Understanding forced marriage protection orders in the UK

ABSTRACT

This article examines the use of Forced Marriage Protection Orders (FMPOs) in England and Wales to determine which framing narratives affect the outcomes of FMPO cases. Forced marriage is marriage without the consent of one or both parties and is legally recognised as a form of domestic violence in the UK that primarily affects women and girls; FMPOs are civil injunctions designed to prevent forced marriage and protect its victims. Although approximately 200–250 FMPOs have been granted annually across Northern Ireland, England and Wales since 2014, little is known about how the legislation functions. This study used a qualitative socio-legal approach to understand the application and interpretation of the law and the broader socio-political context that shapes this process. It analysed 33 FMPO-related judgements, finding that perceptions of culture, consent, disability and victim credibility influenced how evidence was interpreted and how forced marriage was constructed. It also examined case outcomes and found that FMPOs were breached in a substantial minority of cases and that victims with disabilities faced significant barriers to justice. The study makes a number of recommendations to ensure that FMPOs can function effectively, such as providing training for judges and legal personnel and offering greater witness support.

KEYWORDS:

• Forced marriage case law
• coercion
• consent
• culture
• disability
• forced marriage protection orders
• victim credibility

Introduction

A forced marriage is a marriage that takes place without the consent of one or both parties: it includes child marriage as, by definition, children cannot give informed consent (Anitha and Gill 2009). Forced marriage is a serious violation of an individual’s human rights, including the right not to marry; when it happens to children, it constitutes a form of child abuse (Gill and Gould 2020). It places victims at risk of honour killings and commonly leads to a range of serious and enduring harms, including abduction, domestic violence, rape, forced pregnancy and domestic servitude (Anitha and Gill 2015; Gangoli 2011, Gill and Walker 2020).

The policy documents are careful to distinguish between arranged marriages (consensual) and marriages that are forced or coerced (Foreign and Commonwealth Office and Simmonds, M. 2013, Forced Marriage Unit statistics 2018 2019). However, most academics acknowledge that, in reality, there is a spectrum of levels of pressure or coercion in relation to marriage (Anitha and Gill 2009; Enright 2009, Gill and Gould 2020) ranging from physical violence to parental/cultural expectations that children will marry by a certain age that may make young people compliant against their wishes.

Forced marriage is often framed as a cultural problem (Gangoli 2011) or ‘harmful traditional practice’. This contributes to the narrative that certain cultures and places need to ‘modernise’ or develop in order to eradicate forced marriage. However, feminist academics and activists have argued that the practice can also be viewed as a form of domestic or gender-based violence (Anitha and Gill 2009: Gangoli 2011, Gill and Anitha 2011, Chantler 2012, Yurdakul and Korteweg 2013): in other words, forced marriage can be seen as part of broader patriarchal patterns of coercion and control (Anitha and Gill 2009). Considering forced marriage as related to forms of violence and coercion that are not commonly interpreted as purely ‘cultural’ sheds new light on why forced marriages happen and how they can be prevented (Gill and Gould 2020).

The UK Government’s Forced Marriage Unit (FMU) has supported 1200–1400 forced marriage victims and people at risk of forced marriage each year since 2011. This is likely an underestimate, as many people forced into marriage do not, or cannot, contact the authorities. In the UK, forced marriage primarily affects women and girls from particular Black and minority ethnic (BME) communities; the majority are of South Asian origin, with a substantial minority being of Middle Eastern origin (FMU 2018). This crime is highly gendered, with a majority of cases involving women and girls (75% in 2018), while about 5% of cases (5.3% in 2018) involve people with disabilities. Victims are generally young, with the highest numbers in the under-15 (17.7% in 2018) and 18–21 age groups (17.4% in 2018).

Forced Marriage Protection Orders (FMPOs) are a form of civil injunctions first introduced in 2007 with the Forced Marriage (Civil Protection) Act 2007, which applies to Northern Ireland, England and Wales; the Forced Marriage (Protection and Jurisdiction) (Scotland) Act came into force in 2011. Before the introduction of these laws, activists and advocates from communities particularly affected by forced marriage strongly recommended civil forms of protection, rather than criminal sanctions, as the best way to tackle the issue (Patel 2008, Gill and Anitha 2011, Asokla 2018). They argued that many victims would be reluctant to see their families prosecuted, resulting in fewer people seeking help, and that this would be exacerbated by the introduction of any law that would allow the police to press charges without the agreement of the person at risk (Gill and Anitha 2011; Askola 2018). However, following concerns that FMPOs might not be entirely effective due to regular breaches, forced marriage was criminalised in Scotland, England and Wales in 2014 and in 2015 for Northern Ireland.

The focus of this paper is on FMPOs. Although approximately 200–250 FMPOs have been granted each year across Northern Ireland, England and Wales since 2014, little is known about how the legislation is functioning. Based on an analysis of reported court case judgements on FMPOs, this article investigates key themes that emerge from these judgements to understand prevailing legal constructions of forced marriage, and the efficacy of existing remedies.

Methods

Analysing case law can demonstrate whether a law is functioning well in practice and identify any procedural issues. However, analysis of reported case law does have limitations. These cases are often more complex than others or have some novel aspect; indeed, they are reported specifically because they set a precedent. As a result, the sample investigated here may not represent the range of cases brought before the courts. Nonetheless, as the judgements from cases that proceed to higher courts offer a unique insight into the workings of the law, the sample is particularly well suited to the purposes of this study.

Research focusing on judgements often takes a doctrinal approach by first outlining the law pertaining to a certain issue: ‘the primary or even sole aim is to describe a body of the law and how it applies’ (Dobinson and Johns 2007, p. 21). Other studies employ ‘problem, policy and law reform’ methods; analysis focuses on a particular problem and its influence on the law in order to recommend legislative development and improvements (Dobinson and Johns 2007). This study falls within the category of socio-legal research (Dobinson and Johns 2007), as it examines the influence of society on law (and the converse) rather than undertaking a strictly legal analysis. The approach is grounded in the feminist tradition, which seeks to scrutinise judgements in order to identify the ‘social facts’ (Hunter 2015) – underlying assumptions that are accepted as common knowledge and therefore not substantiated – that underpin these judgements (Burns 2012). These assumptions often encode existing social hierarchies and the forms of discrimination that arise as a consequence (Hunter 2015).

Data collection and analysis

Judgement searches were undertaken, using the search term ‘Forced Marriage Protection Orders’, in the 2010–2020 case law and law report sections of four legal databases: BAILII, Westlaw, Lawtel, and Lexis Library. Duplicates and judgements that only referred to forced marriage in passing were removed, resulting in a sample of 33 judgements on cases that had a substantial focus on FMPOs: 30 from England and Wales, two from Northern Ireland and one from Scotland.

Qualitative thematic analysis was used to identify patterns across the judgements, then group these patterns into key themes. Thematic analysis was also used to identify differences across the sample. All judgements were read closely, and a coding framework was then developed based on the research aims and questions. QSR NVivo 12 was used to code each judgement iteratively, with further codes identified during the coding process as new themes emerged.

Research findings

The first section below outlines factors that lead to forced marriage and co-occurring forms of violence, as gleaned from the judgements. Subsequent sections analyse the key themes that emerged from the judgements, including how perceptions of culture, consent, disability, and victim credibility shape interpretations of evidence and constructions of forced marriage. Finally, the paper examines case outcomes, including the kinds of orders made, as well as examples of breaches of FMPOs.

Factors leading to forced marriage

This section explores the factors that often lead to forced marriage, including co-occurring forms of violence in families, such as domestic violence and child abuse. Although forced marriage is often viewed through a cultural lens, the sample demonstrates the significant overlap and commonalities with other forms of family violence.

Physical abuse was the most common co-occurring type of violence, being mentioned in 15 out of the 33 cases: nine cases involved both domestic violence against the mother of the subject of FMPO and physical abuse of the children in the family, while six involved historic and ongoing child abuse/neglect. The different varieties and forms of violence identified in the judgements is best understood in the context of a continuum of violence (Kelly 1988, Gangoli 2011). While forms of abuse such as physical violence, emotional abuse and sexual violence may seem discrete, they often co-occur: ‘forms of sexual violence shade into each other at various points’ (Kelly 1988, p. 67), as do child abuse and domestic violence against the mother.

There has been greater recognition in recent years that children are affected by living in households in which domestic violence takes place: they are more likely to be physically abused themselves, while witnessing domestic violence has deleterious effects on their development. Multiple forms of abuse, maltreatment, and victimisation affect a significant minority of children and are associated with higher rates of trauma and other mental health issues. According to Wolfe (2018, p.833; also see Price-Robertson et al. 2014), ‘children who experience one type of violence are more likely than not to have experienced (or will experience) others’. As a result, Wolfe argues that, excepting work on single-event experiences of violence, victimisation should generally be viewed as ‘an ebb and flow of negative events that creates a lifelong burden’ (Price-Robertson et al. 2013, p. 833). Multiple traumas result in an overload for children, resulting in difficulties in ‘adapting’ to these events (Wolfe 2018).

As a result, 10 of the children and young adults (from eight cases) discussed in the judgements demonstrated externalising and internalising behaviours, including violence, trouble at school, truancy, aggression towards siblings, harmful sexual behaviours, suicidal ideation and/or suicide attempts; this pattern of difficulties was similar to the ‘developmental consequences of victimisation’ (Musicaro et al. 2019, p.83–84) displayed by other groups of young people experiencing victimisation. However, the ‘bad behaviour’ (e.g. entering into a sexual relationship with a consenting partner) identified by parents in the judgements was often used as a further justification for parental attempts to force children and young people to marry.

These findings support feminist theories that forced marriage should be viewed as a form of domestic violence and that, like other forms of domestic violence, it is linked to historically and culturally specific manifestations of patriarchy. By contrast, explanations that view forced marriage primarily as a cultural phenomenon neglect to identify major commonalities between this and other manifestations of patriarchal violence and coercive control.

Formulations of consent and coercion

This section explores the various ways consent was constructed and conceptualised in the judgements. Judges’ definitions of forced marriage and arranged marriage varied across the sample, though most demonstrated an understanding of arranged marriage and forced marriage as existing on a spectrum. Elsewhere, the authors have argued that ‘consent is often constructed in the context of power imbalances and gendered norms, and, crucially, often in the absence of explicit threats’ (Anitha and Gill, 2011, p.54).

The standard of proof in civil courts is based on ‘the balance of probabilities’, which is easier to meet than the criminal standard (i.e. ‘beyond reasonable doubt’). Victims applying to civil courts need to prove that their claims are more likely true than not in order to receive a protective injunction. Thus the onus is on the claimant to prove claims about coercion. In more than half of the 33 cases, high levels of evidence were required to recognise marriages as forced. A more specific breakdown of these cases is given in the analyses below.

The way the judges discussed consent in relation to forced marriage and sexual assault was not affirmative (i.e. based on clear indications of willingness) and did not often take into account the context in which the choice to marry was made. While early definitions of coercion in a legal context centred on physical coercion, the legal definition of forced marriage has recently evolved to include emotional and psychological forms of coercion (FCO and Simmonds 2019). Affirmative models of consent, which are standard in UK law in the context of sexual coercion, go further: the defendant must prove that they acted to secure consent. According to Crown Prosecution Service (2020) guidance, ‘the defendant (A) has the responsibility to ensure that B consents to the sexual activity at the time in question’. However, the evidential burden was (arguably improperly) placed on the claimants to prove that they did not consent to the marriage or sexual act in question.

Of the cases examined in this study, sexual assault was alleged in 13 judgements. Furthermore, the evidential burden was not always underpinned by an affirmative concept of consent that is applied to sexual violence, albeit in criminal courts. All civil courts should apply the same standards: in this instance, they should draw upon the same definitions to establish whether marriage and/or sex acts were entered into by choice and whether the claimant was free to make this choice without coercion, deception or other impediments. At present, the relevant wording in the FMPO Act (section 63a) is ‘a person (A) is forced into a marriage if another person (B) forces A to enter into a marriage (whether with B or another person) without A’s free and full consent.’ If consent implies an active and conscious choice, then it follows that free and full consent should be ascertained before a marriage or sex act takes place, otherwise a civil court should hold that the marriage or sex act is entered into without consent.

