Domestic violence (DV) encompasses acts of physical, sexual, psychological, and/or economic violence that occur within the family, in the domestic space, or between current or former spouses or partners, regardless of current or past cohabitation between the perpetrator and the victim (Council of Europe [CE], 2011; Sani & Ferreira, 2023). Although this type of violence is not limited to a specific gender, it disproportionately affects women (Comissão para a Cidadania e Igualdade de Género [CIG] & Direção Geral de Educação [DGE], 2007; García-Moreno et al., 2005). The World Health Organization [WHO] (2014) estimates that one in three women worldwide will become victims of physical or sexual violence committed by an intimate partner at some point in her life. DV is a problem of epidemic prevalence, and the consequences for the victims can be injury, death, psychological harm, developmental disability or deprivation (Krug et al., 2002). When minors are exposed to this type of violence, they are deprived of their right to a safe and healthy family environment. DV is therefore a pressing problem (Steketee et al., 2019) with adverse physical, emotional, cognitive, and behavioral consequences for children (Artz et al., 2014).

Domestic violence is a prevalent crime in Portugal (Sani et al., 2020) and is viewed as a major social issue requiring government intervention (Centro de Estudos Judiciários [CEJ] & Comissão para a Cidadania e Igualdade de Género [CIG], 2020). As it is considered an urgent crime, law enforcement agencies are required to taking prompt action within 72 h upon receiving a report of DV (Secretaria de Estado para a Cidadania e a Igualdade, 2020). The purpose of this action is to protect and support the victim, gather and preserve critical evidence, assess and establish the legal situation of the perpetrator, and ensure comprehensive and integrated judicial and social intervention. One of the established procedures is a risk assessment conducted by the police, as mandated by Law No. 112/2009, using a specific assessment tool created for this purpose (Sani & Lopes, 2018; Sani et al., 2020).

In the context of child custody proceedings, there has been growing awareness of DV committed against children in recent years (Centro de Estudos Judiciários [CEJ], 2021). Indeed, the Council of Europe Convention on Preventing and Combating Violence against Women and Domestic Violence, which is better known as the Istanbul Convention (CE, 2011), as well as several studies (e.g., Azevedo & Sani, 2017; Gonçalves & Sani, 2015; Hayes, 2012; Sani & Carvalho, 2018; Saunders et al., 2016), have warned about the predicament of children who are exposed to this form of violence and urged European countries to take steps to protect and assist them. In Portugal, there have been significant legislative advances aimed at protecting victims of DV over the last two decades. However, it was not until 2021 that a comprehensive law was enacted (Law No. 57/2021) that explicitly recognizes children and young people up to age 18 who have been exposed to DV as victims. This new law acknowledges that these individuals should be granted victim status, a designation previously applicable only to adult victims who filed a formal DV complaint under Law 112/2009. Nevertheless, the visibility of children as victims of DV is a relatively recent phenomenon, and efforts to protect their rights remain insufficient (e.g., Tomás et al., 2018).

As the sharing of custody between a victim and perpetrator puts the safety of the victim and the children at risk, the magistrates of family courts must carefully consider any allegations of violence in the family (CEJ & CIG, 2020; Davis et al., 2011; Miller & Manzer, 2021). In 2017, through Law no. 24, Portugal amended the Civil Code to promote urgent regulation of parental responsibilities in situations of DV and other forms of violence in a family context. This legislative change acknowledges that a common regime of parental responsibility may be contrary to children's best interests. As such, following an assessment of the risk of violence to victims, coercive measures may be imposed, including the restriction of contact between parents and children. These measures must be communicated immediately to the public prosecutor's office for the urgent commencement of the respective process of regulation or modification of parental responsibilities. Nevertheless, according to an evaluation by the monitoring mechanism of the Istanbul Convention, despite the growing investment in measures to combat gender violence and protect victims, situations involving DV are still not given due weight in child custody proceedings in Portugal, with inadequate communication between criminal and family courts (Group of Experts on Action Against Violence Against Women and Domestic Violence [GREVIO], 2019).

Against this backdrop, and because there is little research on this subject in Portugal, this study aims to analyze how judicial decisions are rendered in child custody proceedings featuring allegations of DV between parents. We seek to investigate whether these decisions comply with the directives of the Istanbul Convention and understand the legal and extralegal factors driving them. Several field studies have shed light on extralegal factors, such as sociodemographic characteristics and personal constructs, that impact the decisions of judges (Kafka et al., 2019; Koublitskaia, 2012; Rachlinski & Wistrich, 2017). We begin with a brief legal contextualization of the matter being studied and a review of prior studies that guide our empirical research study and justify its relevance.

