Figures from crime surveys typically show higher rates of female victimisation, which are often taken as support for a gendered explanation of IPV. However the context of crime surveys must be considered. People, particularly men, do not typically interpret relationship aggression as a criminal behaviour or view their experiences as violent (Straus, 1999; Povey, Coleman, Kaiza, Hoare, & Jansson, 2008). Thus, cueing respondents to think about IPV as a crime is not conducive to accurate reporting and hence, crime survey figures should be interpreted with caution.
Surveys that ask representative community samples of men and women about their experiences in the context of conflict in relationships are scarce (Esquivel- Sãntovena & Dixon, 2012). One exemplar survey that adopted this methodology was the National Family Violence Survey (NFVS; see Straus, Gelles, & Steinmetz, 1980; Straus & Gelles, 1985). The NFVS was conducted in 1975, and again in 1985, with representative U.S. community samples. The surveys measured rates and severity of partner aggression using a systematic measure, the Conflict Tactics Scale (CTS), which is a self-report tool initially developed in the late 1970s and which has since been revised to the CTS2 (see Straus, Hamby, Boney-McCoy, & Sugarman, 1996). Importantly the CTS/CTS2 set IPV in the context of conflict in the relationship (rather than crime or violence), which is arguably more conducive to honest reporting than the aforementioned contexts (Straus, 1999). Respondents are simply asked to report on a range of predetermined behavioural acts that both they and their partner have engaged in during times of conflict with each other.
The CTS2 contains five subscales that distinguish rationale tactics, physical assault, psychological aggression, sexual coercion and injury. Furthermore, minor acts of physical and psychological aggression, sexual coercion and injury are differentiated from more severe forms of these acts, hence less severe acts of physical assault that might not otherwise be considered as constituting IPV (e.g. slapping, pushing, grabbing) are also measured. The behavioural acts listed form clearly defined behavioural categories. Therefore, results can be systematically compared within and across samples. Indeed, this tool allowed the systematic collection of large data sets from which international prevalence and incidence rates have been calculated. This methodology found approximately equal rates (around 12%) of physical partner aggression perpetrated by both sexes in a 12-month period, across time. In terms of lifetime prevalence 28% of respondents reported physical victimisation in 1975 and 22% in 1985.
11.3.2 Rates of Stalking Behaviours
A range of methodologies has been employed to assess stalking (Fox, Nobles, & Fisher, 2011), and akin to the issues previously discussed in relation to estimates of IPV prevalence, even when survey methods have been used, considerable variations in prevalence have arisen. For example, according to the Crime Survey of England and Wales, for the year ending March 2016 one-fifth of women (20.2%) and 9.8% of men reported having experienced stalking since the age of 16. When based on reported victimisation in the last year, the previously identified gender differences remained, with 4.9% of women and 2.4% of men reporting such experiences. As mentioned previously, stalking was defined through a range of repeated explicitly identified behaviours. Based on experience since age 16, 2.8% of men and 9% of women reported being stalked by a family partner.
Data from North America present prevalence estimates that are similar to those reported by the Crime Survey of England and Wales. According to the 2011 National Intimate Partner and Sexual Violence Survey, Breiding, Smith, Basile, Walters, Chen, & Merrick (2014), 15.2% of women and 5.7% of men reported lifetime experience of stalking, and this reduced to 4.2% and 2.1% respectively in the last 12 months. Estimates of stalking prevalence from the Injury Control and Risk Survey between 2000 and 2003 revealed that 4.5% of adults reported ever having been stalked. Women were more likely than men to report this experience (7% vs. 2%). The low prevalence rates reported are likely due to the definition of stalking used. Participants were asked: “Have you ever had someone besides bill collectors or sales people follow or spy on you, try to communicate with you against your will, or otherwise stalk you for more than one month?” and if respondents said yes, they were then asked to select whether the most recent experience was “nothing to be concerned about, annoying, somewhat dangerous or life-threatening” (Basile, Schwan, Chen, & Saltzman, 2006, p. 173). Only respondents who selected the latter two options were then classified as victims of stalking (Basile et al., 2006) in order to reflect behaviours that were both unwanted and fear provoking.
These data broadly support the findings of an earlier meta-analysis of 103 studies, by Spitzberg (2002) who found that the average prevalence of stalking was 23.5% for women and 10.5% for men, and on average, the stalking occurred over a two-year period. The majority of victims across studies were women (75%) and the majority of perpetrators were men (79%). In just under half of cases (49%) the stalking occurred within the context of an intimate relationship. These findings must be qualified by the possibility of response and reporting bias associated with gender, which also affects estimates of stalking victimisation and perpetration in the same way that it affects estimates of partner violence. That is, men are less likely to report their own victimisation, as they are possibly less likely to feel the necessary extent of threat in order for such experiences to be valid (Sheridan et al., 2002; White, Kowalksi, Lyndon, & Valentine, 2002).