In four cases in which the victims themselves alleged that their marriages had been forced, three of their statements – both those given in court and earlier statements made to police – were not taken to be reliable evidence or were not carefully considered by the judges. In Re C (Female Genital Mutilation and Forced Marriage: Fact Finding) [2019] EWHC 3449 (Fam), a Thames-Valley-based Muslim woman of Kenyan/Somalian heritage with a disability (‘C’) claimed that her marriage was forced, but the judge found that her marriage had been arranged. Before making this finding, the judge said she would consider the circumstances of C’s sisters’ marriages, as they were pertinent. One of C’s sisters (‘B’) had said in a police statement that she (B) had also been ‘persuaded’ to marry though she did not want to, but her mother kept ‘telling her’ to get married. However, in her oral evidence in court, B denied this. In relation to C (who was referred to as ‘the mother’ throughout the judgement), the judge ‘concluded that the mother’s marriage to the father was not forced as she claimed’, going on to say:

However, I was not satisfied that this was a love match entirely free from family influence as was suggested. Rather, this struck me as a consensual arranged marriage brokered initially by the maternal grandparents after they met the paternal family at the wedding of [another party] … That the mother’s marriage was arranged rather than forced fits more neatly with the evidence that the mother and father got to know each other on the telephone and met on a couple of occasions before the marriage ceremony.

In West Sussex County Council and another v F and others [2018] EWHC 1702 (Fam), the judge reflected on the grey area between arranged and forced marriage, but decided that the betrothal ceremony that took place in Pakistan involving a 13-year-old Muslim girl of Pakistani heritage (‘N’) and her cousin in his early 20s (‘B’) had been ‘consensual’. As the parents claimed that the girl was not to be married until 18, the judge expressed confidence that she would be able to refuse at that time if she decided not to proceed with the marriage. This was a complex case as the girl (and her siblings) were said by the judge to have exaggerated parts of those earlier/retracted statements. However, given that the children met the threshold for being taken into care, and the parents had taken the daughter in question out of education completely in Pakistan, it is interesting that the judge did not provide evidence for his satisfaction that the parents would respect the girl’s wishes in the future:

In terms of … the distinction between consensual arranged marriages and forced marriages involving serious human rights abuses this was far closer to the arranged marriage end of the spectrum but taking the form of a loose commitment not undertaken with N’s consent but not with the intention subsequently of her being forced to marry B whether she wanted to or not. The nature of the arrangement was such that had it endured to an age when she might legally have married I am satisfied she would have had a say and had she not wished it, it would not have proceeded.

This framing does not seem to consider the significant age difference between the girl and her cousin. Child marriages often involve large age discrepancies, with a recent Turkish study finding a mean age gap of 8.1 years between a child bride and her husband (Kuygun Karci et al. 2020). These marriages often impede the girl’s decision-making ability across many aspects of her life (Raj 2010, Falb et al. 2015, Kuygun Karci et al. 2020). There is also a higher risk of intimate partner violence in such relationships, and an association with negative reproductive and sexual health impacts (Falb et al. 2015). Furthermore, if a girl is betrothed in a public ceremony at an early age, this makes it difficult for her to subsequently refuse to go ahead with the marriage because of the commitments already made by the families and the potential loss of face if the bride-to-be backs out (Kuygun Karci et al. 2020). In cases involving early betrothal, women are generally made to complete the marriage as soon as is practical (e.g. immediately after completing high school or turning 18), which gives them little opportunity to reconsider (see Gangoli 2011,, Chantler 2012). Deception was used in six such cases in the sample, with families taking girls or young women away on the pretext of a holiday or a visit to see sick relatives only to then reveal the real purpose of the trip was marriage.

In terms of cases involving people with disabilities, AB v HT and others [2018] EWCOP 2 involved a London-based woman (‘M’) of Somalian heritage who had both a physical disability and a severe mental illness. She stated that she was forced into marrying her husband (‘MS’).

The judge accepted her aunt’s evidence that M’s father had said he would ‘make her’ get married. He also heard evidence regarding MS being concerned about his immigration status: according to Dr Shaikh, one of the witnesses for MS testified that while in a group of men, MS had asked if he could marry anyone’s daughter to address these worries, to which M’s father replied, ‘my daughter’. Meanwhile, the local authority argued that ‘Little, if any, consideration was given to M’s wishes and feelings.’ Similarly, the Official Solicitor, in their capacity as M’s litigation friend, stated:

A clear picture emerged indicating that AB [the father] wanted to arrange a marriage for M because he considered it to be culturally and religiously imperative. Her wishes as to the choice of a marital partner did not feature prominently in his thinking. M herself was not in any position to make much of an informed choice because she was only introduced to MS a few days before the wedding.

Nevertheless, the judge found that the marriage was not forced:

On a balance of probabilities, I accept HT’s [M’s aunt] evidence that on 15 August 2013 she told AB that M did not want to marry MS and that AB responded ‘I’ll make her’. There is, however, no evidence of any pressure being exerted on M and, in the light of my finding that it is impossible to discern M’s true wishes and feelings on that day, however, I do not find that AB did in fact force M to marry MS.

Taken together, the above examples show that marriages considered by judges to be arranged rather than forced included those where there was limited or no consultation with the bride-to-be, where the bride had no or limited opportunity to meet her future spouse and/or turn down the marriage, and those where the bride’s consent was not sought for the marriage. In four cases of forced marriage and one case of a betrothal ceremony, the brides/brides-to-be were vulnerable: they were either very young or had limited capacity/severe mental illness. In cases such as M’s, where a marriage has already taken place, in order to rule that the marriage was forced because of inability to consent, it must be proven that the bride did not have capacity at the time of the marriage (Mental Capacity Act 2005, s15.1; Mental Capacity Act 2005, s.13, Matrimonial Causes Act 1973, s.12;). This was difficult to prove in M’s case, despite her very low IQ and the fact that she was being involuntarily held in a mental hospital during the court proceedings. C also had ‘very significant learning disabilities’ and, according to the psychologist who assessed her, ‘likely … met the DSM5 criteria for Post-Traumatic Stress Disorder, Depression and General Anxiety with panic attacks’ (CPS 2019).

Despite the law’s recognition of the role emotional pressure plays in consent versus coercion (Anitha and Gill 2009), formulations of consent in civil cases still seem to rely on a passive construction whereby the mere absence of an expression of dissent to marriage at the time of the wedding is deemed to indicate consent, even in the context of vulnerabilities and significant barriers (including learning disabilities, mental illness and deprivation of liberty) to expressing such dissent.

Cultural considerations

Culture emerged as another key theme in the cases. In common with previous research on forced marriage (Anitha and Gill 2014), our analysis demonstrates that culture was framed in a dichotomous manner, with non-Western cultures positioned as backward and traditional and Western ones as progressive and modern. While other studies have found similar framings in interview-based projects and analyses of UK Government policy (Anitha and Gill 2015, Enright 2009), this project is the first to analyse the way culture has been understood in FMPO proceedings.

Many judges were careful not to offend when discussing culture. However, subtle elements in their judgements showed that they viewed forced marriage as linked to culture through tradition. Family histories of forced marriage were seen to indicate a higher degree of risk, whereas love marriages in recent generations were viewed as mitigating risk. However, this arguably avoids assigning an appropriate level of seriousness to cases where parents who had a love marriage themselves sought a forced marriage for their child (e.g. because of perceived poor marriage prospects, perceived bad behaviour, coercive control or changes in the parents’ personal beliefs).

When tradition, culture and lack of choice are perceived as related, it is unsurprising that judges and respondents tend to make claims intended to signpost their professional status and/or allegiance to modern values, including metropolitan living and the importance of education. Such claims were used in several cases in the sample to deny that a forced marriage had taken part and/or portray even the possibility as implausible. For example, in London Borough of Camden and RZ and HZ and DZ and SZ, Neutral Citation Number: [2015] EWHC 3751 (Fam), involving male and female minors of Afghani background living in London, the judge made the following statement:

Finally in the context of it being alleged by the local authority that he intended to force his daughter into marriage the father has produced information regarding his family in Afghanistan, contending that he comes from a well-educated, metropolitan family for whom forced marriage would be an anathema. The father told the court that one of his sisters is a doctor, one is a teacher and one of his brothers is a retired senior army officer.

Meanwhile, in Re A and A, Case No: NE16C00241, involving female minors living in Newcastle but from a religious Kuwaiti background, counsel on behalf of the guardian drew a clear distinction between Kuwaiti and British culture, arguing for two teenage girls to remain with their foster families in Britain rather than returning to Kuwait. His argument, which was referenced in the judgement, drew on the girls’ ‘integration’ into British society and their ‘rejection’ of Kuwaiti culture:

Mr Grey makes the point, on behalf of the Guardian, that the children have in fact rejected the culture and lifestyle afforded by their own family and both have embraced the alternative possibilities provided by foster care in the UK. These children are well integrated into their life in the UK [.]

In A Local Authority – and – (1) M (2) F (3) A, B, C, D, E, F, G (via their Children’s Guardian), Neutral Citation Number: [2018] EWHC 3295 (Fam), which involved male and female minors (all siblings) of Pakistani Muslim background (some with disabilities), the evidence of the father’s prior conviction for assaulting one of his daughters was raised in court. Care orders were sought for the eldest and placement orders for the younger children (E and F), all of whom lived in North West England. G, who resided in Pakistan, was already a ward of the court. When the older daughters were in their mid-teens they became concerned that their father would force them to marry cousins in Pakistan by their twenties: the judge ordered that existing FMPOs be continued. In relation to the assault, the sentencing remarks of the Honorary Recorder referred to the father bringing shame on his community:

You have brought shame on yourself, and you have brought shame on your community by the way in which you treated your daughter. You struck deliberately with a wooden spoon because she would not do her homework. She is 9. That is barbaric behaviour.

This reference to barbarity, which was echoed in the FMPO-related proceedings, is significant because, as Authors (2011) argue, when crimes such as forced marriage are viewed as ‘cultural’, the host nation is positioned as ‘liberal and neutral’ while the ‘othered society is essentialised as atavistic and illiberal’ (Anitha and Gill 2011, p.49; Razack 2004). Abji et al. (2019) refer to such framings as ‘culture talk’ in their exploration of how essentialising cultures in this way ‘reinforces the idea that certain cultures are barbaric’ (p.799) or more prone to encouraging acts of violence. This characterisation by the judge suggests that the father’s violence reflects on the standing of his community rather than just on him as an individual. Such collective attributions of shame are key aspects of the process of racialisation of minoritised people as bearing common moral attributes that all members of that group share. White offenders, as part of the normative category in postcolonial societies (Frankenberg 1993), are seen as offending because of individual pathologies rather than cultural (or gendered) reasons.

In two cases, one involving a Somali woman and the other a Bangladeshi woman, both with disabilities, an expert witness, Professor Rehman (who specialises in Islamic law), made the contentious argument that in Islam it is legally possible for a guardian to marry off someone underage or with a disability. The choice of expert witness has the potential to strongly influence the outcome of any proceedings. For example, in AB v HT and others [2018] EWCOP 2, the judge stated that

in light of the evidence of Prof. Rehman, I do not consider it necessary, appropriate or proportionate for this court to analyse further the lengthy allegations made on HT’s [the aunt’s] behalf concerning AB’s [the father’s] failure to address this issue [of whether the father should have sought to determine whether the daughter had capacity to consent to marry].

These cultural considerations also extended to class. In most of the judgements involving applications for FMPOs to protect highly educated people, their educated families were framed as modern and therefore unlikely to coerce them into marriage. Asking to complete university education is the most common tactic used by young people who wish to avoid or delay marrying (Anitha and Gill 2011), but completing a degree does not necessarily mean a person will be able to successfully resist being compelled to marry. Markers of ‘modernity’ (e.g. high educational attainment, secular values and middle-class/professional status) are no guarantee that a family will not attempt to pursue a forced marriage (Gill and Gould 2020). ‘Modernity’ does not necessarily signify a decline in patriarchy and a respect for women’s rights: indeed, particular forms of violence against women, including forced marriage, have adapted to the modern context.

As demonstrated above, some judges and respondents equated modernity and progress with Britishness and the West, othering cultures where forced marriage is common. This risks assuming that ‘British ideals’ and traditions from the East/Global South are antithetical when, in practice, many immigrants successfully navigate dual identities on a day-to-day basis. However, as analysis of the judgements demonstrates, the foregrounding of culture in forced marriage cases often elides these complexities in favour of a more simplistic and dichotomous framing (Anitha and Gill, 2011).

Disability, the right to marry, and constructions of capacity

Thirteen cases in the sample involved people with disabilities. However, in two of these cases, the disabled person was the mother or sister of the person to be protected by the FMPO (i.e. the claimant). Nine claimants had learning disabilities, four had mental illnesses, one had an acquired brain injury, one was profoundly deaf and another had an autism spectrum disorder; four had multiple disabilities. Their cases were often focused on the rights to marry and/or have sexual relations and the need to balance these rights with mental capacity and the right to be free from coercion in matters of marriage. Some of the cases outside the Court of Protection (mainly in the Family Court) also involved claimants, including children, and family members with learning disabilities.

Judgements in the five cases related to disability heard in the Court of Protection (and Family Division) mostly involved vulnerable adults who did not have the capacity to consent to marriage and/or sexual intercourse or who were found to be borderline in terms of their mental capacity. In previous research, including Gangoli et al. (2009), in researching early/child marriages in the UK, they found that both were more likely if the ‘young person was not performing well educationally or professionally’ (Gangoli et al. 2009, p. 423).

Women with disabilities experience much higher rates of sexual assault than women without disabilities – estimates range from over three times higher (Harris 2018) to four times higher (McGilloway et al. 2020). The right to autonomy and a private life must therefore be balanced with arguments about protection from harm (Onstot 2019). People with disabilities may experience communication issues, which can make it more difficult for them to give consistent and linear accounts to police and in court. Moreover, McGilloway et al. (2020) found that people with disabilities did not always have appropriate sexual education, which increased their vulnerability to sexual assault. Stereotypes concerning people with disabilities include ideas of them as ‘promiscuous’ (McGilloway et al. 2020) and as ‘sexual deviants’ (Onstot 2019); these perceptions can, in turn, lead to increased victimisation. Harris (2018) describes the influence of what she termed ‘the aesthetics of disability’ – this was evident in some of the cases in the sample, where witnesses with disabilities were made to appear in court simply to demonstrate how their disability impaired their ability to consent. Mental illness can also affect how victims and at-risk persons are perceived by the police and courts. Women who engage in ‘risky behaviours’ or have ‘behavioural health challenges’ (including substance abuse) are viewed as less credible (Morabito et al. 2019).

In the judgements, the gendered dimensions of forced marriage for people with disabilities were complex. In four judgements, the need for disability-related care was recognised as one of the key motivations for marriage: Re RS (An Adult) (Capacity: Non recognition of Foreign Marriage) [2015] EWHC 3534 (Fam), The London Borough of Southwark v KA and others 2016 EWCOP 20, XCC v AA and others [2012] EWHC 2183 (COP), and AB v HT and others [2018] EWCOP 2. Two of these claimants were male and two female. In a fifth case (YLA v PM and another [2013] EWHC 4020 (COP)), the wife had significant care needs that her husband was meant to fulfil; however, the husband told his child’s foster parents that he felt ‘duped’ by the woman’s parents with regard to her disabilities, and the court decided that he could not provide effective care for his wife or their small child. The two cases involving women seeking FMPO-related relief reflected previous research findings that parents of men with disabilities regularly seek wives who will care for their sons, as women are far more often expected to fulfil caring roles than men (Clawson and Fyson 2017). Meanwhile, in the three cases involving husbands being found to provide care for women with disabilities, the men neglected their spouses and did not fulfil their anticipated caring roles (YLA v PM and another [2013] EWHC 4020 (COP); AB v HT and others [2018] EWCOP 2); these husbands had their own immigration-related motivations to marry, with both seeking to bolster their case for leave to remain after previous immigration appeals had been denied (AB v HT and others [2018] EWCOP 2; YLA v PM and another [2013] EWHC 4020 (COP)).

Marriages involving two women with severe intellectual disabilities who then became pregnant gave rise to concerns about the ability to understand and, thus, consent to sexual intercourse and/or to the ability to realise that sex can result in pregnancy and/or sexually transmitted infections (YLA v PM and another [2013] EWHC 4020 (COP)). Moreover, children born from marriages where the mother had severe disabilities resulted in involvement with child protection services in three cases on the basis that the mother was at greater risk of physical and/or sexual violence from the father because of the significant power imbalance. In A Local Authority – and – (1) M (2) F (3) A, B, C, D, E, F, G (via their Children’s Guardian), the Children’s Guardian considered the mother to be at risk of domestic violence in light of her cognitive difficulties and believed that she may not have been able to adequately protect her children from harm in light of these risks. Meanwhile, two mothers with particularly severe disabilities also had difficulty performing childcare duties; for example, in YLA v PM and another [2013] EWHC 4020 (COP) the mother was unable to prepare feeding bottles or bathe her infant without significant support. In one case, this resulted in notifications for neglect (Re C (Female Genital Mutilation and Forced Marriage: Fact Finding) [2019] EWHC 3449 (Fam).

While people with disabilities have rights to privacy and to family life, they may be particularly vulnerable to forced marriage as a result of communication difficulties and a perceived lack of economic opportunities combined with the need to secure carers. Societal perceptions of people with disabilities can negatively affect how their witness testimonies are received in court, impacting their ability to convey their full needs for redress and protection.

Assessment of credibility regarding evidential inconsistencies

Two interrelated themes that emerged from the study concerned evidence and witness credibility. The kinds of evidence provided, and perceptions of that evidence, strongly influenced the outcomes of all the FMPO cases examined. In court, in order to seem reliable and coherent, narratives are expected to be consistent and linear. In this study, witnesses whose testimony was arguably unclear, fragmented and/or changeable were characterised as unreliable or untruthful in five cases. A more specific breakdown of these cases is detailed in the analyses below. In three of the cases explored in this study, the judges described four female witnesses as angry or hostile, even when anger was a reasonable response (i.e. in situations of neglect or abuse by a relative). In Tower Hamlets London Borough Council v BB and others [2011] EWHC 2853 (Fam), the aunt of a woman (‘BB’) with learning disabilities, schizophrenia and an acquired brain injury had removed the woman from her living situation with her husband, who was supposed to be acting as her carer but had been neglecting her needs. The judge found ‘it is undoubtedly the case that HT [the aunt] harbours strong feelings about the conduct of both AB [the father] and MS [the husband], and is a strong character who is not slow to express her views and feelings’. He concluded that ‘her [BB’s] wishes and feelings about her Islamic marriage to MS have been substantially influenced by others, in particular by HT [the aunt]’ and therefore he could not make the finding that the marriage had been conducted without consent. However, female witnesses may be damned if they do show emotion and damned if they don’t given that Kaufman et al. (2003) also found that unemotional female witnesses were regularly seen as lacking credibility.

In cases involving physical violence by family in the run-up to the forced marriage or by the husband following the marriage (as 19 of the judgements in this study did), the trauma that some witnesses had experienced may have affected how they recollected particular events. For example, trauma can fragment memory, and fragmented recollections are not conducive to forming the type of clear, linear narratives that courts generally require. Crespo and Fernández-Lansac (2015) found that trauma memories were ‘dominated by sensorial/perceptual and emotional details’ (p.149) such that coherence, organisation and fragmentation of traumatic memories is common. Indeed, fragmentation of memories in rape victims with PTSD has been linked to dissociation at the time of the assault (Herlihy and Turner 2015), while Salmond et al. (2011)found that children and adolescents presenting at an emergency department who were subsequently diagnosed with acute stress disorder reported more trauma symptoms and presented more disorganised narratives. Research also shows that shame and posttraumatic avoidance affected both victims’ accounts and the ways in which disclosures were made (Herlihy and Turner 2015).

A number of studies have shown that ‘impaired voluntary recall of the event’ (Crespo and Fernández-Lansac 2015, p.154) is central to PTSD, while another key diagnostic criterion concerns avoidance symptoms, which have been associated with shorter accounts of the traumatic event in multiple studies (Crespo and Fernández-Lansac 2015). Short narratives can be viewed as evasive by judges, who described more detailed narratives as more credible or authentic in three cases. This accords with research on rape testimonies in courts in Norway, which found that words such as ‘detailed’, ‘nuanced’ and ‘neutral’ were used to describe testimonies that judges perceived as credible (Laugerud 2020).

Factors such as gender stereotypes and perceptions of people with disabilities also influence the perceived credibility of evidence and witnesses. For instance, in their analysis of the perceived credibility of asylum seekers and rape victims, Herlihy and Turner (2015) argued that rape myths impact all stages of the victim/survivor’s interaction with the judicial system as regards the way they and their claims are viewed. Perceptions of credibility are based on a combination of testifiers’ perceived trustworthiness and the plausibility of their accounts (Tuerkheimer 2017; Jones 2002/2019, Epstein and Goodman 2019). Tuerkheimer draws on the work of philosopher Karen Jones, who argues that ‘Testifiers who belong to “suspect” social groups, and who are bearers of strange tales can thus suffer a double disadvantage. They risk being doubly deauthorized as knowers on account of who they are and what they claim to know’ (Jones in Tuerkheimer 2017, p. 14). Credibility also comes into play in police decisions to pursue a case (O’Neal 2019). Claimants are seen as more credible when they tell consistent narratives in different situations (e.g. in police reports and in court), especially when disclosure is prompt (Herlihy and Turner 2015).

Determining which accounts were most reliable was a complex matter for the judges in all the cases examined in this study as witnesses and respondents often gave conflicting accounts. As a result, many, but not all, judges provided a Lucas direction on reasons for lying; this direction acknowledges that witnesses might lie for many reasons and that one lie or inconsistency does not mean that all the person’s claims are false. In one judgement (Neutral Citation Number: [2018] EWHC 3295 (Fam), A Local Authority – and – (1) M (2) F (3) A, B, C, D, E, F, G (via their Children’s Guardian)), a trauma expert was quoted by the judge, which helped strengthen the idea that inconsistencies or confusion in evidence are not necessarily due to deliberate deception. In this particular judgement, Judge Parker of the High Court’s Family Division stated that

the expert said that the concept of a lie is an interesting one. Different narratives at different times can be caused by children’s anxiety. One should think about how memories are formed. When we remember, it is rarely perfect. We remember snapshots of a memory, and put bits in the middle to create the recollection. Traumatic memories are more prone to distortion, the reason for this is that when something difficult happens, our instinct is to block it out … What can also happen if we ask a child to recall memory again and again is that it can become tampered with as a result of retelling. Also, research tells us that if you have a memory, and someone tells you persistently that it is not true, that can make a child question the accuracy of the memory.

This was one of the few instances in the judgements where there was explicit and sympathetic engagement with the impact of trauma on victims’ narratives. More commonly there were issues regarding how the evidence given by young or otherwise vulnerable witnesses was perceived. In some cases, minor inconsistencies meant that witnesses were treated as unreliable. This was particularly an issue in cases where women with disabilities alleged sexual assault. While the judges often considered minor inconsistencies (e.g. issues with counting and numeracy) as arising from severe learning disabilities, minor changes in accounts of sexual assault by the same witnesses in two cases were taken to mean the events had been fabricated. This is concerning given the often fragmented nature of traumatic event recollection. Combined with difficulties in remembering numbers and dates caused by learning disabilities, these issues meant that in five cases, the sexual assault testimonies of people with learning disabilities were discounted, used by the opposing counsel to undermine victim credibility or withdrawn.

In the case of a Thames Valley–based Muslim woman of Kenyan/Somalian background who had both a mental illness and a disability, Re C (Female Genital Mutilation and Forced Marriage: Fact Finding) [2019] EWHC 3449 (Fam), the judge did not accept that the woman had been raped by her husband, or that she had faced more than one incident of domestic violence, because of inconsistencies in the number of times she said she had been raped:

Given the mother’s [i.e., the woman alleging assault] difficulties with understanding numbers, I was less troubled by trivial numerical mistakes in the mother’s accounts as to how many times the couple had sex, but it did seem to me that there was a world of difference between an account claiming that each rape resulted in a pregnancy and one which said that there were more regular rapes. The inconsistency could not be explained away … There were also inconsistencies in the mother’s accounts of what had occurred during the alleged rapes and no account from her in her police interview that she had said to the father on each relevant occasion that she did not consent to sex with him (though she had earlier alleged she had tried to push the father off).

The judge claimed that he

had little doubt that this was not a marriage in which there was mutual sexual satisfaction. Within their sexual relationship, the father had no regard for the physical and psychological consequences of the FGM suffered by the mother and … detected little understanding of the emotional and psychological consequences for the mother of her alleged childhood rape. This careless disregard by the father was however a long way from habitual rape.

The judge found that ‘the father did not rape the mother’ and that her brother ‘did not sexually assault’ her. He expressed this in strong and definitive language. However, in other cases, the judges simply stated that there was not enough evidence, based on the balance of probabilities, to prove that the alleged events occurred. In a context where both perceptions of witness credibility and simplistic narratives – which involve ‘ideal’ or stereotypical victims and perpetrators – shape case outcomes, women with disabilities face particularly serious challenges in providing evidence that the court believes. This may be due to communication difficulties, but trauma-related memory problems of narrative linearity and/or consistency may also play a role, as might general perceptions of people with disabilities.

Breaches of forced marriage protection orders

This final section examines the outcomes of FMPO proceedings in the 33 cases, including the orders made and any breaches of previous orders or any mention of potential breaches of these orders. Most of the cases involved concurrent proceedings and were not solely focussed on forced marriage, particularly in cases concerning minors. In cases where injunctions were made, the only ones with no concurrent proceedings were two cases involving requests for removing passport orders. Care orders, including interim care orders and wardship orders, were the most common orders imposed alongside FMPOs. In nine cases (i.e. almost half of those involving minors) the judges advised the LA to apply for care orders. The children in these cases were often taken into foster care, both because of the risk of forced marriage and because of other factors (e.g. physical abuse and domestic violence). Indeed, wardship orders were made in 12 cases involving minors, allowing the state to more easily remove minors from the care of parents who had previously taken them overseas for marriage purposes. In 12 other cases, the families were already known to social services before being subject to FMPOs: three of these were from Bangladeshi backgrounds, three from Pakistani backgrounds, one from an Afghani background, one from an unspecified ‘North African’ background, one from a Somalian background, one from a Kenyan and Somalian background, one from a British and Algerian background, and one from an unspecified ethnic background.

FMPO orders relating to mobility featured in 18 cases and included passport orders, port alerts or other mobility restrictions, such as a requirement that the person protected by the FMPO not leave the country. Another condition attached in nine cases was for the parents to either return the children in question to the UK, to not remove them from the UK in the first place or to endeavour not to take them to a named country. In the five Court of Protection cases, orders for capacity assessments and capacity declarations were the most common concurrent orders. Orders to institute marriage nullity proceedings or non-recognition declarations due to lack of capacity took place in two Court of Protection cases and two Family Division cases.

Exploring how breaches of FMPOs took place, and under what circumstances, facilitates a greater understanding of the effectiveness of FMPOs. In this study, seven potential breaches of FMPOs and one suspected breach of an FMPO were discussed in the judgements, and these breaches tended to occur in similar ways. Knowledge of these patterns can shape suitable changes that would strengthen existing legislation and procedural processes. Seven potential breaches of FMPOs and one suspected breach of an FMPO were discussed in the judgements. In three cases, breaches occurred when the family was living overseas or after the children had been removed from the UK, as once the people who had breached the FMPO had left the jurisdiction, they were less likely to be apprehended or subject to enforceable sanctions. The judges in these cases gave the respondents (most often the parents) opportunities (usually multiple opportunities) to respond to FMPOs. In two cases, parents were subsequently charged with contempt of court ([2011] EWCA Civ 555 Lydia Erhire – and – E O – I and Bedfordshire Police Constabulary v RU and another; [2013] EWHC 2350 (Fam)). In two other cases, parents were charged with breaches (Brighton and Hove City Council – and – The Chief Constable of Sussex – and MQ – and- FQ – and – CQ, DQ AND EQ [2018] EWHC 3979 (Fam)). However, in the case of Re: K, the claimant subsequently retracted her allegations, and prosecution was thus not pursued (Re K (forced marriage: passport order) 2020 EWCA Civ 190). In another case, the judge maintained that the parents were at risk of arrest for breaching an FMPO as a result of not returning their children to the UK by a stated deadline (West Sussex County Council and another v F and others [2018] EWHC 1702 (Fam)). In a further case, the father was in breach of an order associated with the FMPO to make his daughter available at the British Consulate in Jeddah (Amina Al- Jeffery and Mohammed Al-Jeffery [2016] EWHC 2151 (Fam)). When Amina – who was 21 at the time of the proceedings – was 16, she was told by her father to come to Saudi Arabia, where she alleged that she was physically abused and held captive. One police officer, whose evidence was discussed in the proceedings, stated that she believed people could circumvent port alerts issued as part of FMPOs. The judge responded that he had never seen any proof of this. However, within the sample of judgements, parents whose children were subject to care orders or FMPOs attempted to leave the country. Some who had dual nationality were able to travel on other passports and/or apply for passports from their country of origin for their children. In six cases where FMPOs were issued when parents were overseas, the parents seemed less concerned about breaches. This may have been because they could not easily be made subject to UK law when residing in another jurisdiction.

In the case of 1. A Chief Constable 2. AA and 1. YK 2. RB 3. ZS 4. SI 5. AK 6. MH Case No: LU09F03718, a marriage was contracted despite the court’s orders:

Despite the orders of the court, A [the person requiring protection under the FMPO] went through a form of marriage ceremony at the beginning of October 2009, but that marriage has not been formally registered, and it cannot be by virtue of prohibitive order of this court, which remain [sic] in force.

Religious marriage ceremonies, such as the one in this case, may be the most significant part of the marriage for some couples (Uddin 2018), as it helps cement links to culture, tradition and spirituality (Mustasaari and Vora 2020). Vora (2020) argues that possible motivations for nikaah-only marriages in the UK include male partners wanting to protect their assets in case of future marriage dissolution and/or young people wanting a ‘trial’ marriage that will still be viewed as valid within their religion and community – however, a minority may not realise that the nikaah-only marriage is not legally valid in the UK. Bone (2020) points out that a nikaah ceremony is not analogous to a civil marriage in the UK, as the practices developed from different traditions and hold different meanings, with the nikaah having both civil and religious implications. In many countries, religious and civil ceremonies are both accepted as valid (Uddin 2018): this is not the case in the UK. Thus, in the UK, multiple studies of Sharia courts have found that 23–64% of Muslim marriages are not legally registered (Uddin 2018). The same is true in parts of Northern Africa, where only some marriages are legally registered after the religious ceremony has taken place. This is because a marriage that is not legal may still be seen as formalised by the community once a religious ceremony has been conducted. Therefore, the judge in the above case viewed the religious ceremony as a serious undertaking that contravened the original FMPO, even though it was not a legal marriage in terms of UK law.

The case of Lydia Erhire – and – E O-I (by his next friend) was heard in the Court of Appeal, Civil Division, in 2011. The mother had been sentenced to eight months’ imprisonment for contempt of court regarding the breach of a wardship order. The son, who was 17 years old and in Nigeria after being taken there by his mother, was afraid that he had been taken there to be married and so applied for an FMPO, although he decided to delay serving it to his mother. She was ordered by the court to return her son from Nigeria to the UK; subsequent to this order, she told the court she was complying. However, she gave conflicting instructions to the school and to the boy’s aunt in Nigeria, telling them to ignore the court’s orders to take the boy to be put on a flight back to the UK. The court gave her multiple opportunities to return her son before deciding she was in breach of the order. The appeal court found that the sentence was not ‘disproportionate or excessive’ and the appeal was dismissed.

This body of evidence demonstrates that concerns about parents flouting FMPOs are not misguided. In six cases, evidence was provided to the court that parents had breached FMPOs and associated orders; moreover, breaches of co-existing orders, such as care orders, also took place on three occasions. It is worth considering the ways parents breached or attempted to circumvent orders: by ignoring FMPOs when overseas (Lydia Erhire – and – E O-I (by his next friend) and Brighton and Hove City Council – and – The Chief Constable of Sussex – and -MQ-and- FQ-and-CQ, DQ AND EQ), by taking children out of the country on other children’s passports (Re A (A Child) (Inherent Jurisdiction: Parens Patriae, FMPO and Passport Orders)), by applying for passports from another country where they were eligible for dual citizenship, by applying to the court to discharge passport orders to go abroad (West Sussex County Council and another v F and others [2018] EWHC 1702 (Fam)), and by claiming the ill health of grandparents (Bedfordshire Police Constabulary v RU and another) or a celebration that required taking the person protected under the FMPO abroad for family reasons.

Conclusion

This is the first research to examine judgements relating to FMPOs. As these orders were introduced relatively recently, it is important to understand how they function in practice. The study used a qualitative socio-legal approach to understand the application and interpretation of the law and the broader socio-political context that shapes this process. Based on an analysis of the reasoning behind 33 FMPO-related judgements, this paper contributes to understanding how framing narratives shaped the results of many cases.

In practice, FMPOs were imposed in many cases, especially to facilitate the return of (potential) victims of forced marriage from overseas and to prevent forced marriages from taking place in the UK. However, in other cases this form of protection was not accessible to applicants because of difficulties in securing the evidence required to prove that an FMPO was necessary. This is a particular issue in the civil context, as police may play a greater role in evidence gathering in criminal cases. Although it is easier to provide acceptable proof when subject to the probability standard (civil) rather than the criminal standard (beyond reasonable doubt), in many cases the judges set what seemed to be an unnecessarily high bar for proof. The term ‘probability’ is fraught in domestic and sexual violence cases, as prevalent socio-cultural myths and stereotypes about gendered violence, consent/coercion, learning disabilities, physical disabilities and mental illnesses mean that victims are often not believed by either the public or the courts.

In the cases examined, FMPO proceedings often formed part of care proceedings in the Family Court. Domestic and sexual violence, and physical abuse of children, were commonly raised in the proceedings, but issues of inconsistency and apparent incoherence often led to these claims being dismissed or largely disregarded. Narratives about severe violence, and multiple forms of co-occurring violence, were often disbelieved on the basis that they were seen as unlikely, even though people who experience violence have been shown to generally experience multiple types. Both disability and gender affect perceptions of victim credibility, while trauma and disability can also shape the process of narrative construction when giving evidence in ways that do not match legal expectations of full, linear and consistent accounts.

Meanwhile, legal constructions of coercion in FMPO proceedings still rely on a passive understanding whereby a lack of resistance implies consent. This places the onus on vulnerable victims to provide proof of coercion instead of on perpetrators to provide proof of consent, again bringing in issues of standards of proof in civil versus criminal proceedings. The concept of ‘affirmative’ or ‘active’ consent in relation to sexual activity could be usefully applied to forced marriage, shifting the focus from the victim’s resistance to marriage to the actions and words of the alleged perpetrators, demonstrating whether agreement to the marriage was sought and/or whether it could be obtained if the victim arguably lacked capacity to consent and/or understand any elements of what was involved.

Dominant constructions of Western versus ‘Othered’ cultures were another complicating factor that influenced the application of the law. Lawyers representing respondents sought to position forced marriage as normal within the claimants’ cultures, while many judgements represented forced marriage as ‘backward’, ‘traditional’ or ‘cultural’ and positioned UK culture as liberal, modern and ‘neutral’ in contrast.

The study revealed a need to address the fact that orders were breached in a substantial minority of cases. As breaches tended to follow a handful of clear patterns – for example, parents ignoring FMPOs when abroad, using other passports or dual passports to take their children abroad, or travelling on the pretext of a family health emergency – procedural and legislative developments should be considered to address these breaches and ensure that all people can benefit from equal access to protection under the law. In particular, victims of forced marriage with disabilities experience significant barriers to justice, including inaccessible legal procedures, stereotypes that exclude or discount their testimony, problems accessing legal representation and protection, assumptions that they lack credibility, lack of accessible information and processes, and many more. Access to justice and equal recognition before the law are thus essential to preserving and advancing the rights of people with disabilities, and special measures must be taken to support these individuals and enable them to voice their experiences in ways that are more likely to be viewed as credible. These measures include further training for judges and legal personnel on trauma and disability; training on the impact of trauma and disability on testimonies and witness behaviour; training that dismantles prevalent socio-cultural myths and stereotypes about the role of culture and tradition in specific forms of violence and abuse; and a review of the model of consent versus coercion to explore whether ‘affirmative consent’ would be a more suitable standard in civil courts given that the burden of proof is lower in these proceedings.

Indeed, greater support should be provided so that all witnesses understand the kinds of evidence required in court and what sort of information will increase their perceived credibility (e.g. giving prompt reports to the police and identifying the specific dates and times of events). Meanwhile, judges must develop a stronger understanding of the range of coercive pressures in a family context that often result in conflicting testimonies and retractions of previous statements. Taken together, these measures can ensure that FMPOs are effective legal tools that protect all victims.

Acknowledgments

Special thanks to Professor Emma Hitchings and Dr Mavis Maclean for constructive comments and insightful observations on earlier drafts. The original research reported in this paper was supported by the Nuffield Foundation [grant number JUS/43810].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Nuffield Foundation [JUS/43810].