Legal Framework for Child Custody Determinations

The Portuguese term for a child custody determination is translated as “parental responsibilities,” and it is defined in Article 1888 of the Portuguese Civil Code (PCC). This article states that, in the interests of their children, parents have a duty to ensure their health and safety, provide for their livelihood, direct their education, represent them, and manage their assets. Moreover, Subsection IV, Article 1906—which addresses parental responsibilities in the case of divorce, the legal separation of people and property, and a declaration of nullity or annulment of a marriage—decrees that parental responsibilities must be shared, especially for issues of particular importance to the child’s life. Thus, it is in the child’s best interests for parents or caregivers to collaborate on important issues in the child’s life and promote a close relationship between the child and the other parent. However, Paragraph 2 provides the scope for the court to decide that shared custody may contravene the child’s interests and therefore award sole custody to one of the parents. Thus, Article 1906 acknowledges that sole custody may better suit the interests of the minor when there is a history of DV. Thus, Subparagraph a) defines the circumstances for enacting a constraint measure or supplementary penalty that prohibits contact between parents, and Subparagraph b) presents the criteria for when the rights and safety of victims are considered to be at high risk, such as in cases of child maltreatment or sexual abuse.

DV perpetrated within the context of intimate relationships can take various forms of abusive behavior, such as physical, sexual, emotional, and economic abuse, often perpetrated by men against their intimate female partners with the intention of exerting power and control over them (Anderson, 2013; Jaffe et al., 2003; Sani et al., 2016). There is also a high risk that such offenders will use children as tools to control and coerce their intimate partners (Hayes, 2012; Heward-Belle et al., 2018; Jaffe et al., 2009; Laing, 2016).

A couple’s separation may trigger an escalation in DV, thus amplifying the risk of threatening behavior, emotional abuse, use of physical force, stalking, and, in the most serious cases, femicide (Elizabeth, 2017; Gennari et al., 2018; Hotton, 2001; Jaffe et al., 2003, 2009). Thus, a child’s exposure to DV may not end when the parents separate or divorce. Logan et al. (2006) demonstrated that offenders who have children in common with victims violate victims’ protective orders more often than offenders without children do, placing these women and children at greater risk of revictimization.

Shared custody between a perpetrator and a victim also allows the former to maintain contact that facilitates perpetuation and even escalation of violence (Miller & Manzer, 2021; Saunders & Oglesby, 2016; Walker et al., 2004). Child custody decisions that allow visitation by the perpetrator can also provide opportunities to perpetuate the DV, increasing the risk of child abuse (Brown, 2019; Hardesty & Ganong, 2006; Ignjatović, 2019; Orr et al., 2023).

The literature has suggested that court decisions in child custody cases have not attributed sufficient value to claims of DV, with magistrates tending to opt for shared custody (Davis et al., 2011; Meier et al., 2019; Saunders et al., 2011). Disregarding the risks of shared custody in cases of DV, for the victim or the child, goes against the directives of the Istanbul Convention and the robust literature on the dynamics of DV. This is especially true when we consider the impacts of a child’s exposure to DV, the frequency with which DV between parents cooccurs with direct child abuse, and the risk of perpetuation and escalation of the DV after separation (e.g., James-Hanman & Holt, 2021; Meier et al., 2019; Pranzo, 2013; Walker, 2020). On the other hand, when professionals understand the dynamics of this type of violence, they are less likely to disregard DV allegations and consequently recommend joint custody (Saunders et al., 2016).

In a qualitative study by Kafka et al. (2019) in North Carolina, 20 judges were invited to share judicial anecdotes involving protective measures in DV cases for an analysis of how biases and perceptions shape decisions in these cases. The study revealed that judges often rely on personal constructs to classify situations as genuine or false cases of DV, which can influence the application of protective measures. In various social spheres, including in the domain of judicial decisions, the stereotype of the "ideal victim" noted by Nils Christie (1986) challenges us to reflect on the extent to which judicial decisions are also an expression of social attitudes and values. The personal constructs of individuals—in this study, magistrates—can shape an idealized victim status, which, fairly or unfairly, is determined based on victims’ characteristics, behavior, and/or relationships with offenders (Duggan, 2018).

The literature suggests that magistrates award shared custody in cases involving DV because they tend to doubt or minimize the severity of the violence. For example, they may believe that DV allegations are false or that the child will benefit from a continued relationship with the perpetrator (Davis et al., 2011; Dore, 2004; Hardesty et al., 2015; Miller & Manzer, 2021; Walker, 2020). Saunders et al. (2011) conducted a study in the United States (U.S.) in which questionnaires were distributed to professionals working in custody processes—including 200 magistrates—and reported that disbelief of mothers’ DV claims is related to several false beliefs: i) that mothers making such claims are seeking to alienate their children from the other parent, ii) that DV is not relevant to custody decisions, iii) that children suffer from victims’ reluctance to share custody, and iv) that victims make false claims about violence against their children.