11.3.3 Rates of Intimate Partner Stalking
Within the international literature, it has been claimed that victims of stalkers are most likely to be current or former intimates or spouses (Melton, 2000). Although the Crime Survey of England and Wales reported above suggests that this is not necessarily the most frequently cited form of stalking in general, the apparent gendered nature of intimate partner stalking appeals to researchers of violence against women more broadly. Intimate partner stalking has been isolated as a “special case” of stalking for five main reasons (Logan & Walker, 2009), as shown in Box 11.1.
Researchers vary in the methodology they employ in conducting surveys to determine rates of IPV and stalking. It is therefore difficult to compare rates across surveys, countries and time. Consequently, it is important to carefully consider the methodology used in studies before generalising figures to represent the population at large. Agreed definitions, terms and consistency in methodological approach will allow researchers to produce comparable studies and resultant prevalence and incidence rates. However, considering large scale self-report community studies (e.g. the National Family Violence Survey) it seems likely that an estimate between 20% and 30% for the lifetime prevalence rate for men and women experiencing physical violence is a sensible approximation (Dixon & Graham-Kevan, 2010), and between 10% and 35% for the lifetime prevalence of stalking. In addition, the evidence suggests considerable overlap between these two phenomena.
BOX 11.1 STALKING WITHIN THE CONTEXT OF AN INTIMATE RELATIONSHIP
Relationships in which stalking arises are characterised by a range of violent and abusive behaviours (e.g. Cupach & Spitzberg, 2000; Davis, Ace, & Andra, 2000). Indeed, Douglas and Dutton (2001) have even gone so far as to suggest that the stalking of current or former intimate partners, incorporating psychological and/or controlling behaviours as outlined in Table 11.1, is a form of IPV itself. In a study of 120 male IPV perpetrators, it was found that 30% reported having also stalked their partners (Burgess et al., 1997). In addition it has been found that between 30% and 65% of stalkers had engaged in violence towards the intimate partners that they had more recently stalked (e.g. Kienlen, Birmingham, Solberg, O’Regan, & Meloy, 1997).
It enables the perpetrator to draw upon a wider range of stalking tactics that are influenced by their intimate knowledge of the victim, in particular their knowledge of specific fears, concerns and vulnerabilities (Mohandie, Meloy, McGowan, & Williams, 2006; Sheridan & Davies, 2001).
It increases the likelihood that perpetrators will both threaten the victim and use violence (James & Farnham, 2003; Rosenfeld, 2004). In addition,
it has been found that violence is more likely to be used by stalkers who first threaten to use it, than those who do not (Brewster, 2000). As typified in the case that opened this chapter, stalking is also a risk factor for intimate partner homicide. For example, McFarlane, Campbell, Wilt, Sachs, Ulrich, & Xu (1999) reported that 76% of partner homicide victims had been stalked prior to being killed.
Such stalking is likely to have been initiated during the course of the relationship rather than once the relationship has terminated (Mullen, Pathe, & Purcell, 2000). Depending on the study, between 25% and 80% of female intimate partner stalking victims reported that the stalking started during the relationship (e.g. Hackett, 2000; Logan, Cole, Shannon, & Walker, 2006; Melton, 2007). Stalking during a relationship has been found to lower the likelihood of a woman leaving (Logan, Cole et al., 2006). In addition, however, there is evidence that being stalked after the cessation of a relationship may place a victim at increased risk of ongoing and increased severity violence (Logan, Walker, Jordan, & Campbell, 2004; Melton, 2007).
The occurrence has been found to be associated with greater psychological distress for victims, ranging from anxiety, and post-traumatic stress disorder symptoms (Logan, Walker, Jordan, & Leukefeld, 2006) to symptoms of psychiatric diagnoses and severe depression (Blaauw, Winkel, Arensman, Sheridan, & Freeve, 2002). Moreover, should stalking occur within a previously violent intimate relationship, this has been found to further compound the emotional distress caused to the victim (e.g. Brewster, 2002).
11.4 RISK FACTORS AND THEORIES
Theories serve to explain how it is that a phenomenon occurs and identifies the circumstances and factors that lead to its occurrence (i.e. risk factors). A broad range of risk factors have been implicated in IPV and stalking, and are typically identified through comparing the characteristics of individuals who engage in the behaviour of interest to those who do not. In contrast to the empirical evidence base relating to IPV, the stalking literature is less comprehensive and our knowledge about the characteristics of stalkers arises either through sampling clinical groups who have been arrested and who are likely to have greater levels of psychopathology, or university students.
A useful framework within which to consider the role of risk factors is that provided by the Nested Ecological Model (Dutton, 1985), adapted from Bronfenbrenner’s (1979) Ecological Developmental Model. Within this framