomes to direct future approaches (TSP, 2008). After police interaction, arrest, and investigation, racial disparities persist inside the courthouse.

1.2.2 Racial Disparities in Courts Judges, prosecutors, and defense attorneys each play a unique role in the determination of guilt and sentencing that occurs in court proceedings. Due to the nature of their work, they too are granted extensive discretion in their duties where legal and extralegal factors are often considered. Legal factors are relevant to an individual’s criminal justice proceedings such as a criminal record and the seriousness of the crime. Extralegal factors do not directly relate to one’s experience with the

system and include race, ethnicity, gender, sexuality and class. Idealistically legal factors would solely be considered by criminal justice professionals when making their decisions. However, because race is inextricably linked to every aspect of U.S. society, this is difficult (Walker et al., 2012). Statistics reveal that extra-legal factors significantly impact court outcomes. African Americans are more likely to be sentenced to prison instead of probation than their White counterparts in similar situations (Rhodes et al., 2015). Similarly, the American Civil Liberties Union (2014) found that Black males receive sentences that are nearly 20% longer than White males convicted of the same crimes. Blacks are also 20 times more likely than Whites to be sentenced to life without parole for nonviolent crimes (ACLU, 2014). This may be because prosecutors use their discretion more harshly with Black defendants than

White defendants, being more likely to charge them under legislation that requires longer sentences (ACLU, 2014). Beyond the discretion of courtroom actors, criminal legislation also contributes to existing racial disparities in justice proceedings. According to some scholars, the war on drugs has created legislation designed to target minority groups, particularly African Americans (Alexander, 2012; Steiner, 2001). The 100:1 sentencing disparity between crack and cocaine, for example, led to Blacks being incarcerated at much higher rates than Whites because crack is more commonly found in African American communities (Alexander, 2012; ACLU, 2014; Steiner, 2001). Although Congress passed legislation in 2010 that reduced the sentencing disparity to 18:1, punishment for the drug in different forms is still unequal (ACLU, 2014). Sentencing disparities not only exist between Black and White defendants but also by skin color, finding that medium- and dark-skin Black defendants receive sentences that are 4.8% higher than

Whites; while the sentencing differences between light-skinned Black defendants are not different than their White counterparts (Burch, 2015). Additionally, Fader and colleagues find that the decisions of court actors in juvenile cases also lead to differential outcomes by race and ethnicity in terms of placements for young offenders (Fader, Kurlycheck & Morgan, 2014). Scholars have put forth several suggestions to reduce racial disparities in sentencing including developing culturally competent guidelines, setting up commissions to punish criminal justice actors for adverse behavior and abolishing legislation that perpetuates racial inequalities and other types of injustice (ACLU, 2014; TSP, 2008). In sum, a variety of factors contribute to the racial disparities that exist in court outcomes including legal and extra-legal factors, courtroom actor discretion and criminal legislation. As in policing, the role of criminal justice practitioners in justice institutions are tied to outcomes that have been empirically established as systemic racial/ethnic injustices. Judges, prosecutors and defense attorneys make decisions and behave in ways that contribute to the differences in treatment of the accused based on their racial and ethnic identities. Once, the accused are convicted and sentenced racial disparities persist under correctional supervision.

1.2.3 Racial Disparities in Corrections Racial disparities in corrections are often discussed in terms of the demographic make-up of the incarcerated population, directly relating to the racial disparities found in courts. In 2014, 6% of Black males between the ages of 30-39 were incarcerated, compared to 2% of Hispanics and 1% of Whites in the same age range (Carson, 2015). Carson also showed that African Americans, entered prison at a rate 6 times higher than Whites. Additionally, corrections disparities have been

documented in terms of resources available to inmates. Scholars acknowledge that minorities are more likely to enter a correctional facility with untreated problems than Whites due to lack of resources in neighborhoods. Prisons and jails are lacking the resources necessary to adequately care for the individuals in their custody. This leads to differential treatment and outcomes across racial/ethnic groups (TSP, 2008). To reduce racial/ethnic correctional disparities scholars have suggested increasing the diversity of correctional staff and culturally competent training (TSP, 2008). Sentencing decisions, available resources, programs and other forms of support that incarcerated individuals need to successfully complete their sentences and reenter society are factors that contribute to the disproportionate representation of minority individuals in U.S. prisons and jails.

Although corrections employees are not directly responsible for the demographic make-up of those they supervise, scholarship often acknowledges the critical role they play in determining the experiences and outcomes faced by detainees when they emphasize the importance of proper training. This means, that even in corrections, criminal justice actors and institutions are critical in reversing racially unjust outcomes. Schwirtz and colleagues 2016 New York Times article found that the demographic makeup of prison staff is correlated with racial/ethnic difference in disciplinary actions for inmates. That is, in most New York state prisons minority detainees were disciplined at higher rates and with more impunity than their white counterparts (Schwirtz, Winerip, & Gebeloff, 2016). However, the article details at Sing-Sing Correctional Facility where Black officers are over half of the uniformed staff, disparities in discipline no longer exist. This article suggests that there must be an overrepresentation of people of color within law enforcement organizations for

inmate disparities to dissipate. This aligns with a policing study that stresses having organizations where minorities are a critical mass (approximately 40%) of the workforce can lead to a reduction in racially disparate outcomes, such as police- involved homicides (Nicholson-Crotty, Nicholson-Crotty, & Fernandez 2017). Although diversifying law enforcement officers may seem like a suitable response to the racial disparities seen in the system, the answer is not black and white. Jurik’s 1985 study shows minority correctional officers hold more positive orientations about inmates than their non-minority counterparts. However, work organization characteristics (such as tenure and being assigned to a minimum-security unit) must also be considered to improve negative staff orientations overall (Jurik, 1985). Another study finds that diversifying staff will not guarantee improved attitudes towards prisoners, nor overall beliefs about the correctional system (Jacobs & Kraft 1978). Meaning, the racial/ethnic identity of an officer does not solely predict their performance and opinions on and off the job. Another complication in relying on the demographic makeup of prison staff to counter racial/ethnic tensions behind bars are the high attrition rates that plague correctional facilities. Conflict exists between young, urban, Black guards and executive staff (who are often majority White); White guards struggle with interactions with minority inmates (Jacobs & Grear, 1977).

Another study finds that the most influential factors in determining turnover intent among jail staff are work-related factors such as job satisfaction and work environment (Leip & Stinchcomb, 2013). Taken together, although the race of correctional staff may matter in terms of disciplinary sanctions, for inmates other factors are also critical.

This dissertation goes beyond the traditionally defined extralegal and legal factors that are examined as contributors to racially inequitable outcomes across the system, by focusing on the role of criminal justice actors. That is, in a criminal justice system that has been identified by scholarship as racially unjust, disproportionately impacting minority populations, this dissertation seeks to theoretically and empirically understand the role of criminal justice actors in perpetuating these outcomes.

1.2.4 Challenges in Racial Disparity Research Criminal justice scholars face challenges as they examine racial disparities across the system. As discussed earlier, researchers consider legal and extra-legal factors when exploring racial/ethnic disparities. Legal variables include criminal history, arrest record, and severity of offense. Influential extra-legal factors include race, age and gender. Although idealized notions of the system may conclude that only legal factors should shape criminal justice outcomes, scholarship has found that disparities in sentencing exist even after considering related legal factors, signifying the role of extralegal variables (Rouseau & Baranger, 2017). Studies find that controlling for legal and extra-legal factors does not nullify racially disparate criminal justice outcomes (Brown & Sorensen, 2014; Clair & Winter, 2016; Feldmeyer &

Ulmer, 2011; Yang, 2015). Brown and Sorensen examined data on juvenile offenders from the largest county in Texas and demonstrated Black and Hispanic juveniles were approximately three times more likely to be transferred to adult courts than their non- Hispanic White counterparts, controlling for legal variables such as criminal history and offense type (2014). Another study examined how minority populations influenced sentence length, using racial threat theory (Feldmeyer & Ulmer, 2011). The racial threat hypothesis posits that as the size of racial minority populations - an

extralegal factor - increases, Whites may feel that their social position and power is threatened, leading to an increase in formal social control over that group (Blumer, 1958; Bobo & Hutchings, 1996). Researchers found that while Blacks received more punitive sentences than Hispanics and Whites, their sentences were not related to the Black population in the judicial district (Feldmeyer & Ulmer, 2011). Also, sentences of Hispanics were found to contradict racial threat hypothesis as larger populations of Hispanics within districts led to shorter sentences for Hispanic defendants. In a recent study exploring racial disparities in federal sentences, Yang finds that racial differences exist after sentencing guidelines were invalidated in United States vs. Booker (543 U.S. 220 [2005]) (Yang, 2015). Yang considered factors that explained these racial disparities include the actions of prosecutors and judges. The study showed finding that criminal justice actors use discretion within their positions leading to racial/ethnic disparities (Yang, 2015). Qualitative work also allows scholars to better understand racial/ethnic disparate outcomes. Clair and Winter (2016) interviewed judges in a Northeastern state to better understand their perception of racial disparities in the criminal justice system. While most judges attributed racial disparities to legal factors such as differential treatment by criminal justice actors and criminal history, a few pointed to extralegal factors such as poverty (Clair & Winter, 2016). Therefore, when determining factors that contribute to criminal justice outcomes across the system, actors must be considered as well, particularly for racially disparate findings. Meta-analyses allow researchers to simultaneously examine the nuances of a vast amount of studies, including the theoretical framework, methodological approach, and each variable considered. Because of this, meta-analyses are an effective way of

determining how inconsistencies manifest in literature, particularly in terms of factors that influence racial/ethnic disparities across criminal justice outcomes. In a meta- analysis over 70 published and unpublished studies were synthesized in hopes of better understanding racial and ethnic differences in sentencing. Mitchell’s study showed (2005) race had a statistically significant influence on sentencing outcomes across studies, even when considering contextual variables. Despite accounting for study factors (i.e. methodology and sample size) as well as contextual variables (i.e. race of the defendant), African-Americans were sentenced more harshly than Whites. Although the strength of race’s influence on sentencing outcomes varied across studies, it remained statistically significant (Mitchell, 2005). Critical literature reviews are also an excellent way of taking stock of the trends in scholarship on a topic. The persistence of race is highlighted in Weizter (1996), finding that discrimination in the criminal justice system is indisputable. In a critical review of the literature on racial discrimination in the justice system, the author notes that discrimination is a complex concept to analyze particularly in less visible parts of the justice system – such as corrections. Furthermore, Weitzer acknowledges that studies that conclude race is not important in one part of the criminal justice system may be accurate but ignore the presence of discrimination in other areas. Therefore, it is critical to examine multiple points within the criminal justice system to determine the true nature of racial bias (Weitzer, 1996). The way data are collected can create several challenges that may lead to inconsistency in research findings. Each part of the system (police, courts and corrections) collects data independently and information about race may be inconsistent. For instance, while some institutions allow inmates to self-identify as a

certain race, others allow employees to assign individuals to a racial/ethnic group. This can lead to phenotypic racial/ethnic classifications that fail to consider the less- visible aspects of racial/ethnic identity. Access to data are also problematic for criminal justice researchers. Oftentimes criminal justice records are restricted due to confidentiality, sensitivity of data, and individual rights (both offenders and victims). Researchers may not be privy to the existence of some institutional statistics or given restricted access to some types of data. Therefore, some research questions may only be answered in part, if at all. This is problematic, especially in terms of race relations. Institutions that want to appear unbiased may strategically grant access to specific types of data to maintain that status. Additionally, scholars have found that certain types of methodology such as ethnography (Short & Hughes, 2009) may be invasive to a highly invisible institution, like corrections. Ethnography is a research method that involves observing groups and people as they live their everyday lives within social settings of interest (Emerson, Fretz & Shaw, 2011; Silverman, 2013). It is often appropriate to explore the nuanced behavior of individuals that is taken for granted, such as employee behavior and institutional responses – the focus of this research. Additionally, institutions that have a negative narrative in society, like police who struggle with legitimacy and trust in communities, may also shy away from certain types of research. For example, one study found that law enforcement agencies may deem the idea that racial biases exist within their organization, a way to degrade and insult them (Harris, 2007). This resistance to certain issues or methodologies in research impacts how data are collected – if at all.

Data access challenges have practical implications for research such as unanswered or partially explored research questions. Scholars attempting to study the impact of mass incarceration on children will be restricted if unable to integrate criminal justice and department of education data. Researchers seeking to predict how police officer behavior impacts case processing will be limited if data from courthouses and police agencies are not captured in a way that are compatible for analysis. Ultimately, authors stress that although data management systems often exist in institutional silos, integrating data from different social institutions can result in research that leads to more holistic, evidence-based practice (DeHart & Shapiro, 2017). Along with statistics, justice scholars have also relied on theory to understand the existence of racial and ethnic disparities. Conflict theorists, for example, state that the criminal justice system is an extension of the law whose purpose is to maintain the power of the dominant group by controlling individuals who threaten that power (Walker et al., 2004). In line with the conflict tradition, racial threat theory hypothesizes that when the dominant group feels threatened economically, politically and/or symbolically by a minority group they will act by using formal control to protect their positions (Bonacich, 1972). In other words, the criminal justice system embodies formal control in our society and – according to racial threat scholarship – is used to control minority populations that pose a threat to the majority. In a similar vein, recent scholarship has garnered a lot of attention by powerfully arguing that the criminal justice system is a formal social control mechanism that is being used to oppress people of color, particularly African Americans, from becoming a threat to the

White racial hierarchy that currently exists in the United States (Alexander, 2012; Eitle, D’Alesso, & Stolzenberg, 2002; Van Cleve & Lauren Mayes, 2015). Michelle Alexander’s book The New Jim Crow (2012) argues that the system of mass incarceration is a new form of the racial caste system that has perpetuated in America since its inception. Because the criminal justice system targets, processes, prosecutes, labels and stigmatizes Blacks in America more than any other racial group, Alexander persuasively argues that systemic control over minorities is not a result of individual actions nor cultural tendencies but instead structural design. By controlling minorities through an institution that is purposed with achieving just outcomes, mass incarceration operates colorblind. Colorblind ideology is a framework that conceptualizes race relations in the United States as covert actions and inactions that reinforce and enhance the dominance of Whiteness in society while appearing to be racially neutral (Bonilla-Silva, 2013). Scholars research how colorblind ideology is enacted in society through individuals and institutions to perpetuate the power of the White racial hierarchy in many ways, including scientific methodology and logic in research (Zuberi & Bonilla-Silva, 2013). Taken together, this work takes the perspective that the differences between racial and ethnic groups in criminal justice outcomes are not merely disparity but discrimination.

Scholars have relied on an assortment of theories to understand the existence of racial and ethnic disparities in the criminal justice system including conflict theory (Walker et al., 2012; Mitchell, 2005), racial threat hypothesis (Bonacich, 1972; Feldmeyer & Ulmer, 2011), focal concerns theory (Fader, Kurlychek, & Morgan, 2014), social control theory and colorblind ideology (Alexander, 2012), theory of reasoned action (Ajzen & Fishbein, 1977, 1980), theory of coercive actions (Tedeschi

& Felson, 1994), expectancy theory and organizational theories (Engel, Calnon, & Bernard, 2002). Conflict theorists state that the criminal justice system is an extension of the law whose purpose is to maintain the power of the dominant group by controlling individuals who threaten that power (Walker et al., 2004; Mitchell, 2005). In line with the conflict tradition, the theory of racial threat hypothesizes that when the dominant group feels threatened economically, politically and/or symbolically by a minority group they use formal control to protect their positions (Bonacich, 1972). Feldmeyer and Ulmer (2011) examined the link between racial disparities and federal sentencing finding no support for the theory. The population of Blacks and Hispanics are not correlated with the sentences received, as expected by the racial threat hypothesis (Feldmeyer & Ulmer, 2011). Using focal concerns theory, Fader and colleagues found that Black and Latino youth are less likely to be committed to therapeutic facilities, in comparison to their White counterparts. Expectancy theory, on the other hand, relates racial disparities in the justice system to organizational expectations from the output of workers (Engel et al., 2002). Generally, theory application to racial/ethnic disparities in the justice system leads to mixed findings. For example, Mitchell finds that although conflict theory may assist in determining what influences outcomes; the threat hypothesis needs to be altered to better fit the systemic racial differences (Mitchell, 2005). Additionally, theories that are applied in one area of the criminal justice system often do not translate or are not applied to others (Walker et al., 2004). As discussed above, data access, theoretical and conceptual differences make it challenging to apply theories across the system in a uniform and effective way.

In all, these studies demonstrate the persistence of racial/ethnic biases in the criminal justice system, made evident by empirically controlling for legal and extralegal variables. Importantly, some studies also highlight the key role criminal justice actors play in these outcomes, particularly when actors have considerable amounts of discretion in their positions. Additionally, inconsistencies in literature exploring racial/ethnic disparities in the criminal justice system involve data issues and varying theoretical frameworks.

1.3 Purpose & Significance of Study The purpose of this dissertation is to understand how racial disparities perpetuate within the criminal justice system by exploring the nature of institutional responses to employee behavior. As shown, there are challenges in the racial disparities literature that include lack of data access and multiple theoretical frameworks that do not consider race a central factor, I will offer a single theoretical framework that prioritizes race and criminal justice actors as key parts in understanding how racial disparities exist across the system. In doing so, this dissertation makes a significant contribution to the literature by seeking to understand a systemic racial phenomenon through organizational mechanisms that discipline and praise those tasked with ensuring justice in our society. Offering a single theory that can be applied across the system, this dissertation takes a distinctive united theoretical approach towards understanding racial inequality. By prioritizing the role of race in shaping institutional responses and systemic outcomes, this work significantly contributes to the criminal justice and racial disparity literatures which focus on minority offenders. The following sections offer an overview of the dissertation by chapter, emphasizing the goals and general conclusions.

Chapter 2 reviews literature on criminal justice actor misconduct across police, courts and corrections. This systemic review of the literature also clearly lays out how – or if - criminal justice organizations hold their employees accountable. Theoretical and empirical literature is presented for each part of the system, allowing comparison across institutions.

Police misconduct scholarship continues to debate the structural, cultural and individual causes of officer misconduct. A review of the literature finds that police organizations encourage and protect officers that engage in misconduct through structural mechanisms. Research on prosecutorial misconduct also places a high burden of responsibility on organizations, including wrongdoing related to racial/ethnic disparities. There is significantly less scholarly research on correctional officer misconduct than other criminal justice professionals. Regardless of the invisible nature of total institutions to larger society or the status of the detained, correctional officer misconduct has major implications on outcomes for inmates. In all, this chapter emphasizes how criminal justice actor misconduct and institutional responses impact organizational outcomes. The next chapter develops an original theoretical model to understand racial disparities across the system – the institutional response model of social control.

Chapter 3 presents a single theoretical model that explores how institutions respond to employee behavior which can perpetuate a culture of racial/ethnic inequality, the institutional response model of social control. I begin with a review of relevant theoretical underpinnings that inform the model: Eduardo Bonilla Silva’s colorblind ideology and Donald Black’s theory of law and social control. These two approaches allow for organizational level analysis while acknowledging the important

roles individuals play on social outcomes. Additionally, the role of race is critical in each of these theories. Colorblind ideology conceptualizes race relations in the United States as covert actions and inactions that reinforce and enhance the dominance of Whiteness in society while appearing to be racially neutral (Bonilla-Silva, 2013). Black’s theory emphasizes that social control is applied according to multidimensional social characteristics. Next, the three core elements of the model are described in detailed. Visibility of the action captures levels of awareness and severity. Institutional context variables account for individual-level characteristics acquired on-the-job, such as rank and tenure. Lastly, status of the actor involves different dimensions of identity acquired prior to employment such as race, age and sex. This is followed by an emphasis on the importance of applying the theory to correctional systems. Lastly, this chapter discusses the research setting, a Mid-Atlantic department of correction, acknowledging the importance of context on findings and conclusions. Chapter 4 discusses the mixed method research design. Research questions and hypotheses are based on two main questions: Is there empirical support for IRM? And if so, how is IRM enacted within correctional employment. This chapter also discusses the nature of data access and methods. Data access required a series of approvals from the department of correction and the University of Delaware’s institutional review board. The quantitative sample was randomly selected from the total number of department of correction employees, provided by the human resources department. Qualitative observations were conducted on several training classes, both inside and outside of prison. Univariate, multivariate regression and qualitative analyses were conducted using the statistical software packages STATA 13 and NVivo 11.

Chapter 5 presents quantitative results conducted to test the institutional response model of social control. This mixed methods study prioritizes quantitative analysis, using qualitative findings to support and enrich the empirical findings. Quantitative results are presented in this chapter including univariate, logistic regression and negative binomial regression results. Because institutional responses to employee behavior can be either positive or negative, a series of separate analyses were conducted to understand commendations and sanctions. Quantitative analyses show varying levels of support for all three of the elements of the institutional response model. Logistic regression models find the presence of commendations in an employee’s file is related to institutional context and status of the actor variables. Age and visibility are influential in the number of praise reports employees had. Shifting to disciplinary sanctions, again all three elements of IRM were supported. Models examining the presence of sanctions in files linked visibility of actions and institutional context variables to the binary outcome. Visibility and status of the actor variables are statistically significant in determining the number of sanctions in files. Chapter 6 presents results from qualitative analysis of observational field notes taken over a 4-month period. Observation allows for a more nuanced understanding of how racial disparities can perpetuate within criminal justice organizations through institutional response processes. Data were coded in two phases, open and focused coding. The purpose of the qualitative data is to confirm the quantitative findings and provide details on institutional response processes. Through language, building structure and victim impact, qualitative narratives illustrate how visibility of employee behavior shapes the presence and nature of institutional response. While, distinctions

among correctional employees emphasize how institutional context variables shape MADOC’s responses to behavior. Lastly, the way race and previous experiences of employees influence outcomes in training is illustrated through narratives drawn from field notes. Chapter 6 concludes by linking quantitative and qualitative results to previous literature to determine how racial disparities are perpetuated within criminal justice organizations. Support for each hypothesis is discussed in detail, emphasizing connections between findings, empirical literature and the theoretical underpinnings of the IRM. Chapter 7 summarizes the purpose, contributions and central findings of this dissertation. Next, the limitations of the study are reviewed, particularly in terms of how alterations in data and access can strengthen the study. As a former correctional officer, scholar, and woman of color, I then list various challenges that arose during the research process linked to my identity. Limitations include being required to provide a level of feedback to training staff towards improving classes during the observation period. The concept “status of the victim”, or the multidimensional identities of those impacted by employee behavior, is suggested as a future consideration for scholars interested in this line of work.

Lastly, three policy recommendations are suggested: (1) continuing to consider race as a priority in analyses, (2) using multiple data sources to strengthen conclusions and (3) considering criminal justice actors as central players in criminal justice outcomes. It is further suggested that the institutional response model of social control be applied to other institutions within and outside of the justice system.

Chapter 2

CRIMINAL JUSTICE ACTOR MISCONDUCT ACROSS THE SYSTEM

“One chain, many links.” – Training Instructor

2.1 Who’s to Blame? Who is to blame for the racial disparities in the criminal justice system is a central question. As discussed in chapter 1, much of the previous literature on racial inequalities in the criminal justice system highlights legal and extra-legal factors. Few however, consider the role criminal justice actors play in the perpetuation of racial injustice, focusing instead on aggregate and institutional trends. Should police officers be held accountable for stopping more minorities than Whites? Or should police departments be held accountable because they train staff to react in differential ways to subgroups of the population? Are prosecutors responsible for the disparities in sentences between Hispanics and Whites? Or is the entire court system responsible for not ensuring that all defendants receive adequate defense counsel? Who is to blame for the subpar conditions that plague the correctional system? In other words, should criminal justice actors be held responsible for issues that are rampant across the system? Some scholarship suggests that even if criminal justice actors contribute to racial disparities in across the system this contribution may not be intentional, but instead implicit or unconscious (Fridell & Lim, 2016; Gravett, 2017; Parks & Davis,

2016). Lawrence’s work on unconscious racism, for instance, uses Freudian theories and social cognition to argue that because racism is embedded in our culture,

individuals are sometimes unaware of negative stereotypes that they hold against others (i.e. blacks are lazy or unintelligent). Because these prejudices are significant parts of our culture, they seem like rational ways of perceiving the world (Lawrence, 1987). In short, racially biased perceptions are central to the culture of society, rooted in individuals at the unconscious level. Related, scholars find that stereotypes or learned associations among groups and traits may be so deeply ingrained they persist in the face of contradictory information (Armour, 1995; Davis, 1989). Criminal justice actors are subject to the same cultural biases and stereotypes found in broader society. Implicit racial biases are pervasive (Gravett, 2017). Using the implicit-bias perspective, Fridell and Lim found that police officer biases lead to greater use of force against Black citizens than their counterparts (Fridell & Lim,

2016). Parks and Davis also note that it is imperative for judges to be aware and work to combat their implicit biases. Judges’ words, body language and attitudes expressed in the courtroom are influential in outcomes such as jurors’ perceptions (Parks & Davis, 2016). Implicit bias is now considered a major issue by many law enforcement agencies. The Department of Justice, for instance, made implicit-bias curriculum a part of their regular training for every employee in 2016.

Institutions are also critical in the perpetuation of discriminatory practices by operating in understandings that are embedded in cultural norms through written and unwritten rules, customs, and practices (Lopez, 2004). Lopez (2004) demonstrates this point by examining how Los Angeles courts exclude Mexican Americans from the grand jury pool. This influences how LatinX activists are prosecuted. The study found that although there was evidence of racial exclusion, judges did not act with discriminatory intent. According to Lopez’s theory, when individuals unconsciously

engage in discrimination, they rely on unchallenged, institutional guidelines and perpetuate racial/ethnic injustice. The institutional guidelines are sometimes unwritten and impact an employee’s experiences and identity on-the-job. For example, Yoshino finds that organizations can create dominant group norms that create pressure for minority workers to act in ways that minimize disfavored racial identities and conform to organizational culture (Yoshino, 2006). Whether criminal justice actors intentionally act in prejudice, employee actions shape institutional outcomes. Street-level bureaucrats (SLBs) are defined as “public service workers who interact directly with citizens during their jobs, and who have substantial discretion in the execution of their work (Lipsky, 2010, p. 3).” Criminal justice professionals qualify as SLBs because their jobs require interactions with citizens and they are given authority and discretion. Through their decisions, street- level bureaucrats create informal policy as they interact with clients, often coping with the demands of their jobs through short-cuts (Lipsky, 2010). Within the justice system, scholars cite 6 key points that require employee discretion and may lead to racial disparity: law enforcement, pre-adjudication, adjudication, incarceration alternatives (probation/community service), incarceration and reentry services (TSP, 2008). At each of these points, the decisions and behaviors of employees can have a considerable impact on the experiences of the detained and convicted. Society places a level of trust in criminal justice professionals because they are charged with achieving justice. Criminal justice actors violate this trust by engaging in a range of adverse behaviors, depending on the position they hold. Regardless of its form, misconduct has serious implications for individuals (termination, suspension, etc.) and institutions

(legitimacy, public trust, etc.). Despite these implications, misconduct among criminal

justice professionals is not researched equally across the system, particularly wrongdoing related to racial/ethnic disparities (see Appendix C). Literature on race and criminal justice actors is sparse, but the misconduct literature allows for the examination of criminal justice actors. The next section reviews misconduct among three key groups: police officers, prosecutors and correctional officers. By reviewing this literature, I emphasize the implications of criminal justice employee misconduct and different scholarly approaches.

