Gendered ?: Analyzing Bail Outcomes from an Courthouse

by

Rachel Schumann

A Thesis presented to The University of Guelph

In partial fulfilment of requirements for the degree of Master of Arts in Criminology and Criminal Justice Policy

Guelph, Ontario,

© Rachel Schumann, May, 2013

ABSTRACT

GENDERED BAIL?: ANALYZING BAIL OUTCOMES FROM AN ONTARIO COURTHOUSE

Rachel Schumann Advisor: University of Guelph, 2013 Dr. Carolyn Yule

The relationship between gender and bail is an important yet understudied area of research. Studies that have found a relationship between gender and bail generally overlook important differences that shape how men and women enter into crime and the types of conditions imposed on their recognisances. This study utilizes 115 bail cases from the Provincial

Courthouse in Kitchener, ON to examine the effect of accused gender on bail outcome. Results show that accused gender did influence decisions to grant or deny bail. While almost all accused persons required a surety and/or bail conditions to be released, the regression analysis suggests that women were more likely to be released compared to men. Based on the deep sample exploratory analysis, gender differences emerged around issues of mental health and drug use.

Theoretical and policy implications from this study are discussed as are avenues for future research.

ii

ACKNOWLEDGEMENTS

First, and foremost, I would like to thank my advisor Dr. Carolyn Yule for everything that you have done during the course of this project. Your constant enthusiasm, support, and dedication made this such a great experience, and I am incredibly lucky to have had the opportunity to work with you. Throughout this process, you pushed me to be a better scholar and gave me the confidence to continue in academia, and for that, I am sincerely grateful. I could not have done this without all your guidance, so thank you, thank you, thank you. Sincere thanks also go out to Dr. Dennis Baker. You have given me so much support during my Masters career and I am so happy you were a member of my committee. You provided such thorough, thought-provoking feedback and made valuable contributions to my project; it would not have been the same without you. Thank you to Dr. Myrna Dawson for acting as my external examiner, especially given the short timeframe. The depth of your feedback was astonishing and very much appreciated. You have given me so much to think about moving forward and I am grateful for having the opportunity to get your thoughts on my project. To my amazing parents, Kathy and Ernst Schumann, you have always taught me that hard work and determination pay off, and I owe much of my success to your unwavering ability to never stop believing in me. Thank you so much for everything. I would like to give a big thanks to my partner Mat for all the love you have shown me over the past eight years. I am so lucky to have you in my life and I look forward to our future together. Thanks to my friends (new and old), for all your encouragement during the last two years, it was always nice knowing that I had such a great support system to fall back on (and a shoulder or two to cry on if I needed it). Last, but certainly not least, a very special thanks goes out to my kitty Martin. You literally sat by my side every step of the way and provided much needed cuddles at exactly the right times. My life changed for the better the day we found you!

iii

TABLE OF CONTENTS Chapter 1: Introduction ...... 1 1.1 Bail: The Canadian Context ...... 2 1.2 Study Overview ...... 7 Chapter 2: Literature Review and Theoretical Perspectives ...... 11 2.1 Introduction ...... 11 2.2 Literature Review ...... 12 Gender, Legal and Extra-Legal Controls ...... 13 Gender and Type of Bail Releases ...... 18 Gender and Remand ...... 22 The Intersection Between Gender, Risk and Bail ...... 26 2.3 Theoretical Perspectives ...... 29 Legislative Changes Relating to Bail ...... 29 Risk Management ...... 33 Governmentality ...... 38 Summary of Theoretical Frameworks ...... 41 2.4 Research Objectives ...... 43 2.5 Conclusion ...... 44 Chapter 3: Data and Methods ...... 45 3.1 Introduction ...... 45 3.2 Data Collection ...... 45 Courtroom observations ...... 45 3.3 Sample ...... 48 3.4 Analytic Strategy ...... 50 3.4.1 Thematic analysis ...... 50 Key Measures ...... 52 Additional themes ...... 54 3.4.2 Logistic Regression ...... 55 Key Measures: ...... 55 3.4.3 Exploratory analysis: Deep sample method ...... 59 Matching the accused ...... 60 3.5 Limitations of the analytic design ...... 66 3.6 Conclusion ...... 67 Chapter 4: Results ...... 69

iv

4.1 Introduction ...... 69 4.2 Thematic Analysis ...... 69 Justifications for Granting Bail...... 69 Bail Conditions ...... 73 4.3 Quantitative Analysis ...... 76 Cross-Tabular Analysis ...... 76 Logistic Regression ...... 87 4.4 Deep Sample Exploratory Qualitative Analysis...... 91 Drug/Alcohol Use: Women receiving more restrictive bail releases ...... 92 Mental health and Culpability: Men receiving more restrictive bail releases ...... 97 Employment and Children: Similar bail outcomes but different justifications ...... 101 Similar outcomes and justifications ...... 104 4.5 Conclusion ...... 106 Chapter 5: Discussion ...... 107 5.1 Introduction ...... 107 5.2 Similarities in Justifications Used to Grant/Deny Bail ...... 107 5.3 Emerging Gender Differences in the Number and Type of Conditions Imposed on a Release ...... 113 5.4 Decisions to Grant Bail: Women Released More Frequently but not Treated More Leniently...... 119 5.5 Theoretical Implications ...... 122 5.6 Policy Implications ...... 129 5.7 Conclusion ...... 135 Chapter 6: Concluding Remarks ...... 137 6.1 Limitations and Future Research...... 137 6.2 Conclusion ...... 140 References ...... 142 Appendix ...... 149

v

LIST OF TABLES TABLE 1 Qualitative Female-Male Matched Pairs, Kitchener Provincial Courthouse, 2012 (N= 19)...... 62

TABLE 2 Justifications for Granting Bail, Kitchener Provincial Courthouse 2012 (N= 205)...... 70

TABLE 3 Justifications for Denying Bail, Kitchener Provincial Courthouse 2012 (N= 81)...... 71

TABLE 4 Bail Conditions, Kitchener Provincial Courthouse 2012 (n= 510)...... 73

TABLE 5 Bivariate Relationships: Legal, Charge, Accused, Victim, and Surety Characteristics by Gender of the Accused, Kitchener Provincial Courthouse 77 2012 (n= 115)......

TABLE 6 Bivariate Relationships: Bail Outcome by Legal, Charge, Accused, Victim, and Surety Characteristics, Kitchener Provincial Courthouse 2012 (n= 82 115)......

TABLE 7 Bivariate Relationships: Type and Number of Bail Conditions and Bail Amount by Gender of the Accused, Kitchener Provincial Courthouse 2012 86 (n= 98)......

TABLE 8 Estimated Coefficients of Logit Models Predicting Bail Outcomes Among Adult Men and Women, Kitchener Provincial Courthouse 2012 (n= 83)...... 90

TABLE A-1 Descriptive Statistics: Dependent, Independent and Control Variables, Kitchener Bail Court 2012 (N= 115)...... 153

TABLE A-2 Phases of Thematic Analysis*...... 158

TABLE A-3 Justifications for Granting Bail that Relate to the Surety, Kitchener Provincial Courthouse 2012 (N= 80)...... 158

TABLE A-4 Justifications for Granting Bail that Relate to the Accused, Kitchener Provincial Courthouse 2012 (N= 58)...... 159

TABLE A-5 Justifications for Granting Bail that Relate to Additional Levels of Supervision, Kitchener Provincial Courthouse 2012 (N= 40)...... 159

TABLE A-6 Justifications for Granting Bail that Relate to the Allegations, Kitchener Provincial Courthouse 2012 (N= 17)...... 160

vi

TABLE A-7 Justifications for Granting Bail that Relate to the Victim, Kitchener Provincial Courthouse 2012 (N= 3)...... 160

TABLE A-8 Justifications for Denying Bail that Relate to the Accused, Kitchener Provincial Courthouse 2012 (N= 42)...... 160

TABLE A-9 Justifications for Denying Bail that Relate to the Surety, Kitchener Provincial Courthouse 2012 (N= 24)...... 161

TABLE A-10 Justifications for Denying Bail that Relate to the Allegations, Kitchener Provincial Courthouse 2012 (N= 8)...... 161

TABLE A-11 Justifications for Denying Bail that Relate to the Victim, Kitchener Provincial Courthouse 2012 (N= 6)...... 161

TABLE A-12 Justifications for Denying Bail that Relate to Public Confidence, Kitchener Provincial Courthouse 2012 (N= 1)...... 162

TABLE A-13 Bail Conditions Related to Possessions, Kitchener Provincial Courthouse 2012 (n= 130)...... 162

TABLE A-14 Bail Conditions Related to Residence, Kitchener Provincial Courthouse 2012 (n= 94)...... 163

TABLE A-15 Bail Conditions Related to Communication, Kitchener Provincial Courthouse 2012 (n= 87)...... 164

TABLE A-16 Bail Conditions Related to Mobility, Kitchener Provincial Courthouse 2012 (n= 80)...... 164

TABLE A-17 Bail Conditions Related to Additional Levels of Supervision, Kitchener Provincial Courthouse 2012 (n= 45)...... 165

TABLE A-18 Bail Conditions Related to Curfew/House Arrest, Kitchener Provincial Courthouse 2012 (n= 37)...... 166

TABLE A-19 Bail Conditions Related to Counselling/Treatment, Kitchener Provincial Courthouse 2012 (n= 36)...... 167

TABLE A-20 Bail Conditions Related to Other, Kitchener Provincial Courthouse 2012 (n= 1)...... 167

vii

LIST OF FIGURES FIGURE A-1 Bail Court Checklist...... 149

viii

Chapter 1: Introduction

On June 25, 2012, Cathy, a white female in her 50s was brought before a Justice of the

Peace at the Kitchener Provincial Courthouse for a consent to release bail hearing.1 She was accused of shoplifting and trespassing after she failed to pay for $82.00 worth of groceries. She was a known alcoholic and had a somewhat related criminal record that included theft under

$5000 and two break and enters. In the prisoners box, she appeared old and in disarray. Although there was no one in the body of the court to support her, she did have Youth in Conflict with the

Law2 on her side. Two days later, on June 28, 2012, Glenn, a white male in his late 40s was brought before a Justice of the Peace for a show cause bail hearing. He was charged with breaching his current recognizance after a police officer saw him walking unsteadily near the courthouse. At the time of arrest, the accused had slurred speech and later admitted to the officer that he had been drinking, which was in direct breach of his release order. He had a lengthy criminal record, which included eight failures to comply. He also had a severe alcohol problem and appeared old and dirty in the prisoners box. No one came forward to offer their support, although Youth in Conflict agreed to supervise. In both cases, the presiding JPs granted bail.

While both accuseds were granted bail, they received very different release orders.

Notably, Cathy had a curfew placed upon her release while Glenn did not. In addition to her curfew, the JP prohibited Cathy from attending the place where the offense occurred, placed her under the supervision of Youth in Conflict with the Law, required her to reside at an approved of

1 In Ontario, there are two main bail courts: contested court and consent to release court. A contested hearing occurs when the Crown contests the release of an accused (presumably those who pose a greater risk) whereas a consent to release hearing occurs when the Crown has already agreed to an accused’s release. 2 In Kitchener, Youth In Conflict With the Law is a bail supervision program that helps eligible individuals locate a surety following their arrest or acts in a supervisory role when a surety is absent. Under the umbrella name, Waterloo Region Bail Program, Youth in Conflict supervises accused individuals of all ages, including adults. 1

residence, and directed her to take counselling. On the other hand, Glenn was prohibited from contacting the complainant (related to his outstanding charges) and possessing any alcohol and weapons. He was also required to take counselling, remain under the supervision of Youth in

Conflict, and reside at his current address or at a residence approved of by Youth in Conflict.

When justifying Glenn’s release, the presiding JP emphasized Glenn’s minor attempts to get better, even though his current charge was for consuming alcohol. For Cathy, the Court focused entirely on her alcohol usage, calling it the “cause of her troubles.” Despite having seemingly similar charge and background characteristics, it appears that Cathy had more onerous conditions placed on her release when compared to Glenn. To assess whether this difference is evidence of a broader trend, the current study investigates the role gender plays in decisions to grant or deny bail in Ontario. By comparing accused men and women, this study not only outlines major factors influencing bail decisions more generally, it also provides evidence to suggest that a gender disparity does exist in decisions to grant or deny bail, particularly in terms of the types of conditions imposed on recognizance orders.

1.1 Bail: The Canadian Context

Bail is an important yet understudied area of research. Not only does it provide researchers the opportunity to examine the role of judicial discretion at an early stage in the criminal justice system, it has also been shown to influence post-conviction sentences (Friedland,

1965; Kellough and Wortley, 2001; Williams, 2005).3 In Canada, once an accused is arrested, s/he is normally brought before a Justice of the Peace (JP) to determine whether release is

3 Specifically, it has been found that remanded accuseds are more likely to be incarcerated compared to those who are released on bail (Friedland, 1965; Kellough and Wortley, 2001; Williams, 2005). 2

appropriate pending trial.4 As Myers (2009) describes, “bail court is in the business of estimating and weighing risks as the court must determine if the accused poses an inappropriate risk such that they cannot be released back into the community (130).” In most cases, the onus is on the

Crown to ‘show cause’ why the accused should be detained while awaiting trial and it is generally assumed that less risky accuseds will be released without the use of sureties and conditions.5

Presently, there are three grounds upon which a JP can justifiably deny bail under section

515 (10) of the C.C.C.:

(a) where the detention is necessary to ensure the accused attends court; (b) where the detention is necessary for the protection or safety of the public, including any victim of or witness to the offence, having regard to all the circumstances including any substantial likelihood that the accused will, if released from custody, commit a criminal offence or interfere with the administration of justice; and (c) if the detention is necessary to maintain confidence in the administration of justice.6 If the presiding JP is not convinced that the accused should be detained, he or she must release the accused. There are a number of release options outlined under sec. 515(2) of the C.C.C. for

JPs to use. They can release an accused without sureties, with sureties, with a monetary deposit, or with both sureties and a monetary deposit. They are also entitled to impose any number of release conditions on a recognizance so long as they are related to the purposes of bail (i.e. to prevent an accused from reoffending). These can range from enumerated conditions such as a no communication order to non-enumerated conditions like house arrest. On the other hand, if the

JP is convinced that detention is appropriate under sec. 515(10), then s/he can remand the

4 Alternatively, an accused may be released on a police undertaking upon arrest. This means they do not need to attend court for an official bail hearing. 5 Since the introduction of the Bail Reform Act, Parliament has added a list of exceptions to this rule under Sec. 515(6),which outlines a number of cases where the onus is on the accused to show cause for release. 6 Paraphrased from Sec. 515(10) of the Criminal Code of Canada. 3

accused to custody pending trial. Indeed, JPs have considerable discretionary power when it comes to the interpretation and determination of bail. According to law, bail decisions are based on a balance of probabilities, meaning JPs only have to be satisfied that an accused will fail to attend court or commit additional offenses to deny bail. In this regard, JPs only need to consider the credibility and trustworthiness of the allegations when making decisions about future risks.

As such, they are not so much ‘triers of fact’ as they are ‘triers of risk’ (Myers, 2009).

The procedure of bail is also guided by constitutional laws that limit the use of unreasonable releases and pre-trial detention. Sections 11(d) of the Charter of Rights and

Freedoms, for example, states that any person charged with an offense has the right to be presumed innocent until proven guilty. Specifically, it protects the innocence of an accused until s/he can be proven guilty by a fair and impartial court (Doob and Webster, 2012). It also prevents the state from imposing punishment on the accused until guilt has been established.

Further, section 11(e) of the Charter states that any person charged with an offense has the right not to be denied reasonable bail without just cause. A common misconception is that this right only protects accuseds from being unreasonably detained, however, in R. v. Pearson (1992), the

Supreme Court clarified that this contained two distinct parts: that bail be reasonable and that it not be denied without just cause. Thus, when taken together, sections 11(d) and (e) of the

Charter not only suggest that remand be used as a means of last resort, they also suggest that bail be as least restrictive as possible.

However, rising non-compliance and remand rates indicate that perhaps the guidelines delineated above are not being followed (Doob and Webster, 2012; Myers, 2009; Trotter, 2010).

Certainly, this is the opinion of McLellan (2011), who argues that the current state of affairs is eroding presumptions of innocence and the right to reasonable bail. In particular, the rate at

4

which female adults were charged with failing to comply tripled between 1986 and 2005, increasing from 33 females per 100 000 population to 103 females per 100 000 population (Kong and Aucoin 2008). At the same time, the rate at which men were charged rose 82 percent during this period. This can have several ramifications when it comes to bail, as a failure to comply charge can initiate a reverse onus situation. According to sec. 515(6) of the C.C.C., this means the onus is on the accused to show cause why s/he should not be detained if s/he has already failed to comply with an earlier order. Not surprisingly, this makes bail more difficult to obtain

(McLellan, 2011; Myers, 2009; Trotter, 2010).

There has also been a simultaneous rise in the number of adult men and women sent to pre-trial custody in Canada (Porter and Calverly 2011, Webster et al. 2009). In 1986-1987,

Canada housed roughly 3700 un-sentenced and 12 000 sentenced individuals within provincial/ territorial prisons whereas in 2005-2006, there were roughly 10 700 un-sentenced and 9600 sentenced individuals within these facilities (Boritch, 2008). In Ontario alone, admissions to remand increased 30 percent between 1999/2000 and 2008/2009, with roughly 67 percent of all individuals in Ontario prisons awaiting trial (Porter and Calverley 2011). For both men and women alike, the majority of admissions to remand resulted from non-violent offenses (68 percent) with the most common being failure to comply with a recognizance or bail release order

(Porter and Calverley 2011). As a result, the growth in non-compliance and remand rates make it increasingly important to examine the factors that influence bail decisions for accused men and women to ensure that legal principles and constitutional rights are being upheld.

While non-compliance and pre-trial detention rates are increasing for both men and women, several patterns emerge that prompt the need for additional research on the topic of gender and bail. First, the female remand rate has grown tremendously since the 1980s.

5

Specifically, the female remand population in Ontario grew from roughly 1.4 per 100 000 females in 1981 to about 6.2 per 100 000 females in 2007, meaning almost two-thirds of all female prisoners in Ontario were un-sentenced (Gartner et al., 2009). Thus, while the number of provincially-sentenced women decreased by 8 percent between 1995/1996 and 2004/2005,7 the number of females admitted on remand more than doubled during this period (Gartner et al.,

2009). Between 1980 and 2007, the remand rate for adult men rose from 38 per 100 000 men to

80 per 100 000 men (Webster et al., 2009). While this is a considerable increase, it is evident that the female remand rate increased more dramatically during this time. This is surprising given that women are much less likely to be charged with a serious offense or possess a serious criminal record when compared to men (Kong and Aucoin 2008). In 2005, for example, almost half (47 percent) of all female accused were charged with property crimes, approximately one- third (28 percent) were charged with crimes against the person and 17 percent were charged with obstructing the administration of justice (Kong and Aucoin 2008).

Feminist criminologists also point to the fact that women often have unique circumstances that warrant additional judicial consideration (Chesney-Lind, 2006; Comack,

2008; Daly 1994; Daly and Chesney-Lind, 1988; Franklin, 2008). Unlike men, women generally enter the criminal justice system because of problems related to abuse, poverty, and drug use

(Boritch 2008, Maidment 2008). Lone parenting is also a major predictor of criminal behaviour since it leads to poverty and welfare dependency (Maidment, 2008). Because of these circumstances, women are also more likely to commit non-violent, “survival” based crimes, such as theft or fraud, and are therefore, less likely to have a serious criminal record (Boritch, 2008;

Franklin, 2008). As a result, some scholars (see Edgar, 2004; Gartner et al., 2009; Player, 2007)

7 Nationally, the charge-rate for women (and men) has declined since 1994/1995 (Kong and Aucoin, 2009). 6

question the rise in remand rates for women more generally, especially in cases where women were later acquitted or received a non-custodial sentence. As it stands, women may enjoy fewer protections than men under secs. 11(d) and (e) of the Charter.

While the relationship between gender and bail remains understudied in Canada, some

American studies have found a relationship between gender and bail in which women are less likely to be denied bail in comparison to men (Ball and Growette-Bostaph, 2009; Daly, 1987;

Demuth and Steffensmeier, 2004; Freiburger and Hilinski, 2010; Kruttschnitt, 1984; Kruttschnitt and McCarthy, 1985; Steury and Frank 1990; Williams, 2003). Scholars also note that certain legal and extra-legal factors affect the likelihood of bail differently for accused men and women

(see: Ball and Growette-Bostaph, 2009; Boritch, 1992; Daly, 1987; Channels and Herzberger,

1993; Demuth and Steffensmeier, 2004; Nagel, 1983). Because women are released more frequently than men, some scholars have concluded that women are treated more leniently by the courts simply because they are released more frequently (see: Ball and Growette-Bostaph, 2009;

Jeffries and Newbold, 2003; Kruttschnitt, 1984; Kruttschnitt and McCarthy, 1985; Visher, 1983).

However, others suggest that this literature is lacking in its assessment of men’s and women’s pathways into crime and the types of releases imposed on accused women (which can be just as punitive as pre-trial detention) (Trotter, 2010). Without a clear sense of how these factors shape bail outcomes, it becomes difficult to conclude whether a gender disparity truly exists in bail decisions (Daly 1994; Gartner et al. 2009).

1.2 Study Overview

Given the recent increases in non-compliance and remand rates for both accused men and women and the minimal attention afforded to this topic, it becomes important to assess the role gender plays in judicial decisions to grant or deny bail as well as additional factors influencing

7

the determination of bail more generally. Previous studies that found that women are more likely to receive bail often overlook important gender differences related to charge characteristics that may account for this finding, like the fact that women are normally charged with less serious crimes and are less likely to possess a lengthy criminal record. Further, little research has focused on the scope of bail releases in Ontario and how these may be imposed differently on accused men versus women. As demonstrated at the beginning of this chapter, the extant literature suggests that gender differences likely do exist in bail outcomes. Thus, the purpose of this study is threefold. First, it seeks to outline the types of justifications JPs used to support their decisions to grant or deny bail and the types of conditions placed on release orders overall.

Second, it aims to assess whether there were any similarities and/or differences in the types of justifications and conditions used based on accused gender. Finally, if gender differences are identified, this study intends to explain why this is so.

Chapter two of the current study reviews the literature on bail with a particular focus on gender. Prior research findings on the topic of gender and bail is mixed. While some studies have found evidence to suggest that gender plays a large role in the determination of bail (see: Bernat,

1984; Bottoms and McClean 1976; Freiburger and Hilinski 2010; Kellough and Wortley 2001;

Kruttschnitt 1984; Kruttschnitt and McCarthy 1985; Steury and Frank, 1990), others have failed to find similar results (see: Allan et al., 2005; Goldkamp and Gottfredson, 1985; Gottfredson

1974; Turner and Johnson 2000; Wilbanks). Further research also demonstrates that JPs in

Ontario are no longer following directives listed in the C.C.C., leading to more restrictive bail outcomes for both accused men and women (Doob and Webster, 2012; McLellan, 2011; Myers,

2009; Trotter, 2010). Four theoretical perspectives are also considered that may explain possible gender differences in bail decisions. These include recent legislative changes, risk aversion, risk

8

management, and governmentality. The first two theoretical frameworks focus on the role of penal populism and public interest in decisions to grant/deny bail while the second two look more closely at the role of risk in explaining different bail outcomes for men and women.

Chapter three describes the methodology of the study. It outlines the data collection procedure followed by a description of the resulting dataset, sample, and analytic strategy. The study’s methodological limitations are also discussed. Chapter four presents the results of the analytic strategy, highlighting major themes and key findings that address the main research questions listed above.

Chapter five discusses the findings of the current study. Similar to past research, findings show that women were probably more likely to be granted bail compared to men; however, certain groups of women, notably those who used drugs, were more likely to have restrictive releases compared to their matched male pair. The results also provide support to suggest that

JPs are imposing more onerous outcomes on accused men and women, particularly those who do not fit traditional masculine and feminine ideologies. Potential theoretical and policy implications stemming from these findings are discussed, followed by a review of the study’s limitations and ideas for future research. Overall, this study seeks to shed more light on the bail decision-making process and the role gender plays in decisions to grant or deny bail in Ontario.

While the results of this study support prior research to the extent that women appear to be released more frequently than men, little evidence was found to suggest that women were treated more ‘leniently’ by the courts. Further, the regular use of sureties and release conditions for both men and women alike suggests JPs were not following directives listed in the C.C.C., which may blur the line between crime prevention and punishment.

9

Finally, I present my concluding remarks in chapter six, where I discuss the key limitations of the study and suggest how these limitations provide important avenues for future research on the relationship between gender and bail outcomes.

10

Chapter 2: Literature Review and Theoretical Perspectives

2.1 Introduction

Increases in the female remand rate, both in Canada and abroad, have drawn recent scholarly attention to the topic of gender and bail. Research examining this relationship has often yielded mixed results. While some studies provide evidence to suggest that gender plays a large role in the determination of bail (see: Bernat, 1984; Bottoms and McClean 1976; Freiburger and

Hilinski 2010; Kellough and Wortley 2001; Kruttschnitt 1984; Kruttschnitt and McCarthy 1985;

Steury and Frank, 1990), others have failed to find similar results (see: Allan et al., 2005;

Goldkamp and Gottfredson, 1985; Gottfredson 1974; Turner and Johnson 2000; Wilbanks). The risk literature, in particular, emphasizes how traditional gender norms continue to influence the identification and classification of risk, leading to different bail outcomes for accused men and women (Boritch 1992; Boyd, 2006; Turnbull and Hannah-Moffat, 2009; Moore and Lyons,

2007). Given the complexity of the bail decision-making process, four interrelated theories are discussed that help provide justifications for expecting a relationship between gender and bail outcomes. These include recent legislative changes, risk aversion, governmentality, and risk management. This chapter begins by discussing the literature pertaining to gender and bail, followed by a detailed review of the four theoretical perspectives used to explain potential gender disparities in bail decisions. The last section of the chapter reviews both the research objectives and questions guiding the current study.

11

2.2 Literature Review

Effects of Gender on Bail Decision-Making

The discretionary nature of bail decision-making (Friedland 1965; McLellan 2011) offers researchers a unique opportunity to investigate the existence of potential gender disparities at an early stage in the criminal justice process. In comparison to sentencing literature, however, research on the effect of gender on decisions to grant or deny bail is quite limited. Studies that have examined gender in greater detail suggest that accused females are more likely to be granted bail than their male counterparts (Ball and Growette-Bostaph, 2009; Daly, 1987; Demuth and Steffensmeier, 2004; Freiburger and Hilinski, 2010; Kellough and Wortley, 2002;

Kruttschnitt, 1984; Kruttschnitt and McCarthy, 1985; Steury and Frank 1990; Williams, 2003).

This was supported by a Canadian study that examined over 1,800 criminal cases from two

Toronto bail courts between 1993 and 1994 (Kellough and Wortley, 2002). Using regression analysis, Kellough and Wortley (2002) concluded that gender was significantly related to the determination of bail, as accused women were less likely to be denied bail when compared to accused men. Based on this finding, they argued that moral assessments about the accused continue to play a prominent role in decisions to grant or deny bail.

Beyond decisions to grant bail, there is evidence to suggest that women are also more likely to receive lower bail amounts and have less stringent financial conditions8 placed on their release (see: Channels and Herzberger, 1993; Demuth and Steffensmeier, 2004; Kruttschnitt,

1984; Steury and Frank, 1990). Based on Kruttschnitt’s (1984) study of pre- and post-conviction decisions, women were commonly given lower bail amounts upon release when compared to

8 In the context of bail, JPs may impose certain financial conditions in addition to a set bail amount. For example, a JP may direct an accused to deposit a sum of money or other valuable security (beyond his/her bail amount) prior to being released. 12

men. While this finding may be attributed to the fact that the study only included females who committed non-violent offenses (Kruttschnitt, 1984), Demuth and Steffensmeier (2004) found similar results after analyzing data from 75 bail courts for the years 1990, 1992, 1994, and 1996.

To test whether gender differences truly existed, they looked at the actual decision to grant or deny bail, bail amount, and type of release (financial versus non-financial), and discovered that gender had a statistically significant and direct effect on bail outcome. Like Kruttschnitt (1984),

Demuth and Steffensmeier (2004) found that females were more likely to be released, have lower bail amounts and less onerous forms of release than men. Likewise, Steury and Frank’s

(1990) study showed that gender was significant in decisions to grant cash versus non-cash bail

(Steury and Frank 1990). Taken together, these findings have led some scholars to conclude that women are treated more ‘leniently’ by the courts compared to men (Ball and Growette-Bostaph,

2009; Jeffries and Newbold, 2003; Kruttschnitt, 1984; Kruttschnitt and McCarthy, 1985; Visher,

1983).

Gender, Legal and Extra-Legal Controls

Prior research shows that legal and extra-legal factors affect the likelihood of being granted bail differently for accused men versus women (see: Ball and Growette-Bostaph, 2009;

Boritch, 1992; Daly, 1987; Channels and Herzberger, 1993; Demuth and Steffensmeier, 2006;

Nagel, 1983). Not surprisingly, legal factors, such as offense seriousness and criminal record, have been shown to affect men’s ability to obtain bail differently than women’s (Ball and

Growette-Bostaph, 2009; Boritch, 1992; Channel and Herzberger, 1993; Farnworth and Teske

Jr., 1995). Earlier studies conducted by Nagel (1983), Kruttschnitt and McCarthy (1985), and

Channels and Herzberger (1993) concluded that offense seriousness, prior record, and outstanding charges were the most significant predictors of pre-trial decisions for men, but not

13

women. Similar results were gleaned from Ball and Growette-Bostaph (2009), who found that type of charge (i.e. violent versus property offence) had a greater effect on men than women across all bail decision points. This finding was limited by the fact that they did not control for extra-legal factors that may affect bail, such as employment or family responsibilities. Although there has been some evidence to suggest that employment status is a significant predictor of bail for men, this relationship is usually tempered when legal factors are controlled for. More specifically, men only benefit from having a job if they have a limited prior record, no outstanding charges, and did not commit a serious offense (Jeffries, 2001; Kruttschnitt, 1984).

Studies reporting on the interaction between gender and race consistently find that black and/or Hispanic men are not only the least likely to be granted bail, they are also more likely to receive high bail amounts and restrictive releases compared to white men and women from all other backgrounds (Demuth and Steffensmeier, 2004; Kellough and Wortley, 2002; Roberts and

Doob, 1997; Williams, 2005). When looking at research on female accused, the interaction effect between gender and race on bail is mixed. Some studies have suggested that black women are treated more leniently than white women (Bickle and Peterson, 1991; Freiburger and Hilinski,

2010; Kruttschnitt and McCarthy, 1985; Spohn and Beichner 2000; Demuth and Steffensmeier,

2004), while others report the opposite effect (Ball and Growette-Bostaph, 2009; Crawford,

2000; Steffensmeier, Ulmer et al., 1998). Using existing aggregate data on 2635 pre-trial decisions from a Michigan court, Freiburger and Hilinski (2010) noted that the effect of race and gender on bail was only significant for accused women, with black females being released more frequently than white females. This finding is contrary to those provided by Demuth and

Steffensmeier (2004), who found that while women were less likely to be detained than men across all racial and ethnic groups, white women were the least likely to be detained followed by

14

black and Hispanic women. Likewise, Ball and Growette-Bostaph’s (2009) results indicated that on average, women of other races were six times more likely to be denied bail relative to white women.

A major drawback to some of these studies is that they fail to test for interactions amongst gender and additional control variables, such as children, marital status, and employment. To account for potential interaction effects, studies have looked at the combined effect of gender, marital status, and children on bail outcomes (see: Daly, 1987; Jeffries, 2001;

Kruttschnitt, 1984; Kruttschnitt and McCarthy, 1984). Kruttschnitt and McCarthy (1985) discovered that women who were charged with less serious crimes and had young dependent children were more likely to be released than men and women without children who were charged with similar offenses. In comparison, employment status played a larger role in pre-trial release decisions of men, as those who were employed and had a limited prior record were more frequently released than those without employment prospects (Kruttschnitt, 1984). Daly’s (1987) study on sentencing and bail decisions from New York City and Seattle reported similar results; being married and having children were more important for women than men in predicting bail outcomes. In fact, she found that married women, regardless of whether or not they were mothers, were more likely to be granted bail than single women without children. Further, Visher

(1983) found that women who conformed to traditional gender norms were treated more

‘leniently’ by the courts than men who committed similar offenses. Women who violated these traditions, on the other hand, did not receive similar treatment (Visher 1983).

Visher’s (1983) findings are supported by Daly’s (1987) study assessing the rationale behind gender disparities in court decisions. A major theme that emerged from her interviews with 35 court officials was the importance of family and work in bail and sentencing decisions.

15

Judges and prosecutors in her study consistently reported that men and women defendants with familial or employment responsibilities deserved more leniency because they are “more stable and have more to lose” than those without such responsibilities (Daly 1987, 273). This held truer for female defendants, who were portrayed as the family caretaker. Although judicial decision- makers expressed leniency towards men who were employed, women who had children were viewed more favourably because of the court’s desire to protect families and children from the potentially negative consequences associated with a mother’s detention (Daly, 1987; Demuth and

Steffensmeier, 2004). A recent Scottish report on bail decisions in relation to female accuseds support some of Daly’s findings. Like Daly, Brown et al. (2004) interviewed thirty-six court personnel from eight different courts across Scotland. When asked, the group of prosecutors and

JPs stated that they would be less likely to deny bail if the accused had children (Brown et al.,

2004). Moreover, the group of JPs claimed they would be even less likely to deny bail if the accused was a single parent (Brown et al., 2004). Unlike previous studies, however, the majority of prosecutors and JPs agreed that gender should not influence bail decisions.

Contrary to the above findings, a small body of literature exists that has found little evidence to suggest that women are treated differently (i.e., more “leniently”) by the courts than men. In fact, some scholars argue that once certain legal factors such as offense seriousness and prior record are controlled for, gender has little or no effect on bail outcomes. In this regard, legal factors refer to characteristics that define criminal behaviour or background, which include prior criminal record, offense seriousness, and number of charges. Not surprisingly, prior research demonstrates that JPs often consider a number of legal factors when formulating their decisions, such as number of current charges, history of violence, seriousness of criminal record, and prior failures to comply (Brown et al., 2004; Kellough and Wortley, 2002). Turner and

16

Johnson’s (2006) results showed that after controlling for legal and extra-legal factors, there was no statistically significant difference with respect to bail amounts for accused men and women.

This finding was corroborated by a study that examined bail decisions from a court in an urban

American city, which found that seriousness of the offense played the largest role in the determination of bail regardless of accused gender (Goldkamp and Gottfredson, 1985). Likewise,

Allan et al. (2005) observed 648 bail cases heard in seven different courts across Perth, Australia and concluded that gender had no effect on bail decision-making. Instead, they identified offense seriousness, number of offenses and the presence of legal representation as significant predictors of bail outcomes.

Based on these findings, some scholars suggest that it is not surprising that women are released more frequently than men. Indeed, Daly (1994) argues that studies reporting on gender disparities fail to make gender the focus of their analysis, which results in an oversight of important gender differences “that shape how men and women enter into crime, the contexts in which they are likely to offend, and the variance in the rates and types of offending behaviour”

(Player, 2007: 413). Canadian statistical data confirms that almost half (47 percent) of all female accuseds are charged with property crimes, approximately one-third (28 percent) are charged with crimes against the person and 17 percent are charged with obstructing the administration of justice (Kong and Aucoin 2008). When women are found guilty, they are infrequently involved in serious violent crimes and the majority do not possess an extensive criminal record (Kong and

Aucoin 2008). As such, some scholars maintain that women are not treated more leniently; they simply commit less crime and have minimal criminal records in comparison to men, making them better candidates for release (Daly, 1994; Turner and Hilinski, 2011; Turner and Johnson,

2006).

17

Gender and Type of Bail Releases

Generally, research examining the effects of gender on bail fail to explore the types of release orders imposed on women, making it difficult to assert that women are in fact treated more ‘leniently’ than men. While some American studies maintain that women are often given smaller bail amounts (Kruttschnitt, 1984; Steury and Frank, 1990) and less onerous forms of financial releases (Demuth and Steffensmeier, 2004), it is difficult to draw parallels to the

Canadian system where there are several forms of release a JP can impose on an accused. These include releases without sureties, with sureties, with a monetary deposit or with both a monetary deposit and sureties (C.C.C. sec. 515(2)(b)-(2)(e). As per sec. 515(1) of the Criminal Code, accused persons should generally be released without sureties or a deposit unless the Crown can show why the imposition of such orders is necessary. The duty of the surety, who is often a family member or close personal acquaintance of the accused, “is to ensure the good behaviour of the accused while on bail and to render him if he does not appear to be keeping the peace and good behaviour” (Prowse J.A., R. v. Patko, 2005). In other words, sureties provide a guarantee to the court that an accused will attend all future court dates, follow the conditions of his/her release, and not reoffend while on bail. Depending on the situation, sureties are expected to monitor the accused person’s behaviour and actions by making regular house visits and phone calls. Often, as part of the bail release, the accused is expected to move in with the surety and follow the “routine and discipline” of the household (Trotter 2010). This empowers the surety to impose any additional rules or conditions on the accused as they see fit. If the accused fails to follow these directives, the surety can breach or render the accused for failing to comply with his/her recognizance (Trotter 2010). As a result, it is suggested that JPs only require more risky accused to have a surety in order to be released.

18

Similar to the use of a surety, sec. 515(4) of the C.C.C. outlines a number of enumerated release conditions JPs can give to an accused, which include conditions that restrict the use of weapons or limit communication. In addition, JPs also have the discretionary power to order any added conditions they deem necessary, as per sec. 515 (4)(f) of the C.C.C. These are often more intrusive than those listed in the C.C.C. and include conditions that relate to curfews, medical treatment, and drug/alcohol abstinence (Trotter 2010). Trotter maintains that these types of conditions not only cause “significant deprivation of liberty for the accused,” they also add an additional responsibility for the surety to ensure the accused does not breach (Trotter 2010, 6-

35). Given the restrictive nature of these types of releases, it has been argued that they should be used as a means of last resort, and only when the Crown has shown cause (R. v. Anoussis, 2008).

Presently, minimal research exists about the role and use of sureties and bail conditions in

Canada. A recent study by Myers (2009), however, indicates that both forms of release are becoming more prevalent in Ontario courts for accused men and women alike. Based on a study of 4085 cases heard in eight separate bail courts across Ontario, Myers (2009) found that the accused needed to have a surety in order to be released in 63 to 100 percent of contested release cases. This number remained high in consent to release situations, where it was found that both the Crown and JP required the accused to have a surety in roughly 60 percent of cases (Myers,

2009). When the accused did not have a surety present, Myers (2009) discovered that defense attorneys consistently adjourned the case until a surety could be located. Further, over 50 percent of accused persons who had a contested release were ordered to follow at least five conditions while on bail. This number was similar in cases where the Crown agreed to an accused person’s release (Myers, 2009). Although the C.C.C. limits the use of release conditions, Myers (2009) provides evidence to suggest this directive is not being followed in Ontario. Sprott and Doob

19

(2010) came to similar conclusions in their study of youth bail courts in Toronto. These findings showed that girls were significantly more likely to have treatment and rehabilitative conditions placed on them compared to boys. Sprott and Doob (2010) argued that this was the result of a continued paternalistic concern for girls that was based on notions of proper gendered behaviour

(Sprott and Doob 2010). On the basis of such findings, some scholars have argued that these types of releases are simply the result of risk governance (Lyons and Moore, 2007; Myers,

2009).

Additional research has also suggested that more stringent releases, namely those with sureties and conditions attached, make compliance extremely difficult (Sprott and Myers, 2009).

Indeed, Trotter (2010) argues that while the purpose of these releases is prevention, the consequences for failing to abide by them are often punitive in nature (i.e. arrest, loss of monetary bail amount, and detention). On average, 12 percent of adults charged with an offense are charged with failing to comply with a court order, which is often viewed by JPs and Crown

Attorneys as a disregard for court directives (Doob and Webster 2012). In 2008/2009, 68 percent of adults were remanded for a non-violent offense, the most common being failure to comply

(Porter and Calverly 2011). The rate at which female adults were charged with failing to comply tripled between 1986 and 2005, increasing from 33 females per 100 000 population to 103 females per 100 000 population (Kong and Aucoin 2008). At the same time, the rate at which men were charged rose 82 percent during this period. Overall, the rising number of bail violations for both men and women may indicate that JPs are imposing an increasing number of strict release conditions upon those charged with an offense, presenting a greater risk of non- compliance. However, compared to men, women infrequently commit serious offenses, which questions the rise in female non-compliance rates more generally (Kong and Aucoin, 2008). If

20

Sprott and Doob’s (2010) study is any indication on the types of conditions imposed, it may be that women as well as girls are being forced to engage in treatment-like programs in order to satisfy their bail plan, making it more difficult for them to comply.

Indeed, Sprott and Myers’ (2009) study of a sample of 276 youth bail cases show a

“significant interaction between the duration of the case and the number of bail conditions on whether or not the youth came back into court” (417). In other words, youth who had seven or more conditions attached to their recognizance and whose release lasted longer than 194 days were more likely to breach a condition than youth who had fewer conditions and spent a shorter time on bail (Sprott and Myers 2009). Based on these results, Sprott and Myers (2009) argue that youth released on a recognizance were ‘set up to fail’ due to the conditions placed upon them, finding that released youth were frequently sent to custody after failing to uphold their bail conditions. While this study looked at youth compliance, it is possible that adults are experiencing similar problems when trying to follow their release orders. Accordingly, Doob and

Webster (2012) note that failure to comply charges are simply “the result of criminalizing ordinary behaviour...whereby the criminal justice system produces its own crime” (43).

The risk literature has shown that both the type and number of bail conditions placed upon an accused’s release creates a network of control that may hinder their ability to successfully maneuver their imposed lifestyle changes. Typically under the watch of the police, surety, and/or bail program co-ordinators, the more conditions placed on an accused’s release, the more likely s/he will breach their bail plan and be sent back to court (Hannah-Moffat and

O’Malley, 2007; Moore and Lyons, 2007). Risk studies looking at the role of gender on bail tend to suggest that certain groups of women, in particular, are seen as more risky, which results in more stringent forms of releases. For example, Moore and Lyons (2007) and Boyd (2006),

21

discovered that drug-using women frequently received harsher bail outcomes when compared to other women, because they were viewed as being ‘doubly deviant.’ Similar to Sprott and Doob

(2010), Moore and Lyons (2007) found that typical releases included treatment/counselling-like conditions in addition to restrictions on residency and communication. Accordingly they argue that bail conditions that limited residency and mobility often placed women at risk by removing their ability to leave. If they did attempt to leave, they were faced with the possibility of being sent back to custody for breaching their conditions. Violating a bail condition initiates a reverse onus situation, where it is up to the accused to prove why he or she should be released (Trotter

2010). Rarely does this lead to the pre-trial release of the accused (Myers, 2009). While a woman may be released on a recognizance, the conditions in place may very well send her back to court if she is unable to comply. Without a systematic review of the types of conditions placed on release orders, it is difficult to speculate how these releases may invite a breach. The current study attempts to address this by analyzing the types of releases for accused men and women.

Gender and Remand

Despite the guidelines listed in the C.C.C., remand rates for both men and women have risen dramatically within the last decade. In 1986-1987, Canada housed roughly 3700 un- sentenced and 12 000 sentenced individuals within provincial/ territorial prisons whereas in

2005-2006, there were roughly 10 700 un-sentenced and 9600 sentenced individuals within these facilities (Boritch, 2008). In Ontario alone, admissions to remand increased 30 percent between

1999/2000 and 2008/2009, with roughly 67 percent of all individuals in Ontario prisons awaiting trial (Porter and Calverley 2011). According to the Centre for Justice Statistics, the female remand population in Ontario grew from roughly 1.4 per 100 000 females in 1981 to about 6.2 per 100 000 females in 2007, meaning almost two-thirds of all female prisoners in Ontario were

22

un-sentenced (Gartner et al., 2009). Thus, while the number of provincially-sentenced women decreased by 8 percent between 1995/1996 and 2004/2005, the number of females admitted on remand more than doubled during this period (Gartner et al., 2009). During this same period, the remand rate for adult men rose from 38 per 100 000 men in 1980 to 80 per 100 000 men in 2007.

It is apparent that while this is a considerable increase, the female remand rate increased more exponentially during this time.

The negative impacts of pre-trial custody on both correctional costs and the accused have been well documented. In particular, several studies have found that accused persons who are remanded to custody before trial are more likely to be found guilty than those who are released on bail (Friedland 1965; Kellough and Wortley 2002; Williams 2005). Specifically, defendants who experienced pre-trial detention are more likely to be incarcerated and receive longer sentences than defendants who are released (Friedland, 1965; Kellough and Wortley, 2002;

Williams, 2005). While this finding can presumably be explained, in part, by the fact that riskier accused are remanded, Kellough and Wortley (2002) suggest that pre-trial custody can lead to unfair plea bargains because it assumes guilt prior to trial. Specifically, they found that the chance of coercing a guilty plea was greatly enhanced if the accused was held in pre-trial detention. They also discovered that the Crown was more likely to withdraw all charges if the accused was released on bail.

Evidence also suggests that gender may play a mediating role in the relationship between bail and sentencing decisions. In fact, Williams (2005) found that after controlling for gender, remanded females who were found guilty were significantly less likely to receive a sentence of imprisonment when compared to men. Moreover, she discovered that the courts routinely acquitted women who spent time in pre-trial custody, suggesting major inconsistencies between

23

bail and sentencing decisions that needed further review (Williams 2005). Player (2007) substantiates these claims, arguing that because the majority of remanded women do not receive a custodial sentence (if found guilty), pre-trial custody should be reserved for women who present specific risks.

Issues of overcrowding, inadequate programming, and lack of proper services are all documented problems of provincial prisons and remand centres (Johnson 2002; Manson 2005;

Roberts 2005). Consequently, Roberts (2005) refers to time in remand as “dead time” since inmates have minimal access to the activities normally granted to sentenced offenders such as recreational, vocational, and rehabilitative programs. For instance, Porter and Calverley (2011) found that although offenders who previously spent time in pre-trial custody routinely requested programming for substance abuse, employment, and family visitations when in remand centres, access to these services was quite limited. Accordingly, judges normally consider the condition of remand centres upon sentencing remanded offenders as sec. 719(3) of the C.C.C. permits a sentencing judge to “take into account any time spent in custody by the person as a result of the offence.” According to Manson (2005) and Roberts (2005), these conditions directly result in the amount of credit given to offenders who served a pre-trial sentence. Although the purpose of denying bail is supposed to serve a preventative rather than punitive function, the emphasis on pre-trial credit during sentencing suggests that remand is a form of punishment. In many ways, the lack of programming afforded to those on remand implies that it may be more punitive than post-conviction imprisonment.

Casale (1989) corroborates many of these findings in her study on remanded women in

Holloway, England. She concluded that because remanded women had not yet been handed down a decision by a judge as either innocent or guilty, they were often not allowed to

24

participate in several vocational and recreational programs available in the facility (Casale 1989).

As a result, the majority of women interviewed spent at least 20 hours per day in their cells.

Similar to remanded men, spending time in custody pending trial caused an overwhelming amount of stress for un-sentenced women. Poor sanitation and inadequate clothing made women feel insecure and doubtful about their chances in court (Casale 1989). Several women expressed feeling hesitant about the outcome of their cases because they were not given ample time to prepare and meet with legal aid. Interestingly, Edgar (2004) found that the increasing female remand rate has actually made conditions in female remand centres far worse in comparison to those for men. The pressure placed on centres to respond to the growing remand population has lead to an overall lack of support for women suffering from mental health and substance abuse issues. Women suffering from drug addictions, for example, were often left on their own with no medication to treat the effects of withdrawal.

Aside from the growing remand rate, Cassale (1989) maintains that pre-trial custody can be especially difficult for women with family and children responsibilities. Specifically, Casale

(1989) discovered that the dislocation of remand centres, prison policies, and financial strain all put increasing pressure on accused women when trying to maintain positive relationships with their children. She found that prison policies made it extremely difficult for women to spend time with their children, as mothers were only allowed three 15-minute visits per day with their families. In addition, remanded women had limited access to a telephone, which made it extremely difficult to organize proper custodial arrangements for their children. Based on her interviews with remanded mothers, this separation became a major source of anxiety while in custody (Casale 1989). Although the impact of remand on Canadian women is unknown (as no studies were located that spoke directly about women’s experiences in pre-trial custody), the

25

literature suggests that remand may have a more profound effect on accused women than on accused men. If this holds true, then JPs should consider these effects when formulating their decisions – particularly in cases where the charge would not lead to a custodial sentence.

The Intersection Between Gender, Risk and Bail

According to the risk literature, the prevalence of gender stereotyping regarding appropriate masculine and feminine behaviours plays a large role in the construction of risk for men and women in bail courts. For women, the reliance on traditional gender constructs is based on a belief that they are less rational, less capable of virtue, and less ready to learn than their male counterparts (Frigon 1996; Hannah Moffat and O’Malley, 2007). Women are considered weak, emotionally sensitive, and incapable of engaging in masculine-type crimes (Turner and

Johnson 2000). As Daly (1987) found, the courts expected and assumed women to be responsible, caring mothers and wives who protect their loved ones from risk. Because they were seen as having stronger ties to their community and conventional institutions (Demuth and

Steffensmeier, 2004), Daly found that women who had family responsibilities were shown more leniency by the courts than women who did not.9 In this regard, it has been suggested that these gendered constructs are used to identify and construct female risk, as women who deviate from these feminine ideals are often portrayed as being a greater risk than women who embody these characteristics (Turner and Johnson 2000). In fact, studies have discovered that women are often treated more harshly if they violate these traditional gender norms because they are seen as being

‘doubly deviant’ (Boritch, 1992; Boyd, 2006; Moore and Lyons, 2007; Visher, 1983). Indeed,

Moore and Lyons (2007) found that women who used, sold, or imported drugs, regardless of prior record and offense seriousness, were routinely remanded because they violated both gender

9 The term ‘leniency’ here refers to comments made by the court officials she interviewed. 26

and legal norms. They suggested that women who did not conform to traditional feminine ideals were denied the chivalrous dispositions typically reserved for ‘normal’ women because they were seen as incapable of fulfilling their womanly duties. Unlike men, accused women who were arrested for a drug-related offense had the lowest likelihood of being released than any other type of offense (Ball and Growette-Bostaph, 2009).

As a result, scholars argue that womens’ statuses as both women and mothers form the basis of particular, gendered risks. Research completed by Boyd (2006) and Moore and Lyons

(2007) demonstrates the significance of the ‘good mother’ persona within the criminal justice system. These studies show that judicial decisions to grant or deny a pregnant woman bail regularly revolve around her ability to protect and secure her unborn child from risk (Moore and

Lyons 2007). Often times, JPs and the Crown recommended pre-trial custody in order to manage and prevent potential harms to both mother and foetus, especially if the woman was poor or unemployed (Edgar, 2004; Moore and Lyons 2007). Women in drug treatment courts, for example, were routinely encouraged to motivate themselves by thinking about regaining custody of their children (Moore and Lyons 2007). Accordingly, this positioning of the good mother was central to a woman’s success as the courts used this to judge her progress. In contrast, the courts do not appear to scrutinize accused fathers on their relationships with their children in the same way they scrutinize mothers (Hannah Moffat, 2007). Because there is greater interest in the relationships women have with their children, this becomes a means to manage female conduct deemed to be inappropriate. Boyd (2006) found that female drug users who were suspected of committing an offense were seen as more dangerous than male drug users by the courts since they were assumed to be putting their children at increased risk (Boyd 2006). Not surprisingly,

27

JPs were less likely to grant a drug-using mother bail due to the added risks she presented to her children (Moore and Lyons 2007).

Similarly, a British study on remanded women found that more risky women were generally those who suffered from mental health issues, drug and alcohol dependency, and poverty (Brown et al., 2004). When asked, judges and prosecutors agreed that remanded women were considerably more likely to have a substance abuse issue (Brown et al., 2004). Based on perceptions of accused women, interviewees suggested that these women were typically single mothers between the ages of 18 and 35 who had low levels of education and suffered from a drug and/or alcohol issue. Although the majority of remanded women did not commit serious offenses, they were viewed as being more likely to be repeat offenders due to their drug use, mental health status, etc.(Brown et al., 2004). As such, judges felt they had no choice but to detain them.

While studies have found evidence to suggest that constructs of masculinity, such as employment, improve a man’s likelihood of being released, this relationship is generally overshadowed by the fact that men possess lengthy criminal records and commit more serious offenses. According to Jeffries (2001), the court’s interpretation of men’s legal background is guided by masculine ideals that see men as independent, employed, and devoid of emotion.

Based on her study of pre- and post-sentencing decisions, she found that the courts perceived men as active participants in the criminal subculture, which led to more restrictive outcomes at the pre-trial stage compared to women. This seems to comport with existing statistics, which report that men are charged with more (serious) offenses, have lengthier criminal records, and progressively commit more serious crimes over time relative to women (Kong and Aucoin,

2008).

28

2.3 Theoretical Perspectives

As the foregoing review of the literature suggests, there is little consensus on the effect of gender on bail decisions. While some researchers argue that women are more likely to be released than their male counterparts, others maintain that gender has no predictive value in the determination of bail. The dramatic increase in the female remand rate in recent years has prompted scholars to revisit the question of whether gender is a contributing factor in bail outcomes. As such, this section outlines four theoretical perspectives in relation to gender and bail. The first two theories fall under the umbrella of penal populism in shaping bail decisions similarly for accused men and women, while the second two theories focus on the construction of risk in explaining potential gender differences in bail outcomes.

Explaining Potential Similarities in Bail Outcomes for Accused Men and Women

Legislative Changes Relating to Bail

Judicial decisions to grant or deny bail could be explained, in part, by the current political climate surrounding crime and crime prevention. Recent changes in the determination of bail have made it more difficult to obtain, raising questions about certain fundamental rights and freedoms, particularly the right not to be denied reasonable bail and presumptions of innocence

(McLellan 2011, Trotter 2010). The most recent reform occurred in 1997, when Parliament introduced a new ground on which to deny bail based on public confidence. In conjunction with pre-existing grounds that revolve around risks of absconding or reoffending, JPs now have the power to deny bail in order to maintain public confidence in the administration of justice (Trotter

2010). Unlike the previously existing grounds, sec. 515(10)(c) “permits the detention of an accused person based upon the public’s reaction to the decision to release, free from any concern about the accused person absconding, reoffending, or interfering with the administration of

29

justice” (Trotter, 2010: 3-27). Instead, JPs are asked to consider the following factors when determining whether maintaining confidence in the administration of justice warrants detention:

1) the apparent strength of the Crown’s case, 2) the gravity of the offense, and 3) the circumstances surrounding the commission of the offense. Not only did this legislative change allow Parliament to appear sensitive to the public’s fear of crime, it also provided JPs with an additional tool to deny bail. This was supported by the 2002 decision in R. v. Hall where the majority of the Supreme Court ruled that upholding confidence in the administration of justice was a pressing and substantial concern. Although “public interest” was struck down as a distinct means for which to deny bail, sec. 515(10)(c) remains rooted in notions of public interest as the clause encourages JPs to make decisions based on community ideals. Since R. v. Hall, conditions outlined in s. 515 (4) and 515 (4.3) have been expanded in the name of public interest, making it much more difficult to obtain bail (McClellan 2011, Trotter 2010).

This emphasis on public confidence reflects broader changes that have occurred within the administration of justice in Canada more generally. Crime control is no longer the sole responsibility of the government or criminal justice specialists. Rather, a variety of non-state actors, including members of public, now participate in this process, shifting policy towards prevention, harm reduction, and risk management. Public involvement in the criminal justice system has made citizens more aware and fearful of criminal activity (Garland 2001). Garland and Sparks (2000) argue that this crime consciousness has triggered public anxiety related to the consequences and harms of criminal behaviour, which has left the public unsympathetic towards those more prone to crime. Regular citizens empathize with victims of crime, prompting the creation of “new avoidance and crime prevention routines” aimed at prevention, incapacitation, and exclusion (Garland and Sparks 2000, 16). This new citizen involvement in crime and

30

punishment has left little room for expert opinion, as politicians seek to appease public disapproval at the expense of scholarly research and recommendations (Garland and Sparks

2000).

According to Garland (2001: 350) then, “the new penal ideal is one where the public is protected and its sentiments expressed.” This is accomplished in one of two ways - an adaptive strategy that is aimed at prevention and partnership or a sovereign state strategy that emphasizes control and punishment (Garland 2001; Kemshall 2003). An adaptive strategy occurs when the state acknowledges its limited role in crime prevention and delegates both accountability and responsibility for security to the community (Garland 2001). In turn, the community is tasked with “removing criminogenic opportunities through regeneration and social inclusion”

(Kemshall 2003, 42). Alternatively, a sovereign state strategy stresses a tough on crime approach that targets high risk individuals through “selective incapacitation” (Ibid., 44). In both cases, the government appears to favour populist demands over accused’s right. As Kemshall (2003) describes, this “populist punitiveness” receives relatively little resistance from the citizenry since it not only appears tough on crime, but also appears tough on the causes of crime (Ibid).

In this light, the expansion of s. 515(10) of the Criminal Code has intensified the need to protect community interests. This emphasis might make it difficult for JPs to balance accused rights and community interests given the public’s involvement in criminal justice. The decision in R. v. Hall (2002) has perhaps made JPs more protective of community values in difficult to reach cases. Based on this change, it is possible that JPs in Ontario are simply following precedent established by the Supreme Court in R. v. Hall, contributing to the current rise in remand rates for adult women. Although remand rates for men are increasing as well, the growth

31

for women might suggest that they are no longer given “preferential” treatment when it comes to pre-trial release decisions, resulting in similar outcomes for accused men and women.

Risk Aversion

Since R. v. Hall, some Canadian scholars have exposed disparities between the rhetoric and practice of the new punitive approach to criminal justice (Moore and Hannah-Moffat, 2005;

Myers, 2009; Webster et al., 2009). Certainly, Webster et al. (2009) argue that this discursive tone does not seem to comport with other patterns within the Canadian criminal justice system, such as the overall decrease in the levels of imprisonment. Instead, it could be that JPs are reluctant to release accused individuals due to reputational risks. According to Douglas’ (1992) blame society theory, “the system we are in now is almost ready to treat every death as chargeable to someone’s account, every accident as caused by someone’s criminal negligence, every sickness a threatened prosecution. Whose fault? is always the first question” (15-16). In this regard, decision-makers are regularly held accountable for failing to prevent unnecessary risks (Douglas, 1992). Feelings of anxiety and uncertainty have played a large role in the public’s obsession with holding decision-makers accountable for their choices, especially when these decisions are perceived as risky (Power, 2004). In an effort to appear in control, organizations have become hypersensitive to decision outcomes, which has shifted decision- making power into the hands of the public (Beck, 1992).

One unintended consequence of shifting decision-making power to the public has been that organizations are more preoccupied with managing reputational risks (Power, 2004). To avoid backlash, organizations no longer make decisions that have the possibility of ruining their reputation or legitimacy. Concerns about reputational damage and hyper-awareness of uncertainty have made “risk the modelling ideology of organizations, where a good organisation

32

has come to be equated with being a good risk manager” (Myers, 2009: 129). As such, Webster et al. (2010) maintain that Canada’s growing remand population is largely the product of risk aversion whereby “decisions are either being continually passed along to someone else or are simply delayed by those making them” (99). In bail courts, JPs determine both primary and secondary risks. Based on Power’s (2004) definitions of risk, primary risks are the potential risks caused by the accused if they are released and secondary risks are risks to the courts if the accused reoffends while on bail (Myers, 2009). 10 The fear of making a poor decision may force

JPs to be overly cautious when determining bail for accused men and women alike in order to maintain legitimacy. To minimize secondary risks, JPs may not only detain accused men and women at an increased rate, they may also be more likely to impose stricter bail conditions on those released (Myers, 2009). Based on this theory, it is therefore possible that certain groups of men and women are either being detained or receiving more onerous releases, because of the secondary risks they pose.

Explaining Potential Differences in Bail Outcomes for Accused Men and Women

Risk Management

In the context of bail, risk management provides a useful framework for analyzing why certain behaviours and activities are considered risky. The current use of pre-trial custody and release conditions can therefore also be explained by the growing reliance on risk management models within Western criminal justice systems. Over the last twenty years, risk models have gained increased legitimacy because they attempt to predict and identify risky actions before they occur. Although bail decisions have always revolved around the judgment of risk, decisions

10 Examples of primary risks would be if the accused absconded or reoffended while on bail. An example of a secondary risk would be negative publicity resulting from a decision the media or public did not agree with (such as releasing an accused that the public believed should be detained). 33

may no longer be determined by assessing the character of the individual accused but rather on identifying risk profiles based on information gleaned from the accused’s membership to a particular subpopulation (Hannah-Moffat and O’Malley, 2007). Similar to governmentality, risk classifications are sometimes based on gendered values, which help to explain the potential gender disparities that may exist in bail outcomes.

Generally, risk management seeks to alter the “physical and social habitat in which individuals conduct their lives in order to reduce criminal conduct and increase maximum efficiency” (Feeley and Simon, 1994: 178). According to Beck (1992), this ‘risk society’ developed out of the hazards and insecurities caused by modernization and economic progression, which has created new risks that are increasingly visible and unmanageable. Fear and anxiety associated with risk avoidance now means that individuals are no longer “concerned with attaining something good, but rather with preventing the worst” (Beck, 1992, 49). Beck’s risk society theory, however, has been criticized for overstating the presence of new, under- regulated risks. Rather, it has been argued that it is more about the way risks are managed than the existence of more risks (Douglas, 1992; Kemshall, 2003; O’Malley, 1999). For O’Malley

(1999), governments manage risk by creating fear within the citizenry by playing on the uncertainty of the future. In this context, uncertainty acts as a “distinctive modality of governance that relies on creative constitutions of the future with respect to the dispositions of risk taking and potential harms” (O’Malley, 1999: 461). By creating new risk categories, organizations are able to alter how risk is perceived by the public while appearing to be accountable and in control (Douglas 1992; Power 2004). This form of risk management is often future oriented, predictive, and focused on the harms and costs of certain activities rather than the attribution of blame (Feeley and Simon 1994; Silver and Miller 2002).

34

As argued by Feeley and Simon (1994), this focus on actuarial reasoning is quintessential to the New Penology, which views deviance and crime as normal, natural occurrences. The acquisition of knowledge plays a key role in the ability to manage risk, since it not only creates acceptable standards of risk but it also produces new categories of risk in need of management

(Ewald, 1991; Rose 2000). Using this knowledge, penal administrators have been able to control high-risk populations through preventive techniques, such as incapacitation and imprisonment.

The use of risk models, while having always played a role in the criminal justice system, have become intensified in neoliberal societies since they allow state and non-state actors to identify, manage, and transform risky conduct before it occurs. To ensure risk prevention, high-risk individuals are typically managed via imprisonment.

The advantage of using actuarial techniques within the criminal justice system is that they rarely prompt social resistance because they appear to be more technical, statistical and inclusive rather than moral, exclusive and coercive (O’Malley 1996, 1998). Feeley and Simon (1994) maintain that the aim of the New Penology is not to make individuals morally responsible for their actions; it is simply to regulate groups as “part of a strategy of managing danger” (173).

Further, because of its concern with managing the effects of targeted categories of action, rather than a political or moral concern about the causes or faults of crime, actuarial practices avoid the question of assigning responsibility for the origins of criminal behaviour or “social evils”

(Douglas 1992). In other words, risk has completely removed the moral aspect traditionally connected with deviant behaviour by making it appear natural and objective.

When analyzing risk management strategies and who is targeted, however, it is apparent that the entire process is based on moral values. The creation of risk categories, which is a central component of managing risk, is dependent on ideas of morality and normalcy, targeting

35

all those who fall outside these accepted boundaries. Madriz (1997) suggests that the current discourse presenting crime and criminals as highly risky and dangerous serves to deflect criticism away from the fact that the state has failed to address systemic and persistent issues such as poverty, unemployment, class, gender, and racial inequalities. Instead, the public and the state mistakenly blame people who they perceive to be socially threatening rather than recognizing the failures of the state. Those who are seen as high risks tend to be the most socially disadvantaged: the poor, the unemployed, single mothers, and visible minorities (Madriz, 1997;

Rose, 2000). In this regard, individuals are perceived to be high risk, not because of individual wrongdoings or characteristics, but because of their membership to one of the above groups

(Hannah-Moffat and O’Malley, 2007; Rose, 2000).

Although risk management was once seen as a purely objective and neutral tool used for analyzing probability, its current use in political discourse has moved it in the direction of moral regulation. As Douglas (1992:45) states, risk has been transformed into “new moral questions about allocation.” O’Malley (1996, 1998, 1999) substantiates this thought, insisting that the use of actuarial techniques is the result of moral and political struggles over the meaning of proper individual conduct, which is exemplified by the current, neoliberal understanding of citizenship.

As such, citizenship now entails active participation from all members of society, as individuals are urged to take responsibility and protect themselves from the risks of crime (Rose, 2000).

Those who fail to take responsibility and manage their own risk become excluded from mainstream society, often through incapacitation and imprisonment practices (Ibid., 2000).

It has been suggested that risk management within the criminal justice system has become an essential tool in the management of exclusion in “post welfarist strategies of control”

(Rose 2000). Actuarial reasoning within criminal justice systems has further criminalized and

36

marginalized those who are unable to manage their own risk. It is assumed that these individuals are members of communities that either lack moral character or are ‘anti-moral’ in their disposition. As a result, risk management models ensure community protection by identifying risky individuals based on their association with particular communities and subpopulations

(Pratt et al., 2005). Poverty indicators such as unemployment and low education levels have been cited as major determinants of remand, since these factors not only fall outside the standards of active citizenship but connect an individual to risky subpopulations as well (Trotter, 2010). In this context, dependency on the state reflects an individual’s inability to properly conduct themselves in accordance with social norms.

When examining differences in bail outcomes for accused men and women, it is possible that gendered constructs of risk may be influencing judicial decisions to grant or deny bail.

Indeed, Madriz (1997) argues that womanhood has become a risk category within the criminal justice system because women are normally members of high-risk communities due to issues stemming from mental health, drugs, and poverty. A woman’s inability to engage in mainstream society makes her a risk because she fails to personify ideal feminine characteristics. For example, parenthood is a key risk factor for women because it is a traditionally female-role

(Moore and Lyons, 2007). Compared to fathers, the courts are more interested in the types of relationships mothers have with their children, making this an important site of risk management

(Moore and Lyons, 2007). With this in mind, it is likely that accused mothers are seen as being a greater risk to themselves and their children for not taking active responsibility for their situations. It is therefore possible that these women may be more likely to be denied bail or receive more restrictive bail releases than women who embody traditional feminine ideals. As the literature suggests, women who use drugs and commit a crime are viewed as being ‘doubly

37

deviate’ for not only committing an offense, but for transgressing proper womanly conduct as well (Boyd, 2006).

Masculine ideologies may also affect bail outcomes for accused men, leading to further gender differences in bail outcomes. Prior research has discovered that employment is a key factor in pre-trial release decisions for men. Based on the male breadwinner ideal, judges considered employed men less risky than unemployed men, leading to different bail outcomes

(Daly, 1987; Nagel 1983). It has also been shown that accused men are held responsible for their actions differently than accused women (Jeffries, 2001). Unlike women, accused men are typically seen as being actively engaged in crime, which make them more risky. This finding is not surprising since accused men generally possess a lengthy and/or serious criminal record, and are charged with more severe offenses when compared to accused women (Kong and Aucoin,

2008). As a result, it is possible that accused men, because of their backgrounds in crime, are viewed as being the most risky, followed by ‘unfeminine’ women. Women who continue to embody proper feminine characteristics may be considered the least risky and have less restrictive releases compared to men and ‘unfeminine’ women.

Governmentality

Governmentality provides a useful framework for analyzing differences in how accused men and women are managed by the courts. Broadly speaking, governmentality explores the interplay of discourse, activity, and practice used by state and non-state actors to shape individual behaviour (Foucault, 1991, Murray, 2007). As suggested by Foucault (1991), the state does not force individuals to abide by societal norms; rather it encourages individuals to align themselves with these standards using laws, regulations, and dialogue. To understand how these standards are maintained, it is important to discuss the role of neoliberal discourse in the current

38

political environment. Since the early 1980s, neoliberal frameworks have forged new relationships between governments and their citizens. Influenced by economic rationalities, neoliberalism encourages state cutbacks, privatization, and the deregulation of government controls (Dobrowolsky, 2009). Under this governing philosophy, there has been a dramatic shift in the discursive tone used to describe active citizenship, which is now based on self-governance and responsibility (Scoular and O’Neill, 2007). Limits on governmental activity, in areas like the economy, have encouraged individuals to become more entrepreneurial, competitive and economically rational (Burchelle, 1996).

Despite the fact that neoliberal governments appear to be “rolling back the state,” they have creatively invented new ways to shape social life based on economic enterprise. These techniques of power have encouraged citizens to adopt new practical relationships with themselves and their government based on neoliberal values (Donzelot, 1979). In their attempts to decentralize, governments have allowed individuals and organizations to play a more active role in the operation and administration of government objectives (Donzelot, 1979). The price of this involvement, however, is that citizens must assume active responsibility for these activities.

In fulfilling this role, they are required to conduct themselves in accordance with the appropriate model of action, which includes being self-sufficient, autonomous, calculating and responsible

(Brodie, 1996). Based on this relationship, individuals are urged to become responsible, self- governing citizens who are able to protect themselves and their families from the harms associated with risky conduct.

Biopolitics or biopower, which refers to the “governmental interest in the regulation of individuals and populations,” has become an important tool in identifying and managing those who are unable to conduct themselves in accordance with social norms (Murray, 2007, 163).

39

This is accomplished by obtaining statistical information on the citizenry for the purposes of risk management and security (Ericson and Haggerty, 1997). At the centre of biopolitics is the bionic citizen, an individual so rational, that s/he can regulate his/her conduct by calculating the risks of inappropriate behaviour (Isin 2004; Rose, 2000). In neoliberal societies, the bionic citizen has become a model for active citizenship, as individuals are urged to take responsibility for themselves and their families by minimizing risks in everyday life (Rose 2000). Based on this construction, common technologies of the self include self-regulation, responsibilization, and rationality (Braithwaite, 2000).

This power, as argued by Foucault (1980, 221) “is exercised over free subjects, and only in so far as they are free;” however, this freedom only goes so far as to allow citizens the chance to act in accordance with societal norms and government rationalities. Those who digress from these expected standards become stigmatized for falling outside the nexus of appropriate behaviour and are targeted through disciplinary practices. In this regard, active citizenship is highly exclusionary as those who cannot fulfill the requirements of citizenship are often viewed as the underclass. As a result, the shift from Keynesian to neoliberal frameworks has prompted the development of new political rationalities that project clear distinctions between a majority of people who “can and do ensure their own well being and security through their own active self promotion and responsibility for themselves” and those who cannot (Rose, 2000: 196). Those who fail to employ techniques of the self are typically viewed as ‘non-citizens’ who are incapable of taking responsibility for their own conduct. These individuals are usually the most socially and economically disadvantaged, which include visible minorities, the unemployed, and

(single) women (Madriz, 1997; O’Malley, 1996). Although power is exercised over free people,

40

those who fail to embody appropriate techniques are problematized for their shortcomings. Their inability to take responsibility and control over their situations classifies them as risks.

Consequently, bail decisions may act as a form of targeted governance that aims to responsibilize individuals who fall outside the perceived standards of acceptable behaviour.

Here, acceptable behaviour is not only determined by whether an accused is an ideal citizen, but also by whether s/he embodies proper gender norms. When examining bail decisions, the conditions placed on a recognizance order often attempt to instill a sense of self-control amongst accused individuals (Trotter, 2010). Popular bail conditions include treatment or counselling, abstinence from drugs and alcohol, and curfews (Myers, 2009; Trotter, 2010). This is similar to parole conditions that aim to produce “normative, self governing subjects” by attempting to target root causes of crime by implementing individual-based conditions that will break the cycle of criminal tendencies (Turnbull and Hannah Moffat, 2009). What is considered acceptable behaviour for women, however, is not always the same for men as the courts routinely judge an accused woman based on traditionally feminine traits, such as motherhood (Boyd, 2006).

Women who transgress these ideals are viewed as being out of control. In an attempt to shift behaviour, JPs may order women to seek counselling or treatment as part of their release.

Women who are seen as unable to govern themselves with counselling or treatment may be remanded in an effort to exert further control. Although ideals of masculinity, such as employment, influence bail decisions for men, men are normally seen as risky because of legal factors, such as offense seriousness and prior record, resulting in detention.

Summary of Theoretical Frameworks

Overall, this section outlines four possible theoretical approaches related to gender and bail. While all are conceptually different, they nevertheless stress the underlining role of risk in

41

the criminal justice system. The first two theories highlight the role of penal populism in criminal justice, which helps account for increased remand rates and the use of harsher bail conditions in release decisions for women and men alike. Based on recent legislative changes favouring public confidence, bail may be more difficult to obtain because JPs are trying to protect community interests at the expense of accused rights. Alternatively, it is possible that JPs are simply avoiding making decisions that will risk the court’s reputation and legitimacy. This culture of risk aversion means that JPs may be less likely to grant bail for certain groups of men and women in order to minimize secondary risks.

The last two theoretical perspectives, risk management and governmentality, provide clarity regarding potential gender disparities in bail outcomes by outlining how risk is defined and managed by the courts. Building off the previous two theories, these approaches help to explain the possible effects of gender on bail decision-making. Risk management helps explain why certain men and women are considered more risky than others. Based on current understandings of risk models in criminal justice, risk is defined based on an individual’s membership to a population subgroup with known risk levels. Typically, those who are considered more risky are those who suffer from social exclusion. Women who commit crime are often viewed as being more risky for both their involvement in crime and their digression from socially accepted female behaviour, which may lead to increased remand rates or stricter release orders than in the past. Governmentality, on the other hand, describes the processes used to shape individual behaviour. When taken in the context of bail, governmentality highlights how certain decisions may be used to influence individual conduct. In many respects, this process relies on society’s understanding of masculine and feminine roles, accounting for differential

42

treatment of accused men and women. In sum, criminological and sociological theories provide competing expectations regarding the role of gender on bail outcomes.

2.4 Research Objectives

The purpose of the current study is to explore whether and why gender disparities exist in bail outcomes. The extant literature leads us to expect that accused women are more likely to be granted bail relative to men, particularly if they have familial responsibilities. Women who deviate from this norm, on the other hand, are often viewed as being a greater risk, which can result in either restrictive releases or detainments. While the literature acknowledges the role gender plays when determining bail, gaps remain. Specifically, many of the studies discussed thus far (see: Ball and Growette-Bostaph, 2009; Bernat 1984; Bottoms and McClean, 1976;

Demuth and Steffensmeier, 2006; Hucklesby 1996; Turner and Johnson, 2006) fail to account for important gender differences that shape how men and women enter into crime, the contexts in which they are likely to offend, and the variance in the rates and types of offending behaviour

(Boritch 2009; Daly 1994; Maidment 2009; Player 2007). Without a clear sense of how such factors shape women’s offending, it becomes difficult to conclude whether (and why) gender disparities truly exists in bail decisions (Daly 1994; Gartner et al. 2009). This is further compounded by the fact that minimal Canadian research exists that examines the influence of gender on the types of releases imposed, which makes it difficult to conclude that women are granted leniency (as some scholars would attest - see Ball and Growette-Bostaph, 2009; Jeffries et al., 2003; Kruttschnitt, 1984; Kruttschnitt and McCarthy, 1985; Visher, 1983) by the courts.

As such, this study addresses these gaps by analyzing the extent to which gender shapes decisions to grant or deny bail and the types of releases imposed by comparing men and women

43

charged with similar offenses. This is accomplished by addressing the following research questions:

1.a) In cases in which bail is granted, what are the justifications provided by the justice of the peace? 1. b) To what extent are the justifications for granting bail similar/dissimilar across gender and why? 2. a) In cases in which bail is granted, what types of conditions are placed upon the accused? 2. b) To what extent are the conditions attached to bail similar/dissimilar across gender and why?; 3. a) In cases in which bail is denied, what justifications are provided by the justice of the peace? 3. b) To what extent are the justifications for denying bail similar/dissimilar across gender and why? 2.5 Conclusion

To summarize, I began this chapter by discussing factors that may influence bail differently for accused men and women. Next, I delineated the ways in which sociological and criminological theoretical frameworks can be drawn upon to provide expectations about the possible connections between gender and bail. I noted that while a body of literature has investigated the relationship between gender and bail, in its present state, the nature and direction of this relationship is unclear and in need of further examination, especially in the Canadian context. With this in mind, I proposed how the current study will build upon and extend what is known about the gender-bail relationship by addressing three main research questions. I turn now to Chapter Three in which I discuss the data, sample, and methods used to answer my research questions.

44

Chapter 3: Data and Methods

3.1 Introduction

This chapter describes the data, sample, analytic strategy, and key measures used to examine the relationship between gender and bail outcomes. It begins with an overview of the observational method used to collect the data and the resulting sample. Next, this chapter describes the three main analytic tools included to measure the role gender plays in the determination of bail: thematic analysis, logistic regression, and qualitative deep sample analysis.

Finally, the limitations of this methodological approach are discussed.

3.2 Data Collection

Courtroom observations

To assess the role gender plays in the determination of bail, I collected data by observing bail hearings at a provincial courthouse located at 200 Frederick Street in Kitchener, Ontario.

Geographically speaking, Kitchener is located in Southwestern, Ontario, roughly 120km from

Toronto, and is home to 444, 681 people, of whom 218,860 are male and 225, 830 are female

(StatsCan 2011).11 Observations were carried out on two different days per week throughout the months of May to August 2012. In general, court hearings are fully accessible to all members of the public, including researchers, and thus research ethics board approval was not required for this observational study (TriCouncil, 2010).

In order to examine how gender shapes bail outcomes, I required data with several key characteristics. First and foremost, the data needed to contain sufficient variation in the number of cases in which bail was granted versus denied, including cases heard in contested versus

11 The median age of Kitchener residents is 37 years old (StatsCan 2011). 45

regular bail court. Secondly, the data needed to contain sufficient variation in the gender of the accused, meaning I needed cases involving males and cases involving females. While this may seem like an obvious requirement, it is nevertheless a challenging data issue because men comprise the vast majority of court cases. Finally, prior research suggests that certain legal and extra-legal characteristics influence the probability of being granted bail and need to be statistically controlled for when estimating the relationship between gender and bail outcome

(Daly, 1994; Demuth and Steffensmeier, 2004). Information on such characteristics, including offense seriousness, prior record, and race, was thus necessary in order to assess this relationship.

At the Kitchener courthouse, bail hearings occur in either regular or contested court depending on the position of the Crown. If the Crown Attorney decides to dispute an accused’s release, the hearing will be scheduled in contested court, which is typically reserved for more serious charges. On average, this court hears an estimated three full hearings per day and the JP decides the outcome of the case. Otherwise, all additional matters related to bail, including adjournments, assignment of legal aid and video remands, are heard in regular bail court, which generally results in the accused being released on bail. It is very rare for the JP to disagree with a consent to release term put forth by the Crown in regular court. The JP does, however, have the ultimate say in the types of conditions placed upon an accused. Because this study focuses on the circumstances/justifications around decisions to both grant and deny bail, observations took place in both courtrooms, alternating each day. Based on the differences in Crown positions, it was suspected that observations made in regular court would most likely yield data pertaining to release conditions whereas observations made in contested court would yield data on both releases and remands.

46

In order to compare judicial decisions based on the accused’s gender, a detailed checklist was used to ensure the systematic collection of data for all cases (see Appendix- Figure A-1).

Modelled on similar studies (notably, Allan et al., 2005), the checklist included questions on legal and extra legal factors, Defense and Crown submissions, as well as judicial decisions. It was structured to allow for both closed and open responses, which provided both quantitative and qualitative data for each case. According to Gainey et al. (2000), this type of research design yields more rich and detailed information than a purely quantitative or qualitative piece.

Specifically, while quantitative data on legal and extra legal factors may provide valuable findings, they often fail to offer any context for understanding the potential similarities and differences in bail outcomes and the rationales for these outcomes.

The benefit of this time-intensive data collection process is the potential to acquire rich information that will allow for a more thorough understanding of judicial decision-making concerning bail (Allan et al., 2005). Unlike existing aggregate data or courtroom transcripts, observational research provides an opportunity to examine factors that may influence bail decisions within the courtroom setting. Further, all courtroom transcripts are manually recorded by court reporters in Ontario. To date, the Ministry of the Attorney General (2012) has yet to transition to a paperless, electronic process to organize and store court files. This means that transcripts are normally filed off-site due to space restrictions and may not contain all the relevant information needed for a particular study. While access to transcripts and files are normally free, photocopying fees are extremely expensive and may require the researcher to travel to specific destinations to obtain the necessary paperwork. Due to these restrictions, observational research was deemed the most effective data collection tool.

47

3.3 Sample

In total, 118 cases were observed at the Kitchener Provincial Courthouse during the months of May to August 2012. Of the 118 cases observed, 22 cases were female and 96 cases were male. Three cases were later discarded from this study because they did not have a specified bail outcome. This occurred because the presiding JPs adjourned the case to a later date to provide time to assess their decisions. All discarded cases were males. Names of the defendants, complainants, lawyers, and JPs were omitted from this study to maintain anonymity.

Although this information is fully accessible to the public, it did not add any real value to this study. Instead, cases were numbered 1 through 115 for organizational and reference purposes.

Table A-1 (see Appendix) provides the descriptive statistics on the men and women included in my sample. Overall, 85 percent of cases resulted in a release while 15 percent resulted in detention. Thirty percent of accused individuals had no criminal record while nearly

70 percent had a criminal record. Sixty-three percent of my sample was charged with one or two offenses while 35 percent was charged with three or more offenses. For charge seriousness, over half of accused individuals were charged with a minor crime, roughly one-third were charged with a somewhat serious crime12, and fewer than 10 percent were charged with a serious crime.

Thirty-seven percent were charged with a combination of offenses13, 32 percent were charged with a crime against the administration of justice, 16 percent were charged with a violent crime, nine percent were charged with a property crime, and six percent were charged with a drug- related offense. Of all the cases, 44 percent were a reverse onus situation, meaning the accused had to show cause for release. Drugs and/or alcohol were a factor in 35 percent of cases and

12 Although this is not a legal category, it is used in the present study in order to capture the varying degrees of offense severity. 13 This category refers to individuals who were charged with two or more different types of offenses, such as a failure to comply and possession of a controlled substance. 48

weapons were involved in only nine percent of cases heard. There was at least one victim involved in 43 percent of cases, and 27 percent of victims were female.

When examining characteristics pertaining to the surety, it is apparent that at least one surety was present in 70 percent of cases, Youth in Conflict came forward in 20 percent of cases, and 10 percent of cases had no form of supervision.14 Of the cases that had a surety, 33 percent of sureties reported having a strong relationship with the accused. Forty-four percent of sureties were female and just over half of those were cases where the accused’s mother came forward as a surety. Eighty-seven percent of women and 91 percent of men had the support of either a surety or Youth in Conflict with the Law.

The sample shared similar demographic characteristics to other offender populations. For example, sixty-eight percent of accused persons were white while 32 percent were considered a visible minority. More than half were between the ages of 18-29, roughly one-quarter were between the ages of 30-39 years, and less than 20 percent were 40 years or older (three percent was unknown). When looking at employment status, 39 percent of accused individuals were unemployed, 35 percent were employed either full-time or part-time, and two percent fell in the other category (i.e. retired, parental leave). Employment status was not discussed in 24 percent of cases. Thirty-eight percent of the sample had a problem with either drugs or alcohol and 17 percent had a mental health issue.15 The majority of accused individuals were single, and 37 percent reported having children.

When looking at the demographic differences in accused population based on gender, 34 percent of accused men were considered a visible minority compared to only 23 percent of

14 Of those with a surety (81 in total), 85 percent were granted bail and 15 percent were denied. 15 In many cases, it was not clear whether mental health assessments were based on police opinion, police interviews, or self-reports. 49

accused women. Of the men in my sample, 46 percent were employed either full-time or part- time while 51 percent were unemployed. Accused women were employed in 44 percent of cases and unemployed in 56 percent of cases. Nearly 60 percent of accused men had a known drug and/or alcohol issue while 50 percent of accused women had similar problems. Twenty percent of both accused men and women had a mental health issue. Accused men were married in 20 percent of cases while accused women were married in 44 percent of cases. Forty-four percent of men reported having children while 80 percent of women stated they had children. With these differences in mind, I will now proceed by outlining my analytic strategy.

3.4 Analytic Strategy

Using a mixed methods approach, the current study relied on a concurrent triangulation strategy. This often yields more complete results than a monomethod design since it maximizes the strengths and minimizes the weaknesses of quantitative and qualitative approaches and allows for a broader understanding of the topic under study (Cresswell 2009). Following

Cresswell’s (2009) guidelines, quantitative and qualitative data were integrated by transforming qualitative themes into counts, which were then compared with the descriptive quantitative data collected. This was accomplished using a three-pronged analytic strategy: thematic analysis, logistic regression, and a deep sample qualitative analysis.

3.4.1 Thematic analysis

First, the qualitative data collected through open-ended responses on the bail checklist were analyzed using thematic analysis as modelled by Braun and Clarke (2006) (see Appendix-

Table A-2). According to Braun and Clarke (2006), thematic analysis is a “poorly demarcated, rarely acknowledged, yet widely used qualitative analytic method” (77). This method allows researchers to identify, analyze, and report themes within data and is compatible with both

50

positivist and constructionist paradigms in the social sciences (Braun and Clarke 2006). For this study, themes were defined as such if they captured “something important about the data in relation to the research question,” and represented “some level of patterned response or meaning within the data set” (Braun and Clarke 82). Themes about the justifications for granting and denying bail, as well as the types of conditions placed on the accused, were recorded based on their prevalence within the data set.

To begin, each case was systematically reviewed for the presence of any statement that justified a JPs decision to grant or deny bail. These statements were recorded verbatim from the bail checklist and compiled into two lists: justifications for granting bail and justifications for denying bail. Because many of the justifications were similar across cases, counts were recorded every time a similar statement was made by a new JP to support his/her decision (only 1 count was recorded per decision even if the JP used the statement several times). New justifications or statements were simply added to the list. Once all statements were listed, they were categorized based on major themes (so those with the highest count). Results were then recorded in table format. Each category received a total count, which was the summation of individual statement counts. Using the total counts, categories were organized based on most commonly stated justification category to least commonly stated justification category. Although this was a relatively simple process, it was nevertheless important because it highlighted the range of justifications used to grant or deny bail, as well as the frequency with which they were used.

An identical approach was used to address the types of bail conditions placed on an accused if released. As described above, bail conditions were recorded and counted in all cases that resulted in a bail release. Again, once all conditions were accounted for they were placed

51

into several categories, which were later organized based on the most frequently used category of conditions to the least frequently used category of conditions.

Key Measures

Using existing research as a guide, justifications for granting bail were placed into the following five categories: justifications related to the surety, justifications related to the accused, justifications related to additional levels of supervision/community support, justifications related to the allegations, and justifications related to the victim. The first category used was justifications related to the surety. As stated by Myers (2009), being released under the control of a surety has now become the norm for bail releases in Ontario. In her study of eight bail courts across Ontario, she found that several courts automatically required all accused individuals to have a surety without considering less onerous release options (Myers 2009). Based on this finding, justifications that specifically related to the surety were placed into this category. This included judicial statements such as “surety clearly understands his/her role,” “surety is unemployed and can offer 24hr supervision,” and “surety has a good relationship with the accused.” In addition to surety releases, existing research demonstrates that both legal and extra- legal factors related to the accused also play a large role in the determination of bail (Goldkamp and Gottfredson 1985; Gottfredson 1974; Kellough and Wortley 2002; Mauer 2000; Player 2007;

Roberts and Doob 1997). Accordingly, justifications related to the accused were used as a major category when organizing themes for granting bail. Here, any judicial statement that explicitly referred to the accused was placed in this category. In many cases, this included statements about both legal and extra-legal characteristics like, “the accused had a limited criminal record,” “the accused has taken steps to get better on his/her own,” and “the accused is employed.”

52

A third category was created to reflect justifications relating to additional levels of supervision/ community supervision. This category captured judicial statements that discussed any form of supervision that went beyond the use of a surety such as “the accused is bound by additional bail orders,” “Youth in Conflict will be supervising,” or “the accused will be living under the rules of a shelter.” Further, the category referring to justifications related to the allegations included statements that reflected the current allegations or charges, like “allegations are not serious enough to warrant detention.” The last category, justifications related to the victim, included any statement that referred to the potential involvement of a victim such as “the complainant gave the accused mixed messages.” As an exception, seven cases existed where the

JP failed to provide any justifications for granting bail. This only occurred in consent to release situations when the accused had no prior record and was charged with a minor offense. These cases were placed in their own category titled ‘no justification given.’

Similar to the above, justifications for denying bail were placed into five categories: justifications related to the victim, justifications related to the accused, justifications related to the surety, justifications related to the allegations, and justifications related to public confidence.

These categories were similar to those used for granting bail but contained opposing statements such as “accused has lengthy criminal record,” “surety does not understand his/her role,” and

“the accused is unemployed.” Another difference was the role of public confidence in denying bail. While this did not emerge as a major theme, it has been suggested that the increase in remand rates can be explained, in part, by the introduction of sec. 515(10)(c) (McLellan, 2011).

As such, this category included statements like “detention is necessary to maintain confidence in the administration of justice.” Unlike justifications for granting bail, all cases that were denied bail had some form of reasoning stated.

53

Bail conditions were placed into six categories: conditions related to communication, mobility, residence, curfew/house arrest, additional levels of supervision, and treatment or counselling. Existing research demonstrates that while bail conditions should be as least restrictive as possible, JPs generally do not follow this guideline (Myers 2009; Sprott and Myers

2011; Trotter 2010). Sections 515(4) to (4.3) of the C.C.C. provide a list of enumerated conditions that can be reasonably placed on a release order. These include conditions like forfeiting a passport or reporting to a peace officer at a designated time. However, s.515(4)(f) allows a JP to order any condition s/he sees fit so long as it is “reasonable”. These often include more restrictive terms than those listed in s. 515(4)-(4.3) like curfews, house arrests, and limits on drug/alcohol consumption and driving (Trotter 2010). Given the discretionary nature of this process, categories were made as specific as possible to determine what types of conditions were used most often and whether they were enumerated under the C.C.C.

Additional themes

In order to integrate the quantitative and qualitative data collected using the bail checklist, thematic analysis was also used to identify additional themes that were not highlighted in the previous steps. Unlike the process used to locate justifications and conditions, this analysis was much broader in scope and captured underlying assumptions and ideologies that might have influenced judicial decisions. Because of this, a latent approach was used to analyze themes that were viewed as important in relation to the overall research question. It was during this process that additional themes emerged regarding the accused, the complainant, the surety, and the context surrounding the allegations. These themes were recorded and included with those identified previously.

54

3.4.2 Logistic Regression

The themes produced by the foregoing analysis were transformed into statistically testable variables and inputted into SPSS along with the quantitative data obtained from the bail checklists. Given my goal of modeling the predictors of granting versus denying bail, a logistic regression modeling strategy was performed in order to measure the effect of gender on the bail decision-making process. In logistic regression, the value predicted represents a probability between 0 and 1 that indicates the likelihood that something will happen (McLachan 2004 and

Wickens 1989). It was chosen as an appropriate analytic technique because it allows researchers to use a dependent variable with only two values (ie. granted or denied bail) (Mallery 2011).

Similar to the analysis used by Allan et al. (2005) to determine which variables made a statistically significant contribution to granting or denying bail, a logistic regression model was used in this study to determine whether gender was a significant factor in the likelihood of bail and what additional factors were significantly likely to affect bail decisions. Unknown cases were coded as missing and not included in the analysis, which was carried out using SPSS (see

Appendix- Table 9 for missing data).16

Key Measures:

Dependent variable

Bail outcome. For the quantitative design, the dependent variable used to measure the effect of gender on bail was bail outcome. In this study, bail outcomes were limited to two options: granted or denied bail. Individuals who were denied bail were coded as 0 while those who were granted bail were coded as 1. This design did not include outcomes stemming from

16 While the proportion of missing data limited the types of analyses that could be performed using certain variables (i.e. whether the accused had children and the employment status of the accused), it could suggest that the court was not as preoccupied with knowing this type of information as the literature would suggest. 55

adjournments, since these always resulted in temporary detainments and did not involve a complete hearing.

Independent variable

Gender of the accused. The independent variable was the gender of the accused. As discussed in the literature review, existing research suggests that women are more likely to receive bail if they embody “proper” (i.e., traditional) female characteristics while those who reject notions of femininity are more likely to be denied (ie. those who are deemed to be bad mothers) (Player 2007; Turner and Johnson 2011). In comparison to men, however, women are generally more likely to be released (Freiburger and Hilinski 2010; Kellough and Wortley 2002).

In this study, males were coded as 0 and females were coded as 1.

Control variables

When examining factors that affect bail outcomes, some scholars suggest that extra-legal characteristics of the accused, such as age, social class, and race play a role in the determination of bail (Kellough and Wortley 2002; Kruttschnitt 1984; Player 2007; Roberts and Doob 1997), while others argue that these factors lose their predictive value once legal factors are controlled for (Allan et al. 2005; Willbanks 1996). Legally, JPs can only reasonably deny bail when the accused is considered a flight risk, a risk to public safety, and/or a threat to public confidence in the administration of justice. Given the inconsistency in findings relating to bail, this study controlled for six legal and extra-legal factors identified in previous research as potentially influencing the bail decision-making process.

56

Legal factors

Criminal record. Criminal record has been shown to play an influential role in determining bail (Allan et al., 2005; Brown et al., 2004; Willbanks, 1987) because it is used as a proxy for future behaviour (Player, 2007). In this study, it was coded as 0= no criminal record and 1= has a criminal record. Cases in which the criminal record was unstated were coded as missing.

