Impact of Religion on Restorative Justice among Female Offenders in Trinidad and Tobago

Abstract

Although restorative justice in its variety of practices has garnered considerable academic attention, there remains a gap in the research on the interrelationships between restorative justice, religion and imprisonment, especially among women. This article examines the links between restorative justice and religion in a female custodial setting. Exploratory research conducted in Trinidad and Tobago has tested the hypothesis that religion plays a role in the implementation of restorative justice in prisons in that religious inmates are more likely to have encounters with and/or apologise to their victims and perform reparative activities. However, the present empirical findings do not support this claim.

1. Introduction

Prisons are the dinosaurs of the modern age [1]. Despite serving society’s need for specific deterrence, they fail on practically every positive human indicator. Prisons fail to rehabilitate [2]. They are extremely expensive [3]. They undermine family life by leaving children without parents [4, 5, 6]. They lead to more crime by enabling similarly minded individuals to bond. They foster further antisocial behaviour and distress among inmates [7]. They breed violence as the principal recruitment hubs for gangs [8]. They destroy offenders’ prospects for future life [9]. They guarantee high rates of re-offending [10]. They stigmatise offenders and their families [11]. They suppress psychological and mental growth [12], and they are spiritually bankrupt having lost their moral compass [1].

In issues of crime, law and order, the fundamental mission of religion is to bring the ‘good news to the poor and proclaim release to captives’ (cf. Lk 4:18); the principles of prison ministry include promoting justice, equity and accountability. Such an approach should be guided by wisdom and allow for the possibility of healing, forgiveness and reconciliation for both victims and offenders [1].

Religion has long been a difficult issue in the field of restorative justice because it is unclear how it can be integrated into the conversation. Even while discussing how restorative justice can use religious places or religious human resources, it is concluded that people of faith should never bring their own religious persuasions into practice. Thus, I aimed in the present study to challenge Habermas’s [13] argument that ‘[…] the modern society which tends to rely heavily on an instrumental argument needs religion in order to revive our ability to make a moral argument’.

2. Significance

This work adds a piece to the restorative justice ‘puzzle’ to come closer to understanding the links between the notions of imprisonment, religion and restoration. Specifically, I contribute to knowledge in four distinct aspects that I discuss below.

2.1 The restorative aspect

Restorative justice advocates focus on its potential and capacity to improve satisfaction among individuals who come into contact with the criminal justice system. Restorative justice has been found applicable at different stages within the criminal justice process [14, 15, 16], but does restorative justice have a role to play with persons already serving prison sentences?

Researchers around the world have acknowledged the positive contributions of restorative justice in terms of recidivism, behavioural change, cost-effectiveness and victim awareness [17, 18] and in re-establishing a sense of safety and welfare in the community [19]. For instance, restorative justice delivered in custodial settings could help in creating safer communities ([20], 105). It could maximise the opportunity for communities outside of prison walls to aid in rehabilitating offenders [21] and offering victims the chance to have questions answered prior to prisoners’ release [22]; ([23], 203). From an offender’s perspective, in-prison restorative justice programmes can bring awareness of the harm crimes cause and of their obligation to desist from further unlawful actions in the future [24]. Both secular and faith-based restorative programmes can enable prisoners to take full responsibility for their behaviour and provide opportunities for improving self-esteem and empowerment [25] along with consequential behavioural changes [26, 27].

Nowadays, there is an abundance of empirical social science research in the area of correctional policymaking. However, although criminal justice theorists and practitioners alike have periodically documented the potential of restorative justice in custodial settings, there is a dearth of comprehensive literature on restorative justice in custodial settings in terms of offenders’ personal feelings, inner motivations and individual perceptions [28, 29]. Therefore, this study addresses the lack of quantitative research on the applicability of restorative justice behind bars, exploring its potential influence on and contribution to developing restorative (religious) programmes in prisons. It also furthers the debate about offering restorative opportunities as interesting alternatives within prisons for less serious offenders and offences.