2.1.1 Police Officer Misconduct Police officers are considered the gatekeepers of formal social control (Wolfe & Piquero, 2011). Often the first point of contact between an individual and the system, police officers are given the power to enforce laws. Police officer misconduct can lead to the violation of civil and constitutional rights, civil litigation against the department and damage citizen attitudes towards the police (Skolnick & Fyfe, 1993; White & Kane, 2013). Wolfe and Piquero’s 2011 study defines police misconduct as actions that lead to a formal response such as a complaint, internal affairs investigation, or departmental disciplinary charges. This definition overlooks incidents of misconduct that are handled informally or never brought to the attention of police departments. Many policing agencies define officer misconduct via the acronym FADO which stands for the use of unnecessary Force, abusing Authority, being Discourteous and/or using Offensive language when interacting with citizens (Civilian Complaint Review Board, 1993, Henriquez, 1999; Seron, Pereira & Kovath, 2004; Walker & Wright, 1995). Though instinctively one may deem these behaviors unacceptable, policing is a complex job where discretion is necessary and requires on- the-spot decision making and innovative techniques to solve problems. Law scholars

relate types of police misconduct to wrongful conviction. Examples include inadequate police investigation, coaching witnesses before trial, suppressing evidence that suggests the innocence of the accused (exculpatory), using false evidence and, using pressure to obtain a confession from a suspect (Ramsey & Frank, 2007). Although there is a vast amount of academic research on police misconduct, few empirical studies have been done to understand how and why misconduct manifests (White & Kane, 2013). This may be because researchers lack access to misconduct data and/or because there are multiple definitions of policing misconduct (Kane & White, 2009; Ivkocvic, 2005, Sherman, 1978). Furthermore, when misconduct data are collected it is often ignored reflecting the investment some police departments have in researching injustices to create change (Schwartz, 2010). Because the definitions of police misconduct vary, access to officer behavioral data are restricted and police departments may lack the motivation to change at the organizational level, misconduct can perpetuate through the system. Scholarly approaches to police misconduct also vary greatly.

2.1.1.1 Empirical Approaches to Police Misconduct Some scholars explore police misconduct by empirically measuring the attitudes and experiences of citizens and officers. Procedural justice, or the fairness of the process organizations follow, is key in determining how people perceive police (Lind & Tyler, 1988). Fair treatment by police is more influential than receiving a favorable outcome when determining citizen attitudes about officer behavior (Tyler & Huo, 2002). In other words, if someone feels they were justly treated during their interactions with a police officer they are more likely to say their experience was favorable compared to someone who feels they were wronged. This is true even if

both parties receive a formal sanction (ticket, arrest, etc.). Weitzer & Tuch examined White, African American and Hispanic citizen perceptions of four types of police misconduct using nationally representative data. The study found that each of the four types of police misconduct identified (verbal abuse, excessive force, unwarranted stops and corruption) are shaped by race, personal and vicarious interactions with police officers, exposure to media coverage of police behavior and neighborhood conditions (Weitzer & Tuch, 2004). Results showed that for each type of misconduct Whites were least likely to hold negative views about the police, followed by Hispanics and Blacks (Weitzer & Tuch, 2004). In other words, regardless of type of police officer misconduct, Whites perceived police officers more favorably than Hispanics and Blacks, the ladder consistently having the most negative attitudes towards police. This may be because African Americans and Hispanics are more likely to be exposed to mass media reports about police misconduct (Chermak, McGarrell, & Gruenewald, 2006; Chiricos, Padgett, & Gertz 2000), live in disadvantaged/high-crime neighborhood conditions (Sampson & Bartusch, 1998) and have negative interactions with the police - either personally or vicariously (Adams, 1999). When given vignettes that required citizens to rate the level of seriousness for a violation of FADO, respondents considered both legal and extralegal factors when making their determinations (Seron, Pereira, & Kovath 2004). Although civilians believe that police officers should be professional, they also recognize the need for discretion in policing due to the complex nature of the job. Literature on citizen attitudes about policing demonstrates the importance of considering organizational factors, race, and employee behavior to improve police-citizen relations.

Police officer attitudes about misconduct have also been examined in the literature. One study explored how police officers in different countries evaluated the level of seriousness of instances of police misconduct. Police officers from Croatia, Finland and the United States rated scenarios that described various forms of police corruption and excessive force. Results found that although there were general similarities among all police officers in judgment on police misconduct, there was more homogeneity within countries than between (Ivkovic, 2005b). One major contribution of this work is that police officers’ perceptions of adverse behavior are heavily influenced by the context and culture of their larger environment/institution. In other words, violations of FADO are deemed less severe by police themselves if institutional factors neutralize adverse behavior. White and Kane examined the complex pathways that lead police officers to participate in misconduct, resulting in termination. Police officers were at risk for certain types of misconduct throughout their careers depending on a range of factors including officer race, prior criminal history and military service (White & Kane, 2013). The study finds that it is key for police departments to screen officers for “red flags” across the course of their careers to reduce and/or prevent adverse behavior (White & Kane, 2013). Red flags serve as a structural device that would make high-risk officers more visible, triggering proactive institutional efforts to reduce misconduct. In sum, while individual factors such as officer race and background were relevant, scholars consistently highlight the importance of policing organizations in determining attitudes, perceptions and prevention of officer misconduct.

2.1.1.2 Theoretical Approaches to Police Misconduct Police misconduct scholarship is largely atheoretical, although researchers have started applying theoretical frameworks and empirically testing their validity to assist in a more thorough understanding (Wolfe & Piquero, 2011). “Rotten apple” and “rotten barrel” are two concepts in law enforcement literature that symbolize how misconduct manifests in policing. Rotten apple theorizes that adverse behavior exists at the individual level and rotten barrel considers misconduct to be an institutional problem requiring structural remedies (Sherman 1974). Rotten apple assumes integrity within the institution of policing and is more likely to be the perspective taken by police administrators who are invested in its success, morality and honor (Weitzer & Tuch, 2004). African Americans and Hispanics, on the other hand, are more likely to embrace the rotten barrel perspective (Weitzer & Tuch, 2004) which places blame on the system. Further supporting the rotten apple argument, individual-level explanations of police misconduct are common in the literature because police officers must make the choice to engage in adverse behavior (Manning, 2009). Others feel that it is a mistake to use individual officers as scapegoats (Armacost, 2003) when trying to find the culprit(s) responsible for misconduct as the relationship between organization and individual cannot be easily separated (Wolfe & Piquero, 2011). Pogarsky and

Piquero found that deterrence theory plays a key role in whether police officers choose to engage in misconduct. Results suggest that while all three elements of deterrence theory were influential in police officers’ decisions, certainty and swiftness weighed more heavily than severity of potential punishment (Pogarsky & Piquero, 2004). Applying life course criminology to adverse behavior in policing suggests that violations of rules are more likely earlier in careers though some persist throughout an

officer’s tenure (Harris, 2016). Increases in age, experience, and meaningful social bonds can lead to turning points and desistance from policing misconduct. Police subculture perspective is another theoretical framework used to understand police misconduct (Bayley & Mendelsohn, 1969; Knapp, 1973). “The unique demands that are placed on police officers, such as the threat of danger as well as scrutiny by the public, generate a tightly woven environment conducive to the development of feelings of loyalty” (Skolnick, 2005, p. 302). The tight bond between officers creates an informal code of silence, secrecy, and solidarity which socializes officers not to report each other for violations of rules. However, if officers believe that their institutional management is fair, they will break the code of silence (Westley, 1970; Wolfe & Piquero 2011). In other words, structural policies can trump subcultural influences. Another study finds novice officers, experienced veterans and supervisors were all more likely to deem use of excessive force worthy of severe discipline than those that were moderately experienced and in nonsupervisory positions (Micucci & Gomme, 2005). Police subculture has also been linked to racist attitudes towards minority civilians and excessive force in interactions with the public (Christopher, 1991). Organizational theorists argue that it is vital to consider institutional context when exploring police officer behavior because the culture of the organization can manipulate subordinates into a mental framework that shifts their ways of feeling and thinking (Schein, 1993). Therefore, the organization is ultimately responsible for officer misconduct (Ivkovic, 2005a; King, 2009). Other theories that have been applied to police misconduct include racial threat (Lawton, 2007), social disorganization theory (Kane, 2002; Terrill & Reisig, 2003), and social learning theory

(Chappell & Piquero, 2004). Public demands for a higher level of accountability for

police has led to changes in administration guidelines and growth in police misconduct scholarship (Mastrofski, Worden, & Snipes, 1995).

2.1.1.3 Racial Disparities & Police Misconduct Racial disparities exist not only in police interactions with the public but also within department disciplinary procedures. Minority police officers may be subject to differential and/or harsher treatment than White officers when they are disciplined (Bolton & Feagin, 2004; White & Kane, 2013). Rojeck and Decker’s 2009 study examines data collected by internal affairs of a police department over a 5-year period on police disciplinary procedures including complaints, determinations of guilt and any formal sanctions applied. Analysis reveals that while minority officers are overrepresented in terms of complaints against them (particularly those made internally) they are not found guilty more often than their White counterparts (Rojeck & Decker, 2009). Minority officers are also subject to more internal complaints, meaning those made by fellow police officers (Lersh & Mieczkowski, 2000). Research also suggests that Black and Hispanic officers are more likely to be terminated early in their careers than Whites (White & Kane, 2013). Some scholars claim minority officers have more complaints against them because they are more likely to engage in misconduct (Cohen & Chaiken, 1972). Lot argues that by pushing for increased diversity among officers, police departments have reduced their employment standards leading to a decrease in the quality, integrity and effectiveness of policing (Lot, 2000). However, studies also find that minority police officers are more likely to be assigned to high-crime areas where opportunities to participate in misconduct are greater (Fyfe, Kane, Grasso, & Ansbro, 1998). In sum, research shows minorities are subject to disparate justice regardless of their status as a police officer. Furthermore, institutions

take responsibility for police officer misconduct through indemnification. Merriam- Webster defines indemnify as “to secure against hurt, loss, or damage” or “to make compensation for incurred hurt, loss or damage.” Police officers are nearly always indemnified, hardly ever being held responsible for punitive damages filed against them by complainants, even when they were formally disciplined for their behavior

(Schwartz, 2014). Organizational actions like indemnification symbolize institutional responsibility for police wrongdoing. In sum, there is a vast amount of scholarship on police misconduct including citizen attitudes, police perceptions, theoretical applications, assessments of disciplinary structures and institutional responsibility. Although citizens define misconduct more broadly than police (Lersh & Mieczkowski, 2000), most scholars and practitioners use the acronym F.A.D.O. The debate between structural, cultural and individual level causes of police misconduct continue. Processes such as indemnification link police departments to officer misconduct. That is, although formal policy seems just, police organizations encourage and protect officers that engage in misconduct through structural mechanisms. Organizations are critical in researching police misconduct.

2.1.2 Prosecutorial Misconduct “The prosecutor has more control over life, liberty and reputation than any other person in America (Davis, 1969).” Many consider prosecutors to be the most powerful position in the criminal justice process (Davis, 1969; Dennis, 2007; Gottfredson & Gottfredson, 1988; Ramsey & Frank, 2007). The power of the prosecutor includes the ability to decide whether the accused are charged, what the charges consist of, the terms of plea agreements and the type of resources to invest in a

case. Furthermore, this power is protected by the extensive immunity prosecutors receive by the judiciary, professional organizations and in civil liability (Caldwell, 2013; Dennis, 2007; Polzer, Nhan & Polzer, 2014). The trust society places in prosecutors should be the initial and most essential element to consider when discussing prosecutorial misconduct (Schoenfeld, 2005). There are a range of actions that are considered adverse behavior among prosecutors including inadequate case investigation, the suppression of evidence that would negate the guilt of the accused, inappropriate comments during court proceedings, applying extreme pressure when negotiating a plea, prompting a witness, using false testimony of a witness, overcharging a defendant and improperly selecting a jury based on characteristics such as race, sexuality and religion (Caldwell, 2013; Platania & Moran, 1999; Ramsey &

Frank, 2007). Law scholars frequently refer to Brady violations when discussing prosecutorial misconduct (Burke, 2009; Polzer et al., 2014). Brady violations refer to the ruling of the United States Supreme Court which states that prosecutors may not withhold or suppress evidence that may be beneficial to the accused because it violates due process (Brady, 2011). While some scholars believe adverse behavior among prosecutors is relatively rare (Caldwell, 2013), others recognize the opposite.

“Prosecutorial misconduct is not a rare event, but it often goes undetected, unreported, or no action is taken by the criminal justice system (Polzer et al., 2014, p. 652).” Polzer and colleagues acknowledge that prosecutorial misconduct is often made invisible by criminal justice organizations. Here again, the role of organizations in determining how employee behavior is handled, disciplined and defined is critical.

2.1.2.1 Approaches to Studying Prosecutorial Misconduct Scholars often apply theoretical frameworks to understand prosecutorial misconduct. Like policing, subculture theory is a commonly applied framework that helps explain engagement in prosecutorial misconduct. Prosecutor offices promote an organizational culture that praises winning at all costs. (Polzer et al., 2014). This contradicts initial guidelines put in place by the American Bar Association in 1908 which states: “The primary duty of a lawyer engaged in public prosecution is not to convict, but to see that justice is done (p. 576)”. Today, these ethical standards remain largely unchanged. However, the culture of many prosecutor offices places more value on conviction rates and faults prosecutors if convictions are earned through illegal means rather than sharing culpability (Joy, 2006). The pressure prosecutors face to succeed relates to the criminological theory, anomie. Anomie is a criminological theory that posits criminal/deviant behavior is likely when socially defined goals are unattainable through legitimate means (Merton, 1938). Applied to the legal profession, anomie theory states that because prosecutors are under strain to achieve the goal of conviction but are not always equipped with the means to do so in traditional (i.e. legal) ways, they may break rules (Merton, 1938). When organizational expectations are misaligned with work realities it creates stress on prosecutors that may lead to misconduct. The “theory of prosecutorial misconduct” states that prosecutors violate the standards and laws of their profession when they evaluate the motives and opportunities for misconduct in a way that neutralizes controls that would prevent misconduct (Schoenfeld, 2005). The key to this theory is trust. The structure of the profession and the leeway prosecutors are granted by society through institutional means enables them to conclude misconduct is acceptable. Related, the theoretical

framework techniques of neutralization (Sykes & Matza, 1957) can also be applied to prosecutorial misconduct. When prosecutors conceive that obtaining convictions, rather than achieving justice, is best they may rationalize their illegal/inappropriate behavior (Polzer et al., 2014). Edwin Sutherland coined the phrase “white-collar crime” in 1939 to describe crimes committed by upper class professionals (Sutherland,

1940). Scholars consider prosecutorial misconduct white-collar crime, as they work under less supervision and have elevated levels of discretion which give them opportunity to commit offenses. Furthermore, prosecutors who engage in white-collar crime are also immune from the shame and stigma that society normally associates with criminality due to their social standing (Braithwaite, 1989). Lastly, the lack of negative consequences including the infrequent and relatively minor punishments in place for prosecutorial discipline does not serve as an adequate deterrent for prosecutors, making deterrence theory inapplicable in most cases (Polzer et al., 2014). Scholarship also recognizes that prosecutors are likely to have the same implicit/unconscious biases that plague broader society, particularly in terms of racial/ethnicity (Kang et al. 2011). Unconscious bias therefore, can lead to inequitable outcomes prosecutorial discretion based on subconscious race-based beliefs (Smith &

Levinson, 2011). Although unconscious bias research connects the role of race and criminal justice actors to racially disparate outcomes, it considers racial/ethnic inequity subconscious, rather than purposeful and/or tied to a broader institutional context. Currently in the United States there are 3 ways to reduce, eliminate and/or punish prosecutorial misconduct: (1) judicial sanctions – minor consequences that occur in courtroom proceedings, (2) professional discipline – extremely rare

occurrences that fail to deter prosecutors and (3) civil liability – which are also infrequent due to the protection prosecutors receive from the justice system itself (Dennis, 2007). Scholars also cite the infrequently called upon superseder power, which allows the appointment of a special prosecutor to investigate prosecutorial misconduct (Dennis, 2007). To further reduce prosecutorial misconduct, studies argue for a combination of transparence and external commissions to review and adjudicate prosecutorial wrongdoing while maintaining and acknowledging the independence of the position (Caldwell, 2013). McClelland proposes that the Supreme Court should adopt an absolute immunity policy for prosecutors to end the confusion that currently exists regarding how to prosecute those who engage in professional misconduct (McClelland, 2012). This suggestion, although radical, is meant to eradicate fear in prosecutors from doing their job in achieving justice as well as the false hope defendants currently have in convicting a prosecutor for wrongdoing. The article reasons that the current forms of recourse that are available to those wronged are hollow, because the system is set up to protect prosecutors (McClelland, 2012). In other words, adopting absolute immunity will release prosecutors from constraints and allow true transparency in the discipline of prosecutors: there will be none.

In all, theoretical scholarship on prosecutorial misconduct continues to identify organizations as important in determining how employees are held accountable in the criminal justice system. While scholarly literature regarding the nature, status and ways to improve prosecutorial misconduct are vast, there is much less empirical and theoretical scholarship on the issue than exists for policing. Research on prosecutorial misconduct also considers organizations responsible in shaping responses to adverse behavior. The relationship between prosecutors, organizations, and racial disparities is

clear as the high-level of discretion and protection that the profession affords offers no respite for victims. The next section will review literature concerning the misconduct of an even less explored criminal justice professional: correctional officers.

2.1.3 Correctional Officer Misconduct For correctional officers, institutional deviance is defined as actions committed during working hours that are against agency policy or that violate the law (Worley & Worley, 2011). Types of misconduct that correctional employees engage in include: providing inmates with contraband (drugs, alcohol, and other items not approved by the institution), using excessive force (Johnson & Bridgmon, 2009), engaging in inappropriate sexual activity with inmates (Marquart, Bamhill, & Balshaw-Biddle, 2001) and sexually harassing inmates (Cheeseman & Worley, 2006). However, incidents of correctional officer misconduct are significantly less researched as evident by database search results. Examining the search results in a range of databases; I found that prosecutorial misconduct scholarship is not only more common, but more comprehensive and forgiving than what is found on correctional officers. For example, using the search terms “prosecutor” and “misconduct” resulted in nearly 41,000 documents in Google Scholar. In stark contrast, the same database returned under

2,000 results for the search terms “correctional officer” and “misconduct”. This trend was found repeatedly in a range of databases including Web of Science, Sociological Abstracts and Criminal Justice Abstracts (see appendix C for tables that detail these search results). Many of the articles that came back searches for prosecutorial misconduct often referred to the immense amount of discretion and power held by prosecutors and the lack of accountability (Davis, 1969; Dennis, 2007; Gottfredson &

Gottfredson, 1988; Ramsey & Frank, 2007). Corrections scholarship focused on

opinions of the purpose, status and consequences of correctional work and sexual misconduct in prisons (Beck, Rantala, & Rexroat, 2007; Callahan, 2004; Cook & Lane, 2012; Hartley, Davila, Marquart, & Mullings, 2013; Mahfood, Pollock, & Longmire, 2013). This pattern reflects structural mechanisms both within society to grant more attention to higher status occupations, like prosecutors, while simultaneously permitting them more protection from sanction. Some scholars say that correctional employee deviance is given little attention in scholarship because prisons are total institutions (Goffman, 1961; Worley & Worley, 2011). Total institutions, like asylums and prisons, bureaucratically process people in isolation of normal society; individuals live, sleep, eat and recreate within the confines of the institution (Goffman, 1961). Even fewer studies compare correctional officer deviance to the deviance that occurs in other criminal justice occupations (Worley & Worley, 2011). An example of this is Ramsey and Frank’s 2007 examination of nearly 800 Ohio criminal justice professionals to explore opinions of the existence of errors in the system. The authors surveyed police, prosecutors, defense attorneys and judges but excluded correctional officers from the sample, without justification for doing so (Ramsey & Frank, 2007). This may be because correctional officers are not thought to contribute to the process of achieving justice in the system. The job of correctional officers is to supervise individuals in custody. Correctional officers do not play a role in determining who is incarcerated. However, excluding correctional officers from comparative analysis is a mistake because studies have also found that they shape the perceptions of justice and opinions of their detainees (Molleman & Leeuw, 2012).

The lack of literature on correctional officer misconduct is tied directly to the invisibility of the organization. Corrections institutions are less accessible to and engaged with larger society making issues (officer misconduct), under explored in the literature in contrast to more visible criminal justice institutions. In this way, organizations are again vital in how employee behavior is studied. Furthermore, because the behavior of correctional staff impacts outcomes for inmates, including those tied to race and ethnicity, employee misconduct is critical to investigate.

2.1.3.1 Approaches to Studying Correctional Officer Misconduct While there is a growing body of literature exploring correctional officer opinions and stress levels (Callahan, 2004; Cook & Lane, 2012; Hartley et al., 2013; Mahfood et al., 2013), significantly less is being done in terms of correctional officer misconduct. Occupational deviance, or actions that are in violation of the expected appropriate and ethical behavior that relates to a person’s job, has also been used to measure correctional officer misconduct. One study finds that correctional officers are typically prosecuted for excessive force as opposed to other types of misconduct (Johnson & Bridgmon, 2009). Additionally, the study demonstrated that correctional officers were charged more frequently than police officers from 2001-2006 (Johnson

& Bridgmon, 2009). This is interesting because of brutality directed at inmates by staff is uncommon in the literature (Worley & Worley, 2011) beyond retrospective reviews of the causes of major incidents such as riots (Morris, 1988) or more recently the examination of inmates’ complaints (Calavita & Jenness, 2014). Inappropriate sexual behavior in correctional settings is one growing vein of research concerning the deviant acts of correctional officers. In 2011, 48% of substantiated incidents of sexual misconduct in correctional institutions occurred between a staff person and an inmate,

more than a quarter of these incidents committed by female staff (Beck et al., 2014). Fifty-two percent of staff involved were terminated for their behavior, 43% of staff received another form of institutional discipline and only 6% were either arrested or prosecuted (Beck et al., 2014). Beck and colleagues found that inmates, responding to a Bureau of Justice prison sexual assault survey, held prison staff responsible for nearly 50% of all sexual assaults on inmates over a 3-year period (Beck et al., 2007). Qualitative studies have further explored inappropriate relationships between staff and inmates (Smith, 2012; Worley & Cheeseman, 2006; Worley, Marquart, & Mullings, 2003). Sexual misconduct by staff happens across genders and positions of power. In other words, male and female guards are inappropriate with inmates of either sex. Additionally, male and female inmates are inappropriate with guards of both sexes.

Scholarship and practitioners often shy away from female staff that commit sexual abuse, although they account for a sizable proportion of occurrences, because of the discomfort and disbelief that it causes (Smith. 2012). Attitudinal research on correctional officers show that those who feel that they are cared about by their colleagues are more likely to view the actions of other staff members as deviant, than those that do not feel cared for (Worley & Worley, 2011).

Like literature on other criminal justice professionals, the subculture framework has been applied to correctional staff. Research finds that novice correctional officers are socialized into the guard subculture by more experienced coworkers (Riley, 2000). The bond that is established between correctional officers provides rationalizations that justify certain actions, including types of misconduct, and extreme loyalty to each other (Kauffman, 1988). Here, as in policing, loyalty translates into a code of silence where adverse behavior committed by correctional staff is unreported contributing to a

higher level of misconduct including inappropriate and illegal sexual behavior, theft and brutality (Irwin, 2005; Human Rights Watch Women’s Rights Project, 1996) One officer admitted to fearing isolation and retribution from his colleagues if he reported inappropriate behavior of another officer (Prison Legal News, 2013). It is noteworthy that scholarship linking correctional officers, race and implicit/bias to institutional outcomes is sparse at best. Therefore, the relationship between correctional employees and racial disparities are unexplored relative to their criminal justice counterparts, even at the subconscious level of implicit bias. Prior to the 1960s federal courts relied on the expertise of correctional administrators to handle disciplinary actions within the institution. However, this hands-off doctrine resulted in prisoners’ rights rarely being protected. After the Civil

Rights movement prisoners were considered to have the rights of free individuals by the judiciary (Amicis, 2005). Currently, sanctions for correctional officer behavior include suspension, termination and prosecution (Smith, 2012). However, ultimate determinations of correctional officer sanctions have been a subject of dispute. For example, when a correctional officer was terminated for telling a prisoner to commit suicide by hanging herself, appropriate sanctions were hotly debated by courts and the internal disciplinary system of the California Department of Corrections and

Rehabilitation (Prison Legal News, 2013). Although the officer was initially terminated by the department of corrections, courts reduced the sentence to a month- long suspension. The department of corrections appealed, and the termination was ultimately reinstated. To aide in the fair administration of justice scholars suggest improved training of correctional staff, better supervision by correctional managers and administration, and installing security cameras that will survey and protect

inmates and correctional officers from adverse behavior (Amicis, 2005; Worley & Worley, 2011). Overall, there is significantly less scholarly research on correctional officer misconduct than other criminal justice professions. This may be explained by societal perception of correctional facilities, employees and inmates. Regardless of the invisible nature of total institutions, correctional officer misconduct has major implications. Because at least 95% of all state prisoners will be released at some point (Hughes & Wilson, 2016), it is vital that experiences during imprisonment go beyond punishment and towards restoration. Research has also found that correctional officer demographics impact the racial/ethnic disparities in inmate disciplinary actions (Schwirtz et al., 2016). By acknowledging the considerable influence correctional officers have on inmate realities – including those tied to race and ethnicity – this work brings attention to the importance of examining all phases of the criminal justice system with equal rigor to address systemic problems such as racial disparities. In other words, the phenomenon of correctional officer deviance must be studied with the same intensity as police officers, prosecutors and other criminal justice professionals who are more visible.

In this chapter I reviewed issues in the criminal justice system: racial disparities and professional misconduct. Five major themes have emerged from this research that are critical points for this work. First, misconduct exists at every step of the criminal justice system. Although, police officers, prosecutors, and correctional officers are necessarily equipped with discretion to function in their perspective positions, misconduct occurs. Accountability is critical. Second, police scholarship addresses the connection between criminal justice actors and racial/ethnic disparities

far more than research on prosecutors and correctional officers. Although scholars have made connections between implicit/unconscious bias and CJ actors across the system, the link between race and institutions is largely missing when CJ actors are examined. Third, organizational factors play a critical role in both misconduct and racial disparities. Police departments, prosecutorial offices and facilities of confinement set the culture and regulations that dictate whether certain types of behavior are acceptable. The fourth theme emphasizes that scholarly literature has given less attention to some areas of the criminal justice system than others. Police and prosecutors have received more attention among researchers than corrections. Lastly, the connection between systemic racial disparities and misconduct of criminal justice actors has received sparse attention, if any. While race is tied to police officer misconduct, racial analysis of professional wrong doing for prosecutors and correctional officers is largely missing from the literature. The varied theories used to explain employee misconduct do not align with the institutional nature of the problem of racial/ethnic disparities. That is, because racial/ethnic disparities are an institutional problem, a single theory is appropriate to explore this systemic trend. Additionally, differences in theoretical approaches in studying misconduct emphasizes the need for a single framework that can be applied across the system.

2.2 Filling the Gap This dissertation links racial disparities and the behavior of criminal justice professionals, filling a gap in the literature by offering a single theoretical framework – the institutional response model of social control - to explore institutional phenomena like racial disparities.

Specifically, I argue that the link required to explore racial disparities and misconduct are the organizations themselves. That is, because institutions (i.e. police departments, prosecutorial offices, and correctional facilities) play such a critical role in determining the procedures, environment and policies that govern employees, understanding a systemic issue like racial disparities can only be understood by taking an organizational approach. As noted in the review of the literature, scholars have used a range of theoretical frameworks to explore misconduct. This study proposes and tests a single theoretical framework that connects individual behavior and racial disparities across the system by exploring how institutions respond to the actions of their employees.1 Furthermore, this dissertation contributes to the relatively sparse correctional misconduct scholarship by applying this theoretical model to corrections.

Ultimately, the goal of this dissertation is to develop and apply a theoretical model that explains how systemic racism perpetuates across the criminal justice system, considering criminal justice professionals as central to the organizational-level model. This theory, therefore, is distinct from those applied in prior scholarship as it prioritizes criminal justice actors and race in the conceptual framework. The next chapter introduces the theoretical model and emphasizes the need to research this issue in corrections. Specifically, it details two theoretical frameworks that underpin the

1 It is critical to note that the theory developed explores the role of race in determining institutional responses to employee behavior. This is independent from employment discrimination which considers how employers act on prejudices (race, gender, religion, etc.) in the treatment of employees. While the model considers how institutions may unequally treat employees, the focus is on how these responses are symbolic of broader institutional issues that may lead to systemic phenomena such as racial disparities. Ultimately, the model’s focus on institutional outcomes that impact both criminal justice actors and those processed through the system distinguishes it from employee discrimination.

conceptual relationships proposed in the new model: colorblind ideology and the theory of law and social control. Both theories acknowledge the role of individuals in shaping inequitable organizational-level outcomes. Afterwards, the chapter details the propositions, assumptions and key elements of the institutional response model, later underlining the importance of applying the model to correctional facilities. The chapter ends by discussing the specific department of correction in which this study takes place – a unified system in the Mid-Atlantic.

Chapter 3

THE INSTITUTIONAL RESPONSE MODEL OF SOCIAL CONTROL

“Fair, Firm and Consistent” – Training Instructor Being “fair, firm and consistent” was a frequently repeated motto in observed correctional training classes (see Chapter 4 for details on the nature and extent of the observation portion of data collection). This motto directly links to the Institutional Response Model of social control because if criminal justice agencies respond to employees in ways that are unfair or contribute to a culture of injustice, criminal justice outcomes will reflect these realities. In this chapter, I begin with a review of relevant theoretical concepts that inform the model. Next, the three core elements of the model are described in detailed. This is followed by an emphasis on the importance of applying the theory to correctional systems. Lastly, this chapter discusses the research setting in which a theoretical model will be applied, a department of correction in a Mid-Atlantic state, acknowledging the importance of context on findings and conclusions.

3.1 Theoretical Underpinnings The two major theoretical frameworks that inform the Institutional Response Model of social control are Eduardo Bonilla Silva’s colorblind ideology and Donald Black’s theory of law and social control. Both theories allow for organizational level analysis while acknowledging the important roles individual actors play on social outcomes. The allowance for macro level examination while considering micro level

factors are not found in all theories used to explore racial disparities. For example, differential behavior theories suggest that racial disparities in the criminal justice system are due to group differences in behavior (Blumstein, 1982; Hindelang, 1978; Walker et al., 2012). Several studies have shown that crimes committed by Black individuals are more serious and more frequent than their White counterparts (Mosher,

Miethe, Phillips, 2002; Sampson & Lauritsen, 1997). These studies take individual factors into account, failing to consider the macro-level context. Differential treatment, on the other hand, focuses on the embedded mistreatment of minorities by the criminal justice system. Some scholars claim that the system is inherently biased exploiting and oppressing people of color to perpetuate inequality (Alexander, 2012; Cole, 2000; Marable, Steinberg & Middlemass, 2007; Russell-Brown, 2009). This perspective emphasizes the role of the institutions rather than individual label factors. Race is a principal component of both theoretical ideologies that inform the institutional response model. Previously applied to theories on employee behavior in the justice system do not always make the racial identity of employees a key part of their analysis, though they may find race influential after empirical analysis. For instance, police behavior is often attributed to the situational characteristics of the event including suspects’ traits (age, demeanor), legal circumstances (seriousness of offense) and characteristics of the event (location, witness presence, time of day) (Riksheim & Chermak, 1993). In all, the Institutional Response Model is supported by theoretical works that prioritize the importance of individuals, institutions and racial/ethnic identity, contributing to current gaps in the racial disparity and criminal justice literatures.

3.1.1 Colorblind Ideology Colorblind ideology conceptualizes race relations in the United States as covert actions and inactions that reinforce and enhance the dominance of Whiteness in society while appearing to be racially neutral (Bonilla-Silva, 2013). In his work “Racism without Racists”, Eduardo Bonilla-Silva defines colorblind racism as the new dominate racial ideology in the United States which makes public displays of racism unacceptable but continues to oppress minorities thereby maintaining the racial status quo. The four central frames of colorblind ideology are abstract liberalism, naturalization, cultural racism and minimization of racism. Frames help us understand the way information is interpreted. The most critical frame to colorblind ideology is “abstract realism”, which re-frames racial issues in liberal terms creating an appearance of morality. For example, the idea that jobs should be awarded solely on an individual’s merit, ignoring evidence of systemic, contextual and historical racism. Although this concept may appear just, it actually perpetuates White dominance in society. Naturalization involves rationalizing racism to be a result of natural occurrences, dismissing issues related to race in seemingly organic ways. While research shows that racial residential segregation is rooted in institutional and political practices (Massey & Denton, 1993), naturalization interprets it as the tendency for people to choose to live in areas where people are more like them. Cultural racism uses arguments based on culture, or the perception of a culture, to explain racial issues. For example, this frame takes statements such as “Blacks commit more crime” to justify the overrepresentation of Black people in the criminal justice system. According to this frame, individuals are labelled as more likely to engage in criminality by belonging to a racial group that has accepted deviance as a part of its identity. Negative stereotypes of minority cultures often go unchallenged and

unquestioned by mainstream society. The final frame of colorblind ideology is minimization of racism which suggests that although racism had a major impact on the status of certain populations in the past, modern society has evolved to a point where it is no longer a crucial factor in shaping life experience. This frame allows Whites to acknowledge the significant role racism played within the U.S. while simultaneously accusing individuals who tie present-day issues to race of playing the “race-card”. Each of the four ways of interpreting issues related to race allow individuals to subtly justify and reproduce White dominance without risking the negative stigma associated with being an overt racist. Scholars examine how colorblind ideology is enacted in society through individuals and institutions to perpetuate the power of White racial hierarchy in several ways, including scientific methodology and logic in research (Zuberi & Bonilla-Silva, 2013). White logic is the context that allows social facts to be interpreted in a way that grants objectivity to White perspectives and subjectivity to non-White views. Furthermore, White methods support racial stratification through the preference of tools that produce empirical data, holding White individuals above others (Zuberi & Bonilla-Silva, 2013). Armenta’s 2017 piece uses a colorblind framework to argue that law enforcement agencies racialize the routine activities of

Latinos which leads to the institutional production of immigrant criminality (Armenta, 2017). Scholars also call the combination of two paradigmatic shifts, assessing risk in penology and colorblindness in racial ideology, as the perfect storm for the criminal justice system to serve as a tool of racist practices (Van Cleve & Mayes, 2015). In other words, contemporary penal rationale focuses on punishment and management, classifying dangerous and potentially dangerous offenders. Simultaneously, the

contemporary racial tone of colorblind ideology makes it inappropriate and taboo to overtly link penal objectives to race and ethnicity. The combination of these factors creates an environment where the perpetuation of racial injustice in the system is subtle yet powerful and therefore must be examined critically (Alexander, 2012; Vane Cleve & Mayes, 2015).

In all, the combination of White logic and White methods produces knowledge that is neither objective, neutral nor independent of inherent biases. Instead, disciplines are founded on concepts and methodologies riddled with implicit bias. It is the appearance of neutrality, or the invisibility of these biases, that further solidify colorblind ideology as the racial frame of modern-day U.S. society. The colorblind framework has been applied to the criminal justice system leading to what some scholars consider ideal conditions for the perpetuation of the oppression of racial/ethnic minorities through instruments of justice. My dissertation develops a framework informed by colorblind ideology, understanding that institutions respond to employee behaviors in ways that perpetuate injustice but appears race-neutral. Additionally, the four frames of colorblind ideology inform the qualitative methodology that this dissertation research employs. Acknowledging that racially charged events, exchanges and policies may function in covert ways, qualitative analysis of field notes allow patterns of interactions to emerge. While a single event or interaction may be dismissed in quantitative analysis as an outlier or singular occurrence, a pattern of interactions speaks to a deeper significance such as the covert operation of race in a neutral-appearing institution. Therefore, the colorblind framework allows for a richer understanding of the function of race in institutional

responses to employee behavior, justifying the use of qualitative methodology in this study.

3.1.2 Black’s Theory of Law & Social Control Social control is a widely studied concept (Ross, 1901; Park & Burgess, 1921; Cohen, 1966; Black, 1976; Black, 1984). Since its inception, two distinct definitions of social control have developed. First, social control refers to the vast array of practices that influence people to conform to society’s norms. Second, social control is more narrowly defined as the way people respond to deviance. Donald Black contends that social control can be viewed as both a social norm and a response to deviance (Black, 1976). Black’s theory of law & social control states that the level of law and social control applied to individuals is patterned in a way that is predictable and shaped by status and other characteristics of society (Black, 1984). When outlining how social control can be researched as a dependent variable, Black (1976) emphasizes that society needs to evaluate not only the deviant behavior and its response, but also seek to understand how social control can be predicted and explained as a social phenomenon. He contends that social control can take different forms, styles and be applied in various quantities. By form, Black refers to the mechanism through which a person, group or institution chooses to express their reaction. These mechanisms include a variety of informal and formal actions such as gossip, facial expressions, scolding, compensation and incarceration as determined by a court. Style includes the approach and language each response takes. While Black recognizes that there may be more styles, he identifies four: penal, compensatory, therapeutic and conciliatory. Each style defines and handles deviance differently. For example, penal considers the individual who commits the deviant action as an offender

deserving punishment. Therapeutic, in contrast, views the same individual as a victim in need of rehabilitation. Different still, the conciliatory style focuses on how the offender and the victim can reconcile. Lastly, compensatory focuses on the damages caused by deviance, seeking to make the victim whole. To research social control as a dependent variable, Black states that each of these aspects can be captured quantitatively for analysis. Here, it is important to realize that the form and style of social control does not necessarily have a relationship to how much it is applied (Black, 1984). Instead, Black posits that the level of social control applied is positively related to the level of diversity and status of cultures, groups and individuals. In other words, Black’s theory emphasizes that social control is applied according to multiple characteristics. In this way, the theory of law and social control informs using quantitative methodology in this dissertation. The presence of institutional responses to employee behavior (or whether social control is applied) is analyzed separately from the number of responses an employee receives (or the quantity of social control applied), in alignment with Black’s theory. (See chapter 4 for more details.) It is important to note that Black’s focus on individual social status also relates to intersectionality theory, put forth by Kimberle Crenshaw which considers how social identities such as race, class, gender, sexual orientation and religion, overlap and influence related systems of oppression, domination and discrimination in society (Crenshaw, 1989). In other words, the experiences of individuals in society, including the way social control is applied, can be better understood by considering the multidimensional nature of their social status and relevant oppressive structures.

Intersectionality theory allows for both qualitative and quantitative methodology.

Black ultimately calls for the development of models of social control. These models should facilitate a better understanding of the existence and patterns of social control in different contexts. Alpert and Dunham (2004), for instance, developed authority maintenance theory, providing a specific model of social control to explain how conflicts arise in police-citizen encounters. Not all scholars have developed a model of social control, but instead rely on Black’s central argument to explore responses to behavior in a variety of contexts including the increased scrutiny of scientific misconduct (Hackett, 1994); socially disorganized neighborhoods (Warner, 2007; Cantillon, 2003); and mass incarceration as a form of repression (Oliver, 2008). The model I propose is another response to Black’s call for models of social control to explain how social control is applied to employees in institutional contexts.

The theory acknowledges race as critical in the function of the criminal justice system based on previous literature and the theoretical support of colorblind ideology and the theory of law and social control. The importance of race is critical to the model because its covert yet impactful nature within organizations has often been studied in terms of empirical disparities rather than theoretical explanations of pervasive institutional differences. Colorblind ideology and the theory of law and social control allow for the development of a theory that makes race a priority and allows for a mixed-method study of organizational phenomena.

3.2 The Institutional Response Model of Social Control The Institutional Response Model of social control, hereafter also referred to as IRM, posits that institutional responses to employee behavior are based on three key elements: (1) visibility of the action, (2) institutional context, and (3) status of the criminal justice actor. The framework has three major assumptions based on prior

literature and its theoretical underpinnings. The first assumption is that institutional responses to employee behavior are predictable and measurable as expected by Black’s claim that social control is quantifiable and should be explored as a social phenomenon. Secondly, it is assumed that institutional actors are punished and commended neither arbitrarily nor equally across individuals and actions. That is, there is an inequitable pattern to the extent to which institutions discipline and praise actors, as Black’s theory of law and social control also predicts. The theory goes further to state that this pattern of inequity is predictable based on three variables: visibility of the action, institutional context, and status of the actor. Because employee behavior can be either positive or negative, this model considers institutional responses to be both commendations and sanctions. Lastly, the third assumption of the

IRM is that race is a key factor in how institutions respond to employee behavior. As Black and Bonilla Silva state, social control is measurable and applied unevenly across the population, but institutions strive to appear race-neutral resulting in covert interactions that perpetuate race-based differences within organizations. The Institutional Response Model is outlined in the conceptual pathway below. It is important to note that there may be interactions between these three concepts.

Institutional Response = Visibility of Action + Institutional Context + Status of Actor

The following sections discuss each element of the Institutional Response Model of social control.

3.2.1 Visibility of the Action The first element of the Institutional Response framework “visibility of the action” consists of two parts: (1) how aware others are to the behavior committed and

(2) the severity of that behavior. For example, police officers interact with citizens in

public and correctional officers interact with individuals behind bars. The former may be at a greater risk for institutional response due to differences in visibility, specifically how aware society is to the action taking place. On the other hand, the murder of an inmate in contrast to an unlawful arrest may result in greater punishment for the correctional officer. In this case the severity of the action increases visibility and may impact the likelihood and nature of sanction. 2Although disciplinary sanctions may not always occur, the institutional response model posits that because the actions are more visible there is an increase in the formal response to the incident. Formal responses may include suspension with or without pay, termination and never involve criminal charges. Increased visibility may lead to more accountability for criminal justice actors.

In Discipline and Punish, Foucault begins by describing a profound change in penal justice “the disappearance of torture as a public spectacle (1977, p. 7)”. Foucault writes that when executioners successfully killed criminals they were praised by the watching crowd. But those who failed at their duty and caused the convicted pain and agony were liable to punishment, sometimes from mob violence. “Punishment, then, will tend to become the most hidden part of the penal process (1977, p. 9)”; as it became less visible, society became less aware of the actions of those charged with

2 It is important to note that this theoretical model does not claim that “justice” will be served in any instance where an action is visible, nor that justice is more likely in cases where visibility increased. For instance, cell-phone surveillance of police- citizens interactions does not guarantee sanction for an officer. Several police- involved homicides of Black individuals have not been convicted for their behavior despite the event being recorded, these cases include the deaths of Philando Castile (2016), Alton Sterling (2016), and Eric Garner (July 17, 2014) (Lee & Park, 2017; Yan, 2017).

punishing offenders. Visibility of the action captures the significant shift from public spectacle to secrecy that occurred in penal justice through awareness and severity. Also, central to the visibility element are organizational resources; an institution’s ability to supervise employee behavior is directly linked to the formal response taken, if any. Organizational resources that increase visibility of behavior range from technological equipment, such as video surveillance cameras and automated time managements systems, to the number of staff assigned to certain areas. Without the presence of others serving as witnesses to behaviors, employee actions can go unchecked. Surveillance cameras serve as an example of visibility via technology that requires institutional resources. Video surveillance is a growing way the criminal justice system surveils employees and interactions with the public (Ready & Young, 2015). However, the implementation of cameras is not uniform across the system. The Center for Evidence- Based Crime Policy at George Mason University reviewed literature on body-worn cameras, hereafter BWCs, in policing and the use of BWC evidence in court proceedings. The study found that nearly 93% of the prosecutor offices that have BWC evidence available to them have used it to prosecute private citizens while only

8.3% of offices have used this same evidence to prosecute police officers (Lum,

Koper, Merola, Scherer, & Reioux, 2015). Therefore, although BWC has the potential to assist in courtroom proceedings, the evidence is more likely to be used to prosecute private citizens than police officers. Citizen perceptions of the potential effectiveness of body worn cameras are mixed. A Maryland based study demonstrated that everyday citizens are either supporters or skeptics, the latter believing body worn cameras will not change power dynamics or the structure of police departments (Ray, Marsh &

Powelson, 2017). Similarly, Lum and colleagues showed that prosecutors were skeptics as well, disbelieving BWC evidence would improve the legitimacy or trust between police and citizens. Police officers in contrast are generally supportive of BWCs as they can potentially benefit both citizen and officer behavior (Jennings, Fridell & Lynch, 2014).

Further exploring the relationship between police misconduct and BWCs, Ariel and colleagues conducted a randomized, controlled experiment and showed that the likelihood of force being used doubles without the presence of BWCs (Ariel, Farrar, & Sutherland, 2015). Reedy and Young (2015) demonstrated that BWCs decrease the likelihood of officers conducting stop-and-frisks and arrests yet increases the likelihood of citations and initial encounters. These studies suggest that the use of

BWCs can increase certain aspects of police-citizen encounters. Visibility of the action has also been tied to positive institutional responses through literature examining the impact of organizational response on employee behavior (Bradler, Dur, & Non, 2016; Elwood & Naquin, 2004; Franklin & Pagan, 2006; Larsen, 1993). Scholars examining organizations find that institutions praise and discipline employees to increase productivity overall (Bradler et al., 2016; Franklin &

Pagan 2006). In their 2016 study, Bradler and colleagues examined the impact of praise on employee performance. The researchers praised workers who were most productive completing a data entry task. Results found that when one employee is publicly praised for their work, increasing visibility, productivity increases overall. The increase stemming mainly from employees who were not recognized (Bradler et al., 2016). Praise is a cost-effective way to manage, motivate, and reward employees.

Furthermore, it is most effective when the remaining employees, or those who do not receive rewards, are not penalized (Larsen, 1993). Relating to the theories that underpin the Institutional Response Model, the frames of colorblind ideology explicitly describe how racism functions in covert ways. For instance, institutions often consider employee behavior a reflection of individual choices rather than systematic processes. Employees who succeed on the job are often thought to have done so because of their individual accomplishments rather than other factors. This is tied to the colorblind frame abstract liberalism. However, the idea of merit and individual deservingness being the main determinants of whether an employee is successful at work does not align with the literature, particularly within corrections. For instance, Britton finds that prisons are gendered organizations that favor traditionally masculine traits (Britton, 1997) making men more visible to institutional support, such as the good ol’ boy network, than their female counterparts. Therefore, to counter the prevailing colorblind narrative that individuals are solely responsible for their behavior on-the-job, organizational trends should be examined qualitatively to uncover covert practices. One can better understand the complexities of the way social control is applied by making behavior and responses visible, in hopes of enacting change at the macro-level. The remaining two elements of the IRM refer to the occupational and demographic characteristics of the employee, respectively.

3.2.2 Institutional Context Institutional context, the second element of the Institutional Response Model (IRM) of social control, is an individual-level category for factors that employees accrue as members of an organization. This element has intentionally been developed

as a distinct category from the final portion of the IRM (status of the actor) to further stress the role of the organization on individual-level outcomes. Therefore, institutional context variables consider only those characteristics employees acquire within the context of their current organization. These organizational factors include tenure, rank, assigned duties, location, workload, etc. Donald Black’s theory of law and social control posits that law is applied more when individuals have less rank and are less respected (Black, 1976). That is, an entry-level employee is at a higher-risk to be disciplined than those in higher standing or who are more experienced. Justification for considering institutional context variables extend beyond theory into empirical studies. Organizational commitment literature has investigated the relationship of tenure or experience on-the-job and institutional outcomes with mixed results. Human capital economists consider tenure to be symbolic of knowledge in a specialized area (Elwood & Naquin, 2004). Employees with more specialized knowledge are more valuable and thereby protected by the institution. Because expert knowledge may be acquired by experience, expressed by excellent work, and later rewarded by promotion – tenure and rank are suitable proxies for institutional context at the individual level.

In contrast, surveys of Australian police officers showed that those with more experience were less committed to the organization than their counterparts as they more likely to identify structural problems within the institution (Beck & Wilson, 1997). To improve organizational commitment scholarship suggests employees receive more support and encouragement from supervisors and better management practices overall, both of which require institutional-level change (Beck & Wilson,

1997; Dick 2011). Likewise, a meta-analysis on the relationships between

organizational tenure and job behavior, including counterproductive behaviors, shows that tenure is positively related to some adverse behaviors including aggression and absences unrelated to illness (Ng & Feldman, 2010). Further complicating results, scholars demonstrated a curvilinear relationship between tenure and performance with new and advanced employees producing at lower levels than their mid-career counterparts (Ng & Feldman, 2010). Kaheny and colleagues showed that judges’ decisions are more predictable in the beginning and ending phases of their career; additionally, the size and dissent rate of the circuit each judge was assigned influenced decision-making (Kaheny, Brodie Haire, & Benesh 2008). The findings from this study highlight why variables such as tenure, assigned location, and workload should be considered when exploring work-related behavior.

In sum, research shows the importance of considering institutional context variables such as tenure, rank, and assigned location when examining employee behavior within organizations. Therefore, it is reasonable for the IRM to predict that institutional context factors will shape the way organizations respond to employee behaviors. This also aligns with theoretical expectations and the criminal justice literature more broadly which suggests that individuals with less institutional standing

(i.e. those with less experience) are subject to more social control than their counterparts (Black, 1984; Kutateladze, Andiloro, Johnson, & Spohn, 2014; Walker et al., 2012; Wu, 1997). Furthermore, it’s been shown that institutional context variables, such as assigned duties in policing (Martin 1980; Hunt 1984), are not race-neutral but perpetuate racial/ethnic inequality. In this way, institutional context is important to consider when exploring the racial disparities across the criminal justice system.

3.2.3 Status of the Actor The final element of the IRM is status of the actor, representing variables that individuals acquire outside of their institutional ties. These include demographic traits such as race, sex, educational level and age, all of which are not subject to change based on organizational factors. Although all aspects of social status are important to consider, this research focuses on the racial/ethnic identity of criminal justice actors because of the perpetuation of systemic racial disparities across the system. Racial/ethnic inequalities exist not only for those processed by the system but have also been found to impact criminal justice actors across the system. There are racial/ethnic disparities in the profession of law. The American Bar Association reported that in 2010, 88% of lawyers identified as White. And the profession perpetuates Whiteness with law schools enrolling only 28.5% minority students during the academic year 2013-2014 (ABA, 2015). With racial/ethnic enrollment this low, the likelihood of the population of lawyers being representative of the racial/ethnic diversity of society is slim. Like the legal profession, in 2006 minority correctional staff members comprised 29% of total employees (DiMarino, 2009). This racial/ethnic makeup is grossly disproportionate to the number of minorities who are incarcerated. Although both occupations are mostly White, the demographic makeup of lawyers is considerably Whiter than corrections; lawyers also have a higher social standing (are more valued) in society than correctional workers. Additionally, the education requirements for each profession may relate to the way society views them and the extent to which structural mechanisms will grant them protection. For instance, the required level of education for a prosecutor in the United States is post-secondary, requiring law school. For correctional officers, on the other hand, many agencies accept a high school diploma or its equivalent.

Both underpinning theories of the IRM inform this element. Colorblind ideology acknowledges how organizations – including criminal justice institutions, appear race-neutral but perpetuate racial/ethnic inequality through covert means (Armenta, 2017; Van Cleve & Mayes, 2015). Black’s theory of law and social control goes further stating social control is unequally applied in diverse settings; specifically, minority populations receiving more social control than those in the majority. By acknowledging the social status of criminal justice actors as key in the determination of how institutions that appear race-neutral respond to their behavior, IRM builds on these two conceptual models. Supported by the theoretical frameworks colorblind ideology, law and social control, the Institutional Response Model seeks to explain how social control is applied to criminal justice actors in response to their actions.3 In other words, this model considers the level of social control imposed on a criminal justice employees dependent on the visibility of actions, institutional context and the status of the actor. This dependence is sheathed under a guise of colorblindness and the inequitable distribution of social control. Each of these elements are informed by empirical and theoretical literature, forming a single-theoretical model that can be applied across the justice system. IRM is distinct from other models because it prioritizes the role of criminal justice actors and their racial/ethnic identity in predicting institutional responses to their behaviors.

3 I recognize that criminal justice actors are not always prosecuted/punished for their actions. Instead, the model is purposed to explore how institutions respond in any form, rather than the nature/outcome of that response.

Considering all the empirical and theoretical literature discussed thus far, and towards testing the IRM model focusing on the following series of research questions and hypotheses:

3.2.4 Research Questions

1. Is there empirical support for the Institutional Response Model of Social Control?

a. Does the visibility of an employee’s actions influence institutional responses to employee behavior?

b. Does institutional context influence institutional responses to employee behavior?

c. Does the status of an employee influence institutional responses to employee behavior? Specifically, does racial/ethnic identity matter?

2. How is the Institutional Response Model of Social Control enacted in correctional employment? That is, what are the processes that support the IRM?

a. How does the visibility of employee behavior relate to institutional responses?

b. How does institutional context influence institutional responses to employee behavior?

c. How does the status of an actor influence institutional responses to employee behavior? Specifically, what role does race play in shaping institutional responses, both positive and negative?

3.2.5 Hypotheses Visibility of Action

1. Institutions are more likely to respond to visible actions than less visible actions.

Institutional Context

2. Employees who have acquired more institutional context variables that relay investment in the institution will receive more positive institutional responses to their behavior than their counterparts.

Status of the Actor

3. Black employees, having a minority status, will receive more negative responses to their behavior than their White counterparts.

It is also expected that qualitative findings will reveal processes that align with hypotheses 1-3. “Power is at its peak when it is least visible, when it shapes preferences, arranges agendas and excludes serious challenges from discussion or even imagination (Minnow, 1987).” Because the way institutions respond to employee behavior across the criminal justice system is tied to visibility, I now turn to one of the most racially disparate yet invisible units within the criminal justice system: correctional facilities.

3.2.6 IRM in Correctional Facilities Testing the IRM in a correctional facility is essential because of the demonstrated lack of literature on employee behavior, the actor’s role in perpetuating racial disparities and institutional responses to misconduct. Unlike police departments and courts where officers, attorneys, and judges have been identified as possible culprits in the perpetuation of racial inequity in the system, racial disparities in corrections do not generally consider how correctional employees contribute to the issue. For instance, a recent book entitled Appealing to Justice discusses in detail the grievance system of the California Department of Corrections and Rehabilitation but does not go into detail about possible racial/ethnic disparities (Calavita & Jenness,

2015). The authors justify this oversight by saying that the perceptions and actions of

both inmates and staff towards the grievance process were homogenous, speaking to the strength of the prison as an institution to shape behavior of diverse populations. However, correctional institutions are like other social institutions, filled with same racial/ethnic biases as schools (Hope, Skoog, & Jagers, 2015), police departments (Gelman, Fagan, Kiss, 2012) and courts (Alesina & La Ferra, 2014). In this way, scholars are contributing to the exclusion of the correctional experience from the narrative of criminal justice process and stalling work towards reversing systemic racial/ethnic biases by not examining it as rigorously as the other phases of the system. The exclusion of correctional officers in racial inequality analyses may seem justifiable because prison staff are not responsible for the demographic makeup of the inmate populations they supervise. However, it is important to acknowledge the behavior and racial makeup of correctional employees influences that experiences of inmates, the majority of which are returning to society (Bureau of Justice Statistics, 2018). Additionally, the number, demographic composition, orientation and working conditions of correctional staff correlate with inmate outcomes such as levels of violence, positive attitudes, and satisfaction about prison conditions (Steiner, 2009; Molleman & Leeuw, 2012; Molleman & van der Broek, 2014). Steiner shows that state unemployment, violent crime rates and the proportion of guards to inmates influence the level of inmate assaults in prison (2009). Another study demonstrated that inmates housed in units with staff who hold positive attitudes towards inmates are more likely to perceive their circumstances in a positive light (Molleman & Leeuw, 2012). Correctional staff who are treated well in turn treat inmates well, as measured by perceptions of treatment (Molleman & van der Broek, 2014). The demands of correctional work have also been proven in scholarship to create stressful conditions

that lead to increased rates of stress, job dissatisfaction and burnout (Schaufeli & Peeters, 2000; Garland, Lambert, Hogan, Kim, & Kelley, 2014; Steiner & Wooldredge, 2015) which may also impact staff behavior on-the-job. In sum, the behavior of correctional staff has consequences for themselves and the populations they serve. This work contributes to the understanding of correctional officer behavior as it relates to the perpetuation of racial/ethnic disparities within prisons by focusing on institutional responses to behavior through analyses of original data collected in a Mid-Atlantic department of correction.

3.3 Research Setting This study takes place in a Mid-Atlantic Department of Correction, hereafter MADOC. MADOC is a unified system, meaning that there are no jails to house offenders. Instead, there are a total of 4 adult correctional facilities, 4 community corrections buildings and 2 violation of probation centers that range in security level. Most of quantitative and qualitative data collection occurred in the MADOC administrative building which is nestled in the capital of the state. The two-story brown building sits between an often-crowded parking lot and large grassy area used by cadets for exercise and marching formations. In front of the building, the wind dances with two flags representing the state and United States of America. A handful of parking spots mark positions of distinction, among them warden and employee of the month. Markers of distinction continue as you enter the building from one of three entrances: human resources, correctional attire store, and administrative offices. Contrary to its simple external appearance, MADOC’s administrative building is full of a maze of offices and variety of people in assorted positions. According to the

MADOC’s 2016 annual report, the department employed 1,698 correctional officers

and 256 probation officers who served approximately 7,000 inmates and 17,000 probationers. Although supervising a mostly minority inmate population, nearly 59% of correctional officers are White, followed by 36% Black and less than 1% other races. The pattern in demographics for probation officers follow the same trends as correctional staff. To function at optimal capacity, MADOC needs to employ 1,796 correctional officers, over one-hundred more than were employed at the time of this study. Like many other department of corrections, employee retention and attrition are major concerns. MADOC is an appropriate research setting to test the institutional response model of social control for three main reasons: (1) the consistency of institutional responses to employee behavior; (2) the racial/ethnic makeup of staff and inmates; and

(3) the focus on recruiting and retaining correctional officers. Testing the IRM at MADOC is appropriate because the unified nature of the system allows for a large sample of employees who are subject to the same institutional response system and have been trained to behave in a way that a single organization deems appropriate. In other department of corrections where jurisdiction is divided, the employee population may be trained differently or be subject to differing institutional responses. Related, it was appropriate to collect data at the administrative building, rather than at multiple sites across the state because while facilities may have cultures for institutional response, commendations and sanctions distributed by the administrative building are expected to consistent. Secondly, MADOC is an appropriate place to test IRM because of the demographic makeup of staff and inmates; racial/ethnic disparities are present among both populations in the ways literature would expect. In terms of inmates, minority populations comprise most of the population. However, employees are

mostly White. Furthermore, race-related issues are an important yet sensitive topic for MADOC, like many other institutions. Recent state-wide employee surveys related to diversity and inclusion revealed minority staff felt that race contributed to decreased opportunities for advancement. MADOC respondents reported these issues at higher rates than employees at other state institutions. Therefore, race plays a role in outcomes for MADOC employees, although these interactions may be covert. Finally, MADOC is appropriate to test the IRM because the focus of the criminal justice actor – in this case correctional employees – are at the forefront of priorities for the department due to heightened recruitment and retention issues because of a recent incident. In the winter of 2017, inmates took 5 staff hostages, killing one officer before the 15-hour standoff ended. Although staff attrition rates and officer misconduct were long-time systemic issues within the MADOC, the hostage situation exacerbated problems. Considering the recent incident and heightened tensions between staff, administrators and inmates, this research is timely and relevant to MADOC. Results can provide insight into the training, socialization and disciplinary procedures of correctional staff. Understanding these processes can lead to increased safety for inmates and employees by contributing to dialogue focused on improving MADOC practices. The immediate and most notable institutional response to this fatal incident largely involves an increase in correctional staff and a stronger emphasis on training. Beyond these immediate changes, MADOC also hired an external review team to assess the department and make recommendations towards improving efficiency, effectiveness and safety. The recommendations made by this external committee inform MADOC specific translational implications provided in Appendix B.

This dissertation seeks to address gaps in the literature regarding how criminal justice organizations respond to employee behavior, specifically within corrections. Additionally, this work makes a distinct theoretical contribution by placing criminal justice actors and race at the forefront of analysis and by developing and testing a single theoretical model. This chapter introduced the institutional response model of social control, its purpose, theoretical underpinnings and central elements. Research questions and hypotheses carved from theoretical and empirical expectations were presented. Due to the invisible nature of correctional institutions in contrast to other segments of the criminal justice system, the importance of applying the IRM to corrections, specifically MADOC was also addressed. Chapter 4 lays out the data and methodology for the application of the institutional response model within a Mid-

Atlantic Department of Correction.

Chapter 4

DATA AND METHODOLOGY

“We are not freaking out. We have a plan.” – Correctional Nurse As one nurse said in response to whispers that a contagious infection was causing chaos inside a facility, “We are not freaking out. We have a plan.” This chapter details the mixed methods research design used to test the institutional response model of social control in a department of correction.

4.1 Data Access

To gain access to the Mid-Atlantic Department of Correction, I underwent a series of security and project approvals with various individuals at multiple institutions. Before beginning data collection, the research design and protocol were developed with consultations from my dissertation chair and the head of research at the Mid-Atlantic Department of Correction. It was critical to get the perspective of the research liaison because MADOC data were critical to the success of the project. The original research proposal, presented to MADOC, placed concerns related to race/ethnicity at the forefront. This race dominant framing was discouraged because although interesting and important, it was perceived as ill-timed. As discussed in Chapter 3, a statewide survey distributed to all employees in the state roused concerns surrounding differential experiences based on race/ethnicity. Additionally, there were concerns over correctional employee safety, recruitment, retention and attrition.

Considering these heightened tensions, I was advised that making race such a

prominent feature of the research may increase the likelihood of being denied access. The research design was reframed to make the socialization, experiences and behavior of correctional employees the central part of the project, with race/ethnicity being a crucial factor that would be considered. This modified research design was approved by my dissertation committee and the Institutional Review Board at the University of

Delaware [1029815-1]. Upon receipt of IRB approval, a separate approval process began to gain access to MADOC. This involved several rounds of review by internal MADOC boards which included members of the Employee Development Center, Human Resources Department, Research Team, and the Commissioner. Additionally, I had to submit documents to pass a security background check.4 Ultimately, I received approval for data collection at the end of January 2017. However, the hostage incident, discussed in Chapter 2, prevented data collection from beginning until May 2017. Data collection spanned three months, reaching completion in August 2017.

4.2 Data & Sampling The theoretical model (IRM) proposed in this dissertation required original mixed method data collection. Quantitative data were collected from a random sample of all MADOC employees according to the employee roll as of July 15th, 2017; this master employee list was provided to the researcher via an excel spreadsheet by the human resources department. This spreadsheet was then transformed into an SpSS file,

4 Documents submitted for data collection in the field contain information that would reveal the location of the department of correction and, therefore, are not included in this dissertation.

where a random sample of 300 were selected using their unique employee ID numbers, or 17.6 of the MADOC employee population. Data on employee demographics, sanctions, commendations, and other work-related factors were compiled from individual human resource files and a Mid-Atlantic automated corrections system (MACS). MACS tracks offender information electronically and covers all aspects of the department’s operations. MACS data were collected at MADOC headquarters due to network restrictions protecting sensitive data. The sample size of three-hundred allowed for power in statistical analysis (Green, 1991). Additionally, because data were collected from employee files and the process was time-consuming, three-hundred employees was practical for the author, who worked alone. All identifiable information was kept on a secured USB file.

Quantitative analysis affords an overarching view of the aggregate trends of institutional responses to employee behavior. However, observations of training classes contribute a major insight to this work: exploring processes behind institutional response. This is critical in understanding how the everyday, real-world training of a correctional officer shapes their expectations about appropriate on-the-job behavior and disciplinary process. Training classes give the unique opportunity to understand how the expectations of behavior are introduced to staff. Through observational data, the empirical trends found can be understood in a more detailed way. I conducted over 150 hours of observations of Correctional Employment Initial Training over 3 months. During this time over 100 correctional cadets, officers, instructors and staff members were observed both inside and outside of prison in 5 classes of three different types. Three of the classes observed were Correctional

Employment Initial Training (CEIT) classes which are for new correctional

employees. The remaining classes observed were Refresher Training courses, required courses for current MADOC employees on a range of different topics. Typically, each observed class had around 20 individuals. The demographic makeup of observed classes was consistently mostly white and male. Those classes that were for new correctional employee training had between 1-5 individuals who were training to become correctional counselors. Most observations were of one cohort of correctional recruits. This class graduated with 18 cadets and had an instructional team of 4 correctional employees assigned. The Employee Development Center (EDC), structures classes to integrate a variety of instructors, specialists and others to assist with cadet training. The EDC was under-staffed during the observational period. This resulted in instructional staff being pulled in different directions to meet the needs of the department. This allowed me to observe a number of instructional staff, recruits and correctional officers, noting similarities and differences in style, approach, and interactions. I also observed, shared in, and overheard informal conversations which revealed candid reflections of the training process. The employee training manual and other instructional material were made available to me, providing a textbook perspective on MADOC training. Appendix B provides details on observations of

MADOC training, meetings, and perusing the official training documentation provided. My identity as a Ph.D. candidate observing training classes for dissertation research was made clear to staff and cadets primarily through introductions from instructional staff or myself. These introductions were not always conducted at the time of my initial interaction with training classes resulting in a shift in the way individuals interacted with me. These differences were recorded in field notes. In rare

occasions, I was approached by a MADOC employee who had heard of my research through word-of-mouth. The department allowed me to observe training classes with two main stipulations, maintain the confidentiality of all individuals observed and provide constructive feedback to the instructional team upon request. Anonymity was maintained through two sets of pseudonyms for correctional cadets, staff and instructors. The first set of pseudonyms were created and used by the author during the observation class to refer to individual cadets. Therefore, cadets are aware of their original pseudonyms. Given the relatively small size of the state and MADOC CEIT classes, I created a second set of pseudonyms that the cadets cannot identify. The second set of pseudonyms are used in this dissertation. When requested, I provided informal feedback to instructional staff regarding their approaches to teaching. My recommendations included calling on “quiet” cadets rather than letting a few voices dominate classroom discussion. Although my observations were mainly non-participatory; at times I shared my experiences as a correctional officer alongside instructional staff, to help make points clear. For example, when discussing potential inmate manipulation techniques, I shared a story on the importance of being fair when selecting inmates to serve as helpers with various tasks to avoid potential manipulation and accusations of favoritism.

Additionally, when the class participated in partner activities, I served as a stand-in cadet if attendance was uneven that day. When I shadowed the class in prison, I was given a cadet uniform to minimize my outsider status to inmates. Reactions to my presence, both in “civilian” and “officer” clothing were also noted. Lastly, informal interviews occurred with approximately 50 staff members, initiated by myself or a

MADOC employee. These conversations contextualized observations, clarified points of confusion to people, and increased attention on aspects of cadet training.

4.3 Methodology Individuals were given unique identifiers by the researcher, independent from MADOC information for analysis. STATA version 13 was used to conduct each phase of analysis. The next section details all the variables that are used to capture each part of the IRM.

4.3.1 Dependent Variables: Institutional Responses This study considers institutional responses to individual behavior representing social control over correctional staff. Four dependent variables are examined: presence of commendations in file (Y/N), the number of commendations in file, presence of disciplinary sanctions in file (Y/N) and the number of disciplinary sanctions in file. Commendations in an individual file were formal positive reports regarding behavior that MADOC found favorable. These behaviors include perfect attendance, assisting in a medical emergency, and general praise for excellent performance. Disciplinary sanctions present in the file detail institutional responses to behavior that were deemed unacceptable. These behaviors include unexcused absences/lateness, cyber security violations, failure to report DUI arrests and failing to perform security checks. A total of 800 institutional responses were collected from human resource files. The human resource department and some of the administrative staff were responsible for determining whether institutional responses were given and the nature of the response. Like the overall makeup of the MADOC employee population, the majority of these persons were White. Again, because these data are from the administrative level,

responses – particularly sanctions – that were given were done so according to MADOC policy. This was evident from documentation in employee files that cited the regulation that was violated. Noteworthy, regulation numbers were not listed in commendation documentation. As a researcher, I was not present when determinations for institutional responses were made. Most of the institutional responses were commendations (609), or positive responses; far fewer were sanctions (191). Although most employees had no sanctions in their files, 30% of Black employees did compared to 24% of Whites. This is in stark contrast to commendations, where most White employees had sanctions in their files, though most Black employees did not. The next section discusses the way each element of the institutional response model is measured empirically.

Table 1. Variable List

Variable Category Name of Variable Variable Measurement Dependent Variables Positive Commendations in File Yes=1 Positive Commendations in File Count Disciplinary Sanctions in File Yes=1 Disciplinary Sanctions in File Count Independent Variables Visibility of the Action Victim/Person Impacted Yes=1 Physical Injury/Benefit Yes=1 External Entity Involved Yes=1 Additive Severity Scale 0-3; 3=Most Visible Incidents Continuous (ln) Institutional Context

Multiple Prison Locations Yes=1 Correctional Position Front-Line=1 Tenure Years Status of Actor Sex Female=1 Race Black=1 Age Years Education High School/GED = 1

4.3.2 Visibility of the Action Visibility considers both the severity and awareness of a behavior. Severity of an offense requires the details of actions so that the impact of the behavior can be determined and/or compared to others. Details were noted in individual human resources files and included type of formal response (award, certificate, note of thanks, suspension, warning, etc.) and reasoning for response (faithful service, continued absences, neglect of duty, etc.) While details were collected on the nature of the formal institutional responses to behaviors, these contextual bits of information were not available in all files or for all incidents. The file included copies of letters to officers detailing justifications for the punishment/reward they were to receive in response to a given behavior. However, a thorough explanation was not always given. For example, a file may contain a letter notifying an officer that they will receive 3- day suspension without pay for conduct unbecoming of an officer, with no details on what the adverse behavior entailed. Another file containing a notification that an officer will receive the Warden’s Award for their service, may not explain the specific behavior that merited the award. For this reason, records regarding individual behavior - whether positive or negative - were generally inconsistent in terms of detail. Due to these inconsistencies, three binary variables (Yes=1; No=0) were created to capture characteristics that increased visibility, whether behavior (1) impacted another

individual, (2) caused or improved physical status, and (3) involved an external entity. These three binary variables were then merged to form an additive scale of visibility. This scale ranged from 0-3, where 0 means the actions of a correctional employee were least visible (e.g. excessive lateness to work or multiple perfect attendance certificates). On the other hand, a correctional officer who was praised by a police department (external entity) for medically assisting (physical benefit) civilians (impacted an individual) who were in an accident would receive a 3 on the additive scale. The alphas for the additive commendation and sanction scales are .78 and .71, respectively (Nunnaly, 1978). The presence of a formal response to behavior is representative of a level of awareness but is simultaneously symbolic of a series discretionary of decisions made by criminal justice professionals regarding the appropriate response. In other words, the presence of a formal response to behavior simultaneously implies a level of visibility. Without being aware an officer behaved in a certain way, there would be no formal response – positive or negative. Therefore, a type of visibility can be inferred by the univariate statistics of the dependent variables (see Table 2 in Chapter 5). Additionally, awareness is captured through the number of incidents employees are involved in. Incidents symbolize how much formal interaction an individual staff person has with the organization. Incidents can range from events related to missing/broken equipment to witnessing a physical altercation that happened on-the- job. Therefore, I utilize incidents to represent a general awareness the institution has to the overall behavior of employees. Only 251 employees had valid data for incidents because some were unable to be located in MADOCs electronic system. The total number of incidents an officer was involved ranged from 0 - 2,886, with an average of

204 incidents over a career. Three employees were considered outliers, having participated in over 2,100 incidents. This was an additional 1,000 incidents in comparison to other individuals in the sample. These outliers were removed. The incident variable was logged due to positive skewness (John & Draper, 1980), making the quantitative results easier to interpret.

Ultimately, visibility of an action is a complex concept to capture as inconsistency in data impacts the way the variable can measured. I use incident details and the number of incidents an employee is involved in to capture levels of visibility. I also rely on qualitative data to further illustrate how visibility impacts institutional response.

4.3.3 Institutional Context

Institutional context represents work-related traits that may influence employee status. These variables include rank, number of locations assigned, and tenure. The ranks of all 1,642 MADOC employees were included in the list provided by the Human Resources department from which the random sample was drawn. There were 14 different ranks possible. Only 12 of these ranks showed up in the sample for this study. Most employees were correctional officers, followed by corporal, correctional sergeants, and correctional captains. During analysis this variable was made dichotomous where individuals with the specific rank “Correctional Officer” = 1 and all others = 0. Those with different ranks such as “Correctional Staff Sergeant” and “Correctional Officer Storekeeper” have interactions and experiences with inmates that may be inherently distinct from a front-line correctional officer.

Staff experiences and behavior can also be influenced by the correctional facility they report to. Analysis considers whether an employee was assigned to multiple locations throughout their correctional tenure (yes=1; no=0). Tenure was calculated in years based on the date of hire as listed in each file. The length of time a correctional employee has worked for MADOC is also related to institutional context and many of the variables discussed above. For instance, an employee who has longer tenure with the MADOC has more opportunity to have institutional responses, be promoted beyond Correctional Officer and have participated in incidents. Therefore, the regression models control for tenure either as an independent or exposure variable, where appropriate (Hilbe, 2011).

4.3.4 Status of the Actor

The Institutional Response Model of social control posits that the identity of employees impacts the way institutions respond to their on-the-job behavior. The independent variables that capture these individual traits include: age, race/ethnicity, sex, and level of education. The variable age is measured in years, calculated by year of birth. Racial/ethnic categories were restricted to set options. They include: White, Hispanic, Black, Asian and American Indian. Due to inconsistencies in the way records are kept, an Unknown category was created to account for missing data. Race/ethnic identity was undefined for 8 employees in the sample. Because over 94 percent of the sample were identified as Black or White, the remaining racial categories were excluded from analysis. Note that the number of individuals in the other racial/ethnic categories (16) is too low to power a statistical analysis. This reduced the total sample used for analysis to 276 correctional employees. Sex is a dichotomous variable where 0=male and 1=female. Employee levels of education

were collected via resumes, copies of degrees and vocational certificates. Education levels fell into four categories: GED/high school diploma, some college, associate, bachelor and advanced degree. For clarity in interpretation, these categories were collapsed into a dichotomous variable where high school diploma or equivalent =1 and all else=0. These categories were chosen because the minimum education level for correctional employees at MADOC is high-school/GED equivalent. Quantitative data allows the exploration of the empirical relationship between institutional responses, employees, and their behavior. However, the processes of these relationship can only be explored through qualitative analysis.

4.3.5 Qualitative Because my role as a researcher was made apparent, observational notes were taken actively. Notes from observations and informal discussions with staff and cadets were written by hand in notebooks. In the classroom setting, note taking camouflaged with the natural environment. On occasion, when notebooks were unavailable (e.g. shadowing in prison), I recorded my observations and reactions from each day on a tape recorder. Each day of observation is considered a separate case. Cases were transcribed in Microsoft Word. All documents were transferred into the qualitative software program, NVivo 11. Coding was conducted in two phases: open and focused. In open coding, data are examined for patterns. Patterns that relate to the Institutional Response Model were then coded intensely. Examples of codes found in the data include “Staff Wisdom” – a reference to tips and general advice given by seasoned MADOC staff to cadets; “Stories from the Inside” – specific narratives shared by MADOC employees of their experiences working behind bars; and “Feedback for

Cadet” – describes instances of direct reactions to cadet behavior by training staff.

This process resulted in a total of 1,096 references or phases of interest from the field notes. Qualitative results are reported after quantitative findings to provide nuance, exceptions, and examples.

4.4 Reflexive Note Here, it is important to make a reflexive note regarding my connections to the

Mid-Atlantic Department of Correction and the field of corrections. It is critical for researchers to take a reflexive stance that acknowledge and challenge their positions, interests and beliefs that may influence their research to draw objective conclusions (Charmaz, 2014; Jootun, McGhee & Campus, 2009). Reflexivity in research should not only consider the subjectivity of the researcher and their biography but also their knowledge that may influence access to data, participants and information relevant to the project thereby impacting findings and conclusions (Roberts & Sanders, 2005). Although reflexivity frequently takes place in qualitative research, some quantitative scholarship is also reflexive (Ryan & Golden, 2006). After quantitative study on depression, Ryan and Golden argue that taking a “reflexive approach would not undermine the value of the research study but would add a depth of understanding about how, where, when and by whom data were collected (2006, p. 1198).” They found this to be particularly true when the study focused on sensitive topics that provoked emotional responses from the researchers (Ryan & Golden, 2006). Furthermore, research also shows reflexivity to be critical in mixed-methods studies as a part of efficiently reviewing work and positively contributing to the analysis process and researcher development (Walker et al. 2012). I established a strong research relationship with MADOC, beginning with my

M.A. thesis data collection in 2014. Since that time, I have conducted multiple original

data collection projects in facilities across the state, primarily as the Principal Investigator. My commitment to excellence in corrections research is grounded in my experiences as both a scholar and former corrections officer. This wealth of knowledge affords me unique insights into the profession that bring benefits and challenges to my work. My professional experience as a correctional officer socialized me to interact with other corrections practitioners as someone who had not only studied the profession, but who had experienced it. For example, conversations and questions I ask were sometimes based on comparisons or similarities from my prior employment. Simultaneously, this professional insight could also introduce challenges to the work as my former experiences could taint my perceptions of data collected during the dissertation.

Additionally, I am a young woman of Color, with a small physique researching a field that produces disparate outcomes for women, minority groups, and those perceived as physically inferior. Again, my demographic and physical state could introduce both advantages and disadvantages to my work. As discussed above, correctional institutions are predominantly White and Masculine, often giving professional credibility to individuals who are perceived to be physically capable of handling conflicts with resistant inmates (Britton, 2003). My body is the opposite of all those characteristics. During the data collection process, some correctional staff may be hesitant to sharing experiences and opinions with me because of my outsider appearance. Ultimately, my dissertation work is reflexive to counteract these varied and intersecting layers of identity. Additionally, I rely on multiple data sources and existing literature to ground my findings.

This chapter discussed the mixed-method research design of this project. Finally, I address my personal and professional identity, acknowledging the impact researchers have on the research process. Quantitative results of this study are presented in chapter 5.

Chapter 5

QUANTITATIVE RESULTS

“Heads in the books. Cut the talk. Knock the work out.” -Training Instructor

This work tests the Institutional Response Model of Social Control in MADOC using mixed methods. This study prioritizes quantitative analysis using qualitative methods to enrich empirical findings. The quantitative and qualitative strands of data collection and analysis were conducted independent of each other, integrated only to draw conclusions. Quantitative data explores aggregate trends of institutional responses to employee behavior, illustrating overarching patterns. Qualitative data, on the other hand, provides nuance to aggregate trends via complimentary narratives that aid in understanding how institutional response processes function within organizations. Because institutional responses to employee behavior can be either positive or negative, a series of separate analyses were conducted to understand predictors of commendations and sanctions. This chapter details the results of quantitative analysis.

5.1 Univariate Statistics Three hundred randomly selected correctional employees comprise the quantitative sample. Because race is a principal component of the IRM, employees who were neither Black nor White were excluded from analysis because of their small number (less than 6% of the sample). Missing data in the human resource files further

decreased the sample by 2%. After these exclusionary conditions, a total sample of 276 correctional employees were analyzed. Descriptive statistics are described in Table 2. On average, employees had more positive commendations (2.21) in their files than disciplinary sanctions (.69). Correctional employees had at most 44 commendations in their file. The maximum number of sanctions one employee had in their file was 13. This difference may be because employees who acquire multiple commendations are likely remain on the job. On the other hand, correctional staff who are disciplined by the institution may not be employed long enough to amass large numbers of sanctions in their files. On average, correctional staff were involved in 224 incidents throughout their employment at MADOC. As defined earlier, incidents reflect a level of general formal institutional engagement and awareness, resulting in an increase in visibility of an action. An incident can include an officer’s involvement in various behaviors from missing equipment to witnessing verbal altercations between two individuals. Most correctional employees in the sample behaved in ways that are often viewed not as severe, such as impacting other individuals, involving physical injuries and/or benefits, nor including external entities, such as police departments and attorney officers.

Front-line correctional officers, who engage with inmates on a regular basis, comprised 54% of the sample. The remaining employees held the titles of corporal (15%), correctional sergeant (11%), lieutenant (7%), maintenance officer (4%), correctional captain (3%), food service officer (3%), or worked with canines, in trade or as a correctional staff lieutenant or storekeeper. These diverse positions can potentially offer differing outcomes in terms of what shapes institutional responses to

employee behavior. However, due to low sub sample sizes for each rank outside of front-line correctional officer, this type of examination was not possible for this study. Rank and position of correctional employees may relate to how visible their behavior is as well. Future work is encouraged to consider how differences in rank influence institutional responses to behavior, particularly how institutional responses may change with an employee’s rank. Correctional staff spent 11 years as a MADOC employee on average; the mode for the tenure variable was 17 years. Over their time with MADOC, 42% of employees had been assigned to multiple facilities throughout the system, meaning over half of correctional staff remain at their initial assigned location for their entire career. Twenty-two percent of the sample is female, aligning with the known masculine nature of correctional employment. The average age of the sample is 42 years old, ranging from 21 to 65. Forty-one percent of correctional staff were educated at the level of high-school or GED equivalent, the MADOC minimum for employees. Nine percent held bachelor’s degrees at the time of data collection, while only 1% had advanced degrees (Master, Doctoral, etc.) Correlation matrixes were conducted to assess the potential for multicollinearity in the models and identify high correlation between variables. These matrixes can be found in Tables 5 and 6. All variables were well below .7, meaning that multicollinearity is not an issue, except some of the severity measures. Additionally, to double check for issues with collinearity, variance inflation factors (VIF) were calculated for all variables. All were found to be well-under the standard of 4 (Craney & Surles, 2002; Marquardt 1970) except severity variables. Because the measures of severity were sometimes coded based on the same incident, these findings

make sense. To counter the high correlation of the severity items, the measures are included in separate negative binomial regressions.5 Having established the general descriptive characteristics of the sample, the next section discusses regression results, highlighting the elements of the institutional response model throughout. Because employee behaviors provoke commendations and sanctions, they are analyzed separately yet presented together for comparison.

5.2 Quantitative Results

5.2.1 Logistic Regression Logistic regression allows for the analysis of binary dependent variables, such as whether an employee has commendations or sanctions in their file. Odds ratio are presented for simplification of interpretation. Logistic regression diagnostics were also executed. Classification tables examine how well the models predict the outcome variable. In the case of positive commendations and disciplinary sanctions, the model correctly predicted 73.58% and 73.17% of the cases, respectively.

Table 2. Univariate Statistics

Variable Name Mean Std. Min Max N Dev. Dependent Variables

5 Items that capture severity are inappropriate to include in the logistic regression models because they perfectly predict the binary dependent variables which measure presence of institutional response.

Positive Commendations in File (Count) 2.21 4.81 0 44 276

Disciplinary Actions in File (Count) 0.69 1.76 0 13 276

Visibility of Action Positive: Victim/Person Impacted .18 .38 0 1 276

Positive: Physical Injury/Benefit .13 .33 0 1 276 Positive: External Entity Involved .07 .25 0 1 276

Positive: Additive Severity Scale .37 .82 0 3 276 Disciplinary: Victim/Person Impacted .07 .26 0 1 276

Disciplinary: Physical Injury/Benefit .03 .18 0 1 276

Disciplinary: External Entity Involved .02 .15 0 1 276

Disciplinary: Additive Severity Scale .13 .48 0 3 276

Incidents (Continuous) 224 259 0 1207 251 Institutional Context Multiple Prison Locations (Yes=1) 0.42 .49 0 1 273

Correctional Position (Front-Line=1) 0.54 .50 0 1 276

Tenure* (Years) 11 8 0 38 276 Status of Actor Sex (Female=1) 0.22 .42 0 1 276 Race (Black=1) 0.39 .49 0 1 276 Age (Years) 42 12 21 65 276 Education (High School/GED=1) 0.41 .49 0 1 274

*Exposure Variable

Table 3. Correlation Matrix. Positive Commendations

Variable Response Person Physical External Severity Incidents Multiple Front- Tenure Sex Race Age Edu Name (Count) Impacted Inj/Benefit Entity Scale Locations Line CO Years (H.S.) Response 1.00 (Count) Person .40 1.00 Impacted Physical .31 .83 1.00 Inj/Benefit External .37 .42 .38 1.00 Entity Severity .42 .93 .90 .66 1.00 Scale 90 Incidents .06 .09 .05 .07 .08 1.00 Multiple .02 .03 .04 -.05 .01 .14 1.00 Locations Front-Line -.18 -.24 -.17 -.20 -.24 -.32 -.20 1.00 CO Tenure .43 .23 .20 .23 .26 .14 .15 -.33 1.00 Years Sex -.08 -.05 -.06 .01 -.04 -.07 -.09 .01 -.11 1.00 Race .05 -.00 -.05 -.00 -.02 .02 .09 .01 .00 .23 1.00 Age .31 .28 .20 .10 .24 -.00 .15 -.21 .64 -.08 .08 1.00 Education .02 .05 -.00 .07 .05 .03 .05 .02 .12 -.18 -.07 .15 1.00 (H.S.)

Table 4. Correlation Matrix. Disciplinary Sanctions

Variable Response Person Physical External Severity Incidents Multiple Front- Tenure Sex Race Age Edu Name (Count) Impacted Inj/Benefit Entity Scale Locations Line Years (H.S.) CO Response 1.00 (Count)

Person .50 1.00 Impacted Physical .33 .61 1.00 Inj/Benefit

External .34 .36 .45 1.00 Entity Severity .50 .88 .83 .68 1.00 Scale

91 Incidents .16 .21 .03 -.04 .11 1.00

Multiple .09 .07 .15 .13 .13 .14 1.00 Locations

Front-Line -.09 -.12 -.04 .04 -.06 -.32 -.20 1.00 CO Tenure .17 .18 .07 .03 .13 .14 .15 -.33 1.00 Years Sex -.02 .03 -.03 -.02 -.00 -.07 -.09 .01 -.11 1.00 Race .10 .05 .00 .03 -.04 .02 .09 .01 .00 .23 1.00 Age .17 .08 .05 .06 .08 -.00 .15 -.21 .64 -.08 .08 1.00 Education .19 .02 .06 .14 .07 .03 .05 .02 .12 -.18 -.07 .15 1.00 (H.S.)

Table 5. Logistic Regression Reporting Odds Ratios for Presence of Institutional Response (N=246)

Model 1 Model 2 Variable Positive Sanctions Visibility of Action Incidents (ln) .98(.10) 1.35(.16)** Institutional Context Multiple Locations Assigned 1.11 (.35) 2.51(.82)** Correctional Officer (Front-Line=1) .46** (.14) 1.18(.40) Tenure (Years) 1.10*** (.03) 1.08(.03)** Status of Actor Sex (Female=1) 1.18 (.45) 1.98(.81) Race (Black=1) .65 (.21) 1.02(.34) Age (Years) 1.05** (.02) 1.01(.02) Education (H.S./GED=1) .97 (.31) 1.78(.58) Prob>chi2 .0000 .0000 Pseudo R2 .2164 .1556 Log Likelihood -133.60882 -120.80806 ***p<.001; **p<.01; *p<.05

Table 5 reports the logistic regression models for institutional responses to employee behavior.6 In the first model, commendations in a correctional employee’s

6 Additional models were run because although data were collected from the administrative level of MADOC, correctional staff were placed in various institutions. Each of these institutions may have different sanction and commendation policies. Therefore, there was potential of institution-based effects. Dummy variables were entered in each model; however, the results were not statistically significant. Additionally, multi-level regression models were also conducted to determine if the current assigned facility of correctional staff impacted the way MADOC responded to behavior. These models were not statistically significant. This series of statistical models demonstrate that when determining the existence and extent of institutional responses, institutions do not matter.

human resource file, only Institutional Context (rank and tenure) and Status of the Actor (age) variables are statistically significant. I find support for two of the three components of the Institutional Response Model. Rank and tenure impact positive commendations. The odds of correctional officers having a positive commendation in their file were 54% less than these with other titles. In other words, having increased interaction with inmates decreases the likelihood of receiving formal praise from MADOC. In this way, having a position that limits one’s interaction with inmates may promote the odds of receiving formal commendations. For every additional year an employee’s tenure increases, the odds of having a commendation in file increases by 10%. This makes sense intuitively as employees have more opportunity to receive praise as their tenure increases. Age was the only status of the actor variable that influenced whether an employee received institutional praise. As an employee’s age increases by one year, the odds of having a commendation in file increases by 5%. This relationship between age and praise may be because older individuals have been found to be more attached to their places of employment (Koch & Steers, 1978). Shifting to the presence of sanctions in employee files, results reveal that only visibility of action and institutional context variables from the IRM were influential.

As the number of incidents an employee was involved in increases, the odds of having a disciplinary sanction in their file increased by 35%. In other words, there is a positive relationship between general awareness (i.e. level of formal engagement) to employee behavior and the likelihood of being disciplined. In terms of institutional context variables, being assigned to multiple locations and tenure were statistically significant. Employees who worked at multiple facilities across their tenure with

MADOC had odds 151% greater of receiving disciplinary sanctions in their file than

their counterparts. An employee may be moved from one facility to another to counter disciplinary problems faced at their original institution. Or employees who are moved from one facility to the next may not have the opportunity to build strong ties to their colleagues which may lead to punishable behavior. Lastly, tenure has a statistically significant impact on the discipline of correctional employees. Results show that as an employee’s tenure increases by 1 year the odds of having a sanction in their file also increases by 8%. Like receiving commendations, this positive relationship makes sense because tenure gives more opportunity for institutions to respond to behavior. Overall, institutional context is relevant to both commendations and sanctions but in different ways. Furthermore, visibility of actions has a statistically significant impact on sanctions while status of the actor variables is statistically related to the odds of an employee having a commendation in their file. Tenure is consistently statistically significant across types of institutional response. However, being assigned to multiple prison locations does not impact the odds of receiving commendations, only sanctions. Additionally, working as a front-line correctional officer puts employees at a disadvantage in terms of being positively acknowledged for their behavior, but is not statistically significant for sanctions. This may be because front- line officers execute general duties required to maintain order in prison, such as counting inmates and monitoring inmate behavior 24/7. These tasks may be viewed as mundane or just a part of the job, giving front-line officers less opportunities to be selected for special commendations. Lastly, status of the actor does not influence being sanctioned. This may mean that disciplinary determinations are not influenced by demographics such as race, sex, age and education, an ideal outcome for organizations. However, age is correlated with receiving positive commendations.

The next section of results examines the number of commendations and sanctions in employee’s files. Because the dependent variables are continuous, a different type of regression model is appropriate.

5.2.2 Negative Binomial Regression Negative binomial regression is appropriate when the dependent variable is a count and over-dispersed, meaning the mean is lower than the variance (Osgood, 2000; MacDonald & Lattimore, 2010). This was true for both the number of positive commendations and disciplinary sanctions in an employee’s file. The likelihood-ratio test of alpha for all models were statistically significant, further signifying that negative binomial regression is the appropriate statistical model, rather than Poisson regression (Berk & MacDonald, 2008). Additionally, the negative binomial regression models all use tenure as an exposure variable. It is appropriate to expose the model for tenure because the longer an employee works at MADOC the more they are at risk for having an institutional response in their file, positive or negative (Hilbe, 2011). Presented in Table 6 are a series of negative binomial regressions for commendations and sanctions using different measures of severity. Recall earlier, the binary measures of severity are highly correlated because behaviors that result in an institutional response could be considered severe in multiple ways. For instance, performing CPR on an unconscious inmate is considered “doubly” severe, impacting another person and involving physical benefit to that person. The models are numbered from 1-8 and will be referred to by these numbers throughout this discussion. Models 1 and 5 are the main models because they are considered the most inclusive, including the additive severity scale. Incidence-rate-ratios (IRR) are presented in the results for easier interpretation of the influence of each variable on predicting the number of

institutional responses in an employee’s file. An IRR smaller than 1 signifies negative effects and those greater than 1 indicate a positive effect. Model 1 shows that as an employee’s behavior increases in severity the rate of commendations in file increases by 72%. Therefore, as visibility of employee behavior increases, so does the quantity of praise reports in their file. This positive correlation is consistent in all four commendation models (1-4). Likewise, for every additional year an employee ages, the rate of commendations in their file increases by 2%. The positive relationship between age and number of commendations is consistent except for model 3, where only the measure of visibility is statistically significant. In short, age is relevant in determining the number of sanctions in the employee file. However, when severity is measured solely by whether a person is impacted by employee behavior, age is no longer statistically significant. Models 5-8 display the results for negative binomial regressions examining the number of disciplinary sanctions in an employee’s file. In the main model for sanctions (Model 5) variables from two elements of the institutional response model are statistically significant, visibility of action and status of actor. As employee behavior becomes more severe the rate of disciplinary sanctions in file increases by

166%. This relationship is consistent across all four sanction models, demonstrating how critical severity is in determining the number of sanctions in an employee’s file. Importantly, status of the actor variables (race and education) were found to be statistically significant across sanction models. Model 5 finds that Black correctional employees face sanction rates that are 74% higher than White employees. Likewise, employees with high school diplomas or GEDs face sanction rates 140% higher than

MADOC staff who have higher levels of education. In other words, employees who

are Black or have high school education levels are at a greater risk of having more disciplinary actions in their file that their counterparts.7 These results show that the quantity of disciplinary sanctions within MADOC are not purely based levels of visibility nor institutional context. Instead, status of an actor variables put employees at risk for a greater number of sanctions.

It is also noteworthy that across commendation and sanction models, institutional context variables are not statistically significant. That is, being assigned to multiple locations and being a front-line correctional officer does not impact the number of commendations nor sanctions found in employee files. Tenure was used as an exposure variable, as the logistic regression models provide justification that employees are at a greater risk for receiving institutional responses the longer they work for MADOC.

5.2.3 Quantitative Summary In sum, results from quantitative analyses find varying levels of support for all three of the elements of the institutional response model. Logistic regression models demonstrate that the presence of commendations in an employee’s file is related to institutional context and status of the actor variables

(rank, tenure, and age). The next set of analyses reveal age (status of actor) and severity of behavior (visibility of action) were influential in the number of

7 Because in the negative binomial sanction models race and education were statistically significant, a series of negative binomial regressions were conducted that included an interaction term between these two variables. The interaction term was not statistically significant and are not presented. Based on these models, I conclude that race has an independent, main effect on institutional responses to employee behavior.

commendations reports employees had. All three elements of the institutional response model are supported but differ across type of response (commendations and sanctions). The most consistent predictor of commendations was age of employee, impacting the presence and quantity of commendations in an employee’s file. Turning to disciplinary sanctions, again all three elements of IRM were supported. Models examining the presence of sanctions in files linked visibility of actions and institutional context variables (number of incidents, multiple locations assigned and tenure) to the binary outcome. However, only severity (visibility), race and level of education (status of the actor) were found to be statistically significant in determining the number of sanctions in files. As visibility of employee behavior increase, so does the likelihood of them being sanctioned and being disciplined multiple times. Exploring disciplinary sanction results also found the race and education level of employees put them at risk for more disciplinary sanctions. That is, rather than disciplinary sanctions solely being influenced by visibility of behavior or institutional context, the status of the actor influenced the quantity of sanctions employees received. The next chapter discusses qualitative findings.

Table 6. Negative Binomial Regression reporting IRR coefficients (standard errors) for Institutional Response (N=239)

Variable Positive Positive Positive Positive Sanctions Sanctions Sanctions Sanctions Name Scale1 Person2 Physical3 External4 Scale5 Person6 Physical7 External8 Visibility of Action Severity Measure 1.72(.17)*** 3.42(.72)*** 2.82(.69)*** 2.98(.94)*** 2.66(.71)*** 4.92(2.08)*** 4.65(3.23)* 3.97(2.95)* Incidents .95(.05) .98(.05) .94(.05) .93(.05) 1.01(.10) .10(.10) 1.07(.11) 1.10(.11) Institutional Context Multiple Locations .85(.16) .83(.16) .80(.16) .90(.18) .97(.30) 1.14(.36) 1.09(.36) 1.11(.37) CO (Front-Line=1) .98(.20) .99(.20) .81(.16) .85(.17) 1.58(.49) 1.61(.50) 1.42(.45) 1.36(.44) Status of Actor

99 Sex (Female=1) .80(.20) .83(.21) .78(.20) .80(.21) 1.44(.53) 1.48(.55) 1.39(.54) 1.32(.51) Race (Black=1) 1.04(.21) 1.01(.20) 1.12(.24) 1.01(.21) 1.74(.51)* 1.80(.53)* 1.96(.59)* 1.91(.57)* Age (Years) 1.02(.01)* 1.01(.01) 1.02(.01)* 1.02(.01)* 1.01(.01) 1.01(.01) 1.01(.01) 1.01(.01) Education (H.S.=1) .87(.17) .85(.16) .98(.19) .87(.17) 2.40(.71)** 2.71(.81)** 2.75(.84)*** 2.53(.79)** Prob>chi2 .0000 .0000 .0001 .0010 .0000 .0000 .0005 .0010 Pseudo R2 .0539 .0536 .0398 .0329 .0785 .0764 .0576 .0537 Log Likelihood -377.03332 -377.14536 -382.6384 -385.39352 -224.5972 -225.10384 -229.68359 -230.63246 BIC 808.8313 809.0554 820.0414 825.5517 503.959 504.9723 514.1318 516.0296 ***p<.001; **p<.01; *p<.05 || Exposure: Tenure (Years)

Chapter 6

QUALITATIVE RESULTS

As discussed in Chapter 4, observations of training classes and informal interviews were conducted over a 4-month period, resulting in over 150 hours of observational field notes, memos and notes. These data allow for a more nuanced understanding of how racial disparities can perpetuate within criminal justice organizations through institutional response processes, extending beyond the aggregate trends that quantitative analysis provides. Data were coded in two phases, open and focused coding. Open coding led to the creation of nearly 50 themes that stood out as interesting patterns in the data. Focused coding concentrated on the elements of the Institutional Response Model of Social Control, particularly those that align with the quantitative findings. The empirical analyses found support for all three key elements of the institutional response model of social control – visibility of the action, institutional context, and status of the actor. However, support for these elements varied depending on the type of institutional response (positive commendation or disciplinary sanction) and the nature of the dependent variable (whether a response existed vs the number of responses in employee files). This chapter discusses qualitative analysis which show complimentary results to the quantitative results. Data issues in the race and criminal justice literature make it important to rely on multiple sources of data to strengthen research findings. The cadets and officers in training who were observed are mostly not the same individuals who comprise the

100

empirical sample. There were 4 observed training staff who were also included in the random quantitative data. Research protocol outlined by MADOC prevented these individuals from being formally identified or linked in data collection and analyses. This section will provide themes and narratives from fieldnote data that align with quantitative results and the three elements of the institutional model of social control. A summary of quantitative and qualitative results is presented at the end of this chapter.

6.1.1 Institutional Responses Institutional responses to cadet behavior are decided during “Behavioral Assessment Committee” meetings, informally called BACs. In BACs, instructors, a training administrator and a member of training support staff gather to review the progress of cadets. This working group generally consisted of 4-5 members, 2 of which were permanent (training administrator and member of training support staff) while the other 2 instructors rotated by class. Similar to the larger MADOC population, the training personnel were mostly White, though the training administrator at the time the research was conducted was a Black male with decades of experience. The diversity of the BAC personnel, in terms of race/ethnicity, age, experience, etc., may influence how cadets were sanctioned and/or praise, although training policies were considered a major guide for these decisions. Future research should consider how the diversity of these decision-making committees may influence institutional response outcomes. My access to ongoing BAC sessions was restricted in accordance with personnel concerns expressed in the memorandum of understanding. I was, however, able to debrief with BAC participants after the sessions were completed. During these debriefing talks, I learned that although the BAC determined

101

responses to both positive and negative behaviors, praise was far rarer than sanction. Commendations being a rarer event than sanctions are the reverse of what was illustrated in the quantitative data, where on average employees had over 2 commendations in their files versus .69 disciplinary sanctions. This difference may be because of the nature of correctional training. As the first step in a correctional career, it may be that discipline is a critical part in setting standards for appropriate conduct. That is, the nature of correctional training lends itself to correcting behavior of cadets in order to ensure policies are followed. Correction during training is critical for training in other professions including law enforcement which often takes a stress- based approach (virtual reality) than non-stressed based approach (i.e. academic) to prepare cadets for the demands on-the-job (Hormann, 1995; Reaves, 2009).

This imbalance of responses (more commendations and sanctions) also became apparent in the coding results. Figure 1 depicts the word cloud query for commendations during the observations of training classes. This word cloud is far smaller than that of Figure 2 which portrays disciplinary sanctions. Word clouds are images that depict the frequency of words in each text, where the size of a word relays its frequency or importance. The larger the word, the more frequent/important it is.

Twenty-five sources of observational data from different days were referenced 1,093 times, resulting in 48 different nodes. Positive institutional responses referenced the observational data 14 times. Negative institutional responses, on the other hand, referenced observational data 51 times. Codes for commendations and sanctions only comprised approximately 6% of the total references coded. Observing more sanctions than commendations during training also provides more opportunity to observe the role of race in shaping institutional responses to employee behavior. I draw this

102

conclusion based on the quantitative findings that show race as a consistent statistically significant factor that increases the odds of an employee having multiple sanctions against them.

Figure 1. Word Cloud. Positive Commendations.

103

Noteworthy words that result from the word cloud query on positive institutional responses (see Figure 1) relate to instances of group support and encouragement such as “applauded”, “class”, and “yelled”. This type of group encouragement can also be viewed as a form of informal institutional positive response, where members of the MADOC community support each other to achieve common goals. Strikingly, “negative” came up as an important term in the commendation word query. Again, this confirms the theory that discipline is the central type of institutional response during training.

104

Figure 2 is a result of a data query for words connected to institutional responses to negative behavior, or disciplinary sanction. As expected from the quantitative analysis, “black” was one of the most frequently used words in relation to disciplinary actions observed by the class. Note the increased frequency of the word “black” in observation notes does not tell the story that Black cadets are more likely to have more sanctions in their file. Instead, it tells a complimentary story of specific instances where Black correctional employees were involved in ways that are distinct from their White counterparts. To illustrate the quantitative trends, the following sections give detailed narratives of institutional responses to cadet behavior organized by the elements of the institutional response model of social control. Pseudonyms are used to protect the identities of cadets. While these narratives describe MADOC cadets and seasoned staff members, the focus of this research is on institutional responses to employee behavior, not on the employees themselves.

105

Figure 2. Word Cloud. Disciplinary Sanctions

6.1.2 Visibility of the Action: Seeing Silence The Institutional Response Model posits that visibility influences institutional responses to behavior. Specifically, it claims that the institution must be aware of the behavior to respond and that the degree of severity influences the type of response. Qualitative analysis allows behavior and processes that are invisible to quantitative analysis to emerge as critical. Throughout qualitative analysis behavioral occurrences that were unseen and unheard by those in authority were consistently unaddressed. These invisible occurrences may have gone unnoticed during observations if not for instances that made these events visible. This section discusses how language,

106

building structure, and victim impact can shape institutional responses to behavior by increasing the visibility of an action. Language emerged as a consistent way to control employee behavior and predict institutional responses. Specifically, language that demanded the silence of inmates, cadets, and correctional employees flooded prison walls and were repeated to cadets on a frequent basis. This theme of silence led to 62 references from 21 out of 25 sources of field note data. Individuals were silenced within MADOC 84% of the days of observation. Prison walls were stamped with red and black, block letters that said “NO DISCOURSE” demanding silence from inmates as they moved throughout the prison. In training, cadets were given words of wisdom that silence in prison is a warning sign of pending problems. Training staff advised, “Prisons are loud. If they aren’t loud, something’s up.” Beyond expectations of silence for inmates and tips on the job, silence also came up during training classes. Cadets were often hushed by training staff who used the phrase “lock it up” to demand silence and attention from their class. Ordering incoming correctional staff to be silent in a profession where locks are a critical part of the job stood out among other commands from training staff. On many occasions the absence of authority figures encouraged cadets to talk amongst themselves. At times they would warn each other of training staff’s return to not be caught chatting. Silencing cadets was often used as a form of social control, chitchat being framed as a privilege for the class when performing well. One trainer said “we are just going to go back to no talking” after a day the cadets were unusually rowdy. In this way, the ability to exercise voice in certain situations within MADOC was a form of visibility that can be tied to institutional privilege and control.

107

Language continued to be tied to institutional response and visibility by the use of profanity among correctional staff, particularly cadets. Training administrators discouraged cursing and held standards high, linking the use of profanity to a lack of professionalism on-the-job. Roger was a White male who demonstrated himself to be both a leader and a professional during training. Like many other cadets and correctional staff, Roger occasionally used profanity. One day, Roger’s use of vulgarity was overheard by a member of 5th avenue – the nickname given to members of MADOC’s executive board. Roger received a sanction for his behavior, was given an informal talk and formally apologized to the entire class for his conduct. Although vulgarity took place multiple times, it took a MADOC executive to overhear a curse to provoke an institutional response. This illustrates the influence visibility of behaviors has on formal sanctions. Visibility is also found to be influential in determining institutional responses to behavior during training sessions via the structure of the gymnasium. The MADOC gym serves multiple purposes in the administration building. It houses training for various skills such as marching, defensive tactics and cell-extraction techniques. Graduation ceremony also takes place here. The gym has a grey floor with white streaks. Cushioned chairs line the back wall, interrupted by a water fountain and faux cell – home to two mannequins who are used as inmates during real-world scenarios. Banners honoring correctional employees who served in Operation Iraqi freedom decorate one wall. There are four exits in the gym, leading outside (rarely used) or more offices and classrooms. For morning physical training, cadets run and exercise in the gym on blue mats, with the center of the gym serving as the focal point. Often, correctional

108

employees and cadets arriving to work walk through the gym to other parts of the building during physical training. These individuals, although encouraged to use an alternate route, sometimes stopped and talked to instructors and cadets that were exercising. The presence of these individuals, using the gym as a hallway during training sessions, increased the visibility of cadet behaviors. This increase in visibility often corresponded to more engagement and feedback from a number of training staff members. During defensive tactics, for instance, correctional staff often observed class progress in the gym. When a female cadet fell during her physical training, a passerby paused the training class and instructed cadets to lay down additional blue mats to prevent injury. Additionally, when drill sergeants were integrated in correctional training, over twenty employees came to watch their interactions with cadets. A cadet that had been singled out for failing to hold a plank position by drill sergeants stormed out of the session, pushing through onlookers. Members of the training team followed her and encouraged her to rejoin her peers. Her resistance and withdrawal from training in a highly visible setting provoked an institutional response. After the yelling and intense exchanges were over, one person said, “show’s over” and the group dispersed. The openness of the gymnasium allowed for aspects of training to be visible to witnesses, increasing the opportunity for institutional response.

Layla’s differences from her training cohort were many; these distinctions ultimately led to a situation where the invisible became visible and caused an institutional response. She identified as Lesbian, held a M.A. degree, did not wear a cadet uniform, and was expected to meet a separate set of requirements as a counselor- in-training. These differences made her stand out although she tried her best to meet the standards for a front-line correctional officer. Sonya, a Black female cadet training

109

to be a correctional officer told Layla that she was “not one of us.” Then, turning to their peers, stated “I’m sorry. But she’s not.” Despite all of Layla’s efforts to be one of “us” (physical training, defensive tactics courses, etc.) Sonya directly isolates her from the rest of the training class by verbally emphasizing her differences in front of the entire class. Layla cried. Although there had been many other instances where cadets had mentioned Layla’s differences due to the structure, policies and expectations of the training course, there were no institutional responses until Sonya’s remarks. The female instructional staff pulled all the females in the class to the side and verbally reprimanded them. Following this females-only sanction, a male instructor verbally chastised the entire class for their possible compliance in Layla’s isolation. He emphasized unity and its importance in the profession. In all, Layla’s tears made her continued isolation – both formal and informal – visible in a way that provoked an institutional response.

6.1.3 Institutional Context: “us versus them” The dichotomy “us versus them” represents a mindset where individuals with a similarity make themselves distinct from those lacking that commonality. In other words, groups of people are divided based on a particular characteristic. In the qualitative analysis of the observation notes, “us versus them” quickly emerged as a common theme in the data. Correctional staff were divided from other types of law enforcement officers (police) and citizens. There were also “us versus them” divisions, within the correctional workforce. For example, correctional employees who were not a part of the MADOC administration versus the higher-ups or “5th avenue.” The differences were clear. Members of 5th avenue wore business professional attire to work and had offices in a specific wing of the MADOC administration building.

110

Members of 5th avenue are also culturally and structurally different from those who do not hold administrative position, such as human resources staff, training instructors and emergency response units. Additionally, there are separate unions for those with a lower rank versus those who were ranked Lieutenants or higher. These unions make separate arguments for employee raises. During the observational period, the union representing the lower ranking officers won a raise. Although all MADOC staff were hopeful that this raise in salary would assist in problems with staff retention, those who had not received a raise hoped that they would be next in line. In terms of pay, benefits and other union-related issues, there are distinctions between ranks of MADOC employees, an “us versus them” dichotomy. The varied “us versus them” distinctions directly tie to quantitative findings that illustrated how front-line correctional officers are less likely to have commendations in their file than employees in other positions. That is, the structural and cultural divisions within MADOC have implications for how positive institutional responses are distributed to correctional staff. Separations between correctional staff also extend to the actual structure of the MADOC administration building where most of the observations took place. The building is structured like a maze, full of twists and turns. Although there is some signage in the building, they are relatively small, particularly in comparison to the inmate art that decorates the walls. Visitors and newcomers can quickly get lost without a guide providing step-by-step instructions. Different departments are housed in various areas around the building. It was not uncommon for MADOC employees themselves to need help navigating from place to place. Rank and building structure are two examples of the existence of internal divisions among MADOC staff.

111

Differences among cadets also relate to institutional responses to behavior. Layla’s story demonstrates this point. Layla was one of ten counselors in training across the five classes I observed. Although counselors were in training they only had to participate in a portion of correctional officer requirements. Counselors did not have to participate in physical training, get pepper sprayed, pass the gun-range nor did they have to wear the standard cadet attire. In other words, the training program structured counselors to be different, aka “them”. Layla, however, made herself even more distinct from her counselor counterparts by actively trying to integrate herself into the standards of correctional officer training. Layla wore a white shirt, blue shorts and participated in physical training every morning. She took part in defensive tactics, baton in hand, alongside the correctional cadets. Layla also openly differentiated herself from others verbally, openly identifying as a White Lesbian woman with a master’s degree. Approximately, 1% of correctional employees in the quantitative sample had an advanced degree. Therefore, it is likely that among her classmates, Layla had the highest level of education. Throughout training Layla held ambitious standards for herself, going beyond what was required of her as a counselor-in-training. When peers commended her for being the first to complete the assigned workbook for the training class she replied, “If I can’t do this with a MA degree, I need to step it up.” She also took positions of leadership among her peers both formally and informally. Classes were structured to rotate leaders and assistants daily. Layla was included in this rotation. However, even during unstructured times where order was needed (lots of talking, etc.) Layla was among the trainees who attempted to maintain order by making verbal commands in

112

an authoritative way. These commands included phrases such as “Come on, guys. We’ve been talked to about this before”, “Cut the chit-chat” and “We should be reading in silence”. Structurally Layla was “them”, an outsider who was held to different standards than her counterparts. The skills that she possessed were often praised and admired in training. Her tenacity and work-ethic can be linked to the lack of observed disciplinary actions taken against her, which aligns with the quantitative findings of individuals with education levels beyond a high-school diploma. Those employees who had a high-school education level had higher odds of having multiple disciplinary sanctions in their file. Because Layla had the highest level of education in her class, the lack of disciplinary sanctions against her ties to these results. In sum, distinctions between correctional staff via culture, structure, rank, or education level situates individuals differently within the corrections organization. Status of the actor variables are discussed with a qualitative lens below.

6.1.4 Status of the Actor: Before MADOC Status of the actor emphasizes the relationship between characteristics acquired before joining an organization and institutional responses to behavior. This section highlights three narratives that demonstrate how race and age are linked to institutional responses two status of the actor variables that were also statistically significant in the quantitative analyses. I gifted “Jazz” his pseudonym because he was a soft-spoken, heavy-set White male who seemed to always be joined by another soft-spoken White male. The two of them were a duo, I called “Smooth Jazz.” Like every other cadet, Jazz showed up each morning in a white shirt and blue shorts and participated in physical training. Jazz wanted to be a correctional officer. The requirements to be a correctional officer

113

include passing a physical training test. And each morning, he joined the others and ran around the gym between sets of exercises. Well, he didn’t really run, he shuffled. He shuffled around and around. And when he got tired, he walked. Among the other cadets lifting weights, running, panting, sweating, even singing along to pop billboard hits that played in the background, Jazz stood out in a negative way. He quietly worked his way through the training circuit, often being one of the last to complete it. Jazz garnered attention when instructors realized he was cutting corners to finish faster. One day, I arrived at training and noticed that Jazz was missing. I inquired and discovered that he was no longer employed with MADOC. The details being reduced to the succinct statement: “The job is not for everyone.” And in what seemed like an instant, Jazz vanished from training and from the class narrative.

Alanna was an expressive, heavy-set, Black female with a strong voice. And as is required for every cadet, Alanna showed up early each morning in a white shirt and blue shorts to participate in physical training. Alanna wanted to be a correctional officer, which required that she pass a physical training test. So, each morning, she joined others and ran around the gym between sets of exercises. Well, she didn’t really run – just like Jazz – she shuffled. She shuffled around and around. And when she was too fatigued to run, she walked. Among the other cadets who were doing planks, jogging, joking with passersby, even stopping to take a swig of water from the fountain, Alanna stood out in a negative way. She slowly worked her way through the training circuit, usually being one of the last to complete it. One morning after physical training and showers, all cadets returned to the classroom. The instructors and Alanna did not. And after nearly an hour, the instructors returned. Alanna did not. I asked the instructors what happened and discovered she was let go. Details about the

114

circumstances around her termination, unlike Jazz, became twisted and tangled in a whispered dialogue by both instructors and cadets. Stories implied that Alanna had assaulted an instructor, was arrested and hauled off MADOC grounds by the state police department. It was rumored that she only wanted to become a correctional officer to serve as a contact for someone on the inside, something that is common for a

“person like that”. In fact, cadets even spread gossip outside the department resulting in more disciplinary sanctions for the entire class. The narrative crystallized and dialogue about her termination continued beyond graduation. Although there are several differences in the nuances of these narratives, the institutional responses to both employees failing to be up-to-standard for the physical requirements resulted in termination. Key to this analysis is not only the institutional response (termination) but the differences in the narrative surrounding their departures. Jazz essentially disappeared from training. Instructors and cadets rarely mentioning him or speculating why he left. Alanna’s narrative became a part of a dialogue inside of MADOC. Importantly, Alanna’s narrative was racialized as gossip surrounding her termination included suspected connections to inmates and being someone who has ulterior motives for wanting to become a correctional officer. What type of person was Alanna accused of being? Although, race was not overtly mentioned in any informal discussions, connections to criminality and deceit were only a part of conversations about Alanna, a minority cadet. Another example of how status of the actor impacts institutional response is the training class award of “Outstanding Cadet.” The recipient of this award is decided by the instructional team. Roger was one of the oldest cadets in the training class who worked as a bail bondsman prior to beginning a career as a correctional officer. Roger

115

is a physically fit, White male who approached his training seriously. Observation notes of his training included a mix of institutional responses including informal disciplinary sanctions (for using vulgarity) as well as commendations (for excellent work during an intense situation while shadowing in prison). The instructional team quickly realized that Roger was a leader among his peers, as he consistently encouraged others to meet standards in class. Interactions between White male instructors and Roger quickly became more collegial than instructor-cadet. Smooth, on the other hand, was a soft-spoken, tall, thin White male cadet who also approached his training seriously. Smooth’s father was a correctional officer. From observation, Smooth was Roger’s junior by two decades. Observation notes lacked institutional responses to Smooth’s behavior because his actions often kept him under the radar. The instructional team noted that Smooth’s academic scores were the highest in the class, but little else. As graduation approached, the instructional team had to decide between the two nominated cadets – Roger and Smooth – for the Outstanding Cadet award. Roger won. Roger’s skill set, a former bail bondsman who had previous experience being a leader in intense situations and interacting with offenders, had the advantage over

Smooth who had solely excelled in every aspect of training. That is, Roger’s additional skill set aligned more with the values praised by the MADOC. These skills go beyond what is offered in training. This was acknowledged by a MADOC administrator who told one class that training creates a foundation and they should build their “career upon that foundation. When you (cadets) leave learning starts, believe me.” And other MADOC staff who encouraged cadets to rely on their training

116

despite any deviation they see when they get inside the facility and to seek out positive mentors for guidance and support. Although training classes create a foundation of correctional employees to launch their careers, institutional responses to behavior often relate to abilities and characteristics attained prior to employment or status of the actor variables.

The following section will summarize the quanitative and qualitative findigs in relation to the specific hypotheses that were presented in chapter 3.

6.2 Results by Hypothesis Hypothesis 1. Institutions are more likely to respond to visible actions than less visible actions. The first element of the Institutional Response framework “visibility of the action” consists of two parts: (1) how aware others are to the behavior committed and (2) the severity of that behavior. Awareness is captured by the total number of incidents an employee was involved in which speaks to a general level of formal institutional interaction. This awareness variable only predicted whether an employee had sanctions in their files. The qualitative theme “silence” supports these findings. Previous literature supports the impact of awareness has on institutional responses through examples such as police use of force incidents (Ariel, Farrar, & Sutherland, 2015) and stop-and-frisks and arrests (Reedy & Young, 2015). Severity was captured with a series of binary variables that represented different aspects of behavior including impact on a person, physical injury or benefit and the involvement of an external entity. Severity of employee behavior consistently predicted a higher quantity of institutional response, positive and negative. Prior literature has found that the severity of a behavior does not consistently merit sanctions from criminal justice

117

organizations, as in the deaths of Black men by police officers (Lee & Park, 2017; Yan, 2017). However, some sort of institutional response (investigations, suspensions with/without pay) often occur (Lee & Park, 2017; Yan, 2017). These findings also relate to the theoretical underpinnings of the IRM. Specifically, Donald Black’s theory of law and social control supports a quantitative approach to studying the presence and level of social control as a phenomenon which varies depending on a range of circumstances including (Black, 1984). Although the visibility of action element was strongly supported across regression models, this support varied based on whether the outcome was the presence or quantity of social control. Colorblind ideology demonstrates that certain types of behavior (specifically racism) continue to flourish in society due to its invisible nature despite laws that make inequity based on race unacceptable (Bonilla-Silva, 2013). This occurs through processes that are viewed as natural, or through the naturalization frame (Bonilla- Silva, 2013) and therefore remain unchallenged. Demanding silence from inmates and cadets, responding differently to behaviors when there are witnesses or that impact others all seem to be natural occurrences. However, these seemingly natural patterns based on visibility are directly tied to the inequitable way organizations respond to employee behavior (age, race, and level of education). Layla, for instance, had the highest level of education in her class (master’s degree) and did not receive disciplinary sanctions. Because of these inequities the distribution of social control within MADOC must be examined, including how visibility may vary depending on status of the actor. In short, quantitative analyses found that visibility of the action mattered differently depending on the outcome variable. Specifically, being involved in

118

incidents generally put employees at greater risk for sanction. Severity of behavior consistently influenced the quantity of social control an employee received – positive or negative. That is, when employee behavior impacted another person, caused injury or benefit to someone physically and/or involved an external entity, the likelihood of having multiple institutional responses increased. Qualitative analysis provides confirmation and nuance to these aggregate trends. Through language, building structure and individual narratives awareness and severity of employee behavior shapes institutional response. The continuous silencing of inmates and cadets establishes visibility as a tool of control. On the contrary, cadets being noisy made training staff more aware of their actions, increasing visibility and serving as a pathway to sanction. Roger’s vulgarity narrative further demonstrated this point. Also increasing awareness, the presence of witnesses to physical training in the gymnasium impacted the level and type of response from training staff. Lastly, Layla’s isolation via structure, expectations, identity, attire and finally other cadets led to sanction. Although Layla’s isolation was embedded in the makeup of the training program, and had been referenced countless times by her peer, it was only when the impact of their actions became visible to staff that a formal institutional response occurred.

Institutions are more likely to respond to visible actions than those actions that are less visible, thereby supporting hypothesis 1. That is, visibility of an employee actions is positively correlated with institutional responses to behavior, as predicted by the IRM. This finding is further supported through qualitative findings, previous literature, and the theoretical underpinnings of the institutional response model. Future research should seek out data sources that can capture visibility in innovative ways.

119

Hypothesis 2. Employees who have acquired more institutional context variables that relay investment in the institution will receive more positive institutional responses to their behavior than their counterparts. The Institutional Response Model of social control defines institutional context as individual-level factors that are directly tied to one’s status as an employee (i.e. tenure, position, and assignment). Empirical results found that each of the three institutional context variables (tenure, being assigned multiple locations, and having the position of a front-line correctional officer) were significant for determining risk of having disciplinary sanctions and/or commendations. Qualitative themes found further showed how institutional context, in various forms, influenced responses to behavior. Hypothesis 2, those employees with institutional investment being more likely to receive positive institutional responses, is supported. Front-line correctional officers were less likely to have commendations in their file than staff members. Rank was not statistically significant, however, in the sanction analysis. For disciplinary sanctions, being assigned to multiple locations put employees at a greater risk of being disciplined. Tenure consistently predicted the presence of institutional response in file, justifying its use as an exposure variable in the negative binomial analyses. Shifting to exploring factors that determine the number of commendations or sanctions, institutional context variables were not statistically significant. Observations demonstrate how distinctions among correctional employees emphasize how institutional context variables shape MADOC’s responses to behavior. Captured in the qualitative theme “us versus them”, privileges given to those with higher rank manifested in informal talk (“5th avenue”) and formal policies

(separate unions). These within group divisions were also evident in building

120

structure. The offices of 5th avenue starkly contrast against the cubicles of others. Lastly, Layla one of the few counselor cadets, received less sanctions than her peers. Layla’s advanced education and Whiteness align with quantitative findings that show both to be protective factors against a higher amount of sanctions in employee files. The us-versus-them dichotomy is rampant in the criminal justice literature as it creates a “divide between the alleged ‘deserving’ and ‘underserving’ poor – a divide that all-too-often cuts along racial lines (Dillard & Nielsen, 2015, p. 770)”. The incarceration system is used as a tool in creating these divisions (Alexander, 2010). The divisions created by the perpetuation of this dichotomy intra-criminal justice actors is unhealthy for the success and improvement of conditions in the system. In this project, this dichotomy divided correctional employees themselves such as “5th avenue” versus line staff and the maze-like structure of the MADOC administrative building. Divisions among correctional employees facilitate/inhibit relationships which may impact responses to behavior (Britton, 2003; Grossi, Keil, & Vito, 2012). Institutional context variables such as tenure are reflections of an individual’s organizational investment. Levels of investment are shown to influence other types of correctional employee outcomes including turnover intent, inappropriate and illegal behavior (Griffin, Hogan, & Lambert, 2014; Harne, 2017). Griffin and colleagues asked over 2,600 correctional officers about their intent to leave their jobs at various stages of their career. Commitment to organization was found to be influential in determining intent for job turnover at every tenure category (Griffin et al., 2014). Another study examined factors that could detect law enforcement officers who were likely to be terminated, given a last chance, demoted, resigned, and suspended.

Analyses were conducted on data from human resources, internal affairs, training,

121

insurance claims and counseling. Correctional staff who completed training slowly, had incidents related to sick leave, were late, dishonest, careless or involved in unprovoked violence were at a greater risk for negative behaviors (Harne, 2017). That is, those with less institutional investment were at risk of negative institutional responses, in line with the results of this project.

Black’s theory of law and social control further shows that employees who have acquired more institutional context variables which relay an increased investment in the institution will receive more commendations than those who are less invested. Individuals with more rank, experience on-the-job, and knowledge about the field may have higher social standing and receive less punitive social control than their counterparts (Black, 1984). In other words, having a higher institutional status can serve as a factor that is both advantageous in terms of commendations and protective from sanctions. Related, the abstract liberalism frame of colorblind ideology posits that racism covertly exists in society through deservingness (Bonilla-Silva, 2013). This connects to Bonilla-Silva’s abstract liberalism frame and the belief that people get what they deserve. Individuals who are invested, follow regulations, work hard and accumulate positive institutional context variables merit rewards as is the case with some of the findings in this study.

Hypothesis 3. Black employees, having a minority status, will receive more negative responses to their behavior than their White counterparts. Supporting hypothesis 3, Black employees were found to be at a higher risk of having more sanctions in their file than White employees. Similarly, high-school educated employees were more likely to have multiple disciplinary sanctions in their files, compared to their more educated counterparts. In essence, race and class play a

122

role in how employee behavior is managed. The differential treatment of racial and ethnic minorities in society as well as unequal access to advanced levels of education also manifest within the correctional system via the distribution of punishment to employees. In contrast, race nor education surface as significant variables when determining commendations in an individual’s file. Instead, age was the only demographic trait found to be significant. Older employees being more likely to have commendations in their files. Race and class only matter in terms of punishment not praise. Black’s theory of law and social control anticipates these inequities, positing that those with lower social status, such as minorities and those with high-school levels of education, are at a higher risk of formal social control. This also aligns with studies on policing that find White officers in positions of authority with substantial amounts of discretion and Black officers being more likely to feel criticized on-the-job (Bolton 2003; Dowler, 2005). Furthermore, scholarship tells us that there may be an increased impact of racism and sexism among the treatment of law enforcement officers. Black female police officers face unique structural barriers in the male dominated profession (Martin, 1994). Although sex did not emerge as a significant variable in this analysis, future work should continue to consider its effect on institutional response outcomes. Race was also found to be influential in determining response outcomes in the qualitative analyses via the racialization of narratives. The narrative about Alanna’s termination persisted and referenced “people like her”, taking on qualities of Bonilla- Silva’s cultural racism frame. Cultural racism uses arguments based on culture to explain racial issues (Bonilla-Silva, 2013). According to this frame,

123

criminality/deviance is attached to individuals by belonging to a racial group (Black/African-American) that has accepted deviance as a part of its identity. Alanna’s termination narrative, like other cultural racism examples, went unchallenged and unquestioned because the cultural stereotype of Black deviance in the United States is strong.

Sonya’s isolating comment to Layla brought attention to race. Informal interactions after this incident tied Sonya’s behavior to her background and race. Black women are often stereotyped as hostile, aggressive, loud, and argumentative (Ashley, 2014; Thomas, Witherspoon, & Speight, 2004). This connects to research that finds Black police officers are perceived to be involved in misconduct and are disproportionality processed in disciplinary hearings than Whites, though not always formally sanctioned after investigation (Smith, Johnson, & Roberts, 2014). During adjudication of Black officers, systems tend to rely on formal methods, while informal measures are considered in adjudication processes of White officers (Smith et al., 2014). Empirically, though institutional context variables were influential in determining whether there was a formal institutional response, the quantity and nuanced nature of responses is driven by more personal traits such as race, sex, and education level. This is concerning because it suggests inequality in a social institution thought to be just. The realities of criminal justice officials impact the experiences of offenders (Schwirtz et al., 2016; Steiner, 2009; Molleman & Leeuw, 2012; Molleman & van der Broek, 2014). Because correctional employees are subject to more sanction based on their race it is possible that they continue the systemic differential treatment of minority inmates.

124

Chapter 7 summarizes the main goals of this dissertation and emphasizes the contributions to the literature. Limitations, reflexive concerns, future research directions and translational implications are also discussed.

125

Chapter 7

DISCUSSION AND CONCLUSION

“Silence sometimes hurts more” – Training Instructor The purpose of this dissertation is to explore the nature of racial disparities across the criminal justice system by developing and testing a theoretical model that prioritizes race and the role of the criminal justice actor. Specifically, the institutional response model of social control is applied to a Mid-Atlantic Department of Correction to understand how correctional organizations react to employee behavior. Using mixed methodology, analysis found support for the three elements of the model: visibility of the action, institutional context, and status of the actor. That is all three concepts impacted institutional responses to employee behavior, however, the impact differed depending on the type and level of response. This chapter presents the goals and findings of this work by referencing the institutional response model of social control and relevant previous literature to highlight this study’s major contributions to the race and criminal justice literature. Next, limitations and challenges of the research are addressed. Lastly, conclusions on how institutional responses to employee behavior perpetuate racial disparities in the criminal justice system are drawn. Future considerations for research and broad policy implications of the work are also discussed. Building on Donald Black’s theory of law and social control, this dissertation tests the institutional response model to understand the distribution of social control in the criminal justice system which can perpetuate racial disparities. IRM predicts the nature of social control among employees based on three key elements: visibility of

126

the action, institutional context, and status of the actor. I now return to the research questions posed earlier in this work. Is there empirical support for the Institutional Response Model of Social Control? Quantitative and qualitative analyses found support for all three variables depending on the nature and type of social control that served as the outcome variable. Tested in a Mid-Atlantic department of correction, employees who were in positions with less interactions with inmates, had longer tenure, were older and whose actions were more visible were more likely to have commendations in their files, than their counterparts. Employees were more likely to be sanctioned if they had longer tenure, were front-line correctional officers, behaved in severe ways, were Black and had high-school levels of education. Observational field notes revealed that informal interactions, structural divisions, and language were three main avenues through which elements of the IRM were enacted daily. For example, employees who have previous experiences favored by the institution are praised by training staff. Additionally, the racialization of narratives creates a culture that covertly signals racial meanings, though not overtly stated. These results match up with theoretical expectations of Bonilla-Silva’s colorblind ideology and Donald Black’s theory of law and social control. The institutional response model of social control is informed by these frameworks in both theory and methodology. Black’s theory encourages empirical measurement and prediction of the way social control is distributed in different social contexts (Black, 1984). The institutional response model predicts the distribution of social control among employees within different organizations. This dissertation in particular uses IRM to predict how racial disparities can be perpetuated through the system via the unequal distribution of social control among criminal justice employees. Recognizing

127

that racism often interacts in covert ways, colorblind ideology allows for the examination of daily interactions through a race lens (Bonilla-Silva, 2013). Through qualitative analysis, the colorblind frames add to the richness of analysis and conclusions drawn. In all, this dissertation makes a major contribution to the corrections, racial disparities and criminal justice literatures. By developing and applying a theoretical framework that prioritizes race in the understanding of institutional responses to employee behavior, this project works that the intersection of these three areas. The three elements of the institutional response model and the theoretical underpinnings of the framework allow for a mixed method analysis that focuses on race, even when race operates on a covert level. Additionally, it prioritizes criminal justice actors instead of those who are processed through the system, which is uncommon in previous theoretical approaches. This work focuses on correctional employees, criminal justice professionals who are invisible in the racial disparity scholarship in comparison to their counterparts in policing and courtrooms. Criminal justice actors are critical to understanding processes and outcomes within the system – particularly those tied to race/ethnicity as discretion affords them a level of power that is vulnerable to biases seen in broader society. In spite of these major contributions to the literature, this dissertation also has limitations that should be taken into consideration; these are discussed in the following section.

7.1 Limitations Chapter 1 details issues in the racial disparities and criminal justice literature: defining disparity versus discrimination, data access, variation in theoretical approaches (Reiter, 2014; Watson 2015). Addressing the first inconsistency, this study

128

considers disparities within MADOC as reflections of systemic differential treatment of minorities but does not make claims about discrimination due to the varied nature of the literature. Additionally, I develop and test a single theoretical model to understand racial disparities in the system as they relate to race, criminal justice professionals and organizational responses to employee behavior. Thereby, overcoming the third issue in the literature: variation in theoretical approaches. This study sought to overcome issues with data access by using multiple types of data to strengthen results. Although both quantitative and qualitative data were available for collection and analyses, these data were restricted and remain a limitation of this study. First, the quantitative data were solely at the administrative level. Within

MADOC, there are at least three levels of data that are important to consider when determining institutional responses: micro, meso and macro. Micro-level data consists of asking individual employees about their behaviors and experiences with praise and discipline in MADOC. Meso-level data includes gathering disciplinary and praise files that are kept at each correctional institution. Finally, macro-level data are formal responses at the administrative level. MADOC procedures allow sanctions and commendations to occur at each of these levels, but not necessarily all levels. For example, a supervisor pulling a correctional employee to the side to give a chat about a mistake made is an informal sanction but would not be reported at the meso- nor macro-level. Attendance, on the other hand is a measured behavior that is captured by the administration. Sanctions are mandatory and uniform across employees. This study sheds light on patterns of institutional responses at the macro level. It is likely that many instances of response are not captured in this study.

129

Qualitative data are limited in a comparable way. Observations were only allowed of training classes. Individuals in training classes may behave differently than correctional employee inside prison. Therefore, behavior observed may be different than what is reflected in the empirical analysis. This work focuses on the behavior of the institution, patterns can be drawn at the structural level. The connections between empirical and thematic patterns found in the analysis further support this conclusion. Building on Bonilla-Silva’s colorblind ideology and Black’s theory of law and social control, the quantitative and qualitative results support the institutional response model. However, the elements of IRM were captured in distinct ways because data were extracted from separate sources. As discussed above, the individuals in the empirical sample were correctional employees; those observed were cadets.

Additionally, I was prohibited from pulling human resource files for cadets. Variables of interest such as race, age, education-level, and prior skill-set, were limited to what the researcher could observe. Some variables, such as education-level and skill-set were only able to be collected if this information was openly shared in class. Therefore, I was unable to draw aggregate qualitative patterns for some of the variables and instead had to use case-studies to illustrate patterns found in the empirical analysis. This contrasts with the quantitative data which uses institutional categories to decide the racial/ethnic categories of individuals. Future studies should seek to work with correctional institutions to gather multiple types of data on a larger sample to make even stronger comparisons across methodologies. Next, the data available varied in detail which prevented a richer analysis. For instance, while some instances of disciplinary action included several responses, others only referenced adverse behavior in a single document. An individual being

130

chastised for neglecting their duties may have multiple documents that reflect a series of actions including written warnings, pending investigations, investigative reports, and appeal decisions. Other files may only have the results of the investigation, missing the earlier steps in the sanction process. Similar patterns were found for commendations in files, some individuals having documents at each phase and others only at one phase. The inconsistency in data files made it difficult to empirically analyze the full nature of responses. This study counted each document in a file as a separate response and used qualitative data to exemplify how institutional response processes. Overall, I acknowledge that the study would have benefited from consistent data from each response incident. The hostage situation that took place on February 1st is another possible limitation of the study, particularly the qualitative data. Training staff may have increased their punitive responses to better prepare employees for the possible dangers on-the-job. On the other hand, instructors concerned with high attrition rates and challenges in recruiting efforts may make conscious efforts to praise and nurture cadets, getting them to graduation to fill spots that are sorely needed. Data were collected approximately 4-7 months after the hostage incident. The quantitative data collected examined institutional responses to behavior across the tenure of nearly 300 randomly selected employees. The average tenure of these employees was 11 years of service to the MADOC. Therefore, the likelihood of the patterns found in the empirical analysis being greatly impacted by the hostage situation is slim. Additionally, quantitative data were statistically examined for differences in institutional responses by facility; no statistically significant differences were found.

This may be because the data used in this dissertation were at the administrative level

131

only. The qualitative data, however, may be impacted because the training classes took place in the aftermath of the hostage incident. Employees in the human resources department often remarked that the department “will never be the same”, particularly in terms of recruitment efforts and retention rates. During training, instructors and cadets referenced the hostage incident to highlight the seriousness of correctional work, the need for standards, to warn against complacency and other poor work habits. Although situations like these are impossible to control for, this study used qualitative methods to compliment quantitative findings understanding the possibility for data bias. Future studies are again encouraged to link quantitative and qualitative data whenever possible. Sample generalizability is defined as “the ability to generalize from a sample, or subset, of a larger population itself (Bachman & Schutt, 2014, p. 90)” MADOC is one of the six unified corrections departments in the United States (Vera, 2018). The uniqueness of the system is beneficial for access to state-wide data. However, the findings of this study are not generalizable to differently structured departments. Research applying the Institutional Model of social control should be duplicated in other types of correctional department of corrections. Although variations in access and procedures may cause restrictions or variations in qualitative data collection, duplicating the quantitative research design is possible. Overall, this study can be strengthened through data access improvements. Additionally, micro and macro institutional responses are missing from analysis due to access restrictions. In the future, scholars and practitioners are encouraged to work together to gain access to data that allow challenging research questions to be answered while maintaining the safety and security of facilities.

132

7.2 Challenges as a Researcher As a former correctional officer and current Ph.D. candidate, my relationship with MADOC was positive but complex. As part of the memorandum of understanding, I was expected to make the director of training “immediately aware of any feedback that you have about CEIT (content of training, structure, etc.) both positive and negative. MADOC desires to constantly improve the training experience and to equip correctional officers as best as possible. (p. 2)”8 Being obligated to share my opinions on the shortcomings of training classes while conducting research created a complex situation. To uphold this obligation, I frequently checked in with MADOC’s head of research, training staff, and human resources. Suggestions made included addressing vulgarity in training and in prison, creating activities for cadets to do in their downtime, cold-calling cadets who seemed less engaged in class discussion and updating outdated training resources such as video cassette tapes, workbooks and PowerPoint slides. Additionally, according to the MOU I was to “participate in CEIT classes as an observer only...if at any point your presence becomes disruptive or distracting, access to the classes will be rescinded. (p. 2).” However, this observer role was often blurred in numerous ways including: helping create and act out role-play scenarios, defining key-terms, and sharing personal professional experiences. These instances were always done at the request and/or approval of training staff. The complex, fragile and fluid choices that are faced by both researchers and institutional actors has also been acknowledged in research (Cunliffe & Alcadipani, 2016). Being allowed to assist in these capacities granted me further insight into the training classes, strengthened my

8 The MOU is not provided to maintain the anonymity of MADOC.

133

relationships with participants and granted me opportunities for informal conversations that otherwise may not have been possible. My role as a researcher was also complicated by the way MADOC staff framed my presence. Although granted access and introduced to all MADOC training staff, I was often introduced to cadets and other correctional employees after I had begun observations. This created tensions and rumors among cadets and employees, often getting back to me in the form of disgruntled rumors and dissatisfied statements. Prior to being allowed to make a formal introduction to one training class, cadets approached me during lunchtime and debated whether I was a “snitch or a snake”. A female said, “I really thought you were a snitch at first. Didn’t everybody think she was a snitch?” A male responded, “I didn’t think she was a snitch, I thought she was a snake because snitches talk, snakes just record.” Here, my presence, as well as my notebook, in the classroom shaped a narrative for cadets.9 Cadets in training weren’t the only ones to have strong feelings about my presence. After shadowing the training course inside a MADOC prison, both instructors and cadets told me that the correctional staff inside considered me a mole purposed with reporting back to “5th avenue” or administration. It is important to note that generally perceptions of my presence changed after I was given the opportunity to introduce myself. All individuals seemed more open and interacted with me in a direct way. It was not possible to introduce myself to every person I met (all correctional staff in the prison, employees at the MADOC, etc.), therefore, my framing could not be completely

9 I relayed this exchange to a member of training staff who immediately wanted to sanction the cadet for disrespect. This is the only known institutional response that I interfered with knowingly, as the interaction only happened because of my presence.

134

controlled. Perceptions of my motives further contributed to challenges as a researcher, although my overall data collection experience was positive.

7.3 Further Considerations Another factor to consider in future research includes “status of the victim” which is also supported by Black’s theory of law and social control. The “status of the victim” element accounts for the different social standings of targets of individual behavior. Like the IRM element Status of the Actor, Donald Black’s behavior of law established that crimes committed against inmates or people of marginalized groups such as African Americans will be held in less regard than those committed against a White or higher-status individual (Black, 1976). In this study, the target of employee behavior was not always available due to inconsistent details across HR files and instances where there was no victim involved. However, the qualitative data – specifically in the case of Layla – and previous criminal justice scholarship demonstrate the importance of considering the victim, not only in terms of severity of behavior but also because of their status. The social status of victims is crucial in determining the nature and consequences of interactions with criminal justice professionals and should be considered in future research.

It is also important to note that this dissertation does not consider promotion, pay raises, increase in rank, etc. as types of formal commendations. Future research may consider informal mechanisms (access to social networks that may enhance opportunities for praise) may be better captured in an extensive qualitative study. In future research designs, scholars should compare the way employees and institutions racially/ethnically categorize individuals. Because this study focused on formal responses to behaviors, the data provided (institutionally assigned racial/ethnic

135

categories) were appropriate. However, realizing that race is a complex social construct by also allowing employees to self-identify is important in future work. Lastly, future research is encouraged to consider how race interacts with other variables, such as tenure, rank, and age, to further flesh out the possible nuances in the behavior of race quantitatively. Although the interaction term between race and education was not statistically significant (see footnote 7), it is possible that race interacts with other factors to influence institutional responses. It is suggested that scholars explore these connections when theory and methodology justify it.

7.4 Policy Implications The Institutional Response Model of social control and the results of this dissertation have significant translational implications. Here, I make three major policy suggestions for criminal justice scholars and practitioners to consider in their future work. First, race matters. Although the racial/ethnic composition of criminal justice employees are not referenced in literature as much as the demographic makeup of those being processed through the system, it is important to consider. The racialization of justice in the United States cannot be confined to any one group. That is, all persons who are subject to any form of discipline in the U.S. are at risk for the same patterns and trends in broader society. This dissertation shows that Black correctional employees are at risk for higher quantities of sanction than their White counterparts. Beyond empirical testing, qualitative analysis provided further insight into how interactions that appear race-neutral may take on covert racialized meaning using the frames of the colorblind ideological framework. This suggests that scholars and practitioners concerned with racial/ethnic disparities research dig beyond statistics by

136

examining daily interactions where racialized meanings manifest in ways that may be overlooked. Second, although data access to certain social institutions can be difficult, relying on multiple sources of data can lead to findings and the development of translational implications that are more easily put into practice. Using data to extend, reinforce, contradict or complement each other is a powerful technique that enables researchers to speak to the nuances that practitioners witness daily. A major portion of maintaining healthy research relationships (once established) is presenting policy implications that are grounded in the realities of the organization. It is my suggestion that researchers seek to use multiple sources of data as often as possible, when appropriate, to answer social justice research questions with the complexity and nuance that these topics demand. Third, I recommend that the collateral consequences of the criminal justice system include criminal justice actors. Despite their oaths to enforce the law, these professionals are also being processed by the system. They are subject to the same physical, psychological and structural barriers that the accused, convicted, incarcerated and released face. If they are treated as above the justice system, adverse outcomes will perpetuate, often in ways that are more visible to those in disadvantaged communities than to mainstream society. The justice of the system, therefore, should be examined both externally and internally. It cannot be assumed that the system works in one direction. Practitioners, scholars and everyday citizens need to begin accounting for that. Furthermore, to better understand the interactions between actors, organizations and the perpetuation of systemic phenomena, I suggest the application

137

of the institutional response model of social control throughout other arms of the criminal justice system as well as to other institutions, more broadly. For instance, in what way does the praise and sanctioning of teachers and medical professionals lead to racial disparities in education (Quintana & Mahgoub, 2016) and healthcare (Egede, 2006), respectively? Literature finds that employee experiences shape outcomes for those who interact with the system – all types of systems. Therefore, it is vital to examine how employees are treated.

7.5 Conclusion This dissertation demonstrates correctional officers are managed through a series of decisions that are influenced by incident, individual and institutional characteristics. Institutional context is more influential in determining whether an institutional response occurs at all. Delving further results show that visibility and status of the actor variables (severity, age, race, and education-level) influence the risk of having more responses. Secondly, I find that the racial and ethnic inequalities that exist in the criminal justice system also extend to correctional staff. Black officers are at a higher risk for having multiple disciplinary sanctions in files than White officers. This disparity is not found when examining positive sanctions, in alignment with the disturbing racial/ethnic patterns found across the system. This is particularly important as employee experiences directly influence offender outcomes. Although acknowledging racial/ethnic disparities among criminal justice actors may be uncomfortable for social institutions, it is vital to confront to positively impact outcomes for staff, clientele and broader society. Results of this dissertation led to the development of three general policy implications and others contextualized for the Mid-Atlantic Department of Correction

138

(see Appendix B). Researchers and experts are encouraged to continually acknowledge the role of race in justice procedures within social institutions. The IRM should be applied in different contexts to challenge the colorblind and just distribution of social control among employees. Although generally difficult to access, relying on mixed methodology can serve as an advantage in answering research questions and establishing a healthy relationship between researchers and practitioners. Lastly, the impact of the criminal justice system on its employees must be considered one of the many collateral consequences of the system. More broadly, this dissertation suggests that the way race functions within criminal justice institutions in terms of institutional responses and employee training may lead to employees having different expectations of their interactions with their employer based on their institutional and personal status, as well as the visibility of their behavior. For instance, individuals whose professional and personal identities align with those that MADOC praised, like Roger, may have expectations of support and respect on-the-job. While, Alanna’s experiences during training or those of her coworkers of color who bore witness to the racialized narratives around her termination, may have less positive expectations of MADOC. These variations in employees’ expectations from their employer may manifest in a variety of ways, and differ based on race, level of education, prior law enforcement experience, etc. Ultimately, this dissertation prioritizes race, the criminal justice actor and the institution in perpetuating systemic phenomena such as racial disparities by examining institutional responses in correctional employment. This work matters not only for criminal justice actors who may be treated differently based on their race, level of education, tenure, etc., but also for those persons stopped, arrested, accused,

139

convicted, detained and supervised by the system. In other words, how criminal justice institutions respond to employee behavior may serve as a symbol of organizational realities that impact all who interact with it. In short, racial/ethnic disparate outcomes in the criminal justice system are rampant. These outcomes are tied to criminal justice actor misconduct. Criminal justice institutions set the standard for appropriate behavior by praising and sanctioning particular behavior. Therefore, the system must strive harder to hold itself accountable for all its short comings. If not, it is guilty of injustice - a system of blinded bars.

140

REFERENCES

Adams, K. (1999). What we know about police use of force. In Use of force by police: Overview of national and local data. Washington, DC: National Institute of Justice and Bureau of Justice Statistics.

Ajzen, I., & Fishbein, M. (1977). Attitude-behavior relations: A theoretical analysis and review of empirical research. Psychological Bulletin, 84(5), 888.

Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behaviour. Englewood Cliffs, NJ: Prentice-Hall.

Alexander, M. (2012). The new Jim crow: Mass incarceration in the age of colorblindness. New York: The New Press.

Alpert, G. P., & Dunham, R. G. (2004). Understanding police use of force: Officers, suspects, and reciprocity. Cambridge: Cambridge University Press.

Alesina, A., & La Ferrara, E. (2014). A test of racial bias in capital sentencing. The American Economic Review, 104(11), 3397-3433.

American Bar Association. (1908). Canons of Professional Ethics. Retrieved from http://www.americanbar.org/content/dam/aba/migrated/cpr/mrpc/Canons_Ethi cs.authcheckdam.pdf

American Civil Liberties Union Foundation. (2014). Racial Disparities in Sentencing. Hearing on Reports of Racism in the Justice System of the United States. Submitted to the Inter-American Commission on Human Rights. Retrieved from https://www.aclu.org/sites/default/files/assets/141027_iachr_racial_disparities_ aclu_submission_0.pdf

Ariel, B., Farrar, W. A., & Sutherland, A. (2015). The effect of police body-worn cameras on use of force and citizens’ complaints against the police: A randomized controlled trial. Journal of Quantitative Criminology, 31(3), 509- 535.

141

Armenta, A. (2017). Racializing crimmigration: Structural racism, colorblindness, and the institutional production of immigrant criminality. Sociology of Race and Ethnicity, 3(1), 82-95.

Amicis, A.D. (2005). An Ethical Dilemma in Corrections. National Criminal Justice Reference Service. Retrieved from https://www.ncjrs.gov/pdffiles1/212621.pdf

Armacost, B.E. (2003). Organizational culture and police misconduct. George Washington Law Review, 72, 453-545.

Armour, J. (1995). Stereotypes and Prejudice: Helping Legal Decisionmakers Break the Prejudice Habit. California Law Review, 83, 733.

Ashley, W. (2014). The angry black woman: The impact of pejorative stereotypes on psychotherapy with black women. Social Work in Public Health, 29(1), 27-34.

Bachman, R. D., & Schutt, R. K. (2014). Survey research. The Practice of Research in Criminology and Criminal Justice, 5, 189-234.

Bayley, D. H., & Mendelsohn, H. (1968). Minorities and the police: Confrontation in America. New York: Free Press.

Beck, A.J., Rantala, R., & Rexroat, J. (2014). Sexual Victimization Reported by Adult Correctional Authorities, 2009-11. U.S. Department of Justice. Office of Justice Programs. Bureau of Justice Statistics. Special Report. Retrieved from: https://www.bjs.gov/content/pub/pdf/svraca0911.pdf

Beck, K., & Wilson, C. (1997). Police officers’ views on cultivating organizational commitment. Policing: An International Journal of Police Strategies & Management, 20(1), 175-195.

Berk, R., & MacDonald, J. (2008) Overdispersion and Poisson Regression. Journal of Quantitative Criminology, 24(3), 269-284.

Black, D. J. (1976). The behavior of law. New York: Academic Press.

Black, D. J. (1984). Toward a general theory of social control. New York: Academic Press.

Blauner, R. (1969). Internal colonialism and ghetto revolt. Social Problems, 16, 393- 408.

Blumer, H. (1958). Race prejudice as a sense of group position. Pacific Sociological Review, 1(1), 3-7.

142

Blumstein, A. (1982). On the racial disproportionality of United States’ prison populations. Journal of Criminal Law and Criminology, 73, 1259–1281.

Bobo, L., & Hutchings, V. L. (1996). Perceptions of racial group competition: Extending Blumer's theory of group position to a multiracial social context. American Sociological Review, 951-972.

Bolton Jr., K. (2003). Shared perceptions: Black officers discuss continuing barriers in policing. Policing: An International Journal of Police Strategies & Management, 26(3), 386-399.

Bolton Jr., K., & Feagin, J. R. (2004). Black in blue: African-American police officers and racism. New York: Routledge.

Bonacich, E. (1972). A theory of ethnic antagonism: The split labor market. American Sociological Review, 37(5), 547-559.

Bonilla-Silva, E. (2013). Racism without racists. (4th edition). Lanham, MD: Rowman and Littlefield.

Bradler, C., Dur, R., Neckermann, S., & Non, A. (2016). Employee recognition and performance: A field experiment. Management Science, 62(11), 3085-3099.

Brady’s Mandate. (2011). The New York Times. Retrieved from http://www.nytimes.com/2011/11/13/opinion/sunday/bradys-mandate.html?_r- 2&

Braithwaite, J. (1989). Crime, shame and reintegration. New York, NY: Cambridge University Press.

Britton, D. M. (1997). Gendered organizational logic: policy and practice in men's and women's prisons. Gender & Society, 11(6), 796-818.

Britton, D. M. (2003). At work in the iron cage: The prison as gendered organization. New York, NY: NYU Press.

Britton, D. M., & Logan, L. (2008). Gendered organizations: Progress and prospects. Sociology Compass, 2(1), 107-121.

Brown, J. M., & Sorensen, J. R. (2014). Legal and extra-legal factors related to the imposition of blended sentences. Criminal Justice Policy Review, 25(2), 227- 241.

Bureau of Justice Statistics. (2018). Reentry Trends in the U.S. Retrieved from https://www.bjs.gov/content/reentry/reentry.cfm

143

Bureau of Labor Statistics. (2017). Labor Force Statistics from the Current Population Survey. Retrieved from https://www.bls.gov/cps/cpsaat11.htm

Burke, A.S. (2009). Revisiting prosecutorial disclosure. Indiana Law Journal, 84(2), Article 3, 481-519.

Burch, T. (2015). Skin color and the criminal justice system: Beyond black-white disparities in sentencing. Journal of Empirical Legal Studies, 12, 395–420. doi:10.1111/jels.12077

Calavita, K., & Jenness, V. (2014). Appealing to justice: Prisoner grievances, rights, and carceral logic. Oakland, CA: University of California Press.

Caldwell, H.M. (2013). The prosecutor prince: Misconduct, accountability and a modest proposal. Catholic University Law Review, 63(1), 51-101.

Callahan, L. (2004). Correctional officer attitudes toward inmates with mental disorders. International Journal of Forensic Mental Health, 3(1), 37-54.

Cantillon, D., Davidson II, W. S., & Schweitzer, J. H. (2003). Measuring community social organization: Sense of community as a mediator in social disorganization theory. Journal of Criminal Justice, 31(4), 321-339.

Carson, E.A. (2015). Prisoners in 2014. U.S Department of Justice Officer of Justice Programs Bureau of Justice Statistics. Bulletin. Retrieved from https://www.bjs.gov/content/pub/pdf/p14.pdf

Chappell, A. T., & Piquero, A. R. (2004). Applying social learning theory to police misconduct. Deviant Behavior, 25, 89-108.

Charmaz, K. (2014). Constructing grounded theory. London: Sage.

Cheeseman K.A., & Worley, R.M. (2006). A captive audience: Legal responses and remedies to female inmates and sexual abuse in prison. The Criminal Law Bulletin, 43(2), 439-455.

Chermak, S., McGarrell, E., & Gruenewald, J. (2006). Media coverage of police misconduct and attitudes toward police. Policing: An International Journal of Police Strategies & Management, 29(2), 261-281.

Chiricos, T., Padgett, K., & Gertz, M. (2000). Fear, TV news and the reality of crime. Criminology, 38(3), 755-786.

Christopher, W. (Ed.). (1991). Report of the independent commission on the Los Angeles Police Department. Collingdale, PA: Diane Publishing.

144

Civilian Complaint Review Board. (1993). Civilian Complaint Review Board Appoint Public in Int 0549-1992-A.

Clair, M., & Winter, A. (2016). How Judges think about racial disparities: Disparities: Situational Decision-Making in the Criminal Justice System. Criminology, 54(2), 332-359.

Cohen, A. K. (1966). Deviance and social control. Englewood Cliffs, NJ: Prentice- Hall.

Cohen, B., & Chaiken, J. M. (1972). Police background characteristics and performance. New York: New York City-Rand Institute.

Cole, D. (2000). No equal justice: Race and class in the American criminal justice system. New York, NY: The New Press.

Cook, C. L., & Lane, J. (2012). Examining differences in attitudes about sexual victimization among a sample of jail officers: The importance of officer gender and perceived inmate characteristics. Criminal Justice Review, 37(2), 191-213.

Craney, T. A., & Surles, J. G. (2002). Model-Dependent Variance Inflation Factor Cutoff Values. Quality Engineering, 14(3), 391-404.

Crenshaw, K. (1989). Demarginalizing the intersection of race and sex: A black feminist critique of antidiscrimination doctrine, feminist theory and antiracist politics. University of Chicago Legal F., 139.

Cunliffe, A. L., & Alcadipani, R. (2016). The politics of access in fieldwork: Immersion, backstage dramas, and deception. Organizational Research Methods, 19(4), 535-561.

Data USA. (2015). Bailiffs, correctional officers, & jailers. Retrieved from https://datausa.io/profile/soc/333010/#demographics

Davis, J., & Sorensen, J. R. (2013). Disproportionate minority confinement of juveniles: A national examination of black–white disparity in placements, 1997-2006. Crime & Delinquency, 59(1), 115-139.

Davis, K.C. (1969). Discretionary Justice. Baton Rouge: Louisiana State University Press.

Davis, P.C. (1989). Law as Microagression. Yale Law Journal, 1559, 1561-62.

DeHart, D., & Shapiro, C. (2017). Integrated administrative data & criminal justice research. American Journal of Criminal Justice, 42(2), 255-274.

145

Dennis, A. L. (2007). Reining in the minister of justice: Prosecutorial oversight and the superseder power. Duke Law Journal, 57(1), 131-162.

Dick, G.P.M. (2001). The influence of managerial and job variables on organizational commitment in the police. Public Administration, 89(2), 557-576.

Dillard, P. S., & Nielsen, C. R. (2015). Inmates, education, and the public good: Deploying catholic social thought to deconstruct the us‐versus‐them dichotomy. The Heythrop Journal, 56(5), 769-777.

DiMarino, F. (2009). Recruiting Minority Employees in Corrections. Corrections.com. Retrieved from http://www.corrections.com/articles/21076-recruiting- minority-employees-in-corrections

Donohue, J. J. (2014). An empirical evaluation of the Connecticut death penalty system since 1973: Are there unlawful racial, gender, and geographic disparities?. Journal of Empirical Legal Studies, 11(4), 637-696.

Dowler, K. (2005). Job satisfaction, burnout, and perception of unfair treatment: The relationship between race and police work. Police Quarterly, 8(4), 476-489.

Durose, M.R., Smith, E.L., & Langan, P.A. (2007). Contacts between Police and the Public, 2005. Bureau of Justice Statistics Special Report. Retrieved from https://www.bjs.gov/content/pub/pdf/cpp05.pdf

Egede, L. E. (2006). Race, ethnicity, culture, and disparities in health care. Journal of General Internal Medicine, 21(6), 667-669.

Eitle, D., D'Alessio, S. J., & Stolzenberg, L. (2002). Racial threat and social control: A test of the political, economic, and threat of black crime hypotheses. Social Forces, 81(2), 557-576.

Elwood III, F. H., & Naquin, S. S. (2004). New metrics for employee development. Performance Improvement Quarterly, 17(1), 56-80.

Emerson, R. M., Fretz, R. I., & Shaw, L. L. (2011). Writing ethnographic field notes. Chicago, IL: University of Chicago Press.

Engel, R., Calnon, J., & Bernard, T. (2002). Theory and racial profiling: Shortcomings and future directions in research. Justice Quarterly, 19(2), 249-273.

Fader, J. J., Kurlychek, M. C., & Morgan, K. A. (2014). The color of juvenile justice: Racial disparities in dispositional decisions. Social Science Research, 44, 126- 140.

146

Feldmeyer, B., & Ulmer, J. T. (2011). Racial/ethnic threat and federal sentencing. Journal of Research in Crime and Delinquency, 48(2), 238-270.

Fielding-Miller, R., Davidson, P., & Raj, A. (2016). Blacks face higher risk of drug arrests in white neighborhoods. International Journal of Drug Policy, 32, 100- 103.

Foucault, Michel. (1977). Discipline and punish: The birth of the prison. New York, NY: Pantheon Books.

Franklin, A. L., & Pagan, J. F. (2006). Organization culture as an explanation for employee discipline practices. Review of Public Personnel Administration, 26(1), 52-73.

Fridell, L., & Lim, H. (2016). Assessing the racial aspects of police force using the implicit- and counter-bias perspectives. Journal of Criminal Justice, 44, 36-48.

Fyfe, J. J., Kane, R. J., Grasso, G. A., & Ansbro, M. (1998). Gender, race and discipline in the New York city police department. Philadelphia, PA: Temple University.

Garland, B., Lambert, E. G., Hogan, N. L., Kim, B., & Kelley, T. (2014). The relationship of affective and continuance organizational commitment with correctional staff occupational burnout: A partial replication and expansion study. Criminal Justice and Behavior, 41(10), 1161-1177.

Gelman, A., Fagan, J., & Kiss, A. (2007). An analysis of the New York City police department's “stop-and-frisk” policy in the context of claims of racial bias. Journal of the American Statistical Association, 102(479), 813-823.

Goffman, E. (1961). Asylums: Essays on the social situation of mental patients and other inmates. Garden City, NY: Anchor Books.

Gottfredson, M.R., & Gottfredson, D.M. (1988). Decision making in criminal justice: Toward the rational exercise of discretion. New York: Plenum Press.

Gravett, W.H. (2017). The myth of objectivity: Implicit racial bias and the law (Part 1). Potchefstroom Electronic Law Journal, 20(1), 1-25.

Green, S.B. (1991). How many subjects does it take to do a regression analysis? Multivariate Behavioral Research, 26, 499-510.

147

Griffin, M. L., Hogan, N. L., & Lambert, E. G. (2014). Career stage theory and turnover intent among correctional officers. Criminal Justice and Behavior, 41(1), 4-19.

Higgins, G. E., Vito, G. F., Grossi, E. L., & Vito, A. G. (2012). Searches and traffic stops: Racial profiling and capriciousness. Journal of Ethnicity in Criminal Justice, 10(3), 163-179.

Hackett, E. J. (1994). A social control perspective on scientific misconduct. Journal of Higher Education, 65(3), 242-60.

Harne, J. (2017). Identifying at-risk officers: Can it be done in corrections? NIJ Journal, (278).

Harris, C.J. (2016). Towards a career view of police misconduct. Aggression and Violent Behavior, 31, 219-228.

Harris, D.A. (2007). The importance of research on race and policing: Making race salient to individuals and institutions within criminal justice. Criminology & Public Policy, 6(1), 5-23.

Hartley, D. J., Davila, M. A., Marquart, J. W., & Mullings, J. L. (2013). Fear is a disease: The impact of fear and exposure to infectious disease on correctional officer job stress and satisfaction. American Journal of Criminal Justice: The Journal of the Southern Criminal Justice Association, 38(2), 323-340.

Hartney, C., & Vuong, L. (2009). Created Equal: Racial and Ethnic Disparities in the US Criminal Justice System. National Council on Crime and Delinquency. Retrieved from http://www.nccdglobal.org/sites/default/files/publication_pdf/created-equal.pdf

Henriquez, M.A. (1999). IACP National Database Project on Police Use of Force in C.D. Maxwell et al. (Eds.), Use of Force by Police: Overview of National and Local Data. (NCJ 176330, p. 15-18) Research report jointly published with the Bureau of Justice Statistics. Washington, DC: National Institute of Justice.

Hilbe, J. M. (2011). Negative binomial regression. Cambridge University Press.

Hill, C. P. (2000). Black feminist thought: Knowledge, consciousness, and the politics of empowerment. New York: Routledge

Hindelang, M. (1978). Race and involvement in common law personal crimes. American Sociological Review, 43, 93–109.

148

Hope, E. C., Skoog, A. B., & Jagers, R. J. (2015). It’ll never be the white kids, it’ll always be us”: Black high school students’ evolving critical analysis of racial discrimination and inequity in schools. Journal of Adolescent Research, 30(1), 83-112.

Hormann, J.S. (1995) Virtual Reality: The Future of Law Enforcement Training. FBI Law Enforcement Bulletin (64).

Hughes, T., & Wilson, D.J. (2016). Reentry Trends in the U.S. Bureau of Justice Statistics. Retrieved from http://www.bjs.gov/content/reentry/reentry.cfm

Human Rights Watch Women’s Rights Project. (1996). All Too Familiar: Sexual Abuse of Women in U.S. State Prisons. New York: Human Rights Watch

Hunt, J. (1984). The development of rapport through negotiation of gender in field work among police. Human Organization, 43, 283-296.

Indemnify. (n.d.). Retrieved April 15, 2018, from https://www.merriam- webster.com/dictionary/indemnifying

Irwin, J. (2005). The warehouse prison: Disposal of the new dangerous class. Los Angeles: Roxbury.

Ivković, S.K. (2005a). Fallen blue knights: Controlling police corruption. Oxford, UK: Oxford University Press.

Ivković, S. K. (2005b). Police (mis)behavior: A cross-cultural study of corruption seriousness. Policing: An International Journal of Police Strategies and Management, 28(3), 546-566.

Jacobs, J.B., & Grear, M. P. (1977). Drop outs and rejects: An analysis of the prison guard's revolving door. Criminal Justice Review, 2(2), 57-70.

Jacobs, J.B., & Kraft, L.J. (1978). Integrating the keepers: A comparison of black and white prison guards in Illinois. Social Problems, 25(3), 304-318.

John, J.A., & Raper, N.R. (1980). An Alternative Family of Transformations. Applied Statistics, 29(2), 190.

Jootun, D., McGhee, G., & Campus, H. (2009). Reflexivity: Promoting rigour in qualitative research. Nursing Standard, 23(23), 42-46.

Jennings, W. G., Fridell, L. A., & Lynch, M. D. (2014). Cops and cameras: Officer perceptions of the use of body-worn cameras in law enforcement. Journal of Criminal Justice, 42(6), 549-556.

149

Johnson, B., & Bridgmon, P. (2009). Depriving civil rights. Criminal Justice Review, 34(2), 196-209.

Joy, P.A. (2006). The relationship between prosecutorial misconduct and wrongful convictions: Shaping remedies for a broken system. Wisconsin Law Review, 399-429.

Jurik, N.C. (1985). Individual and organizational determinants of correctional officer attitudes toward inmates. Criminology, 23(3), 523-540.

Jurik, N.C. (1988). Striking a balance: Female correctional officers, gender role stereotypes, and male prisons. Sociological Inquiry, 58(3), 291-305.

Kaheny, E. B., Haire, S. B., & Benesh, S.C. (2008). Change over tenure: Voting, variance, and decision making on the U.S. courts of appeals. American Journal of Political Science, 52(3), 490-503.

Kane, R.J. (2002). The social ecology of police misconduct. Criminology, 40(4), 867- 896.

Kane, R.J., & White, M.D. (2009). Bad cops: A study of career-ending misconduct among New York City police officers. Criminology and Public Policy, 8, 735- 767.

Kang, J., Bennett, M., Carbado, D., & Casey, P. (2011). Implicit bias in the courtroom. UCLA l. Rev., 59, 1124.

King, W.R. (2009). Police officer misconduct as normal accidents: An organizational perspective. Criminology and Public Policy, 8, 771-776.

Knapp, W. (1973). The Knapp Commission report on police corruption. New York, NY: George Braziller.

Kauffman, K. (1988). Prison officers and their world. Cambridge, MA: Harvard University Press.

Koch, J. L., & Steers, R. M. (1978). Job attachment, satisfaction, and turnover among public sector employees. Journal of vocational behavior, 12(1), 119-128.

Kutateladze, B. L., Andiloro, N. R., Johnson, B. D., & Spohn, C. C. (2014). Cumulative disadvantage: Examining racial and ethnic disparity in prosecution and sentencing. Criminology, 52(3), 514-551.