Offense Seriousness. Prior research has identified offense seriousness as a significant predictor of bail (Allan et al., 2005; Goldkamp and Gottfredson, 1985). Accordingly, it was used as a control variable in this study. To help code for this, the Crime Severity Index from the

Uniform Crime Reporting Survey was used as a guide. Based on the Index, minor offenses include charges like theft under $5000, breach of recognizance, mischief under $5000 and assault level 1. Charges related to uttering threats, weapons possessions, and sexual assault level

1, for instance, are considered somewhat serious offenses while charges like robbery, assault level 3 and are deemed most serious. With this in mind, charge seriousness was coded as

1= minor, 2= somewhat serious, and 3= serious.

Bail supervision. According to Myers (2009), accused individuals are being released under the control of a surety at an increasing rate, as JPs in Ontario are frequently requiring an accused to have a surety prior to release. Given the emphasis on the level of bail supervision, this study controlled for surety availability. As stated in the literature review, a surety is a person who guarantees the accused will attend all court appearances, follow all the conditions of his/her release, and not reoffend. In other words, they act as a civilian jailer. This gives the court more confidence that the accused will complete his/her bail successfully. Accordingly, surety availability was included as a control variable in this study and was coded as 0= no, 1= yes, 2=

57

Youth in Conflict with the Law. In Kitchener, Youth In Conflict With the Law is a bail supervision program that helps eligible individuals locate a surety following their arrest or acts in a supervisory role when a surety is absent. Under the umbrella name, Waterloo Region Bail

Program, Youth in Conflict supervises accused individuals of all ages, including adults.

Extra legal factors

Characteristics of the accused. As discussed in the literature review, studies have shown that extra-legal factors pertaining to the accused have a statistically significant effect on bail outcomes (Boritch 1997; Kellough and Wortley 2002; Kruttschnitt 1984; Mauer 2000; Roberts and Doob 1997). Accordingly, both race of the accused and whether the accused had children were controlled for in this study.17 Research has demonstrated that race and/or ethnicity influences judicial decisions to grant or deny bail with black or Hispanic individuals being more likely to be denied bail when compared to white individuals (Demuth and Steffensmeier 2004;

Kellough and Wortley 2002; Roberts and Doob 1997; Williams 2002). In this study, race was almost entirely assumed based on visual inspection of the accused and his/her last name.

However, because of the limited variability in racial backgrounds within the sample, it was appropriate to code race as a dichotomous variable in order to increase the number of cases per category. Thus, it was coded as follows: 0= white and 1= visible minority.

Whether the accused has children has also been shown to influence bail decisions, making it an important control variable. When asked, a group of prosecutors and JPs stated that they would be less likely to deny bail if the accused had children (Brown et al., 2004). Moreover,

17 While it is recognized that employment is also an important control factor when assessing the relationship between gender and bail, it was not used in the current study due to a number of missing cases (employment status was not discussed in over one quarter of cases). Given the already limited sample size of the regression analysis, adding employment status as a control would have decreased the power of the overall logit model and made it difficult to obtain statistically significant results. This is a limitation of the current study. 58

the group of JPs claimed they would be even less likely to deny bail if the accused was a single parent (Brown et al., 2004). While both groups maintained that gender would not matter, the majority of women who enter the criminal justice system are the primary caregiver of their children (Edgar 2004). In this context, children was simply coded as either 0= accused had no children and 1= accused had children. Unknown cases were coded as missing.

Characteristics of the victim. According to Daly (1994), a major drawback of quantitative studies on sentencing is the lack of information pertaining to the victim. Although her study focused on post-conviction sentencing, she found that victim characteristics such as gender and age were important factors in these decisions (Daly 1994). Victim involvement in the allegations was used as a control variable in the current study. Generally, the Crown presented information about the victim during the reading of the allegations. In some cases, however, victim information was either not included or not applicable (i.e. in failure to comply charges). These were simply coded as missing. Further, all information regarding the victim was based solely on the current allegations before the court and not on any outstanding charges the accused might have had. Therefore, this variable was coded as 0= no victim involved and 1= victim involved.

3.4.3 Exploratory analysis: Deep sample method

In addition to the quantitative design, a deep sample qualitative analysis was used to further assess the extent to which bail outcomes were similar or different for men versus women.

This approach matched a smaller sample of 19 pairs of men and women with similar charge and background characteristics. The analysis was guided by the findings of the logistic regression and major findings in the literature that did not emerge as significant in the logistic regression.

This approach was modelled on a similar study completed by Daly (1994), which looked at sentencing differences between men and women. Based on Daly’s (1994) experience, this

59

approach allows for a deeper understanding of the role gender plays in judicial decisions because all other potential explanatory factors are held constant. An additional benefit is that it allows the researcher to take into consideration important differences in the way men and women enter into the crime, which might also influence the judicial decision-making process. While this process allowed for a more nuanced assessment of whether men and women were treated similarly by the court via the outcomes of each case, it was very preliminary in nature. Thus, the themes identified using this technique are not definitive; rather, they are intended to generate discussion about the effects of gender on bail and to identify potential avenues for future research.

Matching the accused

The matching process began by grouping all female cases together since they were the reference group in this analysis. To make the process more manageable, all male cases were organized based on age groupings (18-29, 30-39, 40-49, 50-59, 60+). Using both the results of the logistic regression and existing research findings, the following selection-decision protocol was used to match men and women in the study:

1) charge characteristics (similarities in content, context, and severity) 2) prior record 3) age 4) employment and residence status 5) mental health issues 6) drug and alcohol problems 7) relationship and family status 8) surety available

60

When male-female pairs had similar charge characteristics, those with similar prior records were chosen.18 If men and women were charged with similar offenses and had the same criminal record, those who had the closest ages were selected, and so forth. Throughout the process of selecting cases, bail outcomes were not revealed. This process was used until all 19 female cases had a male matched pair. Table 1 demonstrates descriptive results of the matching process.

Once the cases were matched, the data were re-analyzed using thematic analysis to locate any differences/similarities in outcomes that may be due to gender and to provide possible explanations. The flexibility of thematic analysis means it can be used to reflect reality or to understand subjective meanings of that reality (Braun and Clarke, 2006). For this study, this method was used in order to better understand how risk was constructed within bail court through the types of conditions and justifications given by the judge when an accused is granted or denied bail. Given the subjectivity of this topic, thematic analysis was used as a constructionist method in order to examine the underlining discourses that might have influenced the court’s construction and application of risk across gender.

18 This was based upon similarities in quantity, context, and severity. 61

Table 1: Qualitative Female-Male Matched Pairs, Kitchener Provincial Courthouse, 2012 (N= 19)

Matched Pseudonyms Type of charge Severity Prior Age of Employment Resident Mental Drug/ Relationship Children Surety pair for accused (F/M) of criminal accused status (F/M) status health Alcohol status (F/M) (F/M) (F/M) identificati (F/M) charge record (F/M) (F/M) issues Issues on number (F/M) (F/M) 1 Lynn/Don Fail to attend & Low Minor 30/23 Employed/ Owns No/ No/ Not Single/Single Yes/Not Youth breach probation/ Unemployed home/ Not stated stated in theft under & but has Lives stated conflict/ breach probation prospects with Youth parents in conflict

2 Hannah/ Danger to public High None 20s/20s Employed/ Rents No/No No/ Yes Single/Single Yes/No Mother/ Simon peace, uttering Not stated home/ Mother threats of death, Lives assault, assault with with a weapon & mother possession of weapon/ Assault (x3), uttering threats (x3), mischief under 3 Barb/Harold Possession of Low Minor 30/23 Unemployed/ Rents No/No Yes/No Single/Single No/No Fiancé controlled Employed home/ & substance/ Rents Fiancé’s Possession of home daughte controlled r/ Father substance & possession with the intent to traffic 4 Brenda/ Breach Low Minor 30s/30s Unemployed/ Lives No/No No/Yes Single/ Not No/Not None/ Mateo probation/ Unemployed with t stated stated stated Sister Breach probation mother/ Lives with 62

family

5 Jessica/Mark Fail to attend/ Low None 28/25 Unemployed/ Not Yes/ No/No Single/ No/No Youth Fail to attend (x Not stated stated/ No Common law in 3) Not conflict/ stated Friend 6 Carol/Vince Failure to Low None 36/26 Employed/ Owns No/No Yes/ No Common Yes/No Mother/ comply/Failure Employed home/ law/ Single Mother to comply Owns home 7 Alana/ Bob Assault & Medium None 23/28 Not stated/ Rents Yes/ No/ No Common Yes/ Youth uttering threats Unemployed home/ Yes law/ Yes in (x2)/ Assault Rents Common law conflict/ (x3) & mischief home Mother under 8 Donna/ Possession of Low Minor 50s/30s Employed/ Owns No/No No/No Married/ Yes/ Youth Wally counterfeit Employed home/ Single Yes in money (x 4)/ Fail Rents conflict/ to comply & home None theft under 9 Cathy/Glenn Theft under & Medium Medium 50s/40s Unemployed/ Rents No/ Yes/Yes Not stated/ Not Youth trespassing/ Unemployed home/ No Single stated/ in Failure to Rents No conflict/ comply home Youth in conflict 10 Paula/James Assault (x2) & Medium Severe 27/24 Not stated/ Other/ Yes/ Yes/Yes Single/ Not Yes/No Friend/ breach probation Employed Lives No stated Father (x3)/ Breach with probation, failing family to stop & resisting arrest

63

11 Tiffany/ Breach probation Low Minor/ 20s/20s Employed/ Rents No/No Yes/Yes Common No/No Father/ Graham & failure to None Employed home/ law/ Single Mother comply/ Failure Lives to comply with parents 12 Yolanda/ Assault & assault Low None/ 40s/50s Not stated/ Rents No/ No/No Single/ Yes/Yes Friend/ Henry with a weapon/ Minor Not stated home/ No Divorced Youth Assault Rents in home conflict 13 Catrina/ Theft under (x2) Low None/ 30s/30s Not stated/ Rents Not Not Not stated/ Not None/ Harry & Theft under Minor Unemployed home/ stated/ stated/ Single stated/ None Lives in No Yes No shelter 14 Geraldine/ Theft under & Medium None 40s/20s Unemployed/ Lives No/ Yes/Yes Married/ Yes/ Father/ John failure to Unemployed with Yes Married Yes Mother comply/ Assault, father/ theft under, & Lives resisting arrest with family 15 Maureen/ Possession of Low None/ 30s/26 Unemployed/ Lives in No/ Yes/ Not Not stated/ Not Youth Mario controlled Minor Unemployed shelter/ No stated Not stated stated/ in substance (x4) & Lives in Not conflict/ Possession of shelter stated None stolen property/ Theft under, possession of stolen property 16 Farah/Carl Possession of Medium None/ 20s/20s Unemployed/ Other/ Yes/N Yes/Yes Single/ Not Aunt/ controlled Minor Unemployed Lives o Single stated/ Friend substance, with Not possession with friends stated intent to traffic & failure to comply/ Possession of controlled substance & possession with 64

intent to traffic

17 Holly/ Possession of a Low Minor 20s/20s Unemployed/ Lives No/No Yes/No Single/Single Yes/ No Mother/ Samuel controlled Unemployed with Mother substance & fail friends/ to appear/ Lives Possession of a with controlled family substance & breach of probation 18 Paige/Mike Assault, assault High Minor/ 30s/30s Unemployed Other/ No No/No Common Yes/ Friends/ with a weapon, Severe (on maternity Rents (but law/ Single Yes Friend uttering threats, leave)/ home suspici breach probation Employed ons)/ & fail to comply/ Yes Assault with a weapon, mischief under, utter threats with a weapon and to cause death 19 Sarah/Tyler Failure to Low None 40s/40s Not stated/ Rents No/No No/Yes Married/ Not Not Mother/ comply/Failure Employed home. stated stated/ Brother to comply Other Not stated

65

3.5 Limitations of the analytic design

While the analytic strategy employed in this study adequately addresses the main research questions, it nevertheless has several limitations that warrant mention. First, this study did not produce a random sample. This is not to say that obtaining a random sample is impossible with this method (see Sprott and Myers (2011)), it was just not practical given the scope of this project. As seen in Sprott and Myers’ (2011) study on youth bail conditions, a random sample can be achieved by observing hundreds of cases and using a table of random numbers. Unfortunately, this is beyond the range of this study due to the relatively small sample size (N=115, in comparison to Sprott and Myers’ N= 225). Second, this study was not geographically representative of bail courts across Ontario since observations took place at the

Kitchener Provincial Courthouse only. This limited the generalizability of the findings due to the possibility of variation in judicial bail decisions across the province.

Third, the dataset used contained a small number of accused who were denied bail, which posed difficulties when analyzing the data. For example, the results of the logistic regression were limited because 17 cases resulted in detention, of which only one was female. This meant that without adding gender as the independent variable, the logistic model correctly predicted bail outcomes 82.5% of the time. A larger dataset would have likely yielded more diverse cases.

An additional drawback of the logistic regression modelling technique used for this study was that it did not show whether the control variables used affected bail similarly for men and women. In other words, being a parent of dependent children might be significantly related to bail release for women but not men, whereas being employed might be a significant predictor of

66

bail release for men but not women. This type of analysis would have required a larger sample size that was not feasible in the current study.19

Lastly, the deep-sample qualitative analysis matched men to women based on female criminogenic characteristics. This might have skewed the results slightly since research shows that women are more likely to be charged with less serious offenses and are less likely to possess lengthy criminal records. While feminist sociolegal scholars tend to dismiss comparative studies because “women are measured in relation to the master category, men,” the same concern can be said for comparing men to women. With this in mind, pairing accused men and women proved to be quite challenging since the men in my sample were more likely to have a serious criminal record and past failures to comply. Further, the contexts in which the men and women allegedly committed an offense also differed, making it difficult to conclude with certainty that gender was the key factor shaping bail decisions.

3.6 Conclusion

To summarize, the purpose of this chapter was to outline the methodological approach taken to assess the role of gender on the determination of bail. The data obtained over the course of four months provided a final sample of 115 bail cases, which were analyzed using a mixed methods approach. Overall, three analytic methods were used, including thematic analysis, logistic regression, and a deep sample qualitative analysis. The purpose of each analytic strategy was discussed in detail along with the key measures used. This chapter also discussed the limitations of analytic strategy. While this section outlines some methodological limitations, the

19 This type of analysis would require splitting the sample based on gender, and running separate regressions for both men and women. 67

framework presented in this chapter effectively addressed the objectives of the study.

Accordingly, the following chapter will present the findings of this approach in detail.

68

Chapter 4: Results

4.1 Introduction

This chapter presents the results from my thematic, quantitative, and deep sample exploratory analyses used to address the main research questions guiding this study about the relationship between gender and bail outcomes. It begins by discussing the findings from the thematic analysis, which provides a general overview of the range and frequency of justifications and conditions used by justices of the peace (JP) when granting or denying bail. Next, I outline the results of the bivariate cross-tabular analyses. I then present the results of the logistic regression model, which reveals a more accurate depiction of these relationships, particularly the effect of accused gender on bail outcomes by controlling for other theoretically relevant factors.

Finally, I present the findings of the deep sample exploratory analysis. By performing this multi- method analysis, I am able to advance what is known about whether and how gender shapes bail outcomes for accused individuals in Southern Ontario, Canada.

4.2 Thematic Analysis

Justifications for Granting Bail

Using thematic analysis as a guide, justifications used by JPs to grant bail were recorded based on theme and frequency. Table 2 presents five justification categories used to support a JPs decision to grant bail. The most frequently used category relates to justifications about the surety

(39 percent), followed by those referring to the accused (28 percent), additional levels of supervision/ community support (20 percent), the allegations (8 percent) and the victim (2 percent). In three percent of cases, there were no justifications given.

69

Table 2: Justifications for Granting Bail, Kitchener Provincial Courthouse 2012 (N= 205)20 Type of Justification Number of Times Used (%)21

Justifications related to the surety 39% (80) Justifications related to the accused 28% (58) Justifications related to additional levels of 20% (40) supervision/ community support Justifications related to the allegations 8% (17) Justifications related to the victim 2% (3) No justification given 3% (7)

The three most common types of justifications used by JPs for granting bail focus on the surety, the accused, and having additional levels of supervision or community support. As shown in Table, 39 percent of the justifications used by JPs referred to the surety, making it the most used type of justification overall. Within this category, having a suitable surety appeared to be the most important reason for granting bail (58 percent), followed by the availability of a residential surety (16 percent), and having multiple sureties (14 percent) (see Appendix- Table

A-3). It also appears that JPs placed considerable emphasis on legal and extra legal factors related to the accused when deciding to grant bail (see Appendix- Table A-4). For example, 28 percent of the justifications for granting bail related to the accused, which included whether the accused had a limited, outdated, or no criminal record (45 percent), and whether the accused had taken steps to get better by engaging in treatment or counselling (10 percent).22 JPs also frequently discussed the presence of social support in their decision to grant bail, especially for accused individuals who had either a non-residential surety or no surety at all. Indeed, the most frequently used justification in this category referred to Youth in Conflict with the Law (43 percent), followed by the presence of additional probation orders (23 percent). Having the

20 This total reflects the number of justifications given, not the number of actual cases that were granted bail. These number are not equal because JPs may have provided multiple explanations for each decision. 21 Due to rounding, these numbers might not add up to 100. 22 In some cases, the JP granted bail to individuals who independently sought treatment/counselling for issues like drug/alcohol addictions, mental health issues, and/or anger problems. 70

support of friends/family (10 percent) and living outside Kitchener/ Cambridge/ Waterloo (10 percent) were also important factors in decisions to grant bail (see Appendix- Table A-5).

JPs used justifications related to the allegations against the accused less frequently than the above categories (see Appendix- Table A-6). In eight percent of cases in which bail was granted, JPs referred to the allegations in the following ways: they were not very serious (59 percent), the Crown had a weak case for detention (10 percent), the offense did not occur in front of children (6 percent), and alcohol was a factor (6 percent). JPs rarely cited justifications relating to the victim as a reason for granting bail (see Appendix- Table A-7).

Just as characteristics of the accused and surety were key reasons for granting bail, so too were they influential for denying bail. As Table 3 outlines, the most commonly used justifications to deny bail referred to the accused (52 percent), followed by those related to the surety (30 percent), the allegations (10 percent), the victim (7 percent), and public confidence (1 percent).

Table 3: Justifications for Denying Bail, Kitchener Provincial Courthouse 2012 (N= 81)23 Type of Justification Number of Times Given (%)24 Justifications related to the accused 52% (42) Justifications related to the surety 30% (24) Justifications related to the allegations 10% (8) Justifications related to the victim 7% (6) Justifications related to public confidence 1% (1)

Over half (52 percent) of all justifications used to deny bail revolved around the accused.

JPs referenced many legal and extra-legal factors when supporting their decision to deny bail,

23 This total reflects the number of justifications given, not the number of actual cases that were denied bail. These number are not equal because JPs gave multiple explanations for each decision 24 Due to rounding, these numbers might not add up to 100 71

including whether the accused had a lengthy criminal record (17 percent), whether the accused was a threat to the public (17 percent), whether the accused had previously ignored court orders

(14 percent), and whether the accused had a violent history (10 percent). Interestingly, JPs also looked unfavourably upon the accused if s/he failed to improve their situation or go to treatment/ counselling (12 percent) (see Appendix- Table A-8). The second most common category of justifications for denying bail related to the surety and their abilities to supervise properly. These types of justifications were cited in 30 percent of cases in which bail was denied (see Appendix-

Table A-9). The main concerns about the surety focused on his/her inability to supervise the accused (25 percent) and having a weak bail plan in place for the accused (13 percent).

JPs also denied bail based on the allegations against the accused in ten percent of all cases, which required the JP to make decisions based on the present case (see Appendix- Table

10). In this category, it appears that JPs took into account the seriousness of the allegations (63 percent) and the strength of the Crown’s case (38 percent) most often. Justifications for denying bail that related to the victim/complainant comprised only six percent of all justifications for denial (see Appendix- Table A-11), which included the need to protect the victim’s safety (67 percent), whether the victim sustained injuries (17 percent), and whether the complainant was deemed untrustworthy (17 percent).25 The least frequently used justification for denying bail related to public confidence as it was used to deny bail in only one case (see Appendix- Table A-

12).

25 This statement referred to a case where the complainant testified against her previous statements to police where she said she was highly fearful of the accused and thought he would kill her. In her testimony, she told the court that these statements were taken out of context and that she was not fearful of the accused and hoped that they could continue seeing each other if he was released. 72

Bail Conditions

Table 4 outlines the results of the thematic analysis pertaining to bail conditions. The most popular bail condition placed on a recognizance order restricted possessions (25 percent), followed by those related to residence (18 percent), communication (17 percent), mobility (16 percent), additional levels of supervision/community support (9 percent), curfew/house arrest (7 percent), and treatment/counselling (7 percent).

Table 4: Bail Conditions, Kitchener Provincial Courthouse 2012 (n= 510)26 Type of Condition Number of Times Given (%)27 Conditions related to possessions 25% (130) Conditions related to residence 18% (94) Conditions related to communication 17% (87) Conditions related to mobility 16% (80) Conditions related to additional levels of 9% (45) supervision Conditions related to curfew/house arrest 7% (37) Conditions related to treatment/counselling 7% (36) Other conditions 0.2% (1)

As outlined above, the most commonly used bail condition placed on a recognizance order restricted the possession and consumption of certain items. In this category (see Appendix-

Table A-13), the most frequently used conditions were those that restricted the possession of weapons (44 percent), and prevented the accused from consuming, possessing, and ingesting any drugs (25 percent) and/or alcohol (22 percent).

Eighteen percent of all conditions placed on release orders restricted the type and location of an accused’s residence during his/her release (see Appendix- Table A-14). More specifically, conditions in this category required the accused to live with his/her surety while released (56

26 This total reflects the number of conditions given. All JPs in this study gave out more than one release condition. 27 Due to rounding, these numbers might not add up to 100. 73

percent), and to live at his/her own residence and not move without prior, written approval by

Youth in Conflict with the Law (11 percent). Those who did not have a fixed residence were often forced to reside at a shelter and not change without prior, written approval by Youth in

Conflict with the Law (10 percent).

The third most used type of condition related to limits on communication, which made up

17 percent of all conditions placed on a release order (see Appendix- Table A-15). The most frequently used condition prevented the accused from communicating, either directly or indirectly, with the complainant(s) or witness(es) (72 percent), followed by those restricting an accused from communicating with his/her children (10 percent).

Mobility restrictions also made up 16 percent of all conditions placed on a release order

(see Appendix- Table A-16). Over one-quarter of all conditions in this category prevented the accused from attending any place where the offense allegedly occurred (28 percent), while several other conditions restricted the accused from attending within 50m (1 percent), 100m (15 percent), 200m (6 percent), and 500m (3 percent) of the victim’s/witness’ place of residence, employment, and education. JPs also regularly enforced a condition that allowed the accused to return to their residence to obtain their belongings so long as they were in the presence of a

Waterloo Regional police officer (13 percent).

Less frequently used were conditions relating to additional levels of supervision, which made up only 9 percent of all conditions placed on release orders (see Appendix- Table A-17).

The most frequently used condition in this category ordered an accused to be under the supervision of Youth in Conflict with the Law (44 percent), followed by reporting to Waterloo

74

Regional police weekly (29 percent), and reporting to his/her probation officer within 24hrs (9 percent), 48hrs (24 percent), and 72hrs (8 percent).

Curfew and treatment were the least used types of bail conditions, each totalling approximately 7 percent (see Appendix- Table A-18 and Table A-19). The most used curfew- type condition prevented the accused from going out between the hours of 10:00pm to 6:00am except with a surety (15 percent), followed by house arrest (14 percent). As for treatment/counselling-like conditions, the most frequently used condition mandated the accused to attend treatment or counselling as directed by his/her surety (39 percent), followed by residing at a residential treatment facility (19 percent), and attending treatment as directed by Youth in

Conflict with the Law (17 percent).

To summarize, JPs justified granting and denying bail based on numerous factors. The most used factor supporting a release decision related to the surety, followed by those relating to the accused, additional community supports, the allegations, and the victim. When denying bail

JPs relied on justifications related to the accused most often, followed by those referring to the surety, the allegations, public safety, and the victim. The most popular bail conditions placed on a release order were those that restricted material possessions. Depending on the case, JPs also ordered conditions that were associated with residence, communication, mobility, additional levels of supervision, curfew, and treatment.

75

4.3 Quantitative Analysis

Cross-Tabular Analysis

In order to determine if gender is associated with bail outcomes, I now turn to cross- tabular analysis. Using Pearson’s chi-square as a measure of statistical significance, cross-tabular analyses were performed to assess potential relationships between legal, charge, accused, victim, and surety characteristics with accused gender and bail outcomes. More specifically, I present three cross tabular analyses: (1) between accused gender and legal, charge, accused, victim and surety characteristics, (2) between bail outcome and legal, charge, accused, victim, and surety characteristics, and (3) between accused gender and bail conditions.

The bivariate analysis between gender of the accused, bail outcome, and the control variables, presented in Table 5, reveals that several relationships are significant. Twenty-five percent of accused men possessed a serious criminal record whereas only 5 percent of accused women had a serious criminal record. When compared to accused women, accused men were also more likely to possess a related criminal record (60 percent vs. 27 percent) and have past failures to comply (51 percent vs. 23 percent). In relation to the Crown, women were more likely to get a consent to release (86 percent vs. 60 percent), while men were more likely to have a contested hearing (40 percent vs. 14 percent). The Crown was also much more likely to have secondary ground concerns for men as opposed to women (44 percent vs. 9 percent). On average, women were generally older at 30-39 years (compared to a slightly younger age group for men at

20-29 years) and were also more likely to have children compared to men (80 percent vs. 44 percent).

76

Table 5: Bivariate Relationships: Legal, Charge, Accused, Victim, and Surety Characteristics by Gender of the Accused, Kitchener Provincial Courthouse 2012 (n= 115)28

Female Accused Male Accused Variable p Percent (n= 22) Percent (n= 93) Dependent variable Bail outcome 0.132 Granted bail 95 83 Denied bail 5 17

Legal Characteristics Criminal record of the accused (n= 114) 0.095 Has a criminal record 55 73 No criminal record 46 27 Severity of criminal record (n=110) .007** Low 50 30 Medium 17 High 5 25 No record 46 28 Related criminal record (n=97) 0.011* No 27 9 Yes 27 60 No record 46 31 History of Violence (n= 93) 0.052 No history of violence 82 59 History of violence 18 41 Previous release (n= 111) 0.244 Granted bail 68 80 Not granted bail 32 20 Previous release success (n= 72) 0.793 Not successful 57 67 Successful 7 9 Not applicable 36 24 Past failures to comply 0.017* No failures to comply 77 49 Failed to comply 23 51 Position of the Crown .013* Consent to release 86 60 Seeking detention 14 40 Secondary ground concerns 0.001*** No 91 56 Yes 9 44 Charge characteristics Number of current charges (n= 112) 0.863 2 offenses 2 offenses Offense seriousness 0.856 Low 55 59 Medium 41 34

28 When cell counts fell below the expected count, a Fisher’s Exact Test was used to obtain the Pearson Chi-Square values. Fisher’s Exact Test is used when one or more of the cells has an expected value of less than 5 (George and Mallery, 2012). 77

High 5 7 Type of offense 0.906 Drug 9 5 Against the admin of justice 32 32 Property 9 9 Violent 18 15 Combination 32 39 Reverse Onus 0.787 Not a reverse onus situation 59 56 Reverse onus 41 44

Accused on probation 0.954 Not on probation 68 69 On probation 32 31 Drug/Alcohol a factor 0.745 No 68 65 Yes 32 36 Children present during offense (n=110) 0.391 No 82 89 Yes 18 11 Interactions with police (n= 28) 0.877 Accused did not cooperate 50 54 Accused cooperated 50 45 Involvement of weapons 0.942 No weapons involved 91 91 Weapons involved 9 9

Available Surety Surety availability 0.641 No surety available 14 9 Surety available 64 72 Youth in Conflict 23 19 Residential surety availability 0.226 No residential surety 14 18 Has residential surety 59 70 Not required 27 12 Probable level of supervision 0.272 low level supervision 36 33 medium level supervision 14 30 high level supervision 50 37 Accused characteristics Race/Ethnicity 0.292 White 77 66 Non-white 23 34 Age (n= 112) 0.031* 30-39 years 25-29 years Employment status (n=87 ) 0.877 Employed fulltime 38 42 Employed part time 6 4 Unemployed 56 51 Other 3 Residence status (n= 105) 0.136 Owns home 15 4 Rents home 45 37 Lives in shelter 5 7

78

Lives with friends/family 20 44 Other 15 8 Drug/Alcohol Dependency (n= 102) 0.489 Yes 50 59 No 50 42 Accused in treatment/counselling (n= 101) 0.708 Not in treatment/counselling 40 31 In treatment/counselling 10 10 Not applicable 50 59 Mental health issues (n= 95) 1 Has mental health issues 20 20 Does not have mental health issues 80 80 Taking medication for mental health issues (n=93) 0.767 Taking medication 7 Not taking medication 11 13 Not applicable 89 80 Relationship status (n=95) 0.060 Single 56 71 Married 44 20 Divorced/ separated 9 Children (n= 83) 0.012* Has children 80 44 Does not have children 20 56 Victim characteristics Victim-accused relationship (n= 111) 0.702 No victim 64 54 Significant other 5 16 Family 5 3 Non-family 23 20 Multiple victims 5 7 Victim age 0.697 0-12 yrs 6 3 13-18 yrs 0 3 19-29 yrs 6 14 30-39 yrs 0 4 40-49 yrs 6 4 Multiple 6 11 Not applicable 78 61 Victim Gender (n=112) 0.213 Male 18 8 Female 14 31 Mix 5 8 Not applicable 64 53 Level of fear (n=83) 0.212 Not fearful 7 6 Somewhat fearful 7 Very fearful 16 Not applicable 93 71 Same victim (n=103) 0.425 No 9 12 Yes 5 14 Not applicable 86 74 Surety Characteristics Surety identity 0.353 No surety available 36 28

79

Family 36 40 Non-family 18 8 Multiple 9 19 Surety-accused relationship (n=108) 0.122 Low 14 5 Medium 10 31 High 38 35 Not applicable 38 30 Surety gender 0.704 No surety 36 28 Male 9 18 Female 46 43 Mix 9 11 Employment status (n=93) 0.516 No surety 42 35 Employed 26 41 Unemployed 32 24 Amount of assets (n=88) 0.399 Minimal 22 31 Average 33 21 Substantial 10 No surety 44 37 Residence status (n=94) 0.912 Local 45 50 Not local 15 15 No surety 40 35 General surety experience (n= 82) 0.492 No 28 27 Yes 20 34 Not applicable 53 39 Surety for the accused in past (n=91) 0.291 No 31 23 Yes 6 29 Not applicable 63 48 Surety have children (n=61) 0.291 No 9 6 Yes 18 44 Not applicable 73 50 * p ≤ .05 **p ≤ .01 ***p ≤ .001

In addition to examining the bivariate association between the main independent variable

– gender – and the dependent variable – bail outcome, I also examined the associations between bail and the other variables included in my multivariate analysis as controls. Table 6 presents the results of the bivariate analysis between the control variables and bail outcome. Interestingly, gender was not significantly related to bail outcome at this level. As expected though, compared to those who were granted bail, individuals who were denied bail were more likely to possess a

80

criminal record (94 percent vs. 65 percent), possess a related criminal record (94 percent vs. 44 percent), have a history of violence (92 percent vs. 26 percent) and have past failures to comply

(77 percent vs. 39 percent). When looking at seriousness of the criminal record, 40 percent of those granted bail had a minor criminal record, while 82 percent of those denied bail had a serious criminal record. All individuals denied bail had a contested hearing (vs. 24 percent of those granted bail) and were considered a risk of reoffending (vs. 27 percent of those granted bail).

In comparison to individuals who were granted bail, those who were denied bail were also more likely to be charged with a greater number of offenses (4 vs. 2), more serious offenses

(18 percent v. 4 percent), and a combination of offenses (77 percent v. 31 percent). When looking at probable level of bail supervision, those who were denied bail were more likely to have a low level of bail supervision compared to those granted bail (65 percent vs. 29 percent).

They were also less likely to have a very strong relationship with their surety compared to those granted bail (6 percent vs. 41 percent). In terms of victims, those who were denied bail were also more likely to have multiple victims (28 percent vs. 6 percent) and to have committed an offense against the same victim in comparison to those granted bail (35 percent vs. 7 percent).

81

Table 6: Bivariate Relationships: Bail Outcome by Legal, Charge, Accused, Victim, and Surety Characteristics, Kitchener Provincial Courthouse 2012 (n= 115)29

Variable Granted Bail (n= 98) Denied Bail (n= 17) p Independent Variable

Accused Gender 0.132

Male 79 94

Female 21 6

Legal Characteristics Criminal record of the accused (n= 114) .01** Has a criminal record 65 94 No criminal record 35 6 Seriousness of criminal record (n= 110) .000*** Low 40 Medium 14 12 High 10 82 No record 37 6 Related criminal record (n= 97) .001*** No 16 Yes 44 94 No record 40 6 History of Violence (n= 93) .000*** No history of violence 74 8 History of violence 26 92 Previous release (n= 111) 0.249 Granted bail 76 88 Not granted bail 25 12 Previous release success (n= 72) 0.057 Not successful 59 93 Successful 10 Not applicable 31 7 Past failures to comply (n= 104) .005** No failures to comply 61 24 Failed to comply 39 77 Crown Position .000*** Consent to release 77 Seeking detention 24 100 Secondary ground concerns .000*** No 74 Yes 27 100 Charge characteristics Number of current charges (n= 113) .000*** 2 charges 4 charges Offense seriousness .002** Low 64 24 Medium 32 59 High 4 18 Type of offense .004**

29 When cell counts fell below the expect count, a Fisher’s Exact Test was used to obtain the Pearson Chi-Square values. 82

Drug 7 Against the admin of justice 36 12 Property 8 12 Violent 18 Combination 31 77 Reverse Onus 0.747 Not a reverse onus situation 57 53 Reverse onus 43 47

Accused on probation 0.701 Not on probation 69 65 On probation 31 35 Drug/Alcohol a factor 0.962 No 65 65 Yes 35 35 Children present during offense (n= 110) 0.111 No 89 75 Yes 11 25 Interactions with police (n= 28) 0.099 Accused did not cooperate 46 83 Accused cooperated 55 17 Involvement of weapons 0.156 No weapons involved 93 82 Weapons involved 7 18

Available Surety Surety availability 0.361 No surety available 8 18 Surety available 70 71 Youth in Conflict 21 12 Residential surety availability 0.079 No residential surety 15 29 Has residential surety 67 71 Not required 17 Probable level of bail supervision .002** low level supervision 29 65 medium level supervision 27 29 high level supervision 45 6 Accused characteristics Race/Ethnicity 0.792 White 67 71 Other 33 29 Age (n= 112) 0.093 30-39 years 25-29 years Employment status (n= 87) 0.121 Employed fulltime 42 40 Employed part time 3 13 Unemployed 54 40 Other 1 7 Residence status (n= 105) 0.613 Owns home 8 Rents home 38 41 Lives in shelter 7 6 Lives with friends/family 40 35 Other 8 18

83

Drug/Alcohol Dependency (n= 102) 0.721 Yes 42 47 No 58 53 Accused in treatment/counselling (n= 101) 0.911 Not in treatment/counselling 32 35 In treatment/counselling 10 12 Not applicable 58 53 Mental health issues (n= 95) 0.082 Has mental health issues 18 35 Does not have mental health issues 83 65 Taking medication for mental health issues (n= 93) 0.193 Taking medication 10 25 Not taking medication 5 6 Not applicable 84 69 Relationship status (n= 95) 0.741 Single 69 65 Married 24 24 Divorced/ separated 6 12 Children (n= 83) 0.289 Has children 48 63 Does not have children 52 38 Victim characteristics Victim-accused relationship (n= 111) 0.301 No victim 59 38 Significant other 13 19 Family 4 Non-family 19 31 Multiple victims 5 13 Victim age (n= 97) .025* 0-12 yrs 4 13-18 yrs 2 19-29 yrs 12 20 30-39 yrs 1 13 40-49 yrs 5 Multiple 7 28 Not applicable 68 40 Victim Gender (n= 112) 0.177 Male 10 12 Female 26 35 Mix 5 18 Not applicable 59 35 Level of fear (n= 83) 0.352 Not fearful 5 4 Somewhat fearful 32 5 Very fearful 12 Not stated 27 Not applicable 64 52 Same victim (n= 103) .000*** No 9 24 Yes 7 35 No victim 84 41 Sure ty Characteristics Surety identity 0.867 No surety available 30 29 Significant other 5

84

Family 40 35 Non-family 9 12 Multiple victims 16 24 Surety-accused relationship (n= 108) .006** Low 6 12 Medium 22 53 High 41 6 Not applicable 32 29 Surety gender 0.4 No surety 30 29 Male 14 29 Female 46 29 Mix 10 12 Employment status (n= 93) 0.831 No surety 38 31 Employed 38 38 Unemployed 25 31 Amount of assets (n= 88) 0.9 Minimal 27 40 Average 25 20 Substantial 8 7 No surety 40 33 Residence status (n= 94) 0.793 Local 48 57 Not local 16 7 No surety 36 36 General experience as a surety (n= 82) 0.214 No 29 20 Yes 27 50 Not applicable 44 31 Surety for the accused in past (n= 91) 0.131 No 24 25 Yes 21 44 Not applicable 55 31 * p ≤ .05 **p ≤ .01 ***p ≤ .001

The results from the bivariate analysis between number and type of bail conditions and accused gender are presented in Table 7. Only one relationship was statistically significant in this analysis. On average, accused men were more likely to receive higher bail amounts than accused women ($2500-4999.99 vs. $1000-2499.99). It is also important to note that the relationship between gender and treatment/counselling conditions was statistically significant at the p < .10 level. Here, 43 percent of accused women received a treatment-like condition on their recognizance compared to only 22 percent of accused men.

85

Table 7: Bivariate Relationships: Type and Number of Bail Conditions and Bail Amount by Gender of the Accused, Kitchener Provincial Courthouse 2012 (n= 98)

Female Accused Percent Male Accused Percent Variable p (n= 21) (n= 77) Bail Amount

Bail Amount .04*

$1000-2499.99 $2500-4999.99

Number of Bail Conditions Average number of bail conditions 0.953 5 conditions 5 conditions

Type of Bail Condition (n= 98) Communication 0.354 No 33 23 Yes 64 77 Mobility 0.472 No 38 30 Yes 62 70 Residence 0.6 No 0 1 Yes 100 99 Curfew/House Arrest 0.637 No 67 61 Yes 33 39 Additional Supervision 0.503 No 48 56 Yes 52 44 Treatment/Counselling 0.056+ No 57 78 Yes 43 22 Possession 0.23 No 29 17 Yes 71 83 * p ≤ .05 **p ≤ .01 ***p ≤ .001

In sum, the results of the analyses suggest that gender is not significantly related to bail outcome at the bivariate level. It was shown, however, that accused men were more likely to have a serious and related criminal record, have past failures to comply, and be heard in contested court while accused women were more likely to have children. Those who were denied bail were more likely to possess a criminal record, have a related and serious record, have a history of violence, and have past failures to comply. Compared to those granted bail, individuals

86

who were denied bail were also typically charged with a greater number offenses, more serious offenses, and were more likely to have a weak bail plan and relationship with their proposed surety. They were also more likely to have multiple victims and be involved with the same victim. On average, men were more likely given a higher bail amount while some evidence suggests that women tended to receive more treatment-like conditions on their release orders.

Logistic Regression

Four step-wise logit models were used to examine the effects of accused gender on bail outcome, controlling for theoretically important legal and extra-legal variables (see Table 8). In

Model 1, I included only one variable – gender of the accused – in order to determine the baseline effect of gender on bail outcomes. Model 2 incorporates the effect of legal variables on bail outcome, while Model 3 includes characteristics of the accused. Characteristics of the victim are included in Model 4. Models 1 through 4 thus reveal the relative contribution of various legal and extra-legal factors to bail outcomes.

Absent the inclusion of controls, Model 1 shows that gender of the accused is not statistically significant. In other words, men and women are equally likely to be granted bail when other legal and extra-legal variables are not considered. Model 2 incorporates measures of legal variables such as criminal record, charge severity, and surety availability. The results of this model show that the foregoing legal variables are all statistically significant predictors of bail outcome. When compared to those who had no criminal record and committed a minor criminal offense, individuals with a criminal record are less likely to be granted bail, as are those who committed somewhat serious and serious criminal offenses. Further, when compared to individuals who did not have a surety, those who had a surety were 15 times more likely to be granted bail while those who were under the direction of Youth in Conflict with the Law were 22

87

times more likely to be granted bail. Importantly, the relationship between the gender of the accused and bail outcome remains unchanged in Model 2; that is, with the inclusion of legal variables, women are not significantly different from men with respect to their likelihood of being granted bail.

Model 3 includes the accused’s characteristics, namely their race and whether they have children. After the inclusion of these variables, gender of the accused becomes a statistically significant predictor of bail outcome, such that accused females are more likely to be granted bail when controlling for both legal and extra legal factors. The legal variables in the model – criminal record, charge severity, and surety availability – become slightly stronger predictors of bail than in Model 2 once the accused’s characteristics are added. Although race and children are not significant determinants of bail outcome, they are important factors in this model because they increase the predictive strength of gender.30 When looking at the size of the standard errors in Model 3, however, this effect is likely due to chance (Schroeder et al., 1986).31 In multiple and logistic regressions, large standard errors are normally caused by a small sample size, variables with small variances, or multicollinearity (Berry and Feldman, 1985). Tests for collinearity resulted in Low Variance Inflation Factors (VIFs), which implies that multicollinearity is not an issue in this model. Instead, the large standard errors are likely due to the small sample size and the limited variation between those granted versus denied bail. When taken together, these issues make it difficult conclude with 95% confidence that gender has an influence on bail outcome.

30 In order to test for possible interaction effects amongst these variables and gender, the model was re-run twice using product terms between gender and race and gender and children. Neither term had a statistically significant effect on bail outcome. 31 A standard error (SE B) measures the amount of variability among different b’s “estimated from samples drawn from the same population” (Schroeder et al., 1986, 42). It allows social scientists to make “inferences about how sensitive the estimate of β is to changes in sample composition without taking another sample” (Ibid., 43). Accordingly, a large standard error indicates that minimal changes in observations can result in a large change to the slope of the regression line, limiting the ability to draw inferences to the population parameter. 88

The pattern of results in Model 4, which includes victim involvement, conforms to those presented in Model 3, whereby gender of the accused and the legal variables appear to be significantly related to bail outcome. Although gender of the accused, criminal record, charge severity, and surety availability remain significant, their predictive strength is weakened after controlling for victim involvement. Thus, with the inclusion of theoretically-relevant control variables, this analysis suggests that the relationship between gender of the accused and bail outcome may be significant the p=.05 level. In other words, it is possible that among the accused in my sample, women are more likely to be released on bail than men after controlling for legal and extra-legal factors. Unfortunately, the limited sample size makes it difficult to validate this finding, as the results may simply be due to chance.

89

Table 8: Estimated Coefficients of Logit Models Predicting Bail Outcomes Among Adult Men and Women, Kitchener Provincial Courthouse 2012 (n= 83)

Variables Model 1 Model 2 Model 3 Model 4 B SE B eB B SE B eB B SE B eB B SE B eB Independent Variable Gender of the accused Female accused 1.377 1.076 3.962 2.709 1.395 15.008 2.706* 1.347 14.963 2.804* 1.403 16.508 Legal Variables Criminal record Has criminal record -3.066* 1.270 .047 -3.095* 1.345 .045 -3.051* 1.358 .047 Charge severity Medium severity -3.042** 1.032 .048 -3.373** 1.150 .034 -3.29** 1.69 .037 High severity -4.792** 1.654 .008 -5.415** 1.845 .004 -5.21** 1.874 .005 Surety Available Surety available 2.743* 1.129 15.526 3.034* 1.232 20.776 2.868* 1.214 17.609 Youth in Conflict 3.117* 1.558 22.577 3.359* 1.605 28.768 3.416* 1.658 30.458 Accused Characteristics Race of the accused Non-white .712 .834 2.037 .77 .868 2.161 Children Has children .497 .825 1.644 .137 .864 1.147 Victim Characteristics Victim Involved .249 .874 1.282

Constant 1.262 3.425 3.014 3.03 Log Likelihood 79.108 51.716 50.61 48.069 Wald χ2 18.627 6.02 4.076 3.687 Pseudo R^2 0.043 0.481 0.469 0.496 * p ≤ .05 **p ≤ .01 ***p ≤ .001

90

4.4 Deep Sample Exploratory Qualitative Analysis

The previous section presented the results of both the bivariate and multivariate analyses and revealed which factors significantly affect the relationship between accused gender and bail outcome. Quantitative analysis techniques, such as those used above, help provide a better understanding of the relationships between variables using complex mathematical functions. To capture the subjective realities of court proceedings, a deep sample exploratory qualitative analysis was also used to compliment my quantitative findings. This section displays the results of the qualitative analysis that matched 19 accused men and women based on similar charge and background characteristics. The purpose of this analysis was to explore the possibility of gender disparities in bail outcomes by assessing the types of justifications and release conditions given to the men and women in my sample. Gender was the focus here; however, it is possible that other factors, beyond gender, played a role in bail outcomes. The themes identified in this section are the result of a preliminary analysis that is intended to generate ideas that can be further explored in future research. This was accomplished using thematic analysis.

While this deep sample analysis demonstrates that men and women are granted and denied bail at similar rates, it shows that bail conditions often vary for men and women. This is important because it helps identify potentially gendered discourses that may influence decisions to grant bail. The first eight cases include accused women who received more restrictive bail conditions than their matched male pairs. A potential explanatory factor is the court’s perception of male and female drug/alcohol abuse and the subsequent construction of risk. Pairs nine to thirteen present cases where the accused man received more restrictive bail conditions than their matched female pair. Differences in how the court assesses mental health and culpability for men and women are discussed as a possible reason for accused men receiving harsher bail outcomes.

91

Cases fourteen to sixteen received similar bail outcomes but the justifications for these outcomes differed between the men and women. For women, justifications centred on their familial role and position in the community, while for men, these justifications looked toward maintaining paid employment. Finally, the remaining three cases all had similar justifications and bail conditions placed on their release. For the purposes of this study, three pairs were excluded because of too many differences.

Drug/Alcohol Use: Women receiving more restrictive bail releases

In Canada, JPs are able to impose any reasonable condition in order to mitigate potential risk of re-offending. According to Trotter (2010), mandated treatment/counselling and curfews are the most restrictive condition placed on a bail release because they attempt to correct and punish risky behaviour. In the following eight cases, six women who were suspected or known users of drugs or alcohol received more restrictive bail releases than their male counterparts, and all six had to either engage in some form of treatment or counselling or obey a curfew as part of their release.

Pair 1- Barb and Harold32: Despite the fact that the JPs described Harold’s previous breaches as “disturbing,”33 Barb received a harsher bail release than Harold. In both cases,

Harold and Barb committed minor drug related offenses34 and neither had a serious criminal record. Specifically, Harold was charged with possession with the intent to traffic and possession of a controlled substance while Barb was charged with possession of a controlled substance and possession of drug paraphernalia. Barb had a known addiction to crack cocaine while it was only suspected that Harold had a drug issue. Unlike Harold, Barb was unemployed due to valid

32 Pseudonyms were used to maintain anonymity of the accused individuals in this analysis. 33 The JP in Harold’s case referred to the number of past breaches as disturbing, not the seriousness of the breaches.

92

medical reasons. Although both were granted bail, they received very different bail releases. For

Barb, it was seen as appropriate to place her under a 24-hour house arrest except in the presence of her sureties and to attend treatment or counselling for her drug addiction. In comparison, the

Crown thought a 24-hour house arrest was inappropriate for Harold and instead, placed him on a curfew between 10:00pm and 6:00am. Despite suspicions of drug use, he was not forced to attend treatment. In fact, the Crown geared most of their questions for Harold’s surety towards his employment prospects once released; instead of how well the surety could supervise.

Pair 2- Cathy and Glenn: Both Cathy and Glenn had similar charges in terms of severity and criminal records, although Glenn had more breaches and was heard in contested court. Glenn was charged with a breach of recognizance while Cathy was charged with shoplifting and trespassing after she failed to pay for $82.00 worth of groceries.35 Both were unemployed, lived in rental accommodations and had severe problems with alcohol. Because of their limited ties to friends and family, Youth in Conflict with the Law agreed to act as the surety despite having previous issues with both Cathy and Glenn. Both were granted bail and received similar conditions with one notable difference; Cathy had a curfew placed upon her release while Glenn did not. In addition to putting her on a curfew between 10:00pm and 6:00am, Cathy’s release order also restricted her from attending the place where the offense occurred, placed her under the supervision of Youth in Conflict with the Law, required her to reside at a residence approved of by Youth in Conflict, and directed her to take counselling as directed by Youth in Conflict. On the other hand, Glenn was to forbidden to contact the complainant or possess any alcohol and weapons, and was required to remain under the supervision of Youth in Conflict, take counselling as directed by Youth in Conflict and reside at his current address or at a residence

35 Cathy was already bound by a no-trespass order restricting her from attending this particular grocery store. 93

approved of by Youth in Conflict. When justifying Glenn’s release, the presiding JP emphasized

Glenn’s minor attempts to get better, even though his current charge was for consuming alcohol.

For Cathy, the Court focused entirely on her alcohol usage, calling it the “cause of her troubles.”

Pair 3- Tiffany and Graham: Tiffany and Graham were both charged with minor breaches. Tiffany had a minor criminal record, and although Graham did not have a record, he did have outstanding charges for previous breaches. Graham was employed fulltime, lived with his parents, and had a serious alcohol problem while Tiffany was employed part-time, lived in rental accommodations, and had a severe drug addiction (type of drug not specified). For

Tiffany, the Crown and JP consented to her release because she would be living with her surety in Etobicoke, which removed her from her current situation and prevented her from contacting the complainant – her common law husband. Although both Tiffany and Graham were forced to participate in treatment while released, Tiffany had several additional conditions placed on her release, including a 24-hour house arrest provision. In comparison, Graham was only given a curfew between 10:00pm and 6:00am except for attending work and was forbidden from attending any place where alcohol was sold primarily. According to the JP, restricting him from consuming alcohol would “most certainly” lead to a breach. This same concern was not expressed for Tiffany, as she was restricted from consuming drugs and alcohol outright.

Pair 4- Geraldine and John: In this pairing, Geraldine was charged with theft under

$5000 and breach of recognizance while John was charged with two counts of theft under $5000 and assaulting a police officer. Neither had a criminal record and both were unemployed and living with family. Geraldine and John had reported issues with drugs, and although Geraldine had a more serious addiction, John had bipolar and deliria and used drugs to self-medicate.

John’s mother came forward as his surety and stated she would help him deal with his mental

94

health issues and get him proper medication. Geraldine’s father acted as her surety and told the court that he would be living with her in order to help get her life back on track. While the

Crown consented to both of their releases because they had minimal records and appropriate sureties, only Geraldine was forced to engage in treatment during her release. This was surprising given the fact that John also had issues with drugs and was using them to deal with his mental health issues.

Pair 5- Farah and Carl: Farah was in her 20s, unemployed, and did not have a fixed address. At the time of her bail hearing, she was taking methadone to overcome an addiction to heroin and did not have a criminal record. She was arrested and charged with possession for the purpose of trafficking and breaching her recognizance. According to the Crown, she also had schizophrenia and was bipolar. Carl was in his 20s, unemployed, and lived with friends. He was a known user of marijuana, getting his means via drug dealing. Although he had a criminal record, it was very outdated. He was charged with possession, and possession for the purpose of trafficking after trying to sell marijuana to an undercover police officer. Both Carl and Farah received very strict bail release conditions. Carl was to obey a curfew between 12:00am-6:00am, refrain from drug use or possession, and actively seek and maintain lawful employment.

However, Farah’s conditions were much more intrusive. Not only was she to refrain from using drugs, she was to actively participate in treatment/counselling, report to her family doctor regularly and follow his direction, not to use or possess any cell phones, and obey a curfew between 8:00pm-8:00am.

Pair 6- Holly and Samuel: Holly was in her 20s at the time of arrest and was charged with possession of a controlled substance and failure to appear. She had a limited criminal record, was unemployed, and lived with friends. After her arrest, she admitted to police she was

95

taking methadone to help with her addiction to heroin and had relapsed because her dosage was too low. Holly’s mother came forward as surety and told the court that she hoped to get Holly into a residential treatment program. Samuel was 20 years old and was charged with possession of a controlled substance (Chrystal meth) and two counts of breaching probation. He had a lengthy criminal record, was unemployed, and lived with his family. His mother acted as his surety. Interestingly, neither the Crown nor the JP asked Samuel or his mother about a potential drug problem despite having .1 gram of Chrystal meth in his possession upon arrest. Instead, he was released under the following conditions: reside with surety, do not contact or associate with any persons with a criminal record and abstain from possessing and ingesting any drugs. On the contrary, the JP mandated that Holly reside with her surety, not contact, directly or indirectly, the co-accused, abstain from possessing or ingesting any drugs, attend treatment as directed by surety, and obey a curfew between 10:00pm-6:00am. The major difference between Holly and

Samuel is the use of a curfew. Although Holly was to attend treatment (this can be explained by the fact that Holly admitted to having an issue and Samuel did not), the biggest difference was the use of a curfew for Holly despite having very similar charges to Samuel.

Pair 7- Carol and Vince: In this pairing, both Carol and Vince were charged with breaching their recognizance after contacting the complainant from the original charges. Neither

Carol nor Vince had a criminal record but both had outstanding charges for assault and Carol had previously breached. According to Crown allegations, both Carol and Vince were in violent relationships that went unreported but only Vince’s ex-partner feared for her safety. Both were employed full time and owned their own homes. During Carol’s hearing, the Crown assumed she had a problem with alcohol because she worked at the local brewery. This assumption may explain why Carol needed to have a residential surety while Vince did not. While both Vince and

96

Carol needed to attend anger management as part of their release, Carol was also forced to reside with her surety and attend counselling or treatment for her supposed alcohol issue.

Pair 8- Maureen and Mario: Both Maureen and Mario were heard in regular bail court as both were charged with minor criminal offenses.36 They were both unemployed and lived in a shelter. In this pairing, only Mario had a minor criminal record and Maureen had a problem with opiates. Maureen was released under the supervision of Youth in Conflict with the Law while

Mario was released on his own recognizance. Despite having committed a minor offense and having no prior record, Maureen was placed on a much stricter release. In addition to being under the supervision of Youth in Conflict with the Law and residing at the shelter, Maureen needed to obey a curfew between 10pm-6am. She also could not contact her co-accused; possess any drugs, drug paraphernalia, and cell phones. Although Maureen’s lawyer argued the no cell phone provision made it difficult to look for work, the JP saw it as an appropriate condition. In comparison, Mario only needed to reside at the shelter and not attend the place where the crime was committed.

Mental health and Culpability: Men receiving more restrictive bail releases

The following four pairs are examples of men receiving more restrictive bail releases than their female counterparts. A potential explanation is that the court perceives mental health and culpability differently for men and women. Men with mental health issues seem to be viewed by JPs as being a greater risk to the public, requiring stricter bail conditions to prevent reoffending, while women with mental health issues are viewed as being a greater risk to

36 Maureen was charged with 4 counts of possession of a controlled substance and 1 count of possession of stolen property while Mario was charged with theft under and possession of stolen property. The Courts characterized both offenses as ‘relatively minor’. 97

themselves. The men in these cases also appear to be held accountable for their actions differently than women, which might have justified more restrictive outcomes.

Pair 9- Brenda and Mateo: Both Brenda and Mateo were in their 30s at the time of arrest and were charged with breaching probation. Brenda had a minor criminal record, was unemployed, and lived with her mother. She stopped reporting to her probation officer in 2006 and recently turned herself in. She had a medical issue, which made it difficult to speak. Mateo also had a minor criminal record, was unemployed, and lived with family. He failed to appear at his court date and the police arrested him following a regular traffic stop. The Crown mentioned in passing that he had an issue with alcohol, although it was not discussed in detail. Mateo’s sister acted as his surety while Brenda was released on her own recognizance. Despite having similar background and charge characteristics, Brenda was released with only two conditions while Mateo was released with four. To fulfill her release, Brenda needed to reside at her stated address and report to the Waterloo Regional Police weekly. In contrast, Mateo needed to abstain from drugs and alcohol, reside with his surety, report to the Waterloo Regional Police weekly, and obey a curfew between 10:00pm-6:00am. The biggest difference is the use of a surety and curfew for Mateo but not for Brenda. In this case, the Crown and the JP appeared to be sympathetic towards Brenda due to her medical condition.

Pair 10- Alana and Bob: Alana was 23 at the time of her arrest and was charged with assault and two counts of uttering threats against her common law spouse. She had no criminal record, lived with the complainant in rental accommodations, and had two children under one year old. Based on the allegations, she and her spouse got into a verbal argument over an iPod.

The argument escalated when she pushed her spouse and threatened to kill him and herself with an exacto knife, which she later retrieved from her bedroom. According to her spouse, Alana was

98

suffering from extreme post partum depression following the birth of her daughter (who was 2 months old at the time of the allegations). Bob was in his 20s and was charged with three counts of assault and mischief under $5000. He had a limited record, was unemployed, lived with the complainant and had one child. According to the Crown, he slapped his spouse after accusing her of cheating. He suffered from bipolar disorder, for which he took medication. Bob’s mother acted as his surety and was seen as very appropriate by both the Crown and JP. Alana was released under the supervision of Youth in Conflict, but she had the support of her family who was present in the court.

Both Alana and Bob were granted bail, but from the conditions it is clear that Bob was portrayed as a bigger risk to others while Alana was a bigger risk to herself. For her release,

Alana was not to contact her children except in the presence of her spouse or grandmother, be under the supervision of Youth in Conflict and report there weekly, reside at her grandmother’s residence or at a treatment facility, attend and actively participate in mental health counselling as directed by Youth in Conflict, and not possess any weapons. In comparison, Bob was restricted from contacting his spouse (the complainant) and his children except through an agreed upon 3rd party (normally a case worker at Family and Children Services) and could only attend his residence one time with a Waterloo regional police to collect his belongings. He also needed to reside with his surety and could not possess any weapons. While it may appear that Alana received the harshest release, she was still able to contact her spouse and children easily and accessibly without the need for a 3rd party, despite the allegations. Instead of being a threat to those around her, the Court saw her as a bigger threat to herself and in need of help. Bob, on the other hand, was seen as unstable and at a greater risk of harming his family. Thus, he could not contact his children unless through a 3rd party.

99

Pair 11- Paige and Mike: In this pair, both Paige and Mike were charged with violent offenses and had a contested hearing. Although both had criminal records and convictions for breaching, Mike had a history of violence. Mike was employed fulltime, had a young daughter, and lived in rental accommodations. Mike was charged with assault with a weapon, mischief under, and two counts of uttering threats to cause death against his girlfriend. According to the

Crown, Mike suffered from severe anxiety and depression, which caused him to get aggressive.

Despite telling police she was highly fearful of Mike, Mike’s girlfriend testified during his bail hearing that she was not fearful of him and that she wanted him released. Mike’s long time friend came forward as a surety. At the time of Paige’s hearing, she had just given birth to her daughter and was unemployed (it was not specified whether she was on maternity leave). She had six children and lived in a hotel through St. Mary’s Support Program (a shelter for women). She was charged with assault, assault with a weapon, uttering threats, breach of probation, and breach of undertaking. A witness called police after seeing Paige repeatedly hit her 5-year-old son with a shoe and threaten to “beat the shit out of him.” Both of Paige’s sureties expressed concern that she had extreme post partum depression and needed help. During their hearings, both Paige and

Mike were visibly upset and expressed guilt. However, only Paige was released on bail because she had appropriate sureties. According to the JP, she was not able to contact her children except through FACS or a Family Court order, to attend the place where the offence occurred, or possess any weapons. She was also forced to reside with her surety, attend counselling as directed by her sureties, and sign any medical release forms. In comparison, Mike was denied bail. According to the JP, Mike’s mental health issues made him unstable and at a greater risk of reoffending.

100

Pair 12- Jessica and Mark: Both Jessica and Mark were charged with failure to attend.

Mark did not have a criminal record and lived with his common law wife. He failed to attend court in 2007 and a warrant was issued for his arrest. He was arrested after police saw him having a verbal argument with his common law wife. A female friend acted as his surety. Jessica had a minor criminal record and failed to attend court in 2011. She was on probation at the time.

She did not have a surety and instead, Youth in Conflict came forward to supervise her. The

Crown stated she had a mental health issue but this was not discussed in detail. While both Mark and Jessica were granted bail, Mark received more stringent bail release conditions because he had to live with his surety while Jessica was released on her own recognizance. It is possible that the court considered Jessica’s mental health issue a factor that led to her failing to appear and were thus more lenient in her release. For Mark, they might have viewed him as avoiding ownership and thus in need of greater supervision.

Employment and Children: Similar bail outcomes but different justifications

The men and women included in this section had almost identical bail outcomes. The most notable difference between the men and women in these pairings, however, is how the court justified their decisions. For the women, the court appeared to emphasize the woman’s maternal role and her position within the community when granting bail. For the men, the court seemed more preoccupied with ensuring their employment post-release. In the last pairing, the accused man and woman were denied bail for different reasons despite having similar charge and background characteristics.

Pair 13- Lynn and Don: Lynn was in her 30s at the time of her bail hearing and was charged with failing to attend and breaching probation. She had an outdated, minor criminal record, was employed full time, and owned her own home. She was married with one 10 year old

101

daughter, who was put into FACS upon Lynn’s arrest. Don, on the other hand, was charged with breaching probation and theft under $5000, had a minor criminal record, and had good prospects for employment upon release. Both were released under similar conditions: be under the supervision of Youth in Conflict, reside at stated residence, and do not change until prior approval, do not attend the location of the alleged offense, and report to police/probation officer regularly. However, when crafting Lynn’s release, the JP took into consideration her “roots to the community,”37 and her 10 year old daughter. For Don, the JP focused on his youth and not wanting to interfere with his prospective employment. Despite the fact both were employed, it was seemingly only a determining factor for Don’s release.

Pair 14- Hannah and Simon: In this pairing, both Hannah and Simon were charged with numerous offenses and neither had a criminal record. Hannah was charged with uttering threats of death, assault, assault with a weapon, and endangerment of public peace while Simon was charged with assault, three counts of uttering threats, and mischief under $5000. Hannah was a recent immigrant to Canada, employed fulltime and lived in rental accommodations with the complainant’s niece. Her mother came forward as a surety. Simon, on the other hand, lived with his mother and had a serious issue with drugs and alcohol. His mother came forward as surety.

While the Crown consented to both Hannah and Simon’s release, they only justified their decision in Hannah’s case. Throughout her hearing, the Crown stressed the seriousness of her charges because they allegedly occurred in front of her children. They stated the only reason why

Hannah was considered releasable was because she would be living with her surety in Hamilton.

37 While roots to the community could reflect employment status, the JP made it seem like she was referring to the fact that Lynn had a fixed address. There was actually very little discussion about Lynn’s employment status beyond owning her own business (it was never discussed what kind of business or for how long). 102

There was no regard for the seriousness of Simon’s case despite the fact that he did not fear authority and allegedly said he would “shoot and kill the police.”

Pair 15- Paula and James: Neither Paula nor James were granted bail after being charged with breaching probation, fleeing and assaulting police. Both had lengthy and serious criminal records that included violence and multiple breaches and both had issues with drugs and alcohol.

James was employed full time and lived with family at the time of arrest. His father came forward as his surety despite having issues with James in the past. Based on the Crown’s submissions, James could not be trusted while on bail due to his issues with alcohol, his surety’s inability to properly supervise, and his lengthy criminal record. Paula, on the other hand, did not have stable living arrangements, had borderline personality disorder, and had two children

(neither of whom she had custody of). A female friend came forward as her surety despite only knowing Paula for a couple of months. According to the Crown, there was an increased likelihood that Paula would reoffend while on bail due to her serious criminal record and multiple breaches. In both cases, the JP denied bail. For Paula, the JP stated that the current bail plan did not help address the Crown’s primary and secondary concerns because there was no plan for treatment, she had a history of breaches and violence, and her surety was naive and not well acquainted with her. For James, the JP was concerned that James would be a threat to public safety. The JP emphasized James’ serious criminal record, which included many breaches, the seriousness of the allegations, and the inability of the surety to properly supervise. While many of the justifications were similar, a major reason why the JP denied Paula bail was because there was no treatment arranged for her post-release. This was not a concern for James.

103

Similar outcomes and justifications

In the following four cases (16-19), both the men and women received very similar outcomes and justifications. Unlike the above cases, the majority of men and women in this section did not have problems with drugs or alcohol, they did not have known mental health issues, and they all lived in fixed accommodations.

Pair 16- Donna and Wally: Both Donna and Wally were charged with minor crimes and had minimal, outdated criminal records. In particular, Donna was charged with possession of counterfeit money while Wally was charged with theft under and breach of bail. Both were employed fulltime, had children, and neither had issues with drugs or alcohol. In both cases, the

Crown did not require a surety in order to consent to release although Donna had the support of her family and Wally had the support of Youth in Conflict. Given the circumstances, both Donna and Wally received similar bail releases in terms of restrictiveness. Wally was released on the condition that he remain under the supervision of Youth in Conflict and report there weekly and reside at a residence approved of by Youth in Conflict. Because Donna still had strong ties to

Vietnam, she was to reside at her stated residence and not change without prior approval, surrender her passport within 48hrs to the RCMP, remain in Ontario, report to the RCMP weekly, and not possess any replica currency unless permitted under s.457 of the C.C.C. While

Donna received more conditions, they did not severely limit her anymore than Wally.

Pair 17- Yolanda and Henry: In this pairing, Yolanda was charged with assault and assault with a weapon after throwing a cup of tea at her ex-boyfriend and hitting him in the back of the neck. Henry was charged with assault after he punched his girlfriend in the face.38 Henry did not have a problem with alcohol although it was a factor in the allegations. Yolanda did not

38 In both cases, injuries were minor. 104

have a criminal record, had an adult daughter, and lived in fixed rental accommodations. A female friend came forward upon her behalf to act as a non-residential surety. Similarly, Henry rented his home, had a son, and unlike Yolanda, had a very dated criminal record. Youth in

Conflict came forward to act in a supervisory capacity. Both Yolanda and Henry were released because they had minor charges and would be adequately supervised upon release. They both had fairly identical conditions placed on their recognizance. Yolanda was not to contact the complainant or be within 100m of anyplace known to be their place of residence, employment, and/or education, reside at her stated residence and not change without prior approval, be amenable to the routine and discipline of her surety, and not to possess any weapons. Likewise,

Henry was not to contact the complainant or be within 100m of anyplace known to be their residence, employment, and/or education, to be under the supervision of Youth in Conflict and report their weekly, reside at stated residence and not change without prior approval, abstain from alcohol, and not to possess any weapons.

Pair 18- Catrina and Harry: Catrina was charged with two counts of theft under $5000 when she attempted to shoplift $586.00 worth of groceries from a local grocery store. She had no prior record but did have a similar outstanding charge. The Crown did not require her to have a surety because the “charges were not serious enough to warrant a surety.” Harry was charged with theft under $5000 after he tried to steal a bottle of alcohol from the LCBO. Unlike Catrina,

Harry had a criminal record albeit outdated and minor. He was laid off at the time of arrest, was living in a shelter, and had a severe alcohol addiction. The Crown agreed to his release because of his record and minor charges. To satisfy her release, Catrina needed to reside at her stated residence and not change without prior approval, report to the Waterloo Regional Police every week, and not attend the place where the offenses occurred. Similarly, Harry was to reside at the

105

shelter and not change without prior approval, report to the Waterloo Regional Police every week, not attend the place where the offense was committed, and abstain from possessing and consuming alcohol. Despite his problems with alcohol, the JP did not mandate a treatment or counselling condition.

Pair 19- Sarah and Tyler: In this pairing, both Sarah and Tyler were charged with breaching their recognizance. Tyler was employed fulltime, lived in a motel, and was addicted to crack cocaine at the time of arrest. His brother came forward as a surety. Sarah was married and rented her home. Her mother acted as surety. In both cases, the Crown consented to release because there was a residential surety available. Although Tyler was given a curfew between

11:00pm – 6:00am, they both received similar releases. The curfew order placed on Tyler’s recognizance was relatively lax and did not interfere with his daily routines. Interestingly, he did not need to attend treatment or counselling for his drug issue.

4.5 Conclusion

This chapter presented the results from the thematic, quantitative, and deep sample exploratory analyses. The thematic analysis highlighted the major justifications used by JPs to both grant and deny bail and the most frequently used bail conditions placed on a recognizance.

The quantitative analysis helped further identify key predictors of bail outcomes, while the deep sample exploratory analysis assessed whether there were gender differences in the way men and women were managed by the Courts. The importance of the findings resulting from these analyses will be discussed in the next chapter.

106

Chapter 5: Discussion

5.1 Introduction

The primary objective of this study was to examine whether a gender disparity existed within bail decisions and why. The current study extends the scope of prior research, which focuses mainly on decisions to grant or deny bail, by assessing the types of release outcomes imposed on accused men and women. This chapter outlines the significance of this study’s results by drawing connections to existing literature. It begins by discussing key findings as they relate to the objectives of this study and prior research, followed by a detailed overview of potential theoretical and policy implications.

5.2 Similarities in Justifications Used to Grant/Deny Bail

One of the main goals of this study was to investigate the types of justifications used to grant and deny bail for accused men and women. Overall, the results of the current study show that minimal gender differences existed in the types of justifications used to grant or deny bail.

For both men and women, JPs frequently considered the level of proposed bail supervision when formulating their decisions to grant or deny bail. Unlike prior research, however, there was limited evidence to conclude that maintaining public confidence in the administration of justice played an overwhelming role in decisions to deny bail. Of the cases that resulted in the denial of bail, only one JP justified his decision based on public confidence, making it the least used type of justification to deny bail.39 This appears to be inconsistent with prior research (McLellan,

39 The accused in this case was charged with many firearm related offenses, including trafficking, possession of a restricted firearm, and possession of a firearm with no licence. He was also charged with breaching his recognizance. Given the seriousness of the allegations and strength of the Crown’s case, the JP denied bail in order to maintain confidence in the administration of justice. He also detained the accused based on primary and secondary grounds. At the time, this case garnered the attention of local media because of the involvement of illegal firearms. 107

2010; Stuart and Harris, 2005) that suggests that since the inclusion of sec. 515(10)(c) of the

C.C.C., JPs are more inclined to deny bail for both men and women in order to protect community interests and maintain confidence in the administration of justice in some cases.

These contradictory findings may be attributed to one of three factors: the broader discretionary scope of bail decision-making, the uncertainty surrounding the meaning of public confidence, and the relatively limited number of bail denials in the current study’s sample.

As discussed in the literature review, JPs have considerable discretionary power when deciding to grant or deny bail. Indeed, the addition of sec. 515(10)(c) has further broadened this power by providing JPs with an additional tool to deny bail (McClellan, 2011). Under sec.

515(10)(c), JPs can seek the detention of an accused based upon the perceived public’s reaction to release. To aid in their decision, JPs are urged to consider offense seriousness, the nature and quality of the accused’s alleged conduct, the nature of the community in which the offense took place, and the strength of the Crown’s case. However, according to Trotter (2010), these criteria

“fail to provide much guidance on how sec. 515(10)(c) is to be applied on a case-by-case basis,” as many of them overlap with additional grounds for detention already outlined in the C.C.C.

(Trotter, 2010: 3-3). Because of this, it becomes increasingly difficult to determine the use of the public confidence ground in bail decisions. Given the relative invisibility of judicial decision- making at the bail stage, it is possible that while public confidence did not explicitly influence bail outcomes, it was a guiding factor when denying bail for other reasons, such as an accused’s prior criminal record. In other words, JPs might have justified their decision to deny bail based on accused characteristics but did so in order to ensure confidence in the administration. The inherently discretionary nature of bail decision-making makes it difficult to assess this relationship.

108

While McLellan (2011) assumes that JPs are more than willing to use sec. 515(10)(c) to deny bail, the current study suggests that this may not be the case. In this regard, it is possible that the historical development of this clause has caused confusion around what public confidence actually means and when it should be applied, limiting its use in bail decision-making overall. According to the majority decision in R. v. Hall,

Section 515(10)(c) sets out specific factors which delineate a narrow set of circumstances under which bail can be denied on the basis of maintaining confidence in the administration of justice. . . . [S]ituations may arise where, despite the fact the accused is not likely to abscond or commit further crimes while awaiting trial, his presence in the community will call into question the public’s confidence in the administration of justice. Whether such a situation has arisen is judged by all the circumstances, but in particular the four factors that Parliament has set out in s. 515(10)—the apparent strength of the prosecution’s case, the gravity of the nature of the offence, the circumstances surrounding its commission and the potential for lengthy imprisonment. Where, as here, the crime is horrific, inexplicable, and strongly linked to the accused, a justice system that cannot detain the accused risks losing the public confidence upon which the bail system and the justice system as a whole repose (at para 40).

While the majority claimed that maintaining public confidence was narrower and more precise than the former public interest ground (which was struck down), their decision did little to clarify instances in which JPs should use sec. 515(10)(c) to deny bail (Trotter, 2010). As argued by the dissent, “the problem with s. 515(10)(c) is that, stripped to its essence, its very purpose is to allow subjective fears to form the sole basis by which bail is denied” (Ibid, at para 107). As shown in Table A-12, only one accused was denied bail in order to maintain confidence in the administration of justice. As such, JPs might have avoided using this clause explicitly, due to a lack of clearly articulated guidelines.40 At the same time, however, the limited findings regarding

40 At the same time, the minimal use of the public confidence ground could also be attributed to the fact that the majority decision in R. v. Hall recommended that it be used sparingly in bail proceedings. However, Trotter (2010: 3-47) argues that “if one were not reading Hall very carefully, it would be difficult to spot this prescription, buried in a quote, cited for a much broader proposition.” Thus, he concludes that, “it is questionable whether this passage was intended to have taken on the significance it has assumed.” 109

the use of the sec. 515(10)(c) may also be due to the small number of accused persons denied bail in the current sample. This limits the amount of variation amongst decision outcomes, making it difficult to conclude whether this clause plays a role in bail outcomes. Future research could assess this relationship by obtaining a larger sample of individuals who are denied bail.

Similar to past research, the results of both the quantitative and qualitative analyses show that sureties played a prominent role when determining bail outcomes for accused men and women alike. For example, the results of the thematic analysis show that JPs frequently used statements related to the surety when justifying their decisions to grant bail or deny bail. As shown in Table A-3, justifications related to the surety were used to grant bail most often, especially in cases where the accused had a sufficient surety, a residential surety, or multiple sureties. Based on this, it appears that JPs were relying on sureties to ensure an accused attended future court dates, did not reoffend, and followed all court directives. As Myers (2009) suggests, offloading responsibility onto the surety protects a JP from any negative attention s/he may receive if the accused reoffends while on bail. This is further exemplified by the justifications used to deny bail. For example, JPs regularly denied bail in cases where the accused’s surety was unable to properly supervise or had a weak bail plan in place. While this may not be the only reason for denying bail, it shows that a considerable amount of attention was being placed on the ability of the surety to supervise the accused, even if the accused posed a minimal risk. The quality of the surety was therefore a major focus in bail outcomes for both accused men and women.

According to Trotter (2010), surety releases are one of the most onerous forms of release because the accused is under constant surveillance. Although the C.C.C. recommends using sureties when there is a concern that the accused may abscond or reoffend, the results of the

110

current study show that this directive was not being followed. This is in line with recent studies that found that sureties were an important aspect of most bail outcomes in Ontario (Myers, 2009;

Sprott and Doob, 2011; Sprott and Myers, 2011). In particular, the bivariate analysis illustrates that 87 percent of accused women and 91 percent of accused men had the support of either a surety or the bail supervision program at the time of their hearing. Of those released, 91 percent were under the supervision of a surety or bail supervisor. Moreover, the regression results show that surety availability was a significant predictor of bail outcome, regardless of accused gender.

Accuseds most likely to be released were supported by Youth in Conflict with the Law, followed by those with a surety. Accuseds who were least likely to be granted bail were those who had neither a surety nor a bail supervisor. While it was expected that those with a surety would be granted bail most often, Youth in Conflict was a better predictor of bail releases. Taken together, the results of both quantitative analyses suggest that bail supervision was a key focus in the determination of bail for almost all accuseds in my sample. While surety releases are meant for more risky accuseds (i.e. those who may potentially miss court or reoffend if released), the results show that JPs required less risky accuseds to have a surety without considering less onerous options. It is unlikely that a surety would provide much assurance to the court in cases where the accused already presented a minimal risk. Thus, it is possible that JPs were simply being more cautious in their decisions in order to minimize potential reputational risks.

While the thematic and bivariate analyses found that the majority of justifications used to grant and deny bail were similar for men and women alike, some minor differences emerged during the deep sample exploratory analysis that warrant discussion. In three matched pairs, the court justified their decisions differently for the accused men and women despite having similar outcomes. Notably in two cases, the women were released because of their role within the family

111

and the community while the men were released because of their employment and lack of criminal background. In the third pairing, the woman was denied bail for not having treatment in place while the man was denied bail because of his criminal record. These results are consistent with past research that has found that certain legal and extra legal factors affect bail decisions differently for accused men versus. women; the only difference being they did not have as large an effect as this literature suggests (Ball and Growette-Bostaph, 2009; Daly, 1987; Jeffries, 2001;

Kruttschnitt, 1984; Kruttschnitt and McCarthy, 1985). Although it appears that gender did not substantially influence justifications to grant or deny bail with my sample, a larger sample might have found results more in keeping with the literature. Nonetheless, a gender disparity was most evident in the pairing of Lynn and Don. Although both were employed, it only appeared to be a determining factor in Don’s release. Instead of focusing on Lynn’s employment, the JP cited her

10-year old daughter and her ties to the community as key reasons supporting her release. Like

Ball and Growette-Bostaph (2009), Daly (1987), Jeffries (2001), and Kruttschnitt and McCarthy

(1985), the findings from the current study seem to suggest that children and employment were important bail release factors in a minority of cases; the former being more important for women than for men.

Indeed, motherhood was a common factor for two of the three women in my sample.

While the literature suggests that women with familial responsibilities tend to be released most often because of the court’s preoccupation with child protection (see Daly 1987; Kruttschnitt and

McCarthy; 1985), it is also possible that motherhood can be viewed as a source of risk that is unique to women. In other words, women with children may be considered more risky than women without children in some situations. For instance, the court viewed Hannah as being more risky simply because her offense allegedly took place in front of her children, which could

112

account for the onerous conditions governing her release. Although she had similar release criteria imposed as her matched pair Simon, her release was more restrictive when compared to the other women in the category. According to Hannah-Moffat (2007), the courts do not scrutinize accused fathers on their relationships with their children in the same way they scrutinize mothers. Because there is greater interest in the relationships women have with their children, this becomes a means to manage improper female conduct. It is therefore likely that if

Hannah committed her offense elsewhere, she would not have been considered as risky. Thus, it is important to not only consider how certain factors affect men and women differently, but also how certain factors affect different men and women. Contrary to the thematic and bivariate analyses, the results of these three matched pairs provide some cursory evidence to suggest that gender differences did exist in bail decision making. When taking into account the cumulative effect of all three analyses, however, the results suggest that these differences may not be as pronounced as outlined in the literature, making it difficult to conclude that gender had any substantive effect on the types of justifications used to grant or deny bail.

5.3 Emerging Gender Differences in the Number and Type of Conditions Imposed on a

Release

A second major goal of this study was to examine the different release conditions imposed on accused men and women. Unlike the types of justifications used to grant or deny bail, gender appears to have a greater affect on the types of releases imposed on accused men and women. The results of the quantitative and qualitative analyses yield two key findings: 1) that more restrictive forms of release were popular for those granted bail, and 2) that gender played a more significant role in the types of releases imposed on accused persons. Similar to previous studies (Myers, 2009; Sprott and Doob, 2010), all accused persons in the present study who were

113

granted bail had at least one condition placed on their release. As shown in chapter four, an average of five release conditions were placed on recognizance orders for both men and women, which is low compared to Myers (2009) results. While Myers found that over 50 percent of accused persons released had five or more conditions imposed on their recognizance, those who were released following a contested hearing had upwards of ten or more conditions attached to their recognizance, increasing the overall average. The low number established in the current study could be explained, in part, by the fact the Crown consented to release in 62 percent of cases, reducing the average number of conditions overall. It can be assumed that those who had a consent to release hearing were less risky and therefore required fewer conditions. However, considering release conditions are to be reserved for cases in which the Crown has shown cause, the increasing number of conditions placed on recognizance orders for both men and women seem to suggest that JPs were being more mindful of their reputations when crafting release decisions. Similar to the use of sureties, it is unlikely that imposing more conditions on an accused, who is already less risky by virtue of the allegations or their criminal background, will prevent him/her from missing court or reoffending. Thus, by imposing more release conditions upon accused individuals, it is possible that JPs were attempting to minimize potential risks to their reputations and appear in control.

According to Trotter (2010), JPs should use restraint when imposing bail conditions, making them as least restrictive as possible. Indeed, there are punishments that result from not abiding by a bail condition, such as loss of money, arrest, and even detention (Doob and

Webster, 2012; Trotter, 2010). Similar to pre-trial detention, the actual conditions themselves may be experienced as a form of punishment, as they restrict certain freedoms like mobility, association, and communication. Based on the current study, the most frequently imposed

114

conditions were those that restricted possessions and residency. While some of the conditions under these categories, like prohibiting the possession of weapons, were enumerated in the

C.C.C., the majority of them were not. The most used non-enumerated condition limited the consumption and possession of drugs and/or alcohol. Although JPs may be justified in imposing this condition if the allegations referred to drugs and/or alcohol, it is likely that this type of condition may invite a breach if it is unlikely that the accused could abstain from drugs and/or alcohol while released. The potential link between release conditions and non-compliance rates has caused some scholars to criticize release decisions and urge JPs to take care when crafting release orders (see Doob and Webster, 2012; Trotter, 2010). The frequent use of conditions suggests that this type of consideration was not being taken.

According to the preliminary bivariate analysis, a significant relationship between type of condition and gender does not exist, except in relation to treatment/ counselling-conditions.

While the thematic analysis shows this as being one of the least used conditions placed on a recognizance, the bivariate results show this to be a seemingly popular choice for accused females. Of the women in my sample, 43 percent had a treatment or counselling-like condition placed on their recognizance compared to only 22 percent of men. This finding is consistent with those of Sprott and Doob (2010) who found that girls were significantly more likely to receive a treatment-like condition compared to boys. According to Trotter (2010), treatment is one of the most invasive conditions placed on a recognizance order because it requires some form of rehabilitation. Although JPs may find it necessary to impose this type of condition to prevent an accused from reoffending while on bail, its intrusive nature makes it inappropriate for less risky accuseds (Trotter, 2010). When examining the relationship between gender and secondary ground concerns, women were significantly less likely to be seen as a risk for reoffending while

115

on bail when compared to men. Despite being a less risky group in this regard, women were still more likely to have a treatment-like condition placed on their release relative to men. This finding questions the increased use of treatment-like conditions for women more generally.

Accordingly, it is possible that gender did play a more substantive role when determining bail outcomes than the initial bivariate results indicate. Indeed, Sprott and Doob (2010) attribute this trend to a continued paternalistic concern for girls within the courts. The current study provides little evidence to conclude that this explanation can easily translate to the experience of accused women. In particular, the matched pair analysis shows that only certain types of women had restrictive conditions placed on their releases, notably women who used drugs. This group, regardless of criminal background and charge seriousness, were consistently given more onerous release conditions when compared to their matched male pair. In this regard, releases were generally composed of conditions that forced women to engage in treatment/counselling and obey either a curfew or house arrest. For example, Tiffany and Graham were both charged with minor breaches. Tiffany had a severe drug addiction while Graham had a serious alcohol problem. Unlike Graham, Tiffany had to engage in treatment and obey a 24 hour house arrest, making her release far more onerous than Graham’s release. Although the rationale behind imposing these types of conditions is one of protection and prevention, there were no major differences in the number of accused men and women who used drugs or alcohol. This seems to suggest that women who used drugs were considered more risky compared to their matched male pairs.

At the same time, however, women with mental health problems were given less restrictive releases in comparison to their matched-male pair. In five pairings, the accused man actually received more onerous bail outcomes compared to his matched female pair. Mental

116

health appeared to be an aggravating factor for men in two cases, with one man receiving more onerous conditions and the other man receiving detention. For example, Mike, who suffered from severe anxiety and depression, was denied bail because his mental health status made him unstable and a greater risk for reoffending. Meanwhile, his matched female pair, Paige, was granted bail despite suffering from post-partum depression. This is consistent with prior research

(Ball and Growete-Bostaph, 2009; Jeffries, 2001; Kruttschnitt, 1984) that found that men were viewed as a greater risk to the public compared to women with similar charge and background characteristics regardless of possible mental health explanations. According to Jeffries’ (2001) study, men’s mental health problems are often demonized by the seriousness of their allegations.

As a result, the types of outcomes granted to men, like Mike, demonstrate how mental health issues contributed to them being portrayed as a risk to the public compared to their matched female pair.

The depiction of women’s mental health issues differed substantially from the men in my sample. Specifically, mental health seemed to be a mitigating factor in three release outcomes for accused women. Unlike accused men, these women were portrayed as a bigger risk to themselves and were therefore still able to communicate freely with their children and the complainant without the use of a 3rd party. This difference is notable in the pairing of Alana and

Bob, where Alana’s charges were attributed to her post-partum depression. Despite threatening to kill herself and her spouse, she was able to communicate freely with him and see her children while released. This same treatment did not extend to Bob, who slapped his spouse in the head.

Although he too suffered from mental health issues (bipolar) that could explain his outburst, he was prohibited from seeing his spouse and his children, except through a 3rd-party. In this example, Alana was viewed as a bigger risk to herself resulting in her admittance to the hospital

117

while Bob was a greater risk to his family resulting in restricted communication. Similar to

Jeffries (2001), it appears that mental health status provided an explanation for female criminality, neutralizing risk for Alana, Paige, and Jessica. While this is consistent with past research that found that women were more likely to be seen as a risk to themselves, there was little evidence in the current study to suggest that this would result in more women being remanded to custody (see Edgar, 2004; Player, 2007). Potential explanations for this finding could be that the women in my sample who had mental health issues did not commit very serious offenses and that there was only one woman who was denied bail, making it difficult to assess the types of women denied bail. It is also possible that the types of conditions placed on women with mental health issues lead to non-compliance charges, which may increase their chances of being remanded. However, more research is needed to investigate this link further.

Overall, the number and types of conditions placed on a release may have negative consequences for the accused. While the number of release conditions was similar for accused men and women, several gender differences emerged from the qualitative analysis. Women were slightly more likely to receive a treatment/counselling condition on their release compared to men, even though there was no statistically significant relationship between drug and/or alcohol dependency and gender. This was supported by the results of the deep sample analysis, which found that women with drug issues and men with mental health issues were given more restrictive bail outcomes compared to their matched pairs. Although these conditions are supposed to serve a preventative function, they may be experienced as punishment by the accused. Being forced into medical treatment, for example, may be viewed as a form of coercive control by the state because the accused has to actively engage in some form of rehabilitation in order to satisfy his/her bail plan. In most cases, treatment is not of a determinative length

118

(whereas a prison sentence is), which may seriously impede an accused’s right to reasonable bail.

How an accused perceives his/her release may differ based on gender and individual circumstances. For instance, a non-communication order may be more onerous for an accused woman who can no longer communicate with her child than for an accused man, depending on his role within the family. This makes assessing release orders difficult, as it is hard to speculate how an accused may or may not experience his or her release. As a result, the gendered differences in bail outcomes may be understated.

5.4 Decisions to Grant Bail: Women Released More Frequently but not Treated More

Leniently

Studies that have found a relationship between gender and bail conclude that women are more likely to be released and have lower bail amounts relative to accused men. At the bivariate level, the results of the current study are inconsistent with past research as there was not a statistically significant relationship between gender and bail outcome. More support for this relationship was found at the multivariate level. Although the results of the logistic regression show that women were significantly more likely to be released on bail compared to men (after controlling for key legal and extra-legal factors), the statistical strength of this relationship is inconclusive due to a small sample size (n=115) and limited number of women denied bail.

Nevertheless, the finding suggests that, similar to the literature, gender did play some role in the determination of bail. Indeed, the bivariate relationship between gender and bail amount was statistically significant at p=<.05, suggesting that women were also more likely to receive a lower bail amount compared to men. On average, bail amounts for women were between

$1000.00 and $2499.99 while for men this number was between $2500.00 and $4999.99. When taken together, the results of the bivariate and multivariate analyses seem to imply that women

119

were not only more likely to be released, they were also more likely to be given lower bail amounts.

The apparent gender disparity in bail outcomes has led some scholars to maintain that women are treated more favourably or leniently by the courts at the pre-trial stage (for example, see: Kruttschnitt, 1984; Kruttschnitt and McCarthy, 1985; Visher, 1984). However, the results of the multivariate and deep sample analyses find little evidence to suggest that women received preferential treatment compared to men. This inconsistency can be attributed to two factors.

First, the previous studies tend to overlook the types of releases imposed on accused men and women, which can be just as onerous as pre-trial detention. As shown in the preliminary bivariate analysis, 91 percent of accused men and women were released under the supervision of a surety or bail supervisor and all had an average of five conditions placed on their release. More specifically, the results of the bivariate and deep sample analyses suggest that women were more likely to be released, certain groups of women received more onerous release orders compared to their matched male pairs. This push towards more restrictive releases makes it difficult to assert that women are being treated more leniently compared to men, as the number and type of conditions can lead to punitive action. In particular, the restrictive nature of these releases may be experienced as punishment, making it difficult for women to comply with their recognizance.

Thus, it is difficult to conclude that women are treated more leniently compared to men without exploring the types of releases imposed on accused women.

Second, many of these studies fail to account for important gender differences that shape men and women’s offending behaviour. When looking at male and female criminality, there are a variety of social, historical, and cultural factors that influence how men and women enter into crime (Chesney-Lind, 2006; Daly, 1994; Daly and Chesney-Lind, 1988; Franklin, 2008; Hannah-

120

Moffat, 1995; Maidment, 2008). Unlike men, women generally enter the criminal justice system because of problems related to abuse, poverty, and drug use (Boritch 2008, Maidment 2008).

Another common predictor of female criminality is single parenting as it can lead to poverty and welfare dependency (Maidment, 2008). Because of these circumstances, women are also more likely to commit non-violent, “survival” based crimes, such as theft or fraud, and are therefore less likely to have a serious criminal record (Boritch, 2008; Franklin, 2008). These findings are corroborated by the current study. While the bivariate analysis does not show a statistically significant difference in charge characteristics between accused men and women, it does show that women were less likely to possess a serious and/or related criminal record. Compared to accused men, the Crown was also less likely to have secondary ground concerns or seek detention in cases involving accused women. These results seem to indicate that accused women were a less risky group overall and, therefore, better candidates for release relative to accused men.

As the literature suggests, men are generally more risky than women based on legal characteristics alone (Ball and Growette-Bostaph, 2009; Kong and Aucoin, 2008; Kruttschnitt,

1984). This is consistent with the findings of the current study, which show that men were not only more likely to have a serious and/or related criminal record, they also more likely to be seen as a risk post release when compared to women. These legal differences could partly explain the differences in outcomes for both men and women. For example, both the bivariate analysis and logistic regression illustrate the importance of prior record in bail outcomes, with those having a serious or somewhat serious criminal record being more likely to be detained than those without.

Because men were more likely to possess a serious record, it stands to reason that they were considered more risky and were denied bail more often simply by default. Consequently, it is

121

difficult to conclude that men or women are treated more or less favourably without discussing potential legal factors influencing bail outcomes and the types of outcomes imposed on accused men and women. In general, the results of this study find little evidence to suggest that women are treated more leniently at the pre-trial stage.

5.5 Theoretical Implications

Chapter three described four main theoretical approaches that could explain possible gender disparities in bail outcomes – namely, legislative changes, risk aversion, risk management, and governmentality. Broadly speaking, the results show minimal evidence to suggest that legislative changes played a large role in bail outcomes for accused men and women. Greater support was found for the theory of risk aversion, although this theory was limited in explaining how risk was defined and managed by the courts. To account for the gender disparities that emerged during the quantitative and deep sample exploratory analyses, risk management and governmentality appear to be most appropriate.

In theory, recent legislative changes in the determination of bail seem to suggest that public confidence is playing a greater role in decisions to grant or deny bail. Indeed, the addition of sec 515(10)(c) in the C.C.C. has led some scholars to conclude that JPs are more inclined to favour public sentiments over accused rights, resulting in more men and women being denied bail than in the past (McClellan, 2011; Stuart and Harris, 2005). This explanation does not seem to comport with the findings of the current study, which suggest that maintaining confidence in the criminal justice system played a minimal direct role in the determination of bail. Similar to

Moore and Hannah-Moffat’s (2005) conclusion about the role of punitive action in Canada, there appears to be some disparities between the rhetoric and practice of penal administration in

Ontario. In the context of bail, sec. 515(10)(c) allows JPs to deny bail based on the public’s

122

reaction to release. While politicians and the majority of the Supreme Court emphasized the importance of public confidence in the criminal justice system, this rhetoric does not seem to translate easily into practice as only one JP denied bail for this reason. Unfortunately, these differences were not accounted for by McLellan (2011), who simply relied on bail and remand trends as evidence to suggest that since the addition of sec. 515(10)(c) JPs were less likely to grant bail. Although in theory the emphasis on public confidence in sec 515 (10)(c) makes it easier for JPs to deny bail, this does not correspond with the results of this study. As a result, scholars should take care when using this theory to understand current trends in bail decision- making in Ontario.

