Coercive Control
Coercive Control: Update and Review
Violence Against Women
2019, Vol. 25(1) 81–104
© The Author(s) 2018
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/1077801218816191
journals.sagepub.com/home/vaw
Evan Stark1 and Marianne Hester2
Abstract
Keywords
1Rutgers University, Newark, NJ, USA
2University of Bristol, UK
Corresponding Author:
Evan Stark, School of Public Affairs and Administration, Rutgers University, Newark, NJ, 07102, USA.
Email: starkevan342@gmail.com
The Adaptation of Coercive Control in Criminal Law
The Crisis in Domestic Violence Policing
The particular circumstances that led governments with very different political orientations to reform their criminal laws related to partner abuse are beyond our scope (but cf. Stark, 2016a, 2016b, 2018). However, the decision to craft new law reflected the shared perception that the focus of the existing criminal justice response on discrete, injurious assaults was too narrow to capture the patterns of coercion and control growing body of research and personal testimony showed were experienced by many abused women who seek protection.
The hiatus between women’s lived experience of abuse and the narrow window to this experience afforded by law and policy was politically manifest in growing tensions between women’s wide-ranging and multi-faceted demands for help and the response, in which criminal sanctions were rare and limited to egregious assaults. In their study of policing in Northumberland, Hester (2006) and Hester and Westmarland (2006) reported rates of attrition of more than 96% from police calls for offenses related to domestic abuse to conviction or punishment for any crime. A large proportion of the male offenders (but not the female offenders) were “repeaters.” But because police responded to each complaint de novo, criminal sanctions were no more likely after a man’s 50th offense than his first (Hester, 2006). A similar pattern has been found more recently for sexual assault.
Proof of a large subpopulation of women who were experiencing long-standing abuse was in plain sight in government surveys, even if they too were officially invisible. To critically assess the government’s claim that violent crimes had declined in Britain, Walby, Towers, and Francis (2014) recorded all violence reported to the Crime Survey for England and Wales (CSEW) by female victims, disregarding the official cap of five on the number of reported crimes that was counted. Their review showed that the actual total of violent crimes women reported had been increasing so dramatically since 2009 that it offset the decline in almost all other forms of reported violent crime during this period, including most violent crime against men. This increase in reported violence against women was accounted for not by an increase in the number of women reporting stranger assaults or partner abuse but by the growing proportion of “high-frequency repeat female victims” among the reporting sample (Walby et al., 2014).
The response from government. The gap between women’s experience and the legal response reflected gendered differences in the understanding of abuse. Kelly and Westmarland (2016) draw on interviews to show how their sample of abusive men in counseling embraced the incidentalist definition of domestic violence in law, which they applied to disconnect their violent acts from one another in time and space. This allowed them to minimize their violence as “not that bad” and to support victim-blaming accounts of the larger sources of problems in their lives. By contrast, when women are asked about “abuse” during health surveys in the general population, what is marked is the frequency, duration, and variety of physical and sexual violence they report, not necessarily its physically injurious nature (Black et al., 2011). Large majorities of women in many countries identify multiple forms of restriction, isolation, and control as abusive (Butterworth & Westmarland, 2015/2016; Hester, Donovan, & Fahmy, 2010). Control tactics endorsed by women in international surveys range from the constraints partners place on their time, spending, socializing, dieting, or other facets of everyday living to the psychological or “gaslight games” used to “make me feel crazy.”
To help close the gap between its public policy and women’s experience of abuse, the British Home Office conducted a public consultation in 2012 which concluded by identifying “coercive control” as the best framework for a new “cross-governmental response.” In contrast to the previous emphasis on discrete assaults, the new “Working Definition” (WD) defined Coercive Control to include “Any incident or pattern of incidents of controlling, coercive or threatening behaviour, violence or abuse between those aged 16 or over who are or have been intimate partners or family members regardless of gender or sexuality” (UK Home Office, 2013). “Coercion encompassed psychological, physical, sexual, financial and emotional abuse.” Controlling behavior was defined as “making a person subordinate and/or dependent by isolating them from sources of support, exploiting their resources and capacities for personal gain, depriving them of the means needed for independence, resistance and escape and regulating their everyday lives.”1 From 2012–2014, the new WD replaced more than 20 other, conflicting definitions that guided the funding and delivery of services to abuse victims throughout Britain. But the new definition had no legal standing.