Larsen, A. K. (1993). Employee recognition. AORN journal, 57(4), 909-912.

150

Lawrence III, C.R. (1987). The id, the Ego, and Equal Protection: Reckoning with Unconscious Racism, Stanford Law Review, 39, 317, 323.

Lawton, B. A. (2007). Levels of nonlethal force: An examination of individual, situational, and contextual factors. Journal of Research in Crime and Delinquency, 44(2), 163-184.

Lee, J.C. & Park, H. (2017). In 15 High-Profile Cases Involving Deaths of Blacks, One Officer Faces Prison Time. The New York Times. Retrieved from: https://www.nytimes.com/interactive/2017/05/17/us/black-deaths-police.html

Legewie, J. (2016). Racial profiling and use of force in police stops: How local events trigger periods of increased discrimination. American Journal of Sociology, 122(2), 379-424.

Leiber, M. J., Peck, J. H., & Rodriguez, N. (2016). Minority threat and juvenile court outcomes. Crime & Delinquency, 62(1), 54-80.

Leip, L.A., & Stinchcomb, J.B. (2013). Should I stay or should I go? Job satisfaction and turnover intent of jail staff throughout the United States. Criminal Justice Review 38(2), 226-241.

Lersch, K.M., & Mieczkowski, T. (2000). An examination of the convergence and divergence of internal and external allegations of misconduct filed against police officers. Policing: An International Journal of Police Strategies and Management, 23, 54-68.

Lind, E.A., & Tyler, T.R. (1988). The social psychology of procedural justice. New York: NY: Plenum Press.

Lipsky, M. (2010). Street-Level bureaucracy: Dilemmas of the individual in public services. New York: Russell Sage Foundation.

Lopez, I.A. (2004). Racism on trial: The Chicano fight for justice. Cambridge, MA: Belknap/Harvard University Press.

Lot Jr., J.R. (2000). Does a helping hand put other at risk? Affirmative action, police departments, and crime. Economic Inquiry, 38, 239-277.

Lum, C. M., Koper, C.S., Merola, L.M., Scherer, A., & Reioux, A. (2015). Existing and ongoing body worn camera research: Knowledge gaps and opportunities. George Mason University. Retrieved from http://www.arnoldfoundation.org/wp-content/uploads/Phase-I-Report-Nov-28- 2015-FINAL.pdf

151

MacDonald, J. M., & Lattimore, P. K. (2010). Count models in criminology. In Handbook of quantitative criminology (pp. 683-698). Springer, New York, NY.

Mahfood, V.W., Pollock, W., & Longmire, D. (2013). Leave it at the gate: Job stress and satisfaction in correctional staff. Criminal Justice Studies, 26(3), 308-325.

Manning, P. K. (2009). Bad cops. Criminology & Public Policy, 8(4), 787-794.

Marable, M., Steinberg, I., & Middlemass, K. (2007). Racializing justice, disenfranchising lives: The racism, criminal justice, and law reader. New York, NY: Palgrave MacMillan.

Marquart, J.W., Bamhill, M.B., & Balshaw-Biddle, K. (2001). Fatal Attraction: An analysis of employee boundary violations in a southern prison system, 1995- 1998. Justice Quarterly, 18, 878-910.

Marquardt, D.W. (1970). Generalized Inverses, Ridge Regression, Biased Linear Estimation, and Nonlinear Estimation. Technometrics, 12(3), 591.

Martin, S. E. (1994). “Outsider within” the station house: The impact of race and gender on Black women police. Social Problems, 41(3), 383-400.

Massey, D. S., & Denton, N. A. (1993). American apartheid: Segregation and the making of the underclass. Cambridge, Mass: Harvard University Press.

Mastrofski, S.D., Worden, R.E., & Snipes, J.B. (1995). Law enforcement in a time of community policing. Criminology, 33(4), 539-563.

McClelland, K. (2012). “Somebody help me understand this”: The supreme court’s interpretation of prosecutorial immunity and liability under 1983. The Journal of Criminal Law and Criminology, 102(4), 1323-1361.

Mears, D. P., Cochran, J. C., & Lindsey, A. M. (2016). Offending and racial and ethnic disparities in criminal justice: A conceptual framework for guiding theory and research and informing policy. Journal of Contemporary Criminal Justice, 32(1), 78-103.

Merton, R.K. (1938). Social structure and anomie. American Sociological Review, 3, 672-682.

Micucci, A.J., & Gomme, I.M. (2005). American police and subcultural support for the use of excessive force. Journal of Criminal Justice, 33, 487-500.

152

Minnow, M. (1987). Foreword: justice engendered. (Supreme Court 1986 Term). Harvard Law Review, 101(1) 10-95.

Mitchell, O. (2005). A meta-analysis of race and sentencing research: Explaining the inconsistencies. Journal of Quantitative Criminology, 21(4), 439-466

Molleman, T., & Leeuw, F. L. (2012). The influence of prison staff on inmate conditions: A multilevel approach to staff and inmate surveys. European Journal on Criminal Policy and Research, 18(2), 217-233.

Molleman, T., & Van der Broek, T. C. (2014). Understanding the links between perceived prison conditions and prison staff. International Journal of Law, Crime and Justice, 42(1), 33-53.

Morris, R. (1988). The devil’s butchershop: The New Mexico prison uprising. Albuquerque: University of New Mexico Press.

Mosher, C. J., Miethe, T. D., & Phillips, D. M. (2002). The mismeasure of crime. Thousand Oaks, CA: Sage.

Ng, T.W. H., & Feldman, D. (2010). Organizational tenure and job performance. Journal of Management, 36(5), 1220-1250.

Nicholson‐Crotty, S., Nicholson‐Crotty, J., & Fernandez, S. (2017). Will more black cops matter? Officer race and police‐involved homicides of black citizens. Public Administration Review, 77(2), 206-216.

Nunnaly, J. (1978). Psychometric theory. New York: McGraw-Hill.

Oliver, P. (2008). Repression and crime control: Why social movement scholars should pay attention to mass incarceration as a form of repression. Mobilization: An International Quarterly, 13(1), 1-24.

Omni, M., & Winant, H. (1994). Racial formation in the United States. New York: Routledge.

Osgood, D.W. (2000). Poisson-Based Regression Analysis of Aggregate Crime Rates. Journal of Quantitative Criminology, 16(1), 21-43.

Park, R.E., & Burgess, E.W. (1921). Introduction to the science of sociology. Chicago: University of Chicago Press.

153

Parks, G. S., & Davis, A. M. (2016). Confronting Implicit Bias: An Imperative for Judges in Capital Prosecutions. Human Rights, 42(2), 22. Retrieved from http://link.galegroup.com/apps/doc/A516196626/LT?u=udel_main&sid=LT&x id=a9a1f09b

Petersilia, J. (1983). Racial disparities in the criminal justice system. Santa Monica, CA: Rand Corporation.

Platania, J., & Moran, G. (1999). Due process and the death penalty: The role of prosecutorial misconduct in closing argument in capital trials. Law and Human Behavior, 23(4), 471-486.

Pogarsky, G., & Piquero, A. R. (2004). Studying the reach of deterrence: Can deterrence theory help explain police misconduct? Journal of Criminal Justice, 32(4), 371-386.

Polzer, K., Nhan, J., & Polzer, J. (2014). Prosecuting the prosecutor: The makings of the Michael Morton Act. The Social Science Journal, 51(4), 652-658.

Prison Legal News. (2013). Court Upholds California Prison Guard’s Termination for Telling Prisoner to Hang Herself. Retrieved from https://www.prisonlegalnews.org/news/2013/apr/15/court-upholds-california- prison-guards-termination-for-telling-prisoner-to-hang-herself/

Quintana, S. M., & Mahgoub, L. (2016). Ethnic and racial disparities in education: Psychology's role in understanding and reducing disparities. Theory into Practice, 55(2), 94-103.

Ramsey, R. J., & Frank, J. (2007). Wrongful conviction: Perceptions of Criminal Justice Professionals Regarding the Frequency of Wrongful Conviction and the Extent of System Errors. Crime & Delinquency, 53(3), 436-470.

Ray, R., Marsh, K., & Powelson, C. (2017). Can cameras stop the killings? Racial differences in perceptions of the effectiveness of body-worn cameras in police encounters. Sociological Forum, 32, 1032-1050.

Ready, J. T., & Young, J. T. N. (2015). The impact of on-officer video cameras on police-citizen contacts: Findings from a controlled experiment in Mesa, AZ. Journal of Experimental Criminology, 11(3), 445-458.

Reaves, B.A. (2009). State and Local Law Enforcement Training Academies, 2006. Bureau of Justice Statistics Special Report.

154

Reiter, K. (2014). Making windows in walls: Strategies for prison research. Qualitative Inquiry, 20(4), 417-428.

Rhodes, W., Kling, R, Luallen, J., & Dyous, C. (2015). Federal sentencing disparity: 2005-2012. Bureau of Justice Statistics Working Paper Series. Retrieved from https://www.bjs.gov/content/pub/pdf/fsd0512.pdf

Ricciardelli, R., Maier, K., & Hannah-Moffat, K. (2015). Strategic masculinities: Vulnerabilities, risk and the production of prison masculinities. Theoretical Criminology, 19(4), 491-513.

Riksheim, E. C., & Chermak, S. M. (1993). Causes of police behavior revisited. Journal of Criminal Justice, 21, 353-382.

Riley, J. (2000). Sensemaking in prison: Inmate identity as a working understanding. Justice Quarterly, 17(2),359-376.

Roberts, J. M., & Sanders, T. (2005). Before, during and after: Realism, reflexivity and ethnography. The Sociological Review, 53(2), 294-313.

Rojeck, J., & Decker, S.H. (2009). Examining racial disparity in the police discipline process. Police Quarterly, 12(4), 388-407.

Ross, E. A. (1901). Social control. XX. The vicissitudes of social control. American Journal of Sociology, 6(4), 550-562.

Rousseau, D., & Baranger, J. (2017). Disparities in Sentencing. The Encyclopedia of Corrections. Hoboken, NJ: Wiley.

Ryan, L., & Golden, A. (2006). ‘Tick the box please’: A reflexive approach to doing quantitative social research. Sociology, 40(6), 1191-1200.

Russell-Brown, K. (2009). The color of crime: Racial hoaxes, White fear, Black protectionism, police harassment, and other macroaggressions. New York: NYU Press.

Sampson, R.J., & Bartusch, D. (1998). Legal cynicism and (subcultural?) tolerance of deviance: The neighbourhood context of racial differences. Law and Society Review, 32, 777-804.

Sampson, R., & Lauritsen, J. (1997). Racial and ethnic disparities in crime and criminal justice in the United States. Crime and Justice, 21, 311–374.

155

Sampson, R. J., Winship, C., & Knight, C. (2013). Overview of: “Translating causal claims: principles and strategies for policy‐relevant criminology”. Criminology & Public Policy, 12(4), 585-586.

Schaufeli, W. B., & Peeters, M. C. (2000). Job stress and burnout among correctional officers: A literature review. International Journal of Stress Management, 7(1), 19-48.

Schein, E.H. (1993). Defining organizational culture. In J.M. Shafritz, J.S. Ott, & Y.S. Jang (Eds.). Classics of organization theory. Belmont, CA: Wadsworth.

Schoenfeld, H. (2005). Violated trust: Conceptualizing prosecutorial misconduct. Journal of Contemporary Criminal Justice, 21(3), 250-271.

Schwartz, J.C. (2010). Myths and mechanics of deterrence: The role of lawsuits in law enforcement decision-making. University of California at Los Angeles Law Review, 57, 1024-1094.

Schwartz, J. (2014). Police indemnification. New York University Law Review, 89,855-1005.

Schwirtz, M. Winerip, M., & Gebeloff, R. (2016). The Scourge of Racial Bias in New York State’s Prisons. The New York Times. Retrieved from https://www.nytimes.com/2016/12/03/nyregion/new-york-state-prisons- inmates-racial-bias.html

Short, J., & Hughes, L. (2009). Urban ethnography and research integrity. Ethnography, 10(4), 397-415.

Silverman, D. (2013). Doing qualitative research: A practical handbook. Thousand Oaks, CA: SAGE Publications Limited.

The Sentencing Project. (2008.) Reducing Racial Disparity in the Criminal Justice System: A Manual for Practitioners and Policymakers. Retrieved from https://www.sentencingproject.org/wp-content/uploads/2016/01/Reducing- Racial-Disparity-in-the-Criminal-Justice-System-A-Manual-for-Practitioners- and-Policymakers.pdf

Seron, C., Pereria, J., & Kovath, J. (2004). Judging police misconduct: “Street-level” versus professional policing. Law & Society Review, 38(4), 665-710.

Sherman, L. (1978). Scandal and reform: Controlling police corruption. Berkeley: University of California Press.

156

Skolnick, J.H. (2005). Corruption and the blue code of silence. In R. Sarre, D.K. Das, & J. Albercht (Eds.), Police corruption: An international perspective. Oxford, K: Lexington Books.

Smith, B. (2012). Uncomfortable places, close spaces: female correctional workers’ sexual interactions with men and boys in custody. University of California at Los Angeles Law Review, 59, 1692-1745.

Smith, G., Hagger Johnson, H., & Roberts, C. (2015). Ethnic minority police officers and disproportionality in misconduct proceedings. Policing and Society, 25(6), 561-578.

Smith, R. J., & Levinson, J. D. (2011). The impact of implicit racial bias on the exercise of prosecutorial discretion. Seattle UL Rev., 35, 795.

Steiner, B.D. (2001). The consciousness of crime and punishment: Reflections on identity politics and lawmaking in the war on drugs. Studies in Law, Politics and Society, 23, 185-212.

Steiner, B.D. (2009). Assessing Static and Dynamic Influences on Inmate Violence Levels. Crime & Delinquency, 55(1), 134-161.

Steiner, B., & Wooldredge, J. (2015). Individual and environmental sources of work stress among prison officers. Criminal Justice and Behavior, 42(8), 800-818.

Sutherland, E.H. (1940). White-Collar Criminality. American Sociological Review, 5(1), 1-12.

Sykes, G., & Matza, D. (1957). Techniques of neutralization: A theory of delinquency. American Sociological Review, 22(6), 664-670.

Tedeschi, J. T., & Felson, R.B. (1994). Violence, aggression, and coercive actions. Washington, DC: American Psychological Association.

Terrill, W., & Reisig, M. (2003). Neighborhood context and police use of force. Journal of Research in Crime and Delinquency, 40(3), 291−321.

The United States Department of Justice. (2016). Department of Justice Announces New Department-Wide Implicit Bias Training for Personnel. Retrieved from https://www.justice.gov/opa/pr/department-justice-announces-new- department-wide-implicit-bias-training-personnel

157

Thomas, A. J., Witherspoon, K. M., & Speight, S. L. (2004). Toward the development of the stereotypic roles for Black women scale. Journal of Black Psychology, 30(3), 426-442.

Tillman, R. (1987). The size of the criminal population: the prevalence and incidence of adult arrest. Criminology, 25, 561-579.

Tyler, T.R., & Huo, Y.J. (2002). Trust in the law: Encouraging public cooperation with the police and courts. New York, NY: Russell Sage.

Van Cleve, N. G., & Mayes, L. (2015). Criminal justice through “colorblind” lenses: A call to examine the mutual constitution of race and criminal justice. Law & Social Inquiry, 40(2), 406-432.

Vera. (Retrieved 2018). Incarceration Trends. Retrieved from http://trends.vera.org/about

Voigt, R., Camp, N. P., Prabhakaran, V., Hamilton, W. L., Hetey, R. C., Griffiths, C. M., Jurgeens, D., Jurafsky, D., & Eberhardt, J. L. (2017). Language from police body camera footage shows racial disparities in officer respect. Proceedings of the National Academy of Sciences, 201702413.

Walker, S., Spohn, C., & DeLone, M. (2012). The color of justice: Race, ethnicity, and crime in America. Boston, MA: Cengage Learning.

Walker, S., & Wright, B. (1995). Citizen Review of the Police 1994: A National Survey. Washington, DC: Police Executive Research Forum Publication.

Warner, B. D. (2007). Directly intervene or call the authorities? A study of forms of neighborhood social control within a social disorganization framework. Criminology, 45(1), 99-129.

Watson, T.M. (2015). Research access barriers as reputational risk management: A case study of censorship in corrections. Canadian Journal of Criminology and Criminal Justice, 57(3), 330-362.

Weitzer, R. (1996). Racial discrimination in the criminal justice system: Findings and problems in the literature. Journal of Criminal Justice, 24(4), 309-322.

Weitzer, R., & Tuch, S.A. (2004). Race and perceptions of police misconduct. Social Problems, 51(3), 305-325.

Westley, W.A. (1970). Violence and the police. Cambridge, MA: MIT Press.

158

White, M. D., & Kane, R. J. (2013). Pathways to career-ending police misconduct: An examination of patterns, timing, and organizational responses to officer malfeasance in the NYPD. Criminal Justice and Behavior, 40(11), 1301-1325.

Wolfe, S., & Piquero, A. (2011). Organizational justice and police misconduct. Criminal Justice and Behavior, 38(4), 332-353.

Worley, R.M., & Cheeseman, K.A. (2006). Guards as embezzlers: The consequences of non-shareable problems in prison settings. Deviant Behavior, 27, 203-222.

Worley, R.M., Marquart, J.W., & Mullings, J.L. (2003). Prison guard predators: An analysis of inmates who established inappropriate relationships with prison staff, 1995-1998. Deviant Behavior, 24,175-198.

Worley, R.M., & Worley, V.B. (2011). Guards gone wild: A self-report study of correctional officer misconduct and the effect of institutional deviance on “care” within the Texas prison system. Deviant Behavior, 32(4), 293-319.

Wright, B.R.E., & Younts, C.W. (2009). Reconsidering the relationship between race and crime. Journal of Research in Crime and Delinquency, 46(3), 327-352.

Wu, B. (1997). The effect of race on juvenile justice processing. Juvenile and Family Court Journal, 48(1), 43-52.

Yan, H. (2017). ‘Black Lives Matter’ cases: What happened after controversial police killings. CNN. Retrieved from: https://www.cnn.com/2017/06/26/us/black- lives-matter-deaths-outcomes/index.html

Yang, C.S. (2015). Free at last? Judicial discretion and racial disparities in federal sentencing. The Journal of Legal Studies, 44(1), 75-111.

Yoshino, K. (2007). Covering: The hidden assault on our civil rights. New York, NY: Random House Trade Paperbacks.

Zuberi, T. & Bonilla-Silva, E. (Eds.). (2008). White logic, White methods: Racism and methodology. Lanham, MD: Rowman and Littlefield.

159

TIME IN THE FIELD

Day Time Spent Event Observed 1 8 hours 30 minutes CEIT 2 8 hours 15 minutes CEIT & PnP 3 3 hours 15 minutes CEIT 4 8 hours 25 minutes CEIT & PnP 5 5 hours 45 minutes CEIT 6 3 hours 45 minutes CEIT 7 5 hours 50 minutes CEIT 8 6 hours 50 minutes CEIT 9 1 hour 30 minutes Meeting 10 6 hours 8 minutes CEIT 11 4 hours 45 minutes CEIT 12 7 hours 20 minutes CEIT 13 8 hours CEIT 14 2 hours CEIT 15 9 hours 30 minutes Prison 16 9 hours 30 minutes Prison 17 9 hours 30 minutes Prison 18 8 hours 20 minutes Prison & Refresher 19 8 hours 30 minutes CEIT 20 3 hours 10 minutes Meeting 21 1 hour 45 minutes CEIT 22 4 hours CEIT 23 2 hours 30 minutes CEIT Total Observation Time 132 hours 23 minutes Total Quantitative Data Collection 52 hours Total Meeting Time 4 hours 40 minutes Total Time Reviewing Official 80 hours Documents Total Data Collection Time 269 hours and 3 minutes

160

MADOC POLICY IMPLICATIONS

After the hostage situation on February 1st, 2017, the governor of the Mid- Atlantic state where this study takes place commissioned an independent review of security issues within the prison and the department. The report gives an overview of the procedures within the prison and DOC that may have contributed to the incident. It also makes policy suggestions that would improve and avoid these circumstances in the future. Although the data examined in this dissertation do not stem directly from the hostage situation, the results of this study directly link to the recommendations made in the report. Sampson and colleagues state three mandatory considerations that researchers must include in a translational suggestion: (1) mechanisms and pathways (2) effect heterogeneity and (3) contextualization (Sampson, Winship, and Knight 2013). Below I offer 4 detailed policy suggestions based on findings of this dissertation that link to recommendations made in the independent review guided by the mandatory considerations by Sampson and colleagues.

B.1 Transparency in Response Processes Recommendation 3.3: Create a promotional career ladder with competitive salaries, and merit-based recognition. Recommendation 4.6: DOC should research, identify and implement a performance management system that holds all staff accountable for the

161

implementation of and adherence to policies and procedures, safety and security practices, as well as efficient and effective operations. Recommendations 3.3 and 4.6 both encourage an emphasis on merit-based commendations and appropriate systems of accountability with the DOC. However, neither of these suggestions sufficiently acknowledge the informal and individual traits that often influence how MADOC responds to employee behavior. In other words, effect heterogeneity is not considered in these recommendations. Results from this study demonstrate that commendations and sanctions are based on characteristics beyond merit – specifically visibility of the action and status of the actor. Therefore, it is critical that the MADOC become transparent in procedures of discipline and praise. This requires the acknowledgment that the current system is shaped by factors that are difficult to quantify such as social relationships. Perhaps even more important, it also calls for institutional recognition that employees of color, and individuals with a high- school level of education are at a disadvantage in terms of discipline. As the current recommendations stand, the institutional procedures do not account for varied experiences of diverse staff. Although MADOC must implement uniform standards, if the department is transparent, consistent and certain the likelihood of inherent biases will decrease.

B.2 Visibility as an Objective Tool Recommendation 7.1 With the review and approval of the newly appointed Warden, the Department should purchase all recommended cameras, recorders, and related equipment necessary to adequately cover all of [facility name removed] as recommended by the recent review done by DTI and have the systems installed as soon as possible.

162

Recommendation 7.4 Mid-Atlantic DOC should explore developing a policy and pilot test a Body Worn Camera program with their CERT Team. Based on this experience, the Department could expand the use of BWC to officers who work in buildings where there are higher numbers of incidents and altercations (medium-high and high security).

Recommendations 7.1 and 7.4 suggest an increase in the visibility of inmates and staff through the installation of cameras, recorders, body-worn cameras, etc. to assist in the review and possible prevention of incidents and altercations. Results of this study find that visibility matters in determining if and how institutions respond to employee behavior, particularly whether disciplinary sanctions are imposed. It is my recommendation that an increase in visibility be used an objective tool for distributing praise and punishment among correctional officers. If officers receive both positive and negative feedback from the presence of technology, the narrative around visibility may become routine, expected and -most importantly - objective. I also recommend an increase in visibility beyond staff-inmate interactions, to include staff-staff interactions. That is, an inmate does not have to be present for the commission of an adverse or commendable behavior. As some of the qualitative examples demonstrated, staff-staff interactions can have a major influence on officer experience.

163

B.3 Break the Code of Silence and Bridge the Gap Recommendation 2.4 DOC executive leadership should endeavor to build and maintain strong relationships with correctional officers and administrative personnel throughout the agency. Recommendation 4.4 Break the Code of Silence and bridge the gap between line officers and the corrections administration. Qualitative and quantitative data found divisions among correctional staff based on rank. Front-line correctional officers were significantly at a lower risk of receiving a positive commendation in their file than employees who had more advanced ranking. Qualitative analysis linked structural and informal divisions among staff to an “us vs. them” dichotomy including attire, nicknames (5th avenue), building layout and training expectations. Borrowing phrasing directly from recommendation 4.4, I recommend the unification of correctional officers and administrative personnel through structure, socialization and symbolism. It is recommended that MADOC hold inclusive training sessions, meetings and other formal gatherings, consisting of individuals across ranks and titles. These events would not only offer the opportunity for socialization between ranks but also emphasize unity and teamwork within

MADOC. Additionally, the department should consider holding multiple department and/or institution-wide social events that would increase socialization and act as symbols of pride for the department. Examples of these events include department sports teams, appreciation picnics and support groups. It is my suggestion that the events be held at various times of day/night to be inclusive to individuals assigned to all shifts.

164

B.4 Tenure-Long, Intensive Training Recommendation 5.2 Ensure training topics and hours meet national corrections standards and include real world scenarios. Recommendation 5.4 Develop a Field Officer program Improving training of correctional staff, as discussed in recommendations 5.2 and 5.3, is the last policy implication. As a component of the memorandum of understanding, I spent countless hours observing training classes and making notes on ways to improve the current system. Training classes, instructors and resources often pull their knowledge from corrections associations. Instructors also often developed, implemented and shared real-world scenarios to prepare cadets for their future as correctional employees. However, instructors also admitted that the training offered in these classes was only the foundation for successful careers in the field. They encouraged cadets to continue to seek out positive mentors, role models and helpful resources to master their jobs. It is my recommendation that correctional training be intensive and continuous throughout tenure. Although refresher training currently exists for officers, staffing shortages create challenges in ensuring that correctional staff receive training frequently. The current structure requires officers to report to the administrative office and be relieved of duty. It is my suggestion to train correctional staff at their assigned institutions. This would keep the manpower within facilities in case a security need arose. Furthermore, refresher training should be intensive, requiring examinations, physical training tests and other standards to be met on a continual basis. In other words, training overall could be improved by making the standards for training more rigorous throughout an employee’s tenure.

165

Relating to institutional responses to employee behavior, increasing training also increases visibility of each individual and the opportunity to spot signs of possible misconduct and excellence. Employees that demonstrate weakness or strength in certain areas could have their training specified to their needs. Additionally, leadership training should be conducted for every employee. This would allow transparency in the promotion and sanction processes and help ensure equity for supervisors and staff. In sum, institutional responses to employee behavior must come from transparent processes that consider the both the informal and formal pathways of praise and punishment. To improve transparency in these processes, technological advances should be framed as objective tools that are present to provide evidence for discipline and rewards. It is also recommended that the lines of communication between officers and administrators be extended through structural events promoting better socialization and symbolizing unity. Intensive training courses can bridge this divide, enforcing the standards for correctional employment and increasing opportunities for social ties.

166

Search Engine Results as of 4.26.16

Keywords Database Scholarly Journals

Police Officer + Misconduct Web of Science 140 Prosecutor + Misconduct Web of Science 45 Correctional Officer + Misconduct Web of Science 18 Police Officer + Misconduct Sociological Abstracts 62 Prosecutor + Misconduct Sociological Abstracts 12 Correctional Officer + Misconduct Sociological Abstracts 43 Police Officer + Misconduct Criminal Justice Abstracts 206 Prosecutor + Misconduct Criminal Justice Abstracts 36 Correctional Officer + Misconduct Criminal Justice Abstracts 7 Police Officer + Misconduct Google Scholar ~31,100 Prosecutor + Misconduct Google Scholar ~40,900 Correctional Officer + Misconduct Google Scholar ~1,790 Police Officer + Misconduct + Race Web of Science 21 Prosecutor + Misconduct + Race Web of Science 1 Correctional Officer + Misconduct + Race Web of Science 3 Police Officer + Misconduct + Race Sociological Abstracts 9 Prosecutor + Misconduct + Race Sociological Abstracts 0 Correctional Officer + Misconduct + Race Sociological Abstracts 6 Police Officer + Misconduct + Race Criminal Justice Abstracts 15 Prosecutor + Misconduct + Race Criminal Justice Abstracts 0 Correctional Officer + Misconduct + Race Criminal Justice Abstracts 0 Police Officer + Misconduct + Race Google Scholar ~18,000 Prosecutor + Misconduct + Race Google Scholar ~20,800 Correctional Officer + Misconduct + Race Google Scholar ~1,250

167

168