Instead, the results of the current study provide some support for the theory of risk aversion in explaining bail outcomes. Similar to Myers (2009) and Webster et al. (2009), the frequent use of sureties and conditions in bail decision making can be explained, in part, by the need of JPs to manage reputational risks and avoid blame for potentially risky decisions (see

Douglas’ blame society theory, 1992). At the bail level, JPs are tasked with determining future risk based on a balance of probabilities. The expeditious nature of bail hearings means JPs need to make quick decisions about the credibility of the allegations and often do not get a chance to thoroughly review evidence (Myers, 2009). Without a methodical review of the facts, JPs may be reluctant to make risky decisions due to the potential consequences that may arise if the accused reoffends or absconds while on bail. This could explain the results of both the thematic analysis and logistic regression, which showed that sureties played a significant role in decisions to grant or deny bail. Diverting responsibility ensures that if an accused reoffends or breaches a condition while on bail the blame is directed to the surety or bail supervisor rather than the presiding JP

(Myers, 2009; Power, 2004). This could explain why those supervised by Youth in Conflict were

123

the most likely to be granted bail. The array of resources, programs, and services offered by the bail supervision program works to reduce the risk of an accused once released. As a result, the expected level of release supervision offered by these types of programs can be more easily predicted than that offered by a surety, minimizing any potential release concerns a JP may have.

The desire of JPs to maintain their reputation can also help clarify the extensive use of release conditions found in the current study. When granting bail, JPs imposed an average of five conditions on accused men and women, which was often in addition to some form of bail supervision. In particular, many of the conditions placed on recognizance orders did not seem connected to the original allegations. For example, of the 57 cases that resulted in a no weapons order, only nine involved the alleged use of weapons. For the majority of these cases, it is doubtful that a weapons prohibition would do much in terms of preventing an accused from reoffending or missing court. It is therefore possible that in some cases JPs were simply imposing excessive and unrelated bail conditions in order to give the “impression of manageability” rather than to prevent future risk (Myers, 2009: 145). In fact, recent studies have shown that more conditions placed on a recognizance actually increase the likelihood that an accused will reoffend on bail. Similar to the widespread use of sureties, the large number of conditions placed on a release order seems to suggest that the JPs in my study were more preoccupied with minimizing reputational risks than those presented by the accused. Where this theory is lacking, however, is in its assessment of who is risky and the types of releases imposed.

While it posits that risk is based on individual perceptions of ‘riskiness’ (Doob and Webster,

2012; Myers, 2009), the current study found evidence to suggest that certain groups of accused men and women were considered more risky than others, leading to different outcomes based on gender. As a result, this theory only partly explains the results of this study.

124

To address this gap, I argue that risk management helps clarify the gender differences that were located within my sample. Broadly, Hannah-Moffat and O’Malley (2007) suggest that risk predictions are based on matching an accused person to groups of people with known risk factors. Common risk factors include prior criminal record, temporary housing, unemployment, and low income (Kellough and Wortley, 2002). According to the logistic regression, women were more likely to be released compared to accused men after controlling for legal and extra- legal factors. This can be explained using the risk literature. As outlined in chapter two, accused women are generally considered to be less risky than men because they are charged with less serious offenses and possess less severe criminal records. In many cases, this makes detention inappropriate. However, Myers (2009) argues that despite the fact that “these variables can only be subjectively assessed, they are often used in a manner that suggests they are objective criteria with predictive capabilities” due to presiding neoliberal values. Here, neoliberal and gendered discourses interact to form risk categories that are unique to accused men and women. According to neoliberal citizenship discourse, self-governance and responsibility are essential characteristics that minimize the costs and harms associated with risky behaviour. Although all individuals are encouraged to uphold these values, citizenship is also characterized by a gendered rhetoric that expects women to be responsible for their families and men to be responsible financially (Turnbull and Hannah Moffat 2009). Men and women who deviate from these norms are therefore considered more risky due to their associations with risky populations (Franklin,

2008; Hannah-Moffat and O’Malley, 2007; Moore and Lyons, 2007).

In particular, the results of the deep sample analysis show that women who received more onerous releases were those who used drugs. This corroborates the findings of Moore and Lyons

(2007) who conclude that women who used drugs were normally seen as more risky not because

125

of their criminal history but because of their associations with risky subpopulations. As explained by Madriz (1998), drug-using women are often poor and transgress traditional female norms that expect women to be caring and responsible mothers and wives. Presumably, the results of the current study can be attributed to the fact that women who used drugs were seen as being more risky simply for falling outside the nexus of appropriate behaviour. Indeed, the court attributed Cathy’s charges to her alcohol problem, calling it the “cause of her troubles.” This resulted in her having a curfew placed on her release in addition to counselling. Here, Cathy’s inability to be socially responsible and in control of her own risk led to her receiving a more onerous release compared to her matched pair, Glenn. While feminists would argue that these women are being treated more harshly for “straying from the paths of traditional womanhood”

(Chesney-Lind, 2006; 504), the risk literature suggests that these qualities make women more risky due to their associations with risky subgroups.

Mental health was also identified as a key site of variance where explanations for bail outcomes could be attributed to gendered notions of risk. Specifically, women with mental health issues received less burdensome conditions than their matched male pairs. This finding can be rationalized by the fact that the courts often view female criminality as being rooted in some psychological manifestation (Turnbull and Hannah-Moffat, 2009; Jeffries, 2001). As Jeffries

(2001) explains, “female criminality is integrally intertwined with pathology because it provides another way to reconcile women’s offending within an overriding feminine discourse...this works to refeminise offending women as real women and not really criminal after all” (171&

187). For three women in my deep sample analysis, mental health issues diminished the risk they presented to the public, which limited the number and types of conditions placed upon their release. In these cases, the court seemed sympathetic towards the women, seeing them as victims

126

of their circumstances. This appears to have neutralized their riskiness, making more restrictive outcomes seem inappropriate. Men with mental health issues were treated much differently by the courts, despite having relatively similar issues to the women. In these cases, men were depicted as unstable and large risks to public safety, resulting in more restrictive outcomes. This appears to reference masculine ideologies that view men as inherently dangerous and irresponsible. As Jeffries (2001) points out, accused men are often viewed as being active participants in the criminal subculture, regardless of mental health status. With this in mind, it is possible that JPs relied on masculine and pathological discourses to classify the men as risks, justifying more onerous outcomes. In all cases regarding mental health, it appears that risk management can explain some of the variation in bail outcomes for accused men and women.

A potential critique of scholars using this theory to explain gender disparities in bail outcomes is that they tend to focus solely on accused women without a clearly articulated reference group (for example, see: Boyd, 2006; Moore and Lyons, 2007; Turnbull and Hannah-

Moffat, 2009). In other words, it is not said whether women are more or less risky in comparison to men or other women. From a feminist perspective, this woman-centred framework is understandable. Comparative studies generally use men as the reference group, which assesses female criminality in relation to men. This has the potential to not only reinforce sexist stereotypes that already exist within the criminal justice system, it also works to keep women subordinate (Chesney-Lind, 2006). However, maintaining that risk classifications are unique to women is difficult to do without understanding how masculinity and men’s situations also influence perceptions of risk (Daly and Chesney-Lind, 1988). This is particularly relevant in areas that have received little scholarly attention, like bail decision-making in Ontario. Moving

127

forward, scholars using this theory should clearly outline a reference group (whether it be males, females etc.) or explain why a reference group is not necessary.

Governmentality also offers a useful and complimentary framework when analyzing differences in bail outcomes for accused men and women. While risk management outlines why certain individuals are classified as risks, governmentality explains the ways in which these individuals are managed by the courts (Foucault, 1971). In the context of bail, this theory helps explain why JPs used certain outcomes more frequently than others depending on the accused.

This theory differs from risk aversion in that it suggests that JPs consciously impose particular bail outcomes in order to shape individual conduct (Foucault, 1971; Turnbulll and Hannah-

Moffat, 2009). Thus, it argues that instead of reducing reputational risks, JPs are focused on managing the risks of the accused and instilling a sense of responsibility through different release orders such as house arrest, treatment, and drug/alcohol abstinence. Certainly, this is seen when looking at the types of bail outcomes given to certain accused men and women. Although treatment/counselling and house arrest are key features of some post-conviction sentences, they were common conditions imposed on released orders for female drug users. Similar to sentencing decisions, the selection of these conditions appear to manage female risk by encouraging ‘unfeminine’ women to become more morally and socially responsible (Franklin,

2008). In particular, treatment or counselling conditions reinforce the ideal that women are supposed to be in control, compliant, and accountable for themselves and family (Boyd, 2006;

Hannah-Moffat, 2009). Like Moore and Lyons (2007) found, women who used drugs were often placed in treatment/counselling in order to ‘deal’ with their emotions and to become better mothers for their children. It was assumed that once in treatment, accused mothers would want to quit using drugs in order to regain custody of their children. Indeed, nearly half of all the women

128

in my sample had a treatment condition placed on their recognizance, which suggests that the JPs in my sample were perhaps attempting to shape these women’s conduct by imposing conditions that would reinforce standards of feminine morality and normalcy.

It also appears that masculine discourses influenced how JPs managed risk for accused men. Because men are often portrayed as being actively engaged in crime regardless of possible pathological explanations, five men received more onerous outcomes in the current study. This finding may underestimate the true relationship between risk and outcome since my deep sample analysis matched men based on female criminogenic characteristics, which limited the number of men who were denied bail in the analysis. Nevertheless, it appears that JPs gave accused men more restrictive outcomes in these five cases in order to manage the apparent risks they posed to the public. In these examples, men were more likely to be seen as risks to the public, which resulted in certain types of outcomes such as surety releases, house arrests, geographic restrictions, and in one case, detention. Like with the outcomes for accused women, these also appear to instill a sense of responsibility and accountability by removing the accused from his criminal environment and making him more accountable to his surety. Consequently, governmentality and risk management offer a complementary theoretical framework when explaining the gender disparities that existed in the types of bail outcomes bestowed on accused men and women.

5.6 Policy Implications

The results of the current study offer two main policy implications: the need for substantive equality in bail decision-making and the importance of making clearer distinctions between preventative and punitive bail outcomes. Since the 1970s, feminist criminology has provided new perspectives on female criminality, and has been a driving force in many areas of

129

penal reform. More generally, this field is based on the notion that female criminality stems from a male dominated social, economic, and political system (Chesney-Lind, 2006; Comack, 2008).

In particular, it suggests that the criminal justice system revolves around a masculine framework where the male gender is considered the norm or reference group (Franklin, 2008). A common concern amongst feminist scholars (see Chesney-Lind, 2006; Comack, 2008; Daly 1994; Daly and Chesney-Lind, 1988; Franklin, 2008; Hannah-Moffat, 1995) is whether the unique needs and circumstances of women are being adequately addressed. In this regard, there are two concepts of equality that feminists use when attempting to create a more equitable court system; formal and substantive equality. Formal equality or gender-neutral policies suggest that both males and females are treated equally, regardless of gender, while substantive equality assumes that in order to reach gender equality, certain considerations need to be made based on inherent differences between men and women (Maidment, 2009).

The results of the current study suggest that gender played a role in the determination of bail - not only in decisions to grant or deny bail but in the types of release outcomes as well. This comports with previous research that has found that women are more likely to be released and receive lower bail amounts relative to men. For some individuals, this may seem problematic because it suggests that women are being treated more ‘favourably’ compared to accused men.

Certainly, this has been a common conclusion amongst scholars studying the effects of gender on bail, who state that women are granted leniency at the pre-trial stage (see Ball and Growette-

Bostaph, 2009; Jeffries et al., 2003; Kruttschnitt, 1984; Kruttschnitt and McCarthy, 1985; Visher,

1983). The use of the term ‘leniency’ implies that the courts are expressing compassion towards all accused women, regardless of any potential legal and extra-legal factors that may be influencing these decisions. However, unlike men, women generally enter the criminal justice

130

system because of problems related to abuse, poverty, and drug use (Boritch 2008, Maidment

2008). Lone parenting is also a major predictor of criminal behaviour since it leads to poverty and welfare dependency (Maidment, 2008). Because of these circumstances it has been found that women are also more likely to commit non-violent, “survival” based crimes, such as theft or fraud, and are therefore, less likely to have a serious criminal record (Boritch, 2008; Franklin,

2008). Based on my bivariate analysis, accused women were less likely to have a severe criminal record, were not seen as a significant risk post-release, and were more likely to have children relative to accused men, making them better candidates for bail overall. Thus, it is not that women were treated more leniently; it is that they were a less risky group compared to the men in my sample.

While this may indicate that JPs were more considerate of the unique needs and circumstances of women, it appears that this type of consideration was not granted to all men and women equally. Specifically, women who failed to embody traditional feminine characteristics received more onerous releases compared to their matched male pair. As discussed previously, risk literature suggests that judicial decisions regarding risk are based on sex-role differentiation that sees women as risk-avoiding mothers and men as the risk-taking breadwinner. In many ways, this division is further entrenched by woman-centred penal policies. According to Hannah-

Moffat (1995), the strict division made between males and females within the criminal justice system works to reinforce gender stereotypes, as all women are placed into one homogenous group. In reality, however, women’s experiences are reflective of, and influenced by, different social, economic, and political positions such as their race, class, sexual orientation, and employment status (Hannah-Moffat, 1995; 144). Findings of the current study show that while women were generally more likely to be released on bail and receive less onerous/similar

131

conditions compared to accused men, a group of women who used drugs received more restrictive conditions compared to their matched male pair. It could be argued that, in these situations, JPs did not consider the circumstances and factors that led these women to commit an offense and/or use drugs/alcohol because they did not fit the homogenous female standard. As

Daly (1987) and Kruttschnitt (1984) found, women who were married and/or had children were more likely to be released than other women. With this in mind, JPs may only be considering certain ‘unique’ needs of women (i.e. marriage, children) that fit the standards of proper feminine conduct while ignoring all others. This too is problematic because it simply maintains and imposes standards of feminine morality and normalcy. For substantive equality to truly exist,

JPs need to acquire a broader understanding of women and their different pathways to crime.

This means not only treating men and women differently based on their unique needs, but also taking into account the diverse needs that exist within both gender categories.

JPs should also take great care in the types of releases imposed on accused men and women. Surety releases, for example, can have several undesirable consequences when it comes to bail. First, they can create a backlog within the courts as defense lawyers regularly adjourn cases where the accused is without a surety (see Myers, 2009; Webster et al., 2009). Second, surety releases can make compliance very difficult. As Trotter (2010) argues, surety releases are one of the most onerous forms of release because accused individuals are under constant watch.

Often referred to as the ‘civilian jailer’ by the courts, sureties are not only responsible to ensure an accused follows his/her bail conditions at all times, they are also empowered to impose any additional conditions (within reason) on the accused to ensure compliance. Sureties may place fairly restrictive conditions on an accused as they stand to lose a significant amount of money if

132

the accused violates a bail condition. 41 As such, the use of a surety may be seen as a form of punishment, and when the accused is neither innocent nor guilty, this is extremely troubling.

Third, the high number of conditions placed on a release order (five conditions on average) not only threaten to limit Charter protected rights and freedoms, like the right to be presumed innocent until proven guilty, and freedoms of communication and association, but they can also have indirect effects on the remand rate in Ontario (see Sprott and Doob, 2009). The restrictive nature of many release conditions may make it difficult for an accused to comply with his/her recognizance, leading to a bail violation and additional charges. How an accused perceives his/her recognizance may be influenced by gender-specific needs. For instance, a woman may perceive an outcome to be more cumbersome than an accused man simply due to her circumstances (or vice versa). This blurs the line between prevention and punishment, which may interfere with an accused person’s right to “reasonable bail” and to be “innocent until proven guilty.” This may be particularly relevant for accused women who need to enter into treatment as part of their recognizance. In many cases, the task of finding appropriate medical treatment for the accused falls onto the surety, who may not know what is best for the accused. If an accused refuses to attend, he or she risks violating bail and/or being rendered by the surety.

As stated in the literature review, the number of women charged with violating their bail has risen dramatically in recent years. If this study is any indication of the types of conditions placed on women’s release orders, it is possible that women are having difficulties complying with their releases, and being charged with additional offenses due to breaching the conditions of their bail.

41 Because of the limited research on surety releases, it is hard to say whether sureties are more likely to impose additional conditions on a recognizance. Certainly, the increase in bail violations may suggest that surety releases are fairly lax when it comes to supervising the accused. However, it is equally possible that this increase is the result of sureties rendering the accused for failing to abide by his/her bail conditions or the “routine and discipline” of the household. Hypothetically, it is easier for accused persons to breach their recognizance when they are supervised by a surety or a bail co-ordinator than it is if they were released without this level of supervision. To assess the ability of the surety in supervising the accused, future research could analyze reasons for failure to comply charges. 133

Individuals who are unable to comply with their recognizance risk being charged and arrested with additional offenses, which initiates a reverse onus situation. In these cases, the onus is on the accused to show why his/her release is justified under s. 515(10) of the C.C.C.

According to the decision in R. v. Pearson (at para 144) the purpose of a reserve onus is to ensure the bail system is not “subverted by continuing criminal activity and by absconding accused.” However, this seems to erode presumptions of innocence by assuming some level of guilt. So while many reverse onuses in the current study resulted in a release, it is possible that these releases were more restrictive than a regular show-cause hearing because the accused was already on bail.42 As Sprott and Doob (2009) contend, this may create a revolving door effect whereby accused persons are released only to be brought back to court for not complying with their release orders. It also means that many accused persons, both male and female, are being charged and arrested for minor conduct such as going out past curfew, drinking alcohol, and refusing to go to treatment. Thus, it is possible that some accused persons are being remanded for actions that would not normally result in a prison sentence post-conviction.43 Although credit is given to offenders who have spent time in remand, no compensation is given to those who are acquitted. This is generally the case for women who have spent time in pre-trial custody (Player,

2007; Williams, 2005). Due to the consequences remand conditions can have on accused men and women (see literature review), this is extremely problematic.

Prior to release, JPs should consider whether the condition is rationally connected to allegations and whether the accused is able to reasonably comply. Trotter (2010) argues that many conditions imposed on an accused have nothing to do with the actual allegations and do

42 More research is needed in this regard. 43 The literature would suggest that this is a major consequence of reverse onus situations (see McClellan, 2011; Myers, 2009; Trotter, 2010). Although the bivariate analysis found that many reverse onuses resulted in release, there was not enough evidence to show this as a significant finding. Thus, it is still likely that reverse onus situations increase the likelihood of detention. 134

little to prevent future risk. These types of conditions can create secondary risk channels and make it more difficult for an accused to comply with his/her recognizance. For example, Farah

(Pair 17) was prohibited from using and possessing a cell phone as one of her release orders, with no rationale as to how this was connected to her allegations (trafficking and breaching her recognizance) or how this reduced her chances of reoffending while released. With many people replacing their home phones with cell phones, this condition could make it difficult for Farah to contact help ( in an emergency), potentially placing her in a dangerous situation. If she did use a cell phone to call for help, she could be seen as being in direct violation with her recognizance.

Without a clear purpose, these types of conditions simply invite bail violations and place accused persons at risk. Further, JPs should also place a great deal of attention on whether an accused could reasonably comply with his/her recognizance. For instance, imposing a no-alcohol order on an accused with a severe alcohol problem may be inappropriate if the JP feels the accused would be unable to abstain from consuming alcohol. In these cases, the chances of breaching are increased (Trotter, 2010). Trotter suggests that in cases where the JP feels that a condition is necessary to prevent future offenses but the accused could not reasonably comply, the accused should be remanded. However, if a prohibition on drugs/alcohol is not essential in preventing future risk, then the JP should simply not include it on a release order. This would not only prevent individuals from breaching their recognizance, it would also reduce the punitive aspects associated with some bail releases.

5.7 Conclusion

To summarize, the results of this study suggest that a gender disparity does exist in decisions to grant or deny bail. Although sureties and conditions were key aspects of bail releases for all accused persons, certain groups of men and women received more restrictive

135

outcomes than their matched pair. In particular, women who used drugs and men with mental health issues were considered more risky than their matched pair, prompting JPs to impose more onerous bail outcomes. On a whole, women were more likely to be released relative to men but this is likely attributed to the fact that the women in my sample had minor records and were not a significant risk for reoffending. Thus, there is little evidence to suggest that women received

‘preferential’ treatment in bail court. The implications of the results show that substantive equality has not been met and that the current use of sureties and conditions is blurring the line between prevention and punishment.

136

Chapter 6: Concluding Remarks

6.1 Limitations and Future Research

I conclude this study by discussing the key limitations of the current study, and suggest how these limitations provide important avenues for future research on the relationship between gender and bail outcomes. First, a small sample size limited the empirical findings. With a sample of 83 accused, the results of the logistic regression are inconclusive. Although the third and fourth logit models show gender as being significant at p<.05, the large standard errors suggest this relationship could be due to chance. This makes it difficult to conclude with 95 percent confidence that gender was not only a significant predictor of bail outcome but that women were also more likely to be released compared to men. This is not to say, however, that gender was irrelevant when determining bail outcome. The results do fall in line with prior research that has found that women are more likely to be released relative to men, which suggests that it is likely that gender did play some role in the current study. Certainly, gender differences did emerge during the deep sample analysis. However, in order to report statistically significant results with small standard errors, I would have needed to obtain a considerably larger sample. Given the time restraints placed on the current study, this was unachievable.

Similarly, the findings of both my quantitative and qualitative analyses were also restricted due to the limited number of women in my sample (n= 22). This is a common challenge and concern with most criminological studies analyzing gender since women commit crime less frequently than men (Kong and Aucoin, 2008). In 2007, for example, only one in five people arrested in Ontario were women, meaning they are rarely seen in court as an accused

(Kong and Aucoin, 2008). This affected the current study as only 22 accused women had a bail hearing during my time in court, and only one woman was denied bail, which seriously limited

137

the types of comparisons I could make regarding gender and the denial of bail. Future research is therefore needed to assess whether and how differences exist in decisions to deny bail for accused men and women. This type of study could be expanded to include a comparison amongst women granted and denied bail to see what factors predict these outcomes differently. However, comparative studies are always challenging, particularly when comparing women and men because they risk reinforcing the divide between males and females. In my view, these types of studies are needed in order to understand how certain factors affect men and women differently and whether a phenomenon is unique to women only.

Further, this study focused entirely on judicial decisions to grant or deny bail, which possibly overstated their role in determining risk. While JPs have the ultimate decision-making power, their decisions are heavily influenced by the input of the Crown, the Defense, and police reports. In particular, the Crown plays an important role in the construction of risk as they determine the type of bail hearing and recommend certain release outcomes. These decisions are based on the level of risk presented by the accused and are normally made prior to the hearing.

The expeditiousness of bail hearings and the limited amount of evidence available to the JPs means they rely heavily on these considerations when formulating their decisions. The same can be said for the defense’s portrayal of the accused. If, for example, the defense depicts his/her client as a loving mother, the JP will use this when deciding how much risk the accused presents if released. Without a more thorough investigation of how this type of input influences judicial decisions to grant or deny bail, it is difficult to assert that JPs are the main constructors of risk.

It is also possible that JPs are seeing more risky accused persons simply because they are already being held in custody. According to the C.C.C., police have the ability to release an individual on an undertaking upon arrest. These undertakings are similar to a bail release in that

138

many of them have conditions attached. Presumably, less risky individuals are released on a police undertaking while more risky individuals are brought into custody for a bail hearing. With this in mind, a considerable about of risk construction is done outside of the courts. These decisions are highly discretionary and are normally invisible to higher scrutiny, which means gendered notions of risk may play an even larger role at the time of arrest. As such, it is important that future research examines how risk is classified by police and whether any gender differences exist at this stage, as these decisions may influence the types of people seen by the courts and subsequent bail outcomes.

More research is also needed to examine the relationship between non-compliance and remand rates in Ontario. While research has found that the number of conditions is related to non-compliance rates, little research has been conducted on the impact of non-compliance rates on the increasing remand rates for accused men and women. Indeed, this study suggests that the rising remand rates for women could perhaps be attributed to the types of releases imposed on accused women, particularly those who are forced to engage in treatment or obey a curfew. As stated in the literature review, this can have several negative consequences on the accused since remand centres and provincial prisons suffer from overcrowding and inadequate programming. It is important to recognize reasons for the growth in remand rates in order to make sound policy suggestions. It is equally important to understand how men and women experience their releases and they are shaped by gender. Further insight would not only provide a better sense of why men and women violate their bail, it may also help JPs in crafting future release orders. Although these releases are preventative in nature, how men and women experience these releases may make them feel more like punishment. This is obviously problematic since they are considered neither innocent nor guilty at the pre-trial stage.

139

6.2 Conclusion

Compared to the sentencing literature, few studies have explored the relationship between gender and bail in detail. This is curious given the affect bail decisions have on post-conviction sentencing. Several studies that have investigated gender in the context of bail find it to be a significant predictor of bail outcomes, generally concluding that accused women are more likely to be released relative to men. Like prior research, the results of this study show that gender did play some role in the determination of bail, as women were more likely to be released in comparison to men. However, unlike past studies, this study did not find evidence to suggest that women were being treated more leniently by the courts at the pre-trial stage. As a whole, the women in my sample were less likely to have a serious criminal record and were less likely to be seen as a risk for reoffending, which made them better candidates for release overall. Further, gender variances emerged during the quantitative and deep sample analysis, which indicate that gendered notions of risk did play a role in the determination of bail. In particular, a large proportion of female drug users received more onerous releases compared to their male counterparts. Like the risk literature would suggest, these women were considered doubly deviant for transgressing both feminine and legal norms, which justified more restrictive releases. Men with mental health issues received similar outcomes as they were viewed as being a greater risk to the public relative to women with mental health issues. Without analyzing the types of outcomes placed on accused men and women, it becomes difficult to conclude whether a gender disparity truly exists in bail decisions. These results also suggest that JPs were not considering the diverse needs and circumstances of all men and women in the present sample.

For substantive equality to be achieved in bail outcomes, the diverse needs of men and women needed to be taken into consideration. This is particularly important since the types of outcomes

140

placed on accused men and women could be perceived as punishment, which infringes their right not only to reasonable bail, but also to be presumed innocent until proven guilty.

141

References

Allan, A., M. Giles, D. Drake, and I Froyland. “An Observational Study of Bail Decision Making.” Psychiatry, Psychology, and Law 12, no. 2 (2005): 319-333. Ball, Jeremy and Lisa Growette-Bostaph. “He Versus She: A Gender Specific Analysis of Legal and Extralegal Effects on Pre-Trial Release for Felony Defendants.” Women & Criminal Justice 19 (2009): 95-119. Beck, Ulrich. Risk Society: Towards a New Modernity. Translated by Mark Ritter. London: Sage Publications, 1992.

Bernat, F.P. “Gender Disparity in the Setting of Bail: Prostitution Offenses in Buffalo, NY 1977- 1979.” In Gender Issues, Sex Offenses, and Criminal Justice: Current Trends, edited by S. Chaneles, 21-47. New York: Haworth Press, 1984. Bickle, Gayle and Ruth Peterson. “The Impact of Gender-Based Family Roles on Criminal Sentencing.” Social Problems 38, no. 3 (1991): 372-394. Boritch, Helen. “Women in Prison in Canada.” In Marginality & Condemnation, edited by Carolyn Brooks and Bernard Schissel, 404-432. Nova Scotia: Fernwood Publishing, 2008. Boritch, Helen. (1992). “Gender and Criminal Court Outcomes: An Historical Analysis.” Criminology 30, no. 3 (1992): 293:325. Bottoms, A.E. and J.D. McClean. Defendants in the Criminal Process. London: Routledge & Kegan Paul, 1976. Boyd, Susan. “Representations of Women in the Drug Trade.” In Criminalizing Women, edited by Gillian Balfour and Elizabeth Comack, 131-149. Halifax: Fernwood Publishing, 2006. Braithwaite, John. “The New Regulatory State and the Transformation of Criminology.” In Criminology and Social Theory, edited by David Garland and Richard Sparks, 47-71. Toronto: Oxford University Press, 2000. Braun, Virginia and Victoria Clarke. “Using Thematic Analysis in Psychology.” Qualitative Research in Psychology 3, no. 1 (2006): 77-101. Brodie, Janine. “Restructuring and the New Citizenship.” In Rethinking Restructuring: Gender and Change in Canada, edited by Isabella Bakker, 126- 141. Toronto: University of Toronto Press, 1996. Brown, Kevin, Peter Duff, and Fiona Leverick. “A Preliminary Analysis of the Bail/Custody Decision in Relation to Female Accused.” Scottish Executive Social Research, 2004. Burchelle, Graham. “Liberal Government and Techniques of the Self.” In Foucault and Political Reason, edited by A. Barry, T. Osborne and N. Rose, 19-37. Chicago: University of Chicago Press, 1996.

142

Casale, Silvia. Women Inside: the Experience of Women Remand Prisoners in Holloway. London: Civil Liberties Trust, 1989. Channels., N.L., and S.D. Herzberger, S.D. “The Role of Gender in Pretrial Decision Making.” In Female Criminality: The State of the Art, edited by C.C. Culliver. New York: Garland Publishers. (1993).

Chesney-Lind, Meda. “Patriarchy, Crime, and Justice: Feminist Criminology in an Era of Backlash.” Feminist Criminology 1, no. 1 (2006): 2-26. Comack, Elizabeth. “Feminism and Criminology.” In Criminology: a Canadian Perspective (6th edition), edited by Rick Linden, 183-215. Toronto: Nelson Education Ltd., 2008. Crawford, Charles. “Gender, Race, and Habitual Offender Sentencing in Florida.” Criminology 28, no. 1 (2000): 263-280. Cresswell, John. Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. London: SAGE Publishing, 2009. Daly, Kathleen. Gender, Crime, and Punishment. London: Yale University Press, 1994. Daly, Kathleen. “Neither Conflict Nor Labelling Nor Paternalism Will Suffice: Intersections of Race, Ethnicity, Gender and Family in Criminal Court Decisions.” Crime & Delinquency 35 (1989): 136- 168. Daly, Kathleen. “Structure and Practice of Familial-Based Justice in a Criminal Court.” Law & Society Review 21, no. 2 (1987): 267-290. Daly, Kathleen and Meda Chesney-Lind. “Feminism and Criminology.” Justice Quarterly 5, no. 4 (1988): 497- 538. Demuth, Stephen and Darrell Steffensmeier. “The Impact of Gender and Race-Ethnicity in the Pretrial Release Process.” Social Problems 51, no. 2 (2004): 222-242. Dobrowolsky, Alexandra. “Introduction: Neoliberalism and After.” Women and Public Policy in Canada, edited by Alexandra Dobrowolsky, 1-25. Toronto: Oxford University Press, 2009. Donzelot, J. The Policing of Families. New York: Pantheon Books, 1979. Doob, Anthony and Cheryl Marie Webster. “Back to the Future? Policy Development in Pre-trial Detention in Canada.” In Canadian Criminal Justice Policy, edited by Karim Ismaili, Jane B. Sprott and Kim Varma, 31-57. Oxford University Press, 2012. Douglas, Mary. Risk and Blame: Essays in Cultural Theory. New York: Routledge, 1992. Ericson, Richard and Kevin Haggerty. Policing the Risk Society. Toronto: University of Toronto Press, 1997. Edgar, Kimmett. Lacking Conviction: the Rise of the Women’s Remand Population. London: Prison Reform Trust, 2004. 143

Ewald, Francois. “Insurance and Risk.” In The Foucault Effect: Studies in Governmentality, edited by G. Burchell, C. Gordon and P. Miller, 197-211. Chicago: University of Chicago Press, 1991 Farnworth, Margaret and Raymond H.C. Teske. “Gender Differences in Felony Court Proceedings: Three Hypotheses of Disparity.” Women & Criminal Justice 6, no. 2 (1995): 23-44. Feeley, Malcolm and Jonathan Simon. “Actuarial Justice: The Emerging New Criminal Law.” In Futures of Criminology, edited by David Nelken, 173-202. London: SAGE Publications, 1994. Foucault, Michel. Discipline and Punish. New York: Vintage Books, 1979. Foucault, Michel. “On Governmentality.” Translated by Rosi Braidotti. Ideology and Consciousness 6 (1979): 5-21. Foucault, Michel. Power/ Knowledge: Selected Interviews and Other Writings, 1972-1979. New York: Pantheon, 1980. Foucault, Michel. “Politics and the Study of Discourse.” In The Foucault Effect: Studies in Governmentality, edited by G. Burchell, C. Gordon and P. Miller, 53-73. Chicago: University of Chicago Press, 1991. Franklin, Cortney. “Women Offenders, Disparate Treatment, and Criminal Justice: a Theoretical, Historical, and Contemporary Overview.” Criminal Justice Studies 21, no. 4 (2008): 341- 360. Freiburger, Tina and Carly Hilinski. “The Impact of Race, Gender, and Age on Pre-trial Decisions.” Criminal Justice Review 35, no. 3 (2010): 318-334. Friedland, Martin L. Detention Before Trial: A Study of Criminal Cases Tried in the Toronto’s Magistrate’s Court. Toronto: University of Toronto Press, 1965. Frigon, Sylvie. “A Gallery of Portraits: Women and the Embodiment of Difference, Deviance and Resistance.” Post-Critical Criminology, edited by Thomas O’Reilly Flemming, 78- 111. Scarborough: Prentice Hall Canada, 1996. Gainey, Randy and Rodney Engen. “Modeling the Effects of Legally Relevant and Extra-legal Factors Under Sentencing Guidelines: the Rules have Changed.” Criminology 38, no. 3 (2000): 1207-1230. Garland, David and Richard Sparks. “Criminology, Social Theory and the Challenge of Our Time.” In Criminology and Social Theory, edited by David Garland and Richard Sparks, 1-23. Toronto: Oxford University Press, 2000. Garland, David. The Culture of Control In Contemporary Society. Chicago: University of Chicago Press, 2001.

144

Gartner, Rosemary, C.M. Webster, and A. Doob. “Trends in the Imprisonment of Women in Canada.” Canadian Journal of Criminology and Criminal Justice 51, no. 2 (2009): 169- 198. Goldkamp, John and Michael Gottfredson. “Bail Decision Making and Pre-Trial Detention: Surfacing Judicial Policy.” Law and Human Behaviour 3, no. 4 (1979): 227-249. Gottfredson, Michael. “An Empirical Analysis of Pre-trial Release Decisions.” Journal of Criminal Justice 2 (1974): 287-304. Hannah Moffat, Kelly. “Gendering Dynamic Risk: Assessing and Managing the Maternal Identities of Women Prisoners.” In Gendered Risks, edited by Kelly Hannah Moffat and Pat O’Malley, 229-247. New York: Routledge-Cavendish, 2007. Hannah-Moffat, Kelly. “Feminine Fortresses: Woman-Centred Prisons?” The Prison Journal 75, no. 2 (1995): 135-164. Hannah-Moffat, Kelly and Pat O'Malley. “Gendered Risks: An Introduction.” In Gendered Risks, edited by Kelly Hannah-Moffat and Pat O’Malley, 1- 31. London: Routledge Cavendish Publishing, 2007. Hesse-Biber, Sharlene. Mixed Methods Research: Merging Theory with Theory. New York: Guilford Press, 2010. Hochstedler- Steury, Ellen and Nancy Frank. “Gender Bias and Pre-Trial Release: More Pieces of the Puzzle.” Journal of Criminal Justice 18 (1990): 417-432. Hucklesby, A. “Bail or Jail? The Practical Operation of the Bail Act 1976.” Journal of Law and Society 23, no. 2 (1996): 213-233. Isin, Engin. “The Neurotic Citizen.” Citizenship Studies 8, no. 3 (2004): 217-235. Jeffries, Samantha. Gender Judgements: An Investigation of Gender Differentiation in Sentencing and Remand in New Zealand. Christchurch: University of Canterbury, 2001. Jeffries, Samantha. 2001. “Gender Judgements: An Investigation of Sentencing and Remand in New Zealand,” TASA 2001 Conference, University of Sydney. Jeffries, Samantha, Garth Fletcher, and Greg Newbold. “Pathways to Sex-Based Differentiation in Criminal Court Sentencing.” Criminology 41, no. 2 (2003): 329. Johnson, Sara. “Custodial Remand in Canada, 1986/1987 to 2000/2001.” Juristat 23, no. 7 (2002): 1-16. Katz, Charles and Cassia Spohn. “The Effect of Race and Gender on Bail Outcomes: a Test of an Interactive Model.” American Journal of Criminal Justice 19, no. 2 (1995): 161- 184. Kellough, Gail and Scot Wortley. “Remand for Plea: Bail Decisions and Plea Bargaining on Commensurate Decisions.” British Journal of Criminology 42 (2002): 186-210.

145

Kemshall, Hazel. Understanding Risk in Criminal Justice. Philadelphia: Open University Press, 2003. Kiselbach, Daniel. “Pre-Trial Criminal Procedure: Preventative Detention and the Presumption of Innocence.” Criminal Law Quarterly 31 (1989): 168- 196. Kong, Rebecca and Kathy Aucoin. “Female Offenders in Canada” Juristat 28, no. 1 (2008): 1-16. Kruttschnitt, Candace. “Sex and criminal court dispositions: An unresolved controversy.” Journal of Research in Crime and Delinquency 12, no. 3 (1984): 213-232.

Kruttschnitt, Candace and Daniel McCarthy. “Familial social control and pretrial sanctions: Does sex really matter?” Journal of Criminal Law and Criminology 76, no. 1 (1984): 151-172.

Madriz, Esther. Nothing Bad Happens to Good Girls: Fear of Crime in Women’s Lives. Berkely: University of California Press, 1997. Maidment, Madonna. Doing Time on the Outside: Deconstructing the Benevolent Community. Toronto: University of Toronto Press, 2008. Manson, Allan. “Pre-sentence Custody and Determination of a Sentence (or how to make a mole hill out of a mountain.” Criminal Law Quarterly 49 (2005): 295-350. McLellan, Myles. “Bail and the Diminishing Presumption of Innocence.” Canadian Criminal Law Review 15 (2011): 58-74. Moe, Angela. “Women, Drugs, and Crime.” Criminal Justice Studies 19, no. 4 (2006): 337-352. Moore, Dawn and Kelly Hannah-Moffat. “The Liberal Veil: Revisiting Canadian Penality.” In The New Punitiveness: Trends, Theories, Perspectives, edited by John Pratt, D. Brown, M. Brown, S. Hallsworth, and W. Morrison, 85-100. Portland: Willan Publishing, 2005. Moore, Dawn and Tara Lyons. “Sentenced to Treatment/Sentenced to Harm: Women, Risk, and the Drug Treatment Courts.” In Gendered Risks, edited by Kelly Hannah-Moffat and Pat O’Malley, 183-204. New York: Routledge-Cavendish, 2007. Murray, Karen Bridget. “Governmentality and the Shifting Winds of Policy Studies.” In Critical Policy Studies, edited by Michael Orsini and Miriam Smith, 161-184. Vancouver: University of British Columbia Press, 2007. Myers, Nicole. “Shifting Risk: Bail and the Use of Sureties.” Current Issues in Criminal Justice 21, no. 1 (2009): 127- 147. Nagel, I.H. “The Legal/Extra-legal Controversy: Judicial Decisions in Pre-trial Release.” Law and Society Review 17, no. 3 (1983): 481-515. O’ Malley, Pat. “Governmentality and the Risk Society.” Economy and Society 28 (1999): 138- 148.

146

O’Malley, Pat. “Risk and Responsibility.” In Foucault and Political Reason, edited by A. Barry, T. Osborne and N. Rose, 189-209. Chicago: University of Chicago Press, 1996. O’Malley, Pat. “Risk, Power, and Crime Prevention.” In Crime and the Risk Society, edited by Pat O’Malley, 71-97. Portland: Ashgate Publishing Ltd., 1998.

Player, Elaine. “Remanding Women in Custody: Concerns for Human Rights.” Modern Law Review Limited 70, no. 3 (2007): 402- 426.

Porter, Lindsay and Donna Calverley. “Trends in the Use of Remand in Canada.” Juristat 2011: 1-28. Power, Michael. The Risk Management of Everything: Rethinking the Politics of Uncertainty London: Demos, 2004.

Pratt, John, D. Brown, M. Brown, S. Hallworth and W. Morrison. “Introduction.” In The New Punitiveness: Trends, Theories, Perspectives, edited by John Pratt, D. Brown, M. Brown, S. Hallsworth, and W. Morrison, 1-30. Portland: Willan Publishing, 2005.

R c. Anoussis, [2008] Q.C.C.Q 8100 (CanLII) R. v. Hall, [2002] 3 S.C.R. 309, 2002 SCC 64

R. v. Patko, [2005] B.C.C.A 183 (CanLII)

R. v. Pearson, [1992] 3 S.C.R. 665, 1992 SCC 22173 Roberts, Julian. “Pre-trial Custody, Terms of Imprisonment and the Conditional Sentence: Crediting ‘Dead Time’ to Effect ‘Regime Change’ in Sentencing.” Canadian Criminal Law Review 9, no. 2 (2005): 191-213. Roberts, Julian and Anthony Doob. “Race, Ethnicity, and Criminal Justice in Canada.” In Ethnicity, Crime and Immigration: Comparative and Cross-National Perspectives, edited by M Tonry. Chicago: University of Chicago Press, 1997. Rose, Nikolas. “Government and Control.” In Criminology and Social Theory, edited by David Garland and Richard Sparks, 183-209. Toronto: Oxford University Press, 2000. Scoular, Jane and Maggie O’Neill. “Regulation Prostitution: Social Inclusion, Responsibilization, and the Politics of Prostitution Reform.” British Journal of Criminology 47 (2007): 761- 778. Silver, Eric and L. Miller. “A Cautionary Note on the Use of Actuarial Risk Assessment Tools for Social Control.” Crime and Delinquency 48, no. 1 (2002): 138-161. Spohn, Cassia and Dawn Beichner. “Is Preferential Treatment of Female Offenders a Thing of the Past? A Multisite Study of Gender, Race, and Imprisonment.” Criminal Justice Policy Review 11, no. 3 (2000): 149-184.

147

Sprott, Jane B and Anthony Doob. “Gendered Treatment: Girls and Treatment Orders in Bail Court.” Canadian Journal of Criminology and Criminal Justice 52, no. 2 (2010): 427- 441. Sprott, Jane B and Nicole Myers. “Set up to Fail: The Unintended Consequences of Multiple Bail Conditions.” Canadian Journal of Criminology and Criminal Justice 53, no. 4 (2011): 404-423 Steffensmeier, Darrell, Jeffrey Ulmer, and John Kramer. “The Interaction of Race, Gender, and Age in Criminal Sentencing: the Punishment Cost of Being Young, Black, and Male.” Criminology 36, no. 4 (1998): 763-798. Stuart, Don and Joanna Harris. “Is the Public Confidence Ground to Deny Bail Used Sparingly?” (2004) 21 C.R. (6th) 232. Trotter, Gary T. The Law of Bail in Canada: Third Edition. Toronto: Carswell, 2010. Turnbull, Sarah and Kelly Hannah-Moffat. "Under These Conditions: Gender, Parole, and the Governance of Reintegration."British Journal of Criminology 49 (2009): 1-20. Turner, K. B and James Johnson. “The Effect of Gender on the Judicial Pretrial Decision Amount Set.” Federal Probation 70 (2000): 1-8. Visher, Christy. “Gender, police arrest decisions, and notions of chivalry.” Criminology 21 (1983): 5-28. Webster, Cheryl Marie, Anthony Doob, and Nicole Myers. “The Parable of Ms. Baker: Understanding Pre-Trial Detention in Canada.” Current Issues in Criminal Justice 21, no. 1 (2009): 79-102. Williams, Marian. “The Effect of Pre-trial Detention on Imprisonment Decisions.” Criminal Justice Review 28, no. 2 (2003): 299-316.

148

Appendix

Figure A-1: Bail Court Checklist

Date:______Court type: Regular Contested Judge Name:______Gender: Male Female Accused Gender: Male Female Accused Age:______Race/Ethnicity:______Legal Factors Charge:______Severity: High Medium Low Criminal Record?: Yes No If yes, state criminal background: ______Has the accused been granted bail previously?: Yes No If yes, what were the conditions/ any violations?: ______Employment status: Employed (full time) Employed (part time) Unemployed Other: ______

149

______Residence: Fixed address (Owned/Rented) Shelter Live with friends/family Other: ______Type of legal representation: ______Drug/Alcohol Dependency?: Yes No If yes, explain: ______Mental Health Issues?: Yes No If yes, what?: ______Physical Appearance: ______Relationship Status: Single Married Common Law Divorced Other: ______Children?: Yes No Pregnant If yes, how many:______

150

If yes, do they have custody of their children?: ______What is the bail set at?:______Surety Available?: Yes No If yes, who is the surety?: ______If no, why not?: ______Reverse Onus?: Yes No Crown Submission: ______Defense Submission: ______

151

______Granted Bail?: Yes No If yes, what are the conditions?: ______If no, what are the reasons given: ______

152

Additional Notes: ______

Table A-1: Descriptive Statistics: Dependent, Independent and Control Variables, Kitchener Bail Court 2012 (N= 115)

Variable N Percent (%)

Dependent Variable Bail outcome Granted bail 98 85 Denied bail 17 15

Independent Variable Accused gender Male 93 81 Female 22 19

Control Variables Legal Factors Criminal record of the accused No criminal record 35 30 Has a criminal record 79 69 Not stated 1 1 Severity of criminal record Low severity 37 32 Medium severity 15 13

153

High severity 23 20 Not stated 5 4 Has no criminal record 35 30 History of violence No history of violence 60 52 History of violence 33 29 Not stated 22 19 Previous release Granted bail 86 75 Not granted bail 25 22 Not stated 4 4 Previous release success Not successful 47 41 Successful 6 5 Not stated 43 38 Not applicable 19 17 Past failures to comply No failures to comply 57 50 Failed to comply 47 41 Not stated 11 10 Number of current charges 1 charge 41 36 2 charges 31 27 3 charges 14 12 4 charges 13 11 5 or more charges 14 12 Not stated 2 2 Offense seriousness Low 67 58 Medium 41 36 High 7 6 Type of offense Drug 7 6 Against the admin of justice 37 32 Property 10 9 Violent 18 16 Combination 43 37 Reverse Onus Not a reverse onus situation 65 56 Reverse onus 50 44

Accused on probation Not on probation 79 69 On probation 36 31 Drug/Alcohol a factor No 75 65 Yes 40 35 Children present during offense No 96 84 Yes 14 12 Not stated 5 4 Interactions with police Accused did not cooperate 15 13 Accused cooperated 13 11 Not stated 87 76

154

Involvement of weapons No weapons involved 105 91 Weapons involved 10 9

Surety availability No surety available 11 10 Surety available 81 70 Youth in Conflict 23 20 Residential surety availability No residential surety 20 17 Has residential surety 78 68 Not required 17 15 Probable level of supervision low level supervision 39 34 medium level supervision 31 27 high level supervision 45 39

Extra-Legal Factors Accused characteristics Race/Ethnicity White 78 68 Black 11 10 Asian 3 3 First Nations 3 3 Middle Eastern 5 4 Mexican 5 4 Other 10 9 Age 18-24 yrs 27 24 25-29 yrs 37 32 30-39 yrs 27 24 40-49 yrs 14 12 50-59 yrs 5 4 60 + yrs 2 2 Not stated 3 3 Employment status Employed fulltime 36 31 Employed part time 4 4 Unemployed 45 39 Other 2 2 Not stated 28 24 Residence status Owns home 7 6 Rents home 40 35 Lives in shelter 7 6 Lives with friends/family 41 36 Other 10 9 Not stated 10 9 Drug/Alcohol Dependency Yes 44 38 No 58 50 Not stated 13 11 Accused in treatment/counselling Not in treatment/counselling 33 29 In treatment/counselling 10 9

155

Not stated 14 12 Not applicable 58 50 Mental health issues Has mental health issues 19 17 Does not have mental health issues 76 66 Not stated 20 17 Taking medication for mental health issues Taking medication 5 4 Not taking medication 12 10 Not stated 22 19 Not applicable 76 66 Relationship status Single 65 57 Married 14 12 Common law 9 8 Divorced/ separated 7 6 Other 20 17 Children Has children 42 37 Does not have children 41 36 Not stated 32 28

Victim characteristics Victim-accused relationship Significant other 15 13 Child 2 2 Parent 2 2 Sibling 0 0 Other family 0 0 Friend 0 0 Stranger 8 7 Other 15 13 Not stated 4 4 Multiple 7 6 Not applicable 62 54 Victim age 0-12 yrs 3 3 13-18 yrs 2 2 19-29 yrs 13 11 30-39 yrs 3 3 40-49 yrs 4 4 Multiple 10 16 Not stated 18 9 Not applicable 62 54 Victim Gender Male 11 10 Female 31 27 Not stated 3 3 Mix 8 7 Not applicable 62 54 Level of fear Not fearful 5 4 Somewhat fearful 5 4 Very fearful 11 10 Not stated 32 28

156

Not applicable 62 54

Surety Characteristics Surety identity No surety available 34 30 Significant other 5 4 Friend 10 9 Mother 26 23 Father 8 7 Sibling 6 5 Other family 5 4 Employer 1 1 Other 20 17 Surety-accused relationship Low 7 6 Medium 29 25 High 38 33 Unknown 7 6 Not applicable 34 30 Surety gender No surety 34 30 Male 19 17 Female 50 44 Mix 12 10 Employment status No surety 34 30 Employed fulltime 29 25 Employed partime 6 5 Unemployed 19 17 Retired 3 3 Other 2 2 Not stated 22 19 Amount of assets Minimal 26 23 Average 21 18 Substantial 7 6 No surety 34 30 Not stated 27 24 Residence status Local 46 40 Not local 14 12 No surety 34 30 Not stated 21 18 General experience as a surety No 22 19 Yes 26 23 Not stated 33 29 Not applicable 34 30 Surety for the accused in past No 22 19 Yes 23 20 Not stated 24 21 Not applicable 46 40

157

Table A-2: Phases of Thematic Analysis*

Phase Description of the Process Transcribing data (if necessary), reading and re- 1. Familiarising yourself with your data: reading the data, noting down initial ideas.

Coding interesting features of the data in a 2. Generating initial codes: systematic fashion across the entire dataset, collating data relevant to each code. Collating codes into potential themes, gathering all 3. Searching for themes: data relevant to each potential theme.

Checking if the themes work in relation to the 4. Reviewing themes: coded extracts and the entire dataset.

Ongoing analysis to refine the specifics of each 5. Defining and naming themes: theme and the overall story of the analysis tells; generating clear definitions and names for each theme. The final opportunity for analysis. Selection of 6. Producing the report: vivid, compelling extracts examples, final analysis of selected extracts, relating back on the analysis to the research question and literature, producing a scholarly report of the analysis. * Table taken from: Braun, Virginia and Victoria Clarke, “Using Thematic Analysis in Psychology,” Qualitative Research in Psychology 3, no. 1 (2000): 87.

Table A-3: Justifications for Granting Bail that Relate to the Surety, Kitchener Provincial Courthouse 2012 (N= 80) Type of Justification Number of Times Used (%)

Sufficient surety available 58% (46) Residential surety available 16% (13) Multiple sureties available 14% (11) Surety was surety in the past and called 4% (3) police/was effective Surety clearly understands their role 3% (2) Surety has set up/thought about treatment or 3% (2) counselling for the accused Surety is honest and well meaning 1% (1) Surety is unemployed and can offer 24hr 1% (1) supervision Accused and surety have a good relationship 1% (1)

158

Table A-4: Justifications for Granting Bail that Relate to the Accused, Kitchener Provincial Courthouse 2012 (N= 58)

Type of Justification Number of Times Used (%)

Limited, dated, or no criminal record 45% (26) Accused has taken steps to get better on his/her 10% (6) own (treatment/counselling) Accused is employed (only for men) 9% (9) Accused is honest and intelligent 5% (3) Accused is young and has prospects for 5% (3) school/work Accused has pre-existing medical conditions 4% (2) beyond his/her control Court is sympathetic to accused’s MH issues 4% (2) (only for women) Accused had a bad childhood/past 2% (1) Accused was forthright in admitting he/she had 2% (1) problems and wanted to get better Accused is a recent immigrant 2% (1) Accused needs to take care of wife and kids 2% (1) Accused would not do well in prison 2% (1) Accused has had no convictions since 2009 2% (1) Accused has never breached 2% (1) Accused will obey 2% (1) Accused has ties to the community 2% (1) Accused needs medication 2% (1) Accused has a young child 2% (1)

Table A-5: Justifications for Granting Bail that Relate to Additional Levels of Supervision, Kitchener Provincial Courthouse 2012 (N= 40)

Type of Justification Number of Times Used (%)

Youth In Conflict with the Law is supervising 43% (17) Accused is bound by additional probation 23% (9) orders Accused has the support of other family 10% (4) members/friends Accused will be living outside KWC while on 10% (4) bail Accused lives in a shelter 8% (3) Accused is under the watchful eye of FACS 5% (2) Accused is bound by additional bail orders 3% (1)

159

Table A-6: Justifications for Granting Bail that Relate to the Allegations, Kitchener Provincial Courthouse 2012 (N= 17)

Type of Justification Number of Times Used (%)

Allegations are not very serious 59% (10) Weak case of detention 29% (5) Alcohol was a factor 6% (1) Allegations did not occur in front of children 6% (1)

Table A-7: Justifications for Granting Bail that Relate to the Victim, Kitchener Provincial Courthouse 2012 (N= 3)

Type of Justification Number of Times Used (%)

Complainant gave mixed messages/ some 67% (2) blame attributed to victim Victim is highly protected by his/her own 33% (1) family

Table A-8: Justifications for Denying Bail that Relate to the Accused, Kitchener Provincial Courthouse 2012 (N= 42) Type of Justification Number of Times Given (n=/%)

Accused has a lengthy criminal record 7/ 17% Accused is a threat to public safety 7/17% Accused has no regard for court orders 6/ 14% Accused has made no attempt to improve 5/ 12% situation or go to treatment/counselling Accused has a violent history 4/ 10% Accused is dishonest and cannot be trusted 3/ 7% Accused would reoffend if released 3/ 7% Accused has unstable MH issues, which 2/ 5% increase chances of reoffending Accused has no regard for the police/going to 2/ 5% jail Accused has no respect for anyone 1/ 2% Accused’s past behaviour is best predictor of 1/ 2% future behaviour Accused has previous breaches 1/ 2%

160

Table A-9: Justifications for Denying Bail that Relate to the Surety, Kitchener Provincial Courthouse 2012 (N= 24) Type of Justification Number of Times Used (%)

Although surety is well meaning they cannot 25% (6) properly supervise Weak bail plan 13% (3) Surety doesn’t understand the seriousness of 8% (2) the offense Accused cannot be released without a 8% (2) residential surety Surety was a surety in the past and was not 8% (2) successful Surety is evasive and confusing- making 4% (1) excuses for accused Surety is unemployed and not appropriate 4% (1) Female surety works in the adult film industry, 4% (1) which is a direct link to the criminal subculture Surety has limited means 4% (1) Surety has small children so they cannot 4% (1) supervise Surety is too young- limited authority 4% (1) Accused doesn’t have a surety 4% (1) Surety does not know the accused very well 4% (1) Surety is not credible 4% (1)

Table A-10: Justifications for Denying Bail that Relate to the Allegations, Kitchener Provincial Courthouse 2012 (N= 8) Type of Justification Number of Times Used (%)

Allegations are serious 63% (5) Crown has a strong case 38% (3)

Table A-11: Justifications for Denying Bail that Relate to the Victim, Kitchener Provincial Courthouse 2012 (N= 6) Type of Justification Number of Times Used

Volatile relationship between the accused and 67% (4) complainant= concerns for victim safety Complainant sustain injuries 17% (1) Cannot take complainant’s testimony in court 17% (1) seriously

161

Table A-12: Justifications for Denying Bail that Relate to Public Confidence, Kitchener Provincial Courthouse 2012 (N= 1) Type of Justification Number of Times Used (%)

Detention is necessary to maintain confidence 100% (1) in the administration of justice

Table A-13: Bail Conditions Related to Possessions, Kitchener Provincial Courthouse 2012 (n= 130) Type of Condition Number of Times Used (%)

Do not possess any weapons 44% (57) Do not consume, possess, or ingest any drugs 25% (33) unless in accordance with a medical prescription Do not consume/possess alcohol or attend any 22% (29) place where alcohol is dispensed primarily Do not possess any drug equipment/ 4% (5) paraphernalia Do not possess any debit or credit cards that 0.8% (1) are not in your name Surrender firearms and firearm licence within 0.8% (1) 96hrs Do not have access to or possess any computer 0.8% (1) with internet Do not posses any tools known to be used to 0.8% (1) commit a theft Do not apply for any firearms and firearms 0.8% (1) licence Do not possess any replica currency unless 0.8% (1) permitted under s. 457 of the C.C.C

162

Table A-14: Bail Conditions Related to Residence, Kitchener Provincial Courthouse 2012 (n= 94) Type of Condition Number of Times Used (%)

Reside with surety and be amenable to the 56% (53) routine and discipline of that residence Reside at your stated address and do not move 11% (10) without prior, written approval of Youth In Conflict Reside at the shelter and be amenable to the 10% (9) routine and disciple of that residence and not move without prior, written approval from Youth In Conflict Reside at your stated address and do not move 8% (7) without prior, written approval of the Court/Police Report any changes to surety’s address to 4% (4) police within 24hrs Reside with your father/ residence approved of 4% (4) by Youth In Conflict and be amenable to the routine/discipline of that household Reside at the shelter and be amenable to the 3% (3) routine and disciple of that residence and not move without prior, written approval of the Court/Police Reside with surety or at a place approved of by 2% (2) surety in writing Reside at any residence approved of by surety 1% (1) and report that address to police within 24hrs Do not attend your residence except one time 1% (1) to pick up your belongings – you can only reside their if you provide your probation officer proof of residence

163

Table A-15: Bail Conditions Related to Communication, Kitchener Provincial Courthouse 2012 (n= 87) Type of Condition Number of Times Used (%)

No communication, either directly or 72% (63) indirectly, with complainant(s)/witness(es) Do not contact, communicate or associate with 10% (9) children, either directly or indirectly, except through Family and Children’s Services, Legal Counsel or a mutually agreed upon 3rd party Do not communicate, contact, or associate with 7% (6) any co-accused Do not possess any cell phones 6% (5) Do not contact any persons under the age of 16 3% (3) year old except in the company of your surety Do any means of personal communication 1% (1) Do not associate with any persons stated by 1% (1) your surety

Table A-16: Bail Conditions Related to Mobility, Kitchener Provincial Courthouse 2012 (n= 80) Type of Condition Number of Times Used (%)

Do not attend any place where the allegations 28% (22) were said to have been committed (i.e. bank, shopping mall, grocery store etc.) Do not be within 100 meters of 15% (12) complainant’s/witness’s place of work, residence, and education Report to your probation officer within 48hrs 14% (11) to provide a copy of recognizance and thereafter as required Do not attend current place of residence except 13% (10) one time to get your personal belongings in the company of a Waterloo Regional Police officer Do not be within 200 meters of 6% (5) complainant’s/witness’s place of work, residence, and education Do not attend Waterloo Region except for 5% (4) court appearances or travelling through Surrender your passport to the Waterloo 4% (3) Regional Police within 48hrs Do not drive or operate a motor vehicle 4% (3) Do not attend any place where persons under 4% (3) the age of 16 normally go ie. schools, malls, 164

daycares, parks etc. except in the company of your surety or with the child’s parents present Remain in the province of Ontario 3% (2) Do not be within 500 meters of 3% (2) complainant’s/witness’s place of work, residence, and education Do not attend any place where alcohol is 1% (1) dispensed primarily Do not be within 50 meters of 1% (1) complainant’s/witness’s place of work, residence, and education Surrender your passport to the RCMP within 1% (1) 48hrs

Table A-17: Bail Conditions Related to Additional Levels of Supervision, Kitchener Provincial Courthouse 2012 (n= 45) Type of Condition Number of Times Used (%)

Be under the supervision of Youth in Conflict 44% (20) with the Law, be amenable to their rules/directives and report their on a weekly basis Report to the Waterloo Regional Police once 29% (13) every week Report to your probation officer within 24hrs 9% (4) to provide a copy of recognizance and thereafter as required Report to your probation officer within 72hrs 8% (3) to provide a copy of recognizance and thereafter as required Report to the RCMP once every week 2% (1) Report to the Waterloo Regional Police 2% (1) everyday Advise your surety of each place of 2% (1) employment prior to going Be amendable to the rules put in place by your 2% (1) surety Report any changes in employment and 2% (1) residence status to police within 24hrs

165

Table A-18: Bail Conditions Related to Curfew/House Arrest, Kitchener Provincial Courthouse 2012 (n= 37) Type of Condition Number of Times Used (%)

Obey a curfew between 10:00pm-6:00am 16% (6) except in the company of a surety Obey a curfew 24hrs/day except in the 14% (5) company of your surety or for previously scheduled medical appointments Obey a curfew between 10:00pm-6:00am daily 11% (4) Obey a curfew 24hrs/day except in the 8% (3) company of your surety Obey a curfew 24hrs/day except in the 8% (3) company of your surety or for lawful employment Obey a curfew between 10:00pm-6:00am 8% (3) except in the company of a surety or for lawful employment Obey a curfew between 11:00pm-6:00am 5% (2) except for lawful employment, travelling to/from work or with a surety Obey a curfew between 7:00pm-6:00am unless 3% (1) in the company of your surety or with written permission by your surety Obey a curfew between 11:00pm-6:00am 3% (1) except for lawful employment or in the company of your surety Obey a curfew between 8:00pm-6:00am- no 3% (1) exceptions Obey a curfew between 9pm-7am by 3% (1) remaining at the shelter- no exceptions Obey a curfew between 9:00pm-6:00am except 3% (1) for lawful employment, travelling to/from work, or with surety Obey a curfew between 11:00pm-11:00am by 3% (1) remaining in your residence except in the company of your surety M-F Obey a 24 hr curfew on weekends except in the 3% (1) company of your surety or for lawful employment Obey a curfew between 12:00am-6:00am by 3% (1) remaining in your home except with prior, written approval by Youth In Conflict and have such document on your persons at all times Obey a curfew between 6:00pm-8:00am 3% (1)

166

Obey a curfew between 8:00pm-8:00am except 3% (1) with surety Obey a curfew between 12:00am-6:00am 3% (1) except for lawful employment or with surety

Table A-19: Bail Conditions Related to Counselling/Treatment, Kitchener Provincial Courthouse 2012 (n= 36) Type of Condition Number of Times Used (%)

Take any counselling or treatment as directed 39% (14) by your surety Reside with surety or a residential treatment 19% (7) facility and be amenable to the routine and discipline of that residence Attend any treatment or counselling as directed 17% (6) by Youth In Conflict Attend any treatment/counselling as directed 6% (2) by surety and sign a release form to give information to surety so they can monitor your progress Obey a 24hr curfew except with surety or at a 6% (2) residential treatment facility Obey a curfew between 5:30pm-8:00am unless 3% (1) for lawful employment, attending residential treatment, or with a surety Obey a curfew between 8:00pm-6:00am unless 3% (1) for lawful employment, counselling or in the company of your surety Continue seeing your psychiatrist 3% (1) Obey a 24hr curfew except for employment, 3% (1) seeking employment, attending school, in the company of your surety or other approved of persons or for drug treatment program Report to your family doctor regularly and 3% (1) follow his directives

Table A-20: Bail Conditions Related to Other, Kitchener Provincial Courthouse 2012 (n= 1) Type of Condition Number of Times Used (%)

Actively seek and maintain lawful employment 100% (1)

167