After the new WD was announced, abuse-related calls to police in England spiked, rising more than 31% between 2012 and 2015. Police reported they were feeling “overwhelmed” by the new demand (HMIC, 2015). HMIC (2014, 2015) reports also noted another disturbing trend. The police response in England was symptomatic of the revolving door created when the same women sought help multiple times over many years for assaults, sexual attacks, and other coercive or oppressive acts that were not, in themselves, sufficiently grave or injurious to merit a serious crime charge or conviction under the existing approach. As these families became “well known” because they turned up repeatedly on call logs and arrest reports, police responded in increasingly perfunctory and even punitive ways, blaming victims for their abuse or insisting that “everyone is going to jail” and so arresting non-offending women along with abusive partners. Government reports (HMIC, 2014) traced this poor response to shortcomings in professional police leadership, morale, and training.
U.K. police forces were already expected to administer the Domestic Abuse, Stalking, and Honour Based Violence (DASH) risk assessment in cases of domestic abuse (Robinson, 2010), a specialized tool similar to the Campbell DAS.2 The DASH helps ration the allocation of scarce justice resources according to a woman’s relative risk as determined at the scene. Only cases which score “high risk” on the DASH merit further police involvement via the assignment of an Independent Domestic Violence Advisor (IDVA) and the review of subsequent “safety plans” by Multi-Agency RiskAssessment Teams (MARACs) at monthly meetings. Other cases are referred to a refuge (shelter) or encouraged to seek further service (Stark, 2016b).
The provision of an IDVA could make interventions more effective than in the past (Coy & Kelly, 2011). But the overall approach merely underlined the stark reality that continued investments in case management and “self-help” were required to protect women from men whom police acknowledged posed the highest risk. Women’s “space for action” expanded post-separation or when they entered a refuge. But it contracted thereafter, as their struggles resumed, with the offending partners still at large. SharpJeffs, Kelly, and Klein (2018) reported,
Removing themselves from the immediate control of an abusive man was . . . only the first step. Over 90 per cent experienced post-separation abuse, which interfered with both being and feeling safe. The limited effectiveness of criminal and civil law enforcement required women to undertake a huge amount of “safety work.” . . . The prevalence of perpetrator manipulation of statutory agencies post-separation also revealed the importance of factoring ongoing support needs into responses to domestic violence. (p. 13)
Information collated from the DASH questionnaires made the need for an enhanced criminal justice response apparent. Aggregating the historical data revealed that longstanding physical and sexual abuse, threats, stalking, isolation, and numerous instances of control comprised the context for a large proportion of current complaints in all risk levels (Stark, 2018). A tool designed to calculate women’s current risk had inadvertently helped to open a window into the larger predicament they faced in abusive relationships.
This was the background in December 2015, when the British Parliament took the unprecedented step of making “coercive or controlling behavior in a family relationship” a criminal offense (under s.76 of the Serious Crime Act of 2015) in England/ Gibraltar and Wales, carrying up to 5 years imprisonment.3 There were several highprofile prosecutions under s.76 in the spring and summer of 2016. But by the end of August, there were just 59 convictions in all of England for “coercive and controlling behavior” (Hill, 2016). Public awareness of the new understanding of coercive control became more widespread after The Archers (2016), a hugely popular radio soap, made “gas lighting” and other elements of the coercive control of Helen Titchener by her husband Rob a long-running theme. Unable to withstand his relentless tormenting, Helen “snapped” and stabbed her husband. She was arrested and charged.4 The story made the front pages of the tabloids and a fundraising page put up in Helen’s name raised nearly 200,000 pounds. When Helen was tried and acquitted because of the history of Rob’s abuse by a jury, prominent jurists, government officials, and media figures, even the Prime Minister’s Office, offered a supportive statement. Section 76 was used to illustrate the sort of legal protection that might have prevented the stabbing.
S.76 makes no reference to gender, only covers current partners, and requires a victim to show “fear” or “distress.” In defending the new law, the Solicitor General repeatedly referenced the Home Office Definition which includes physical and sexual abuse as components of coercive control. But elsewhere, he insisted that s.76 only be applied to stalking and other forms of “psychological abuse,” leaving open the question of which existing laws to apply to the concurrent patterns of ongoing violence, sexual assault, intimidation, and control (Wiener, in press).
In marked contrast to England, the Scottish Executive adopted a gendered definition of domestic abuse in 2000, which explicitly positions it as both a cause and a consequence of gender inequality.5 With this background, Scotland’s prosecution service, the Crown Office and Procurator Fiscal Service (COPFS), and Police Scotland, introduced a raft of changes that favored a “consistent and robust investigative, enforcement and prosecution approach” (Police Scotland & COPFS, 2013). The new offense in Scotland represents one of the most radical attempts yet to align the criminal justice response with a contemporary feminist conceptual understanding of domestic abuse as a form of coercive control.
On February 2, 2018, by unanimous vote (119-0), the Scottish Parliament passed the Domestic Abuse Act of 2018.6 It extends to persons living separately, including current or former partners, and references the offender’s “reasonable” understanding that his behavior will frighten or otherwise harm the targeted partner, rather than proof of those effects by the victim. The offense of coercive and controlling behavior was considered in Scotland, but rejected in favor of a new offense intended to capture a range of behavior that is already criminal, as well as financial abuse and other elements identified in the Home Office WD, which might not be captured by existing laws within a single offense. The Domestic Abuse (Scotland) Act of 2018 sets a maximum penalty for conviction in the High Court of 15 years, with exposure of children to the abuse an aggravating factor in sentencing.
Critical response. The new offenses in England and Scotland were meant to supplement current legal responses to partner assaults by enhancing the probability that the subset of chronic abusive partners would be identified and charged with a broad range of their offenses. Proponents argued that any drop in the population of “repeaters” responsible for dozens, even hundreds, of service contacts over years could justify the added resources needed to properly identify, investigate, and prosecute these serious crimes (Stark, 2018). Giving justice professionals a robust legal tool could relieve their frustration with “failed” intervention, help shift their attention from victim safety to offender accountability, and so remove an important context for victim-blaming. The new law would also facilitate a corresponding shift among community-based services from “safety work” to “empowerment.” Incorporating women’s experiential definitions of abuse into criminal law would also broaden the perceived legitimacy of legal remedies, particularly among groups who lacked access to resources or other alternatives. Conversely, the extent to which the new offense echoes the range of issues advocates/specialists are already addressing creates an important basis for linkage between statutory and voluntary sectors (Stark, 2018; Wiener, in press)
Debate about these new laws in the legal and criminological literature has centered on their timing, the relative wisdom of using “more law” to manage partner abuse, and the prospect that they will be enforced. Critics have worried that even the little progress made in policing the most common partner assaults will be lost if the emphasis shifts from incidents to “patterns” or from physical injury to such intangibles as “control” and “subordination” (FitzGibbon & Walklate, 2017; Walby & Towers, 2017). The mistaken association of coercive control with “psychological abuse” also risks leaving “real” partner violence outside the crime’s spectrum, not merely isolated assaults (Burman & Brooks-Hay, 2018; FitzGibbon & Walklate, 2017; Padfield, 2016; Walby & Towers, 2017). A related set of issues involves what might be termed “governmentalism,” that is, the interplay of the form or substance of a particular law or policy and the constraints imposed through its implementation/enforcement (Douglas, 2015; Tolmie, 2018). From this perspective, the minimalist English/Welsh offense (s.76) makes an interesting contrast to the more comprehensive Scottish crime.
Whether the new laws will affect positive changes in the lives of abused women and their children or further confound a response which even critics of the new law agree is currently ineffective will be shaped, in part, by greater clarity about the categories of persons and acts covered by the legislation and by how openly experiences are shared by those implementing the very different approaches in Scotland, England/ Wales.
Early critics of s.76 warned that broadening the definition to “psychological abuse” could open the door for men to claim they were being “controlled” by female partners and justify yet further inattention to simple partner assaults occurring outside the context of coercive control. A variant of this argument highlighted the incident-specific crime victim data mustered to make the case that partner assaults on women were increasing. Walby et al. (2014; Walby & Towers, 2017) worried that the power of the assault data she and her colleagues had mustered to challenge the Government’s claim that violent crime had dropped would be dissipated if these assaults were aggregated historically to highlight a subclass of targeted “victims.” The incident-focused criminal law still contemplates an act that transpires at an ascertainable instant, meaning that, in theory, one can capture the commission of an archetypical crime in a photograph. It is asking too much, critics suggested, to expect police and courts to replace the transactional crime norm with a victim model predicated on showing “a continuing course of conduct” (Douglas, 2015; Padfield, 2016; Tolmie, 2018).
S.76 posed a particular challenge to a justice culture attuned to equating the seriousness of assaults with visible injuries while ignoring or minimizing the subjective outcomes now being emphasized such as “fear,” “serious alarm and distress,” “subordination,” and the “disruption of daily activities.” Critics warned that evidencing the less tangible elements of coercive control would push police and the justice system more generally outside their comfort zone while requiring even more experiencebased testimony from victims than before (Tolmie, 2018). To this end, Padfield (2016) suggests, the new offense “simply increases the difficulties facing police and prosecutors when deciding what charge or what charges to lay” (p. 149). The dilemma posed by Padfield (2016) echoes a warning commonly voiced in this arena: that criminal justice professionals are disinclined to engage in policy initiatives on this issue.
The particularity of coercive control, the strategic ways in which a specific abuser individualizes his abuse based on his privileged access to personal information about his partner, could also prove elusive for policing. A related challenge is how to distinguish the “coercive” element from the normative ways in which men micro-manage women’s enactment of gender roles (such as how they cook or clean) by default, simply because they are women (Bishop & Bettinson, 2018).
It is too soon to determine whether the new offenses will lead to promised improvements. The number arrested under the offense is increasing, but fewer than 700 cases have been prosecuted under s.76 in England since its inception.7 Even if these men committed thousands of assaults and other “bad acts” that might not have surfaced in the past, they accounted for a tiny fraction of the hundreds of thousands of cases involving partner abuse reported in this period. The vast majority of police forces have brought at least one case under the new law, though a handful of departments are disproportionately represented.
Stark’s (2018) qualitative review of 18 cases prosecuted under s.76 concluded that the law was being correctly applied to historical patterns of abuse that included multiple elements of coercion and control; that prosecutors were tapping repetitive physical and sexual abuse to show the “serious effect” required by the law; and that the vast majority of convicted offenders have received significant time in prison, up to a maximum of 6.6 years.8 Only one woman has been charged thus far under s.76 and there has been no noticeable change in the way other cases of partner abuse are handled, as critics warned. However, none of the prosecutions of s.76 reviewed appear to have included sexual assaults or stalking, even though the concurrence of these crimes with other elements of coercive control is well known and because the Solicitor General made a point of emphasizing that the new law would protect current partners from stalking.
By specifying multiple elements of coercion and control including violence and sexual assault, extending coverage to former partners, stiffening the top sanction, and shifting the weight of evidence to the perceived intent of the offender, the Scottish bespoke offense responds to the more serious shortcomings critics identified in s.76. It remains to be seen whether Scotland will devise means of implementation as innovative as its new offense.
A certain amount of resistance by police, Crown Prosecution Service (CPS), and the judiciary to an offense as broad ranging as s.76 was anticipated. But the selective prosecutions we are witnessing may also reflect the steep learning curve by which police, the CPS, and the judiciary are becoming acclimated to an offense that challenges conventional thinking about the nature of criminal behavior and about what constitutes a remediable harm. Even so, simply giving public voice to the experiencebased “wrong” of coercive control has already transformed the narrative representation of male partner abuse in the various media to emphasize insults to dignity and personhood rather than only physical injury. Establishing whether such changes are more than ephemeral and monitoring their dissemination across the landscape are important challenges to researchers.
Toward a New Research Agenda
Efforts to frame an appropriate legal response to coercive control should not discourage sustained work to flesh out the original construct, whether as part of a useable typology or as a stand-alone paradigm of partner abuse. At a minimum, qualitative and quantitative research is needed to clarify the interplay of violent, merely coercive, and psychological dimensions of this form of abuse in different population and relational contexts; specify which elements of coercive control, either separately or through their combination, elicit which outcomes and for whom; determine which elements/effects are contingent on preexisting status vulnerabilities (such as inequality) and which are relationship or context-specific; and map the survival, coping, resistance, and accommodation strategies as victimized partners (and children) craft “space for action” in the face of tyranny. The most obvious evidence of “control” is provided by abusive tactics, such as “he monitored my time” or “denied me money.” But in the most vulnerable populations—undocumented women or women of color, for instance—individual deprivations are confounded by economic inequalities, cultural bias, and institutional barriers that have yet to be integrated into the model of harm, a process that Ptacek (1999) called social entrapment.9
Should we restrict coercive control to the most devastating type of partner violence (Johnson, 2008); limit its use to nonviolent means of psychological abuse (Crossman, Hardesty, & Raffaelli, 2016), as s.76 does; or define it by its intent and consequence, dominance/subjugation, whatever means are deployed, as in the new Scottish offense? Coercively controlling violence has been shown to be more frequent and severe than other types of violence (Johnson, 2008) and more likely to persist after separation (Ornstein & Rickne, 2013). Other studies have demonstrated, however, that a pattern of coercive controlling behaviors can precede, motivate, or increase the likelihood of violence in relationships (Graham-Kevan & Archer, 2008; Hardesty et al., 2015) and that the dynamics and consequences of coercive control without violence appear similar to those of coercive control with violence (Crossman et al., 2016). Similarly, Dutton, Goodman, and Bennett (1999) found that adverse mental health consequences of violence were independently associated with coercive controlling behaviors.
The remainder of this review addresses four current concerns: differentiating theeffects of violent and controlling tactics in coercive control, coercive control in lesbian and gay relationships, measuring coercive control, and children’s experience of coercive control.
Violence and Control
Stark (2007) defines coercion as “the use of force or threats to compel or dispel a particular response” (p. 228), while control refers to “structural forms of deprivation, exploitation, and command that compel obedience indirectly” (p. 229). When coercion and control occur together, he argues, the result is a “condition of unfreedom” (p. 205) that is experienced as entrapment.
Most cases of coercive control include physical and/or sexual assaults, but a significant proportion do not (Lischick, 1999, 2009). Anderson (2008) reported that 3% of the women responding to the National Violence Against Women Survey (NVAWS) were experiencing high levels of control, but no violence. Control may continue unabated long after violence has ended. In a national sample in Finland, Piispa (2002) found that a population of older women who had not been physically assaulted for an average of 10 years reported higher levels of fear, depression, and other symptoms of abuse than younger women who were still experiencing partner assaults. This raises a question about the importance formerly ascribed to violence relative to “control” and nonviolent coercive tactics in eliciting entrapment as well as the many other negative outcomes linked to abuse.10
Legal scholars (Tolmie, 2018) have highlighted a dilemma posed by findings like these. The research reinforces a core claim of Stark (2007), that the level of “control” independently increases a woman’s risk of serious and fatal injury (e.g., postseparation violence) because it undermines her autonomous capacity to resist or escape abuse. Nonetheless, this work has the potential to undercut understandings of coercive control in some contexts. For example, during separation, particularly where control is high but there was not much physical violence in the relationship, there is a tendency to assume a case involves a more “benign” type of violence, “separation engendered violence” (Jeffries, 2016, p. 14). Conversely, a similar assumption that the violence involved is of the more benign type could lead to a decision not to prosecute coercive control on the facts, even if the offense of coercive control is viewed as the appropriate charge. In this instance, as pointed out by Douglas (2015), one of the effects may be to exclude some very valid experiences of domestic violence from criminalization.
Hardesty and colleagues (Crossman et al., 2016; Hardesty et al., 2015) have focused on the dynamics in relationships that make control “coercive” as distinguished from other control dynamics (e.g., conflict dynamics, unrealistic expectations, gendered ideologies), even where violence is minimal or nonexistent. They used the DominanceIsolation subscale of Tolman’s Psychological Maltreatment of Women Inventory (PMWI) to measure “control” in a subsample of divorcing mothers (n = 147) who met specific criteria related to experiences of violence and coercive control. In subsequent interviews with the mothers, researchers identified the most common pattern of control as “constraint through commitment, where one or both partners constrained the other or themselves to maintain their marriage and manage parenting consistent with social conventions of heterosexual marriage.” A minority of the reported situations of control involved “constraint through force.” In contrast to upholding cultural conventions of marriage to maintain the family in the first pattern, in constraint through force, husbands “used social norms regarding gender and family as a tool for depriving wives of the freedom to be themselves and liberty to do what they wanted.” According to the women, controlling behaviors were not isolated to particular conflicts or situations; rather, they were present early in the relationship and increased over time. Women shared feeling imprisoned by their partners—literally, in their homes, and while they were physically separated from their former partners for 1–34 months (M = 8.2, SD = 6.7 months).
This work reinforces our belief that coercive control by a male partner is the most common and most devastating context for women’s help-seeking because of abuse. But coercive control is not a type of violence. In perhaps as many as a third of all cases, fear, constraints on autonomy, belittlement, and other facets of abuse elicit “entrapment” without any notable incidents of violence; prior level of control predicts post-separation violence/sexual assault and fear of mediation. These findings support the importance given to the “level of control” in risk models of femicide (Glass, Manganello, & Campbell, 2004). Even where violence is an element of coercive control, it may no longer be present or it may not have been the most salient element of the abuse, as far as female victims are concerned.
Coercive Control in Same-Sex and LBGT Relationships
Coercion and control may also be used by same-sex partners and by or against transgender males or females or others whose marginalized status gives them little social power to draw on in response. The largest random population survey to date in the United States, the NVAWS (Tjaden, Thoennes, & Allison, 1999), found that the highest rates of rape, physical assault, and/or stalking were committed by men living with women (30.4%), followed by men living with men (15%), women living with women (11%), and women living with men (7.7%). Gender differences also defined the types of coercive control reported, with men far more likely than women to threaten and commit physical and sexual assault/coercion against other men, while lesbians reported being made to do housework and more forms of emotional abuse (Donovan & Hester, 2014). A well-designed study of 184 self-identified gay men and lesbians found that men and women were equally likely to engage in coercive control and that almost 3 times as many couples engaged in “mutually violent control” (12.5%), where both partners used violence and control tactics, as engaged in coercively controlling violence (4.4%) (Frankland & Brown, 2014).
Renzetti (1992), in her pioneering work on domestic violence in lesbian relationships, argues that a gender inequality and power analysis can be applied to same-sex contexts. Despite a lack of preexisting gender power roles to constrain them, she found that who perpetrated the violence and abuse was still determined to a significant extent by unequal power and power relations. Indeed, she reported, the greater the disparity of power, the more severe the physical and psychological abuse.
Stark (2007, 2018) focused on coercive control as an extension of “gendered inequality” among heterosexual partners. For Stark (2007), the coercive control of women in heterosexual relationships is distinguished from the use of coercive control in institutional or other relational settings by the degree to which its scope, substance, duration, dynamics, societal significance, and individual consequences are inextricably bound to persistent forms of sexual inequality and discrimination. While Stark expects that sexual inequality and heteronormativity will shape abuse in all relational settings to some extent, he argues that the knowledge base is lacking to determine the scope, dynamics, or consequences of coercive control in LBGT relationships.
Johnson (2006) also hesitates to apply models based on heterosexual domestic abuse to same-sex contexts. His rationale is that lesbian or gay violence does not take on patriarchal family values, and therefore does not exhibit “intimate” or “patriarchal” terrorism.
From quite a different angle, a series of authors focusing on individual psychological approaches to gay men’s intimate relationships argue that the unequal gender power model does not apply at all to same sex domestic violence. Such relationships should be deemed “gender neutral” (Island & Letellier, 1991) or involving difficulties in conflict resolution and attachment fears “rather than the intent to control one’s partner” (Stanley, Bartholomew, Taylor, & Landolt, 2006, p. 31).
Other critics reject the application of the gender inequality model to LGBT intimaterelationship abuse from a post-modern perspective. Based on research with gay men, Kwong-Lai Poon (2011) argues that we should “move away from the abstract, but fixed notions of victims and perpetrators while allowing us to see multiple and sometimes contradictory aspects of their personality” (p. 124). Ristock (2002) also rejects the “victim” and “perpetrator” binary found in heterosexual contexts. Based on interviews with 102 mostly lesbian women, she identified “a range of contextual factors that surround abusive relationships” (Ristock, 2002, p. 57), including contexts of LGBT “invisibility” and of “normalisation.” “The existence of these contexts may increase the probability of experiencing or committing violence. . . . But this does mean they cause violence or that individual women in such contexts make risky partners” (Ristock, 2002, p. 57).
In their more recent work on the Canadian Indigenous Two Spirit LGBTQ people, Ristock, Zoccole, Passante, and Potskin (2017) have broadened their emphasis on the contextual factors that shape abuse experiences to include societal forces. They found that experiences of domestic abuse not only were situated and impacted by structural factors such as poverty, unemployment, and homelessness—which featured heavily in the lives of Indigenous communities—but also bore the imprints of racist policies of the state.
Taking a more positive view of the relevance of the “gender inequality and power” model, Donovan and Hester (2014) argue that the societal context of the heterosexual family, “patriarchal” and heteronormative sex/gender unequal values, also form a backdrop for and are also likely to infuse lesbian, gay, bisexual, transgender, and queer (LGBTQ) relationships. As a consequence, an analytical framework that sees gender inequality as key has some relevance to understanding domestic abuse in same-sex relationships. Donovan and Hester’s (2014) U.K. LGBT community study, involving a national survey of 746 LGBT individuals and follow-up interviews, showed that coercive control is a feature of same-sex relationships and is experienced in similar ways across same-sex sexualities, whether lesbian, gay male, bisexual, or male or female homosexual. Severity and impact of domestic violence and abuse (DVA) were found to increase with the number of events and especially where coercive control—a combination of more than one form of abuse (emotional, physical, and/or sexual)—was used by the perpetrator. Contrary to what others have predicted, “mutual” abuse and situational couple violence did not characterize the violence and abuse being reported. There were some gendered differences of note between those survivors who were abused by male perpetrators, regardless of their sexuality. Here, physical violence and physically coercive sexual violence were more typical. Abusive partners in female, same-sex relationships more typically used emotional violence and emotionally coercive sexual violence. The survey showed that men in same-sex relationships were significantly more likely than the women to experience forced sex, refusal to recognize safe words, and refusal to comply with requests for safer sex. They were also significantly more likely to experience financial abuse. The survey also showed that women in same-sex relationships were significantly more likely to want to try harder to please a partner as a result of experiencing abusive behaviors. Women were also more likely to be parents, and this was used as an aspect of theabuse by female partners.
Donovan and Hester’s (2014) follow-up interviews also highlighted gendered aspects in same-sex relationships. Gender inequality norms are both confounded and reinforced in abusive relationships. What may be termed “relationship rules” can be seen to characterize the gender inequality norms of heterosexual love in so far as the abusive partner is the key decision maker, setting the terms for the relationship (traditionally associated with masculinity in men), and the survivor is positioned as responsible for the emotional life and care of the abusive partner and the relationship (traditionally associated with femininity in women). However, by unhooking these behaviors from embodied women and men, we can unpack how abusive partners in same-sex relationships exhibit both stereotypically feminine and masculine behaviors. Thus, Donovan and Hester (2014) argue that relationship rules exist in abusive relationships across sexuality and gender.
The authors mentioned here agree that “intersectionality” is an important facet in any analysis of domestic abuse in LGBT intimate relationships, as the structural discrimination and oppression associated with positioning as LGBT, or in relation to ethnicity, race, and other identity statuses, is also used by abusers in their repertoire of abuse (Donovan & Hester, 2014) and enables us to take into account wider experiences, meanings, and interventions related to domestic violence (Renzetti, 1992).
Measuring Coercive Control
Early critics warned that broadening the definition of partner abuse would put comparability at risk and diminish the public impact of “hard” data on violence and injury (Gordon, 2000). Walby and Tower (2017) raise similar objections to measuring patterns of coercive controlling behaviors on government crime surveys. For the sake of comparability and objectivity, they argue, the focus should be retained on formal legal definitions of physical violence causing physical harm or threat of harm and on sexual violence such as rape, that are more often considered criminal offenses. They also fear that the gender and violence configurations they identified will be lost if all assaults against a given victim are grouped with concurrent violent and other offenses.
While defining physical violence may seem relatively straightforward, otheroffenses such as sexual violence may be less so. For instance, in England and Wales, the offense of sexual violence is linked to consent (whether or not the victim consented to the event), and interpretation is subject to gendered bias (Hester & Lilley, 2016). Using physical injury as a key measure of severity also has limitations, as some of the most injurious violence involves women using weapons against their male partners in protection and/or self-defense (Hester, 2013; Johnson, Leone, & Xu, 2014). In the context of coercive control, meanwhile, the cumulative effect of repeated low-level assaults can be more devastating than injurious, but isolated attacks (Stark & Flitcraft, 1996).
Another approach to measurement has been to tease out evidence of coercive control from preexisting data sets. Johnson et al. (2014) reworked their analysis of the NVAWS survey using data on past, rather than current, intimate relationships to provide an operationalization of coercive controlling and situational couple violence, their rationale being that as current perpetrators fear exposure and victims fear retribution from their abuser, both groups are likely to be underrepresented in general population samples. In their re-analysis, severity was assessed via a Severe Violence Scale based on the items they deemed to have been conventionally identified as severe violence. A Coercive Control Scale was constructed from a subset of nine survey items that dealt with non-violent control tactics used by the respondent’s partner. Using cluster analysis involving the Coercive Control Scale with a threshold of five revealed two clusters: high control, which the authors equated with coercive controlling violence, and low control, which they equated with situational couple violence. While confirming patterns in previous findings (Graham-Kevan & Archer, 2003; Johnson, 2006, 2008), the results were deemed more accurate, showing that 22% of women experienced coercive controlling violence from ex-husbands, while only 5.4% of men experienced coercive controlling violence from their ex-wives. Situational couple violence was perpetrated more equally by men and women (7.4% of ex-husbands, 3.9% of exwives). Coercive controlling violence was found to involve a wider variety of acts of violence, more frequent violence, and more injuries and psychological distress than situational couple violence. It should be noted, however, that impact data in the NVAWS were limited to physical injuries sustained in the most recent incident of partner violence.
As Myhill (2017) points out, focusing on coercive control is crucially important in devising measurements because that provides and reflects the “multiple tactics of coercion and control employed by primary perpetrators” (p. 39) of domestic violence. Focusing only on events that are deemed criminal offenses, as Walby and Towers do, may exclude the many other threats that make up ongoing coercive control and have detrimental impacts. While Johnson et al. (2014) include a broad range of behaviors, their focus constrains them to ignore the estimated 25–30% of coercive control situations where physical and sexual violence play little or no role (Hardesty et al., 2015; Lischick, 2009), or are of secondary importance (Beck & Raghavan, 2010).
In the United Kingdom, the CTS approach also underpins the CSEW, where it is used in the interpersonal violence module to assess frequency of domestic violence. Despite limitations regarding impact, the CSEW interpersonal violence module includes questions that may reflect the impact of coercively controlling behavior. Myhill (2015) re-analyzed the CSEW data using these questions to provide a measure of severity and typology of coercive controlling violence for intimate partners. Respondents were characterized as having experienced coercive control if they said their partner had both “Repeatedly belittled you to the extent that you felt worthless” and “Frightened you, by threatening to hurt you or someone close to you” (p. 362), deemed to reflect abuse that was ongoing, denigrating, perceived as threatening, and had caused a degree of fear. By contrast, all other respondents who reported physical violence or acts of emotional or psychological abuse were classified as having experienced only situational violence. Myhill’s analysis found that nearly a third (30%, n = 791) of the women who reported some form of DVA were found to experience what could be termed coercive controlling violence and about one in 20 men (6%, n = 52) could be classified similarly (as compared to general prevalence figures from the 2008-2009 CSEW which estimated 24% of women and 12% of men had experienced nonsexual DVA). Experience of coercive controlling violence was found to involve more severe and more frequent physical violence, and was more likely to persist over time, than situational violence. Myhill (2015), using past and current relationship data, thus ends up with similar, albeit slightly higher, results regarding exposure to coercive controlling violence to those of Johnson et al. (2014) using past relationship data.