2.2 The remand aspect

Over 2200 people currently incarcerated in Trinidad and Tobago await trial. By law, anyone appearing before a court to face charges is entitled to a presumption in favour of bail, unless they are charged with serious offences such as murder or manslaughter, rape or drug trafficking, as applicable to the female section of the Golden Grove prison. However, unlike sentencing, which is proportionate to the seriousness of the offence, bail decisions can be based on the perceived risk that the defendant poses, including intimidating victims, committing further offences or failing to appear for trial. Although it is an accepted truism that time on remand is a punishment with detrimental effects that go beyond the loss of liberty, many are currently sent to prison, both charged and on remand, for offences that would not necessarily merit a custodial sentence. Table 1 presents rates and percentages of prisoners on remand in Trinidad and Tobago.

Year	Number in pre-trial/remand imprisonment	Percentage of total prison population (%)	Pre-trial/remand population rate (per 100,000 of national population)
2000	1475	31.8	116
2005	1573	42.2	121
2008	1595	41.9	121
2011	2098	52.1	155
2015	2235	60.9	164

Table 1.

Prisoners on remand in Trinidad and Tobago.

Latest available data; more information available from: http://www.prisonstudies.org/country/trinidad-and-tobago.

International experience has taught that sending people to prison is neither always sufficient nor always effective punishment [30]. Although in response, programmes that could be classified as forms of restorative justice [21, 31, 32, 33] are being slowly delivered in prisons around the world, relatively few of them have appeared within correctional settings or involve or target remanded offenders. What we often encounter are programmes that bring together unrelated or ‘surrogate’ victims and offenders to address the harm from criminal acts. To date, all restorative justice initiatives exclusively target already convicted offenders, take place during offenders’ incarceration [34] and do not affect offenders’ sentences [18]. Given the alarming rates and proportions of prisoners on remand compared with those being sentenced on one hand and the all-inclusive nature of restorative justice on the other, I aimed in this study to stretch the above arguments to all imprisoned individuals for achieving greater applicability and more profound results.

2.3 Religion and restorative justice

There has been little formal research on the relationship between religion and restorative justice [35], but there is a long history of religious programmes in prisons. However, except for specialised Christian studies, the crucial relationship between spirituality and criminal justice has been largely overlooked [36]; recent examples of pilot studies on how religion can support restorative justice initiatives come from Sierra Leone [37], the United States [38] and New Zealand [39]. Notwithstanding the above, there have been only a few empirical studies on religion and restorative justice in prisons (e.g. [40], conducted in Belgium). As Workman ([41], 2) argued, there remains a lack of social science research in the area of corrections and the influence of religion on criminal justice policymaking.

Meanwhile, there are compelling reasons to scrutinise the relationships between religion (restorative), justice and punishment (imprisonment), especially when most of the active restorative justice programmes within prisons often rely on and are delivered by prison-based ministries that promote transforming the offender’s character by developing religious ethics and moral values [17]. Past researchers have characterised restorative justice as a public theology imperative [42] or as a unique opportunity for religious engagement [43], highlighting the contribution of religion in restorative justice-based to confront and address offenders; efforts have addressed traumas inflicted by clergy sexual abuse [44] and attempted to counter Islamic radicalisation in prisons [45]. Bearing all these in mind, it is intellectually irresponsible to ignore the faith factor in tackling social problems, especially when empirical evidence suggests that religious belief can make a positive difference ([41], 3).

The present research aims to fill this scientific gap. This article contributes to the knowledge base on how the theology of justice contributes to ongoing discussions on restorative justice. It is also a response to my personal desire to express the views of the imprisoned women whose attitudes could change by participating in such programmes and who could find hope, healing and restoration. This research is not intended to be exclusionary as it is not written from a specific faith perspective, so when I mention religion or religiosity here, I am referring to all denominations. I adopted this approach because religion is not the only legitimate vehicle of restoration in prison; in fact, restorative justice has application beyond particular dogmas, and its principles are deeply embedded in all major religions. Although among all religions, Christianity is undoubtedly dominant both in restorative justice and in prison settings. In the context of countless in-prison restorative justice programmes around the world, it should be realistically acknowledged that the religious part of this research is but an entry into the vague area of prison-religion-restorative justice. Nevertheless, I do provide a way of indicating certain patterns of practice that apply to the intersections of these notions to comprise an important part of the broader picture.

2.4 The feminist aspect

In recent decades, the number of women caught in the criminal justice system has significantly risen, yet our policies remain mostly based on empirical research conducted with male subjects [46]. While the increase in numbers of female inmates has resulted in marginal improvement in services for incarcerated women, the rapid increase has not been matched by a commensurate increase in programmes that are specifically tailored towards women’s needs [47]. Many feminist scholars have criticised the risk and need assessments that are generally used to categorise inmates in terms of recidivism and rehabilitation potential for having little relevance and applicability to women [48]. This lack of gender-sensitive factors has reasonably left many scholars questioning the validity of these gender-neutral or ‘gender-blind’ assessments [49, 50, 51, 52, 53, 54, 55], which in reality reflect male normative standards.

Given the above, the continuous influx of women into the criminal justice system makes this research particularly timely, and developing restorative justice programmes in correctional settings requires a theoretical understanding of the framework around female criminality. With this research, I sought to partially fill in this void in restorative justice literature by examining the context of female offenders. Although there is some overlap between the factors that predict men’s and women’s involvement in restorative justice programmes, the lack of data specific to women has resulted in a global ‘woeful dearth’ [56] of programmes tailored for women.

Generally, women commit crime less often than men, and their entries into criminal behaviour—as well as their backgrounds and experiences—tend to be very different. For example, international research shows that women who commit crime are typically a victim first and an offender second (idem). Women entering the criminal justice system present different risks and needs than do their male counterparts, and they often face issues that men do not such as lower self-esteem [57], more mental health problems linked to drug and alcohol misuse [58, 59], greater risk of self-harm and suicide [60], more intense trauma due to psychological and physical abuse [61] and more social stigma [62]. Women also often struggle with high levels of guilt, stigma and shame linked to the breaching of traditional gender ideals [63].

To date, there has been very little in-depth consideration of the impact of gender in prisons, although a recent qualitative study conducted by Masson and Österman [64] indicates that gender does play a role in restorative conferences. However, while this study addresses important issues in delivering restorative justice programmes, I with the present research take one step back and focus on the preliminary perceptions of female offenders that should be considered in designing such programmes. In this context, widening the feminist lens on restorative justice [65] becomes crucial. For women to reap the full benefits of restorative justice, recognising the complexities and differences in circumstances and vulnerabilities between women’s and men’s pathways into crime is pertinent to understanding how to rehabilitate and, ultimately, restore women who commit crimes.

3. The current study

3.1 Hypotheses

This study aims to describe and understand the roles of offenders’ perceptions on religion in an attempt to apply restorative justice in custodial settings by investigating the multilateral relationships between restorative justice, religion, reparation, apology, community and remorse. The main hypotheses of this study concentrate on the possible role of religion in the future development of in-prison restorative justice programmes as follows:

1. Hypothesis 1 (H1): Religious prisoners¹ feel more apologetic than do irreligious offenders and consider it important to explain the circumstances surrounding their offending to actual and/or surrogate victims (Q7, Q9, Q10 & Q11).

2. Hypothesis 2 (H2): Religious prisoners show more recognition of the impacts of their crimes and are more willing than irreligious offenders to have personal encounters with victims and/or engage in reparative activities (Q7, Q8, Q12 & Q13).

3.2 Sample

I collected the data for this quantitative study from Golden Grove prison in Arouca, Trinidad. Since my target study sample was all the current prisoners at the time I was collecting my data, randomisation of the sample was not needed; detainees on remand count in the total prison population, so they were also in the target sample for the present research. I did not choose Golden Grove penitentiary at random; it is by far the largest maximum-security penitentiary in Trinidad and Tobago that houses both convicts and detainees on remand.

Initially, I administered the study questionnaire (described below) as a pre-test to identify any inconsistencies and ambiguities. Once the randomly assigned prisoners in the pre-test confirmed that all questions were clear and understandable in all languages, I distributed the paper questionnaires by hand to all eligible prisoners along with a brief explanation about the nature and the objectives of the study. I grouped prisoners who had difficulty reading but still wanted to take part in the research by section/unit and gave them the necessary assistance (translating, simplifying and reading aloud) that would enable them to voluntarily and anonymously respond to the questions. I collected the questionnaires myself, so the women had the opportunity to ask for any further clarifications.

3.3 Design and structure

From an operational viewpoint, I used a quantitative methodology to fulfil my study aims, using a questionnaire because it was the most appropriate tool for collecting large-scale numerical data that could depict ‘objective’ reality [66]; questionnaire data allow for statistical analysis that remains unbiased by the researcher’s values [67, 68]. The questionnaire was divided in two sections (see Annex): the first contained close-ended demographic questions (Q1 and Q2) and the second contained open- and close-ended restorative justice questions (Q3 to Q13). Keeping in mind that both types of questions bear dangers and rewards, the close-ended questions, being more easily analysed, were ideal for my demographic questions and other ‘simple’ questions because they take less time to complete [69]; in contrast, the open-ended questions allowed respondents to provide more information including their feelings and attitudes towards and understanding of the subject [70]. Further, open-ended questions eliminated the possibility of one typical response error in that it was not possible for survey respondents to disregard the wording of the questions and, for instance, simply choose the same letter for every multiple-choice question or just choose ‘no’ for every yes-no question [71]. They allowed me to gather in-depth information by providing offenders the opportunity to discuss what was important to them irrespective of confirming any of their answers to the close-ended ones. Using both types of questions best allowed me to fulfil my research aims.

3.4 Data coding

Independent variables: Along with basic demographic and background variables, age, gender and of course religion², I included several other independent variables in the regression models: past personal victimisation, type of crime committed and current participation in any prison programme. Prisoners were asked to describe their relationships with God and whether they considered themselves to be religious or not (‘yes’, ‘no’ and ‘I don’t know’)³, and women who said yes were asked to specify their religion (Christianity, Hinduism, Islam, Orisa⁴ and ‘other’, meaning Judaism and Rastafarianism for this study’s purposes). Towards the aim of identifying influential factors from the respondents’ backgrounds, I also asked them to indicate why they were sent to Golden Grove and how long their sentences had been along with how much time they had left to serve.

Dependent variables: Ten questions measured the inmates’ perceptions of the concept of restorative justice. Four of the items measured the participants’ general willingness to meet with (surrogate) victims and/or their families and their motivations for doing so, and other questions related to whether the participants would be willing to apologise and/or would participate in work in prison that would benefit the victim or the local community. These were binary variables where 1 = yes and 2 = no, and all responses of ‘no reply’ were treated as missing. In addition, an open-ended question asked the respondents about their individual reasons for being willing to meet with victims if they had the opportunity (Table 2).

	N	Minimum	Maximum	Mean	Std. deviation
Status	79	1	2	1.59	.494
Age	77	1	5	2.57	1.141
Nationality	71	1	3	1.24	.492
Prison_Record	78	1	2	1.26	.439
Times_Imprisoned	19	1	4	1.89	.875
Sentence_Length	30	1	8	4.97	1.790
Time_Left_To_Serve	28	1	6	3.25	1.713
Crime_Victimization	78	1	2	1.56	.499
Victimization_Type	30	1	8	3.57	2.431
Reason_For_Imprisonment	69	1	11	4.61	1.526
Programme_Participation	75	1	2	1.27	.445
Personal_Social_Skills	53	1	2	1.57	.500
Employment_Skills	52	1	2	1.60	.495
Prison_Ministry	52	1	2	1.38	.491
Education	52	1	2	1.37	.486
Religiousness	78	1	2	1.15	.363
Religion	67	1	6	1.66	1.250
Impact_of_Offence	78	1	2	1.10	.305
Individual	71	1	2	1.77	.421
Individuals_Family	71	1	2	1.59	.495
Own_Family_Friends	71	1	2	1.23	.421
Community	71	1	2	1.73	.446
Myself	71	1	2	1.30	.460
Victim_Offender_Mediation	53	1	2	1.34	.478
VOM_Explanation	41	1	11	5.37	3.207
Apology_Restoration	58	1	2	1.21	.409
Explanation_Personal_Victims	53	1	2	1.26	.445
Explanation_Surrogate_Victims	51	1	2	1.25	.440
Crime_Effects_Emotional	68	1	2	1.13	.341
Crime_Effects_Physical	68	1	2	1.74	.444
Crime_Effects_Financial	68	1	2	1.41	.496
Prison_Work_Participation	76	1	2	1.11	.309
Repair_Harm	65	1	2	1.58	.497
Improve_Relationship	65	1	2	1.40	.494
Feel_Good_Useful	65	1	2	1.40	.494
Get_Me_Out_Cell	65	1	2	1.63	.486
Gain_Skills	65	1	2	1.25	.434
Keep_Me_Busy	65	1	2	1.49	.504
Prevent_Reoccurance	65	1	2	1.92	.269
Valid N (listwise)	0

Table 2.

Descriptive statistics of all variables.

4. Data analysis

I assigned unique numbers to all the prisoners to ensure confidentiality. I then conducted basic analyses to produce frequencies, percentages and appropriate measures of association for all the questionnaire items and non-parametric, bivariate logistic regressions with all categorical variables when necessary. Due to the relatively small size of the available and collected data, I did not perform any multivariate parametric modelling. I grouped the individual responses to the open-ended questions to produce categories of like answers that could then be all scored via the statistical software SPSS (v.25).

5. Results

5.1 Demographic and covariant questions

At the time of my visit, 84 female inmates were held in Golden Grove. All inmates were eligible to take part in the study, and 79 agreed to participate, giving an exceptionally high return rate of 94%. At the time of my visit, 32 inmates had been convicted, 47 were on remand, 4 declined participation and 1 inmate was unable to take part because she was confined in isolation (likely for misbehaving). By nationality, 66 (83.5%⁵) inmates were from Trinidad and Tobago, and another 13 (18.3%) were from other countries, primarily in the Caribbean⁶. By age, most of the women were between 29 and 39 years old (n = 33, 42.9%), followed by those between 40 and 50 years old (n = 18, 23.4%) and 18–28 years old (n = 11, 14.3%). Moving from the basic demographics to the experiential questions, I asked the respondents if they had been incarcerated in the past and if so, how many times and 74.4% (n = 58) of the total sample had a prison record; over three-quarters, 78.9% (n = 15), had already been imprisoned one to three times. Figure 1 graphically presents the relationships between the women’s ages and their prison records.

Figure 1.

Population pyramid of age by prison record.

At the time of my visit, 55 (73.3%) female inmates were participating in at least one in-prison programme, most pursuing some type of education organised by the prison ministry (computers, language: n = 33, 63.5%; religious content: n = 32, 61.6%). Other programmes included employment (mainly business start-up, gardening, hairdressing and sewing) and personal and social (e.g. music and dance) skills development. Further, I asked the detainees what charges they were serving their time for at the time of my visit and whether they themselves had ever been victims of a crime, along with what kind of victimisation they had experienced if the latter answer was yes. More than one in three female prisoners at the Garden Grove penitentiary in Arouca, Trinidad (n = 11, 36.6%), had experienced some form of physical violence, including sex offences (rape). Other inmates had been victims of robberies and larcenies (n = 10, 33.3%) and drug-related offences (n = 4, 13.3%). Figure 2 visually presents the types of crimes the study participants had been the victims of according to their statuses as convicted or remanded prisoners.

Figure 2.

Respondents’ victimisation types by their prisoner status.

The prevalence of drug trafficking among female offenders was also reflected in the criminalisation trend. The two most common reasons for the women’s imprisonment were drug trafficking and murder/manslaughter: n = 29 (42.6%, Figure 3).

Figure 3.

Study participants’ reasons for imprisonment by status as convicted or remanded.

In terms of religion, the vast majority (n = 66, 84.6%) of the female inmates in this study considered themselves to be religious. Focusing on the official religions in Trinidad and Tobago, 74.6% were Christians (including Catholic, Orthodox, Baptist, Jehovah’s Witness and Seventh-Day Adventist), 10.4% (7) were Muslims, 7.5% (5) followed the Orisa faith and only 3% (2) were Hindus. The remaining self-proclaimed religious prisoners were either Rastafarian or Jewish.

In continue, we tested H1 and H2 against demographic covariants and found no positive correlations between prior victimisation and criminal activity (p = .118⁷) or between types of victimisation (p = .953), like verbal or physical violence, robbery, drug trafficking/use, and sexual abuse. That is, previous victimisation experience was not a contributing factor to the reasons for current imprisonment. Religiousness also did not significantly correlate with any of the covariant variables (recidivism: p = .182; type of crime committed: p = .573), and length of prison sentence had no influence on self-reported religiosity (p = .322). As could have been expected, the only significant bivariate correlation was between prison record (previous incarceration) and length of current imprisonment (p = .026).

5.2 Restorative justice questions

A central tenet of restorative justice is re-establishing relationships, in which process apology plays an important role. Overall, an unexpectedly high number of female prisoners (n = 58, 73.4%) admitted that they were willing to ask for forgiveness. From those 58 respondents, 42 also considered themselves to be religious, providing a statistically significant correlation between perceived religiousness and willingness to apologise to victims (p = .028). That is, religion appeared to have a positive impact on female inmates’ inclination to seek forgiveness from their victims. On the contrary, religion did not appear to influence their inclination to explain their offences either to their actual victims (p = .892) or to surrogates (p = .973). Rather, the prisoners’ individual inclinations to apologise as means to restore relationships with their victims were related to their desire for personal encounters or mediation (p = .015) but only with their actual victims (p = .009). Figure 4 presents the correlations between the respondents’ religiousness and their stances on apologies and explanations to their victims.

Figure 4.

Subgroup plots A.

I asked open-ended questions to learn the prisoners’ views on the possible effects of their crimes and to understand their motivations behind their involvement in any future victim-offender mediation (VOM) or the performance of reparative work (see Annex). It is worth mentioning that although the majority of participants (n = 70, 89.7%) were able to acknowledge the impacts of their crimes, they mainly recognised the effects on their own family members or friends (n = 55, 77.5%) or on themselves (n = 50, 70.4%) rather than on victims (n = 16, 22.5%) or on the community at large (n = 19, 26.8%; Figure 5).

Figure 5.

Study participants’ recognition of the impacts of their crimes on others.

This conclusion was partially verified by the answers regarding their motivations behind any eagerness to work while in prison. Again, although 89.5% (n = 68) of the participants reported being willing to undertake any restorative work in prison, they were mainly driven by their desire to gain new skills and improve their relationships with their victims while feeling good and useful; the pious inmates were neither more nor less inclined to perform restorative work, and religiousness was not significantly related to prison work participation that would benefit victims or the community (p = .282; Figure 6).

Figure 6.

Respondents’ reasons for engaging in prison work according to their religiousness.

Meanwhile, just as the study participants were willing to participate in prison work, they also showed keen interest in victim-offender mediation (VOM). Nearly half of the respondents held positive attitudes towards meeting their victims in person, to clarify their crimes (n = 10, 24.4%), express sadness and even condolences (n = 7, 17.1%), or ask for forgiveness (n = 6, 14.6%). Interestingly, nine inmates admitted having already attempted to contact their victims but had not succeeded and believed that their victims did not wish to reply to their petitions (n = 9, 22%). Some of the prisoners who were negative towards VOM were afraid that their victims would want to take revenge or felt too guilty and ashamed to face them (cumulative n = 3, 7.3%; Figure 7).

Figure 7.

Study participants’ explanations related to VOM by their religiousness.

Finally, I tested the study hypotheses by highlighting any disparities in answers provided by religious and non-religious (or unsure) females; Figure 8 shows that I found no significant correlations between the inmates’ religiousness and their ability to understand the impacts of their offences (p = .367) or to participate in mediation meetings (p = .823). That is, the religiosity or lack thereof of the participants in this study was not influencing whether they would engage in a restorative justice programme behind bars. Figure 5 presents the correlations between the respondents’ religiousness and their stances on VOM, their willingness and desire to participate in prison work, and their abilities to understand the impacts of their offences (Figure 8).

Figure 8.

Subgroup plots B.

6. Conclusions

With the present research, I sought to demonstrate the interrelationships between religion, the ideals of restorative justice in custodial settings and individual inmates’ perceptions in practice. This exploratory study aims to uncover possible relationships between restorative justice, religion and imprisonment by allowing female inmates to express their views and opinions, raise concerns and criticisms, and bring their unique insights to the restorative justice debate.

6.1 Limitations

No empirical research is free of limitations, and I discuss those here before I address any conceptualisation of our hypotheses. Empirical research findings are not formal proofs of facts but rather support or reject hypotheses. However, there is always uncertainty in the results [74], which was expressed in this study as statistical probabilities.

Initially, I note that Trinidad and Tobago is one country and that it could very well have different social norms and moral values from those of other cultures or social spheres. It is possible that institutionalised females across the Caribbean region share commonalities, but it is not necessarily the case that these findings for Trinidadian women are representative of the beliefs of all or even most female prisoners. Keeping in mind the exceptionally high percentage of the women in this study who themselves had experienced first-hand violence and victimisation in their lives, as well as the cultural and ethnical diversity of Trinidad and Tobago, it would have been scientifically imprecise to claim that these results were generalisable and that the present sample represented female prisoners beyond the Golden Grove population. In addition, given the relatively small sample and population and the nature of the questions, I urge drawing conclusions with extreme caution. Owing to conceptual and structural limitations related to constraints on generalisability, I could not test complex theoretical models from the existing literature; however, my work here does aid in formulating more specific hypotheses for future investigation.

A final limitation of this study relates self-selection bias in the collection of the empirical data. This problem inherent to all restorative justice research is that no assignment of study participants can be truly random; participants are necessarily self-selected into those who wish to participate in some form of restoration and those who do not. That is, restorative justice is by its nature a voluntary process that automatically creates a group of participants (here prisoners) who could individually and in general be more motivated or inclined to engage in restorative justice activities than a given prison’s total population. This phenomenon could have skewed the data gathered here and threatens both the internal and external validity of the findings [75].

However, selection bias is minimal if not inapplicable here for three reasons: firstly, all prisoners had the opportunity to participate in this research by completing the study questionnaire; secondly, although it is true that prisoners who volunteer are by definition different from inmates who do not volunteer (and these differences are often not easily understood), there was no prior indication that the prisoners who participated in the research would necessarily have volunteered for restorative justice and thirdly, I did not evaluate an already implemented restorative justice programme but rather aimed to collect inmates’ opinions and desires towards the aim of tailoring an in-prison restorative programme specifically in and for Trinidad and Tobago.

6.2 Final thoughts

Bearing in mind the perceptions of the female inmates at Golden Grove prison in Arouca, Trinidad, combined with my reflections based on observation and personal experience, my primary recommendation from would be less retribution–more research. The culture of violence that currently prevails in Trinidad and Tobago is not being effectively addressed, and the ephemeral plans that focus on retributive crime prevention do not seem to provide any desired effect. The notion of restorative justice in Trinidad and Tobago is still in its infancy due to many unmet challenges and ignored risk factors. To meet these challenges, we must link restorative justice to attempts seeking to address the deeper causes of crime such as poverty, social exclusion and inequalities, family disintegration, poor parenting skills, lack of quality education, poor housing and, as already mentioned earlier, the proliferation of drugs, gangs and guns. For restorative justice to work effectively, all stakeholders—police agents, judges, probation/parole officers, politicians, religious leaders in particular, etc.—need to “buy in” to the concept and refrain from calling the hangman in response to every crime⁸.

There is also a demand for continuing the discussion on restorative justice policy, which commenced in 2014 under the aegis of the Ministry of Justice. Such a policy should not restrict restorative justice implementation to solely criminal acts, prisons or offender rehabilitation; rather, it needs to be holistic and identify those who have been harmed, which relationships have been broken and what needs to be done to repair the harm. Undeniably, programme providers should pay attention to the match between the programme content and the characteristics of their potential participants, and further research on restorative justice and religiosity in women’s female prisons is needed to more efficiently distribute resources to better serve the needs of women, who are often both victims and offenders within the criminal justice system.

In addition, better understanding the types of inmates who are most likely to avail themselves of voluntary faith-based programmes would help policymakers and programme providers plan these services. The issue of who volunteers for restorative justice is crucial for evaluating whether and on what basis inmates might benefit from religious programmes. Following my initial argument, religion would seem to be an important motivator in seeking forgiveness, but I found no such positive relationship in the present study between religiosity and restorative justice, specifically female prisoners’ willingness to engage in mediation or reparative work. This leads me to conclude that religion should be used with caution in the design, delivery and dissemination of restorative justice programmes for female prisoners in Trinidad and Tobago.