In an analysis of 1,137 court cases in the U.S. in which the mother alleged DV and the father did not allege parental alienation, Meier et al. (2019) reported that the courts accepted the DV allegations in only 45% of cases. Moreover, in cases where the mother claimed DV and the father claimed parental alienation, this percentage decreased to 37%. The study also revealed that in cases where the court credited the DV allegations, some mothers lost custody to the offending parent (29% of the analyzed cases) when the court believed that the mother was trying to alienate her children from their father. Thus, DV victims are essentially trapped, facing negative consequences regardless of their actions (Sani & Pereira, 2020). If they remain in the abusive relationship, they are accused of not protecting their children, whereas leaving the relationship often puts them at greater risk of retaliatory violence. Indeed, when they report DV and are reluctant to share custody with the perpetrator, they risk being accused of fabricating the allegation to further parental alienation (Saunders & Oglesby, 2016). Importantly, although some claims of DV in child custody processes are indeed false, the denial and minimization of DV by perpetrators are much more common (Jaffe et al., 2003). Therefore, a mother’s behavior may be perceived as seeking to manipulate a child to alienate the other parent when her intention is in fact to protect the child (Kirchesch & Sani, 2023; Sani, 2006). In addition to the violence that mothers are subjected to, they face a justice system that prioritizes family mediation and shared custody (Laing, 2016) and justice professionals who tend to doubt their allegations (Bow & Boxer, 2003; Forssell & Cater, 2015; Hardesty et al., 2015; Macdonald, 2016; Sani, 2006). The literature therefore suggests that there is a discrepancy between the empirical evidence on the dynamics of DV and the outcomes of child custody cases, which reveals the skepticism of family courts toward allegations of DV, thus contributing to its perpetuation (Koshan, 2023; Walker, 2020; Wiegers, 2023). On the other hand, in a study carried out in six U.S. states (Delaware, Florida, Kentucky, Massachusetts, Minnesota, and Rhode Island), Morrill et al. (2005) demonstrated that receiving training on DV increased magistrates’ likelihood of awarding exclusive custody to victims. Another study by Ogolsky et al. (2023) revealed that sole custody was more prevalent when instances of DV were officially recorded and involved third-party intervention. Collaboration between various service providers (government or private) in such cases proved advantageous in ensuring the child's safety and well-being.

Among the purposes of the Istanbul Convention (CE, 2011) is the need for "law enforcement agencies to effectively co-operate in order to adopt an integrated approach to eliminating violence against women and domestic violence" (see article 1, paragraph e). To this end, professionals who work with victims or perpetrators of violence against women and domestic violence must have specialized knowledge on the prevention and detection of such violence, equality between women and men, the needs and rights of victims, and how to prevent secondary victimization (see article 15, paragraph 1). Law enforcement agencies should also encourage training in coordinated interinstitutional cooperation to enable comprehensive and appropriate management of referrals of cases involving violence falling within the scope of the Istanbul Convention (see Art. 15, paragraph 2).

The literature suggests that although the harmful effects of exposure to DV are well documented and the Istanbul Convention recognizes that shared custody may not be in children’s best interests, magistrates continue to opt for arrangements with visitation rights or shared custody. The real reasons for this phenomenon are unknown, but it has a significant effect on the lives of children, adolescents, and their families, even potentially deviating from international guidelines. To serve all citizens, it is important to investigate and gather credible evidence on the arguments that justify these judicial decisions to determine the best course of action in these cases.

Given the scarcity of national studies analyzing judicial decisions in child custody cases with alleged DV between parents, the following research question arises: To what extent do judicial decisions in child custody cases with alleged DV between parents comply with the directives of the Istanbul Convention?

We adopted a qualitative methodology to answer the above question because it would help us understand each participant’s perspective and attribute some meaning to the problem being investigated (Creswell, 2018). Specifically, we sought to: a) characterize child custody proceedings involving allegations of DV; b) determine the legal and extralegal factors that underlie court decisions; and c) understand how the parties perceive the potential impacts of sharing custody in cases of DV.

Method

The present qualitative study, which was descriptive in nature and transversal in character, was based on interviews with a sample of Portuguese judges who were authorized to participate by the Superior Council of the Judiciary. Below, we outline the method of the research, which was that recommended by Goldberg and Allen (2015).

Participants

The sample was constructed via a nonprobabilistic snowball sampling process (Browne, 2005), given its specificity and relative inaccessibility, subject to the inclusion criterion of being a judge in a Portuguese family court. Six Portuguese family court judges, with average ages between 39 and 49 years, participated in the study (see Table 1). The judges have a solid academic background in law (e.g., divorce, child custody, adoption) and experience in adjudicating family-related cases and associated complex issues, such as DV. Table 1 presents additional sociodemographic data on the participants (identified as P1 to P6), such as the geographical region of the court where they work, their years of experience and any previous training they may have in victimology. The sample was delimited according to a process of theoretical saturation (Fontanella et al., 2008) and was closed when there was a redundancy of data in relation to the testimonies already collected.

Table 1 Socio-demographic characteristics of the sample
Full size table

Instruments

Research data were collected through the use of a sociodemographic form to characterize the participants by age, gender, time served as a magistrate in family court, and training in the field of victimology. A semistructured interview guide was also prepared, comprising 17 (open and closed) questions on various topics. These included professional experience (e.g., “I would like you to tell me a little about your experience in the field of child custody, particularly when DV is alleged”), the characteristics of child custody cases with allegations of DV between parents (e.g., “In the event of a criminal case of DV between the parents of a child, is this information transmitted to the child custody case for that child?”; “In such cases, what procedures are generally adopted?”), the factors that underlie judicial decisions in such cases (e.g., “In cases with alleged DV, do you tend to decide for or against visitation rights for the alleged perpetrator? What factors do you consider?”), and perceptions about the potential impacts of shared custody when there is DV between the parents (e.g., “To what extent does shared custody in cases of DV affect the child?”).

Procedures

The study was approved by the Ethics Committee of Fernando Pessoa University on May 9, 2022 (ID: FCHS/PJU 289/22–2). First, a sociodemographic form, interview guide, and informed consent form were prepared on the basis of a literature review of the subject. Next, a letter was sent to the President of the Superior Magistrate’s Court for authorization and dissemination to the various family courts around the country. This letter was emailed to the Support Office for the Vice President and Members of the Superior Magistrate’s Court together with the interview guide and informed consent form. Three magistrates volunteered to participate by responding to the Support Office, which then informed the researchers by email. The researchers then contacted the participants and arranged interviews, two of which were in person at the respective courts and one through Skype. The remaining three participants contacted the researchers directly by email after the original participants informed them of the study. The last three interviews with these participants were conducted through Skype. The face-to-face interviews took place on the premises of the respective courts, and we coordinated the online interviews by sending the Skype session link to the e-mail address provided by the participants.

The interviews, carried out by the first author of this article, lasted 45 min on average and were audio-recorded, transcribed, and printed for further analysis. To analyze the transcript, we used Braun and Clarke’s (2006) thematic analysis method, which entails approaching the data inductively, describing patterns or themes present in them, and accessing the participants’ meanings and experiences through these data (e.g., Brooks et al., 2015). In this study, the data analysis process took place in six phases (Braun & Clark, 2006): First, we familiarized ourselves with the data, listening to, reading, and rereading the interviews to gain a deep understanding of them; in the second phase, we coded the data, grouping them according to similar patterns of meaning; in the third, we generated themes through grouping codes and formulating a scheme with key patterns present in the data; in the fourth phase, we reviewed the generated themes to determine whether they fit within the coded data and were sufficiently distinct from the codes; if not, we would return to the previous phase; in the fifth phase, we defined the themes, summarizing each one and assigning a representative designation of its meaning that allowed for the development of the final scheme; in the sixth and final phase, we interpreted the narrative and wrote the results and conclusions obtained from the analysis of the thematic scheme.

To ensure the validity of the results according to the intended objectives and the type of research being conducted (Levitt et al., 2017), we carried out investigator triangulation through analysis by two researchers, who debated the categories until they agreed upon the final composition.

Results

Table 2 summarizes the themes that emerged from the analysis of the interviews with family court magistrates.

Table 2 Result of the thematic analysis of the interviews
Full size table

To support the results, some excerpts from the interviews are presented below, with the participants identified by the letter P followed by their identification number.

Child Custody Process Involving an Allegation of DV Between Parents

This theme is presented through two subthemes, which refer to the (in)visibility of the cases to the judges [A1. perceived prevalence] and the exchange of information between different courts [A2. communication between courts] when judges must decide on child custody in cases of allegations of domestic violence].

Although there are no official statistics on the perceived prevalence [A1] of allegations of DV between parents in child custody cases, most of the magistrates considered it to be a frequent phenomenon [P2: “It is not the rule … but there are cases frequently. I cannot give you a number.” P6: “I do not know the statistics either, so everything I say in this regard will be fallible, but it is quite frequent. Let’s say 15% to 20% of situations.”]. All the participants noted that those alleging DV in such processes are almost always women [P3: “Most of the time, those who allege domestic violence are women, without a doubt, without a doubt, without a doubt.” P5: “I just remember one situation a few years ago in which the victim was the male partner. Usually the victim is female.”].

Regarding communication between the courts [A2], the participants stated that their decisions in child custody cases with allegations of DV between parents are influenced by delays in the criminal processes related to the DV and poor communication between the criminal and family courts. Consequently, they must often make custody decisions before the criminal process is concluded and therefore lack sufficient information about the DV [P3: “It often happens that the sentence comes very late in relation to the moment when we have to determine custody.”]. A magistrate said:

P4: Because now one thing that is happening a lot … is that the criminal process is running, and I have the custody process to run, but the criminal process will take much longer than my process. Therefore, I will decide one thing long before the crime is decided. I run the risk of… the aggressor, who is presumed innocent of course, will be acquitted, and I will have suspended him from visits for three years.

Determining the Legal and Extralegal Factors Driving Judicial Decisions

The judges' decision-making involves several factors [B1. risk assessment; B2. agreement between parents; B3. false allegations; B4. children's opinions; and B5. Istanbul Convention] with different focuses and implications.

A risk assessment [B1] of violence perpetrated against women and their children’s consequent exposure to this violence is one of the factors underlying the magistrates’ decisions. Thus, when the assessment reveals high risk, magistrates tend to prefer arrangements with supervised visits or third-party intermediaries to prevent direct contact between the victim and the perpetrator. Nevertheless, the analysis of the interviews revealed that it is rare for them to decide on an arrangement without any visits by the offending parent [P2: “There are parents who, despite the conviction, are at a high risk of committing violent acts again, so we choose more for supervised visits to see if they will be maintained or not.”]. Another magistrate put it this way:

P1: That is, it if I feel that there may be some risk there... I really try to make the visits happen with the presence of another person, so in the moments of contact, they are with another person and... if it is a situation in which I realize that the child does not have a great bond with the father, well, it’s a shorter visit, an afternoon or morning period, until I have more data, so later we can move on to a weekend arrangement or even, eventually, to an alternate residence.

All the participants stated that they always try to establish agreement between the parents [B2] in a custody process, even when there is an allegation of DV. Additionally, they noted that mothers rarely question contact by offending parents, and as a result, there is an agreement about the sharing of custody in most cases [P2: “The mother, who evokes, never questions the child’s contact with the father … and there is rarely an imposition by the court against the mother’s will. From my experience, I would say it is almost impossible.” P3: “In conferences, the objective is not to appreciate evidence, and it is not to see who is right—it is to hear one and the other and seek agreement, try to get them to agree”].

The quest for agreement arises from the notion that children benefit from contact with both parents, and the participants said they assess the extent to which one of the parents facilitates the child’s contact with the other as a factor when deciding custody, even when DV has been alleged [P3: “See to what extent one parent facilitates contact with the other or whether one parent does not facilitate and even hinders contact with the other parent”; P5: “One criterion for granting custody is to give preference to the parent who facilitates and encourages the child’s contact with the other parent.”].

Although the effort to achieve agreement and the legal criterion of facilitating contact are determining factors in judicial decisions, the interviews revealed that most of the magistrates do not consider parents who expose their children to DV good parents because this in itself is a form of child abuse [P1: “…one may not behave the same way as a father or as a husband, but we cannot dissociate the parent from the husband who hits, and insults, resulting in him also abusing his child.”]. Another magistrate put it this way:

P4: The answer is definitive. I cannot believe that the individual is a good father. An aggressor cannot educate if they cannot educate themselves and cannot be an adequate parent. … cannot convey family values, common sense, wisdom, stability. Hence, when a child is exposed to DV, they will have this as their future family model besides being traumatized for the rest of their life. … They have heard their mother being called all sorts of things, belittled, so even their concept of women and mothers is affected when there is a lack of respect at home. They will not respect others.

None of the participants denied considering the possibility of false allegations [B3] of DV in custody proceedings. According to the interviews, false allegations are made in this context to secure benefits, such as procedural rapidity and exclusive custody for the mother. Furthermore, the silence of victims and support for the child’s contact with the offending parent are perceived as revealing data for the veracity of allegations [P4: "…because there are some women who know how to use the system … the benefit, in the case of children, is to retain the exclusive custody of the child … because they are the silent victims … they do not complain.”]. One magistrate said:

P2: I will say this: In majority terms, usually in the processes I have, it is the mother who evokes and often calls into question the contact. It could be wrong, but when it happens … the experience I have is that generally these claims are untrue. … When there are no residence or contact questions, I would say that the probability that they are true is very high and that they often end up in convictions.

The participants agreed that it is important to consider the children’s opinions [B4] regarding maintaining contact with both parents, as long as the children are mature enough to express their views. Additionally, when children reject one parent, the law allows only limited consideration of their opinions, even when maintaining contact with an alleged perpetrator of a crime may pose a risk to a child [P6: “Children, quite simply, reject the figure of one of the parents, but even so, the law does not give much scope to the court to accommodate the child’s will. This acceptance can be important, and it can be decisive in protecting her well-being”]. A magistrate shared the following:

P1: If the child is over twelve years old, we mandatorily must listen to them. If he is a little younger, let’s say ten, eleven, eight, nine, we can also listen to him and understand what kind of relationship the child has with the father and if it is in the child’s interest to maintain that contact.

Most of the participants revealed that the “Istanbul Convention” [B5] influences their decision-making in custody cases with allegations of DV, albeit indirectly because the Convention triggered related legislative changes [P6: “In Portugal, this resulted in a series of legislative changes, at the level of the criminal procedure law, at the level of the penal code, and at the level of the civil code and the general arrangement of the civil tutelage procedure.”]. One participant expressed the following:

P3: There is a media program, a media agenda for this topic, and therefore I would not say directly but rather indirectly. Therefore, it triggered, in fact, a set of legislative and media actions, which have an influence. Um, it’s in the sense that we are being alerted to the issue of victims of violence, isn't it?

Impact of Shared Custody in the Context of DV

In the approach to the impact of shared custody, two subthemes emerged [C1. escalation of violence; C2. instrumentalization of children], both of which are assumed to be negative.

Most of the judges expressed the belief that in the most serious cases, the escalation of violence [C1] occurs after a couple’s separation, thus representing a dangerous situation for the mother and her children [P6: “As a rule, in more serious cases of DV, the violence reaches its peak after the separation. In fact, the child may be exposed to even more DV after separation.” P3: “There are levels of conflict that increase with separation …while they were living together, there was verbal and physical abuse, and afterwards, once the couple had separated, those escalated to threats and stalking, so he could take control of the woman’s life.”].

The magistrates also reported frequent instrumentalization of children [C2], which is when the offending parent uses the children to obtain information about the mother and perpetuate violence against her [P1: “It is when a person, who is in fact aggressive and is unable to establish an affective bond with their child, when they see the child as an instrument to affect the mother…”; P4: “the child is always an object used by one parent to affect another person”].

Discussion

Our analysis of the interviews revealed that DV claims are common in custody proceedings, as suggested in the literature (e.g., Meier et al., 2019). This highlights the importance of prosecutors working in the criminal justice system and magistrates working in the family courts area to communicate effectively, so that DV allegations receive appropriate consideration in the custody process (CEJ & CIG, 2020; Koshan, 2023; Wiegers, 2023). For example, if one parent has a history of DV, it may be considered a factor in determining custody arrangements and visitation rights. Similarly, if a parent has been found guilty of abusing or neglecting his or her child, that can be a strong factor in custody arrangements. The court prioritizes the safety and well-being of the child, and a history of violence or abuse against a child can weigh heavily against the offending parent in custody proceedings. However, the analysis also revealed that this communication is ineffective, as a custody process is usually concluded long before a criminal prosecution for DV concludes. Thus, during the custody process, information about the DV is scarce, and the parent is presumed innocent. This is a proven pattern in international studies and evaluations (GREVIO 2019; Koshan, 2023; Wiegers, 2023), with serious implications for the lives of many victims (Matos et al., 2019).

Judicial decision-making in custody processes in relation to alleged DV between parents is based on legal and extralegal factors. Judges tend to decide cases involving personal characteristics in a manner that aligns with their own individual traits and personal meaning (Kafka et al., 2019; Rachlinski & Wistrich, 2017). As Saunders (2007) noted, a range of extralegal factors must be considered in decisions on custody, visitation, court coordination, safety planning or court orders to reduce the risk of harm to children and their caregivers.

Our analysis verified that the judges consider risk assessments for DV and, in high-risk situations, tend to prefer arrangements with supervised visits or child access facilitated by a third party to avoid direct contact between the perpetrator and the victim. Nevertheless, as the potential impact of exposure to DV and children’s safety are not considered, perpetrators retain visiting rights rather than temporarily or permanently losing custody (Davis et al., 2011; Saunders et al., 2011). Additionally, our analysis revealed that most cases are decided by the parents mutually agreeing on custody, with the alleged DV victim rarely questioning the child’s continued contact with the alleged perpetrator. This search for agreement on the part of the magistrate relates to another criterion that influences judicial decisions, namely, the extent to which one of the parents facilitates the child’s contact with the other parent. The law posits that it is in the child’s best interests for parents to communicate about important issues in the child’s life and maintain some proximity between them (Article 1906 of the PCC). The analysis of narratives revealed that this criterion supersedes the magistrates’ own perception of parenthood, as they consider exposure to DV a form of child abuse. Therefore, the criterion of maintaining contact between the child and both parents was revealed as the dominant factor, even if continued contact with the father may put the child's safety at risk (Callaghan et al., 2018; Hunter et al., 2018; James-Hanman & Holt, 2021). In addition, when a mother facilitates contact between the child and the alleged offender, it is perceived as an indicator of the veracity of the DV allegations (Meier et al., 2019). In contrast, a mother’s pursuit of sole custody is considered indicative of a false claim (Saunders & Oglesby, 2016).

These findings seemingly indicate that the victim's status largely depends on the idealization constructed by the magistrate about victims' standard behavior. The judges seem to assume that a victim does not create obstacles and remains silent about their suffering so as to fit Nils Cristie's ideal victim status and therefore deserving of it (Cristie, 2018). This construction by the magistrates has serious implications for victims' positioning in the justice system and especially for the justice system's stance toward victims, revealing how subjective perspectives on personal responses to victimization can, outside the law, influence the magistrates' decisions.

The implications can be quite negative for victims and their children, especially if decision-makers have little training in victimology and do not triangulate information with other organizations. In Ogolsky et al. (2023), sole custody was more common when DV was documented and involved third-party intervention, and this intercommunication between institutions was an asset in guaranteeing children’s protection. For this reason, it is crucial for family court magistrates to undergo proper training on DV. Depending on the dynamics and types of violence, certain allegations that arise in these cases may or may not be corroborated or refuted by cross-referencing information provided by various institutional parties. Importantly, there is great experiential variability in victimization and contact with justice; that is, victims react differently to their abuse and experiences in court. For some victims, separating from the offender may reduce risk, whereas for others, it may mean increased danger, with other emerging forms of violence (e.g., stalking, death threats). Therefore, a victim's silence does not necessarily mean agreement; on the contrary, it may mean fear or control. The lack of specialized training in victimology among most participants in the study may explain their tendency to view allegations in such cases as straightforward and easily evaluated.

According to the Convention on the Rights of the Child, as ratified by Presidential Decree Nr. 49/90 September 12, children have the right to participate in judicial proceedings involving decisions about their lives when their age and maturity allow for it. In this sense, Walker (2020) warns of the need to guarantee the right of children to participate in family court decisions about their custody according to their age, maturity, and wishes. In Portugal, Article 1906 of the PCC allows for the child to be heard in child custody proceedings. We found that all the magistrates are concerned with listening to children’s opinions about maintaining contact with both parents when their ages allow for it, thus fulfilling what is required by law and recommended by the literature. Although there is no empirical basis for the benefit of forcing children to maintain contact with a parent (Saunders et al., 2016), our analysis of the interviews confirmed that even in cases where a child has expressed a desire not to maintain contact with the father, the judges still have difficulty deciding to give exclusive custody to the mother. Although Article 1906 of the PCC provides for children to be heard, it also states that decisions about child custody should favor agreed-upon contact with both parents and custody sharing. As such, maintaining contact between the child and both parents was found to be a priority factor for children, as also demonstrated by MacDonald (2016). The final determining factor driving the magistrates’ decisions was the influence of the Istanbul Convention. Our analysis revealed that these professionals believed that the Istanbul Convention triggered a series of legislative changes that indirectly influenced their decisions. This perception corroborates the assessment of GREVIO (2019) that Portugal has seen increasing investment in measures to prevent gender-based violence and protect its victims despite not implementing the recommendations of the Istanbul Convention for child custody proceedings with allegations of DV.

The Istanbul Convention is a European norm underlying national legislation. When responding to questions, judges incorporate (to a greater or lesser extent) these principles when making pronouncements on custody cases involving allegations of DV, as well as the factors on which they base their decisions or whatever they can do to defend the victim and prevent DV. All the sections of the interviews were related to the Istanbul Convention, but only the last question explicitly mentioned it. Based on the answers given, we can conclude that the Istanbul Convention is a ratified legal instrument of reference but is not binding for judicial decision-making, which tends to be influenced by extralegal factors, such as the judges’ perceptions of prevention and remediation involving risk to victims, opinions on parenting in the context of DV, and ideas of what is in the best interests of the child without the need to hear from the child in cases involving phenomena such as DV.

The end of a violent intimate relationship is often associated with escalation to threatening behavior, stalking, and emotional abuse (e.g., Elizabeth, 2017; Saunders & Oglesby, 2016). Although only one magistrate in our sample had been trained in victimology, most of the others demonstrated awareness that the end of a violent relationship could trigger an escalation in violence and put the safety of women and children at risk. This aligns with the work of Saunders et al. (2016), who reported that in a sample of 200 magistrates, most (75%) had this knowledge about DV.

Additionally, our analysis of the interviews revealed that these professionals felt that an offending parent may use the children to obtain information about the mother and perpetrate further violence against her. This perception accords with the extensive literature showing that shared custody in the context of DV can help perpetuate violence against women (e.g., Ignjatović, 2019; Miller & Manzer, 2021; Saunders & Oglesby, 2016). It also highlights how offenders often use their children as tools for controlling and coercing their intimate partners (Hayes, 2012; Heward-Belle et al., 2018; Jaffe et al., 2009; Laing, 2016).

Conclusion

This study explored the factors that underlie judicial decisions in child custody cases involving allegations of DV between parents and sought to determine whether these decisions comply with the Istanbul Convention’s directives. Contrary to the Convention’s recommendations, the results indicate that magistrates tend to prefer shared custody arrangements and visitation rights for the offending parent. Despite the well-documented harmful effects of exposure to DV internationally, decisions are not always congruent with the need for prevention and protection of children involved in divorce cases with allegations of DV. Decisions are strongly conditioned by the presumption that maintaining contact with both parents is in children’s best interests, even when there is alleged or proven DV between parents. They are also strongly influenced by the provisions of Article 1906 of the PCC, which favors agreement between parents in child custody cases. Thus, our analysis of the interviews revealed that most decisions derive from mutual agreement about shared custody, such that most mothers do not question their children’s continued contact with the alleged or proven offender. Moreover, once there is agreement between the parents about custody, it is very rare for a court to deliver a decision that contradicts it. Female victims of DV also find that they have to support their children while maintaining contact with the perpetrator or risk having their allegations of DV doubted. This attitude from magistrates puts victims in a catch-22 situation: if they do not report the DV, they will be unprotected and unable to secure exclusive custody, but if they do report it and demonstrate a desire for sole custody, their claims tend to be discredited as an attempt to gain an advantage in the proceedings. On the basis of their responses to victimization, which are dependent on whether they have any training in victimology and whether they have cross-referenced data from other instances, judges develop subjective opinions on the veracity of DV claims and the legitimacy of parents’ victim status, and there are undeniably extralegal factors that influence their decisions.

The presumptions that children will benefit from continued contact with both parents and that determination should favor agreement between the parents in cases of DV contravene the recommendations of the Istanbul Convention because they overlook the empirical evidence that violence is likely to be perpetrated and escalated after separation, as well as the risk that the offending parent will abuse the child directly. In addition, our analysis of the interviews revealed that the law allows little scope to deviate from these presumptions. The only exceptions provided are contained within Article 1906 of the PCC, which allows exclusive custody to be awarded to one parent in cases of DV when the rights and safety of victims are at high risk, such as in situations of mistreatment or sexual abuse of the children involved. Nevertheless, although our analysis showed that the magistrates regard DV as a form of child abuse, their decisions continue to favor shared custody.

This study therefore makes a modest contribution to understanding the need to effectively combat violence against women and DV in general, as well as protect victims. More and better communication is also needed between prosecutors working in the criminal courts and magistrates working in the family courts. In addition, magistrates must give due weight to DV allegations and act in a way that protects victims, who are generally women and children. To facilitate this, these professionals should be trained in victimology, specifically the dynamics of DV, to understand the impacts of exposure to this type of violence and recognize it as a form of child abuse. As children are more likely to suffer emotional, physical, and sexual abuse from the offending parent, shared custody and visitation rights for the offending parent, even with the mother’s agreement, can endanger the safety of both the mother and the children. The family courts have a duty to act on the empirical evidence regarding the dynamics of DV and address any allegations of violence before considering other presumptions about children’s best interests, such as the belief that children should have continued contact with both parents. Such in-depth knowledge of the specifics of DV is essential to ensure that family court decisions do not place victims at greater risk of violence or compromise children’s best interests. Furthermore, magistrates must ensure that children are involved in child custody proceedings not only by listening to them but also by taking their opinions seriously and respecting their right and ability to participate in decisions that will impact their lives.

As this study is qualitative with a limited convenience sample, its results cannot be generalized. Despite delimitation by theoretical saturation, which in itself is an excellent indicator of discursive redundancy, qualitative studies with samples of judges are relatively difficult to carry out, resulting in small samples. In future studies, larger samples, even for more in-depth analysis of the subject, would be desirable.

This study attempts to convey the impact and complexity of DV, to which many children are exposed, even following their parents’ separation or divorce. Future studies must address judicial decisions in child custody cases in which DV between parents is alleged or proven, in terms of understanding both the factors driving these decisions and the impact of these decisions on victims, especially women and children. In this context, it would also be worthwhile to conduct not only qualitative but also quantitative studies, which are based on statistical data concerning the frequency at which allegations of DV occur in custody cases and what custody arrangements are agreed upon when the DV is alleged. Finally, research should evaluate the effects of Portugal’s Directive nr. 5/2019 of April 12, which introduced procedures aimed at overcoming the inadequate communication/cooperation between criminal and family courts, including implementation of an experimental model structure that aims to provide the judges, magistrates, and Public Prosecution Office personnel with standardized guidelines for action in relation to DV.