Hester, Jones, Williamson, Fahmy, and Feder (2017), using a survey on DVA behaviors and impact with men in a clinical setting (family medicine clinics), including questions on coercive control and extensive questions on impact, found an even smaller proportion of heterosexual men might experience what could be termed coercively controlling violence, with 4.4% of 707 men in the sample experiencing this form of DVA.
A final challenge to measurement worth noting involves what might be termed the “embeddedness” of coercive control, that is, the fact that the multiplicity of tactics are deployed in the context of relationships that last a nontrivial time, often over many years. To the extent that coercive control has a duration, it stands with a small subgroup of criminal offenses (harassment and stalking are other “ongoing” offenses) whose measurement has more in common with a comparable class of public health problems (such as HIV-AIDs) than with conventional crimes like assault or robbery that end shortly after they begin and are much like the flu using the public health analogy. As sociologists, we lack ready tools to grasp a social behavior that, without clear beginning or end, nonetheless has a demonstrable impact that is cumulative over time and across social space on a class of victims whose lives and liberties become severely constrained
Children Experiencing Coercive Control
The new understanding of coercive control extends to how we understand the nature, causes, dynamics, and consequences of child harms in abusive situations as well as to how we gauge children’s responses. Based on the known overlap of domestic violence with coercive control and child maltreatment, we can anticipate that coercive control extends to children in a sizable proportion of cases. Counted alongside forms of direct abuse are the many ways in which children are exposed to the coercive control of their mother, used as pawns in control strategies, or are “weaponized” as instruments of the coercion and control. Even when children are direct targets, we consider them “secondary” victims. This is not because the harm they suffer is “collateral damage” or of secondary importance—it is not—but rather because the children are almost always being harmed when, why, and how they are to subordinate the mother. While the strategic logic operating here need not deter emergent child rescue, it highlights the importance of managing coercive control as a spectrum of interrelated harms stemming from a single source.
Early work in the field linked children’s experiences of abuse to their “witnessing.” In an analysis of information from interviews with 26,229 adults in five states, the Centers for Disease Control and Prevention (CDC) estimated that children had witnessed domestic violence in 16.3% of U.S. households, a figure acknowledged to be conservative (Black et al., 2011). Fathers were the largest group of identified perpetrators in these cases and, along with non-cohabitating boyfriends, accounted for 78% of the intimate partner violence incidents witnessed by children, with the percentage of male perpetrators increasing with the level of violence. Although the more global term “exposure” captures the multiple receptors through which children process trauma, witnessing accounts for 65–86% of all exposure.
A common if inexact proxy for coercive control is so-called “psychological” abuse, usually defined to include a range of nonviolent abusive behaviors, many of which, such as isolation and control tactics, are only indirectly related to individual psychology. One of the few surveys to include “psychological” abuse as well as physical violence was a nationally representative telephone survey of the victimization experiences of 4,549 youth, aged 0–17, conducted between January and May 2008. Created by the U.S. Office of Juvenile Justice and Delinquency Prevention, the NatSCEV reported that the proportions of children exposed to psychological and physical abuse were roughly equivalent, both during the same year (5.7 vs. 6.6%) and during their lifetimes (16 vs. 17.9%) (Hamby, Finkelhor, Turner, & Ormrod, 2011). While there is no way to know whether the same youngsters were exposed to both types of abuse, this seemed highly likely from the similarity in proportions. At the least, the survey underlined the importance of considering both physical and psychological abuse as potentially convergent, if not always concurrent, dynamics.
Only a handful of researchers have examined children’s experiences specifically in the context of coercive control (Callaghan et al., 2018; Katz, 2016). This work has produced two findings that offer promising new directions: that children are common targets of coercive control alongside and independently of their mother, and that children exhibit a similar range of strategic/adaptive agency in response, including forms of resistance.
Useful Numbers
Male victims of Domestic Violence can contact Amen on:
Monday -Friday 9am-5pm
- St Luke’ s Hospital 056 778 5000
- Care Doc Kilkenny 1850 334 999
- Garda Kilkenny 056 777 5000
- Garda Carlow 059 913 6620
- Citizens Information 0761 077 910
- Kilkenny LA Housing 056 779 4900
- Child-line 1800 66 66 66
- Rape Crisis Centre 1800 478 478
- Carlow Woman’s Aid 1800 444 944
- Cuan Saor Refuge/Clonmel 1800 576 757
- Oasis House Waterford 051 370 367
- Wexford Women’s Refuge 1800 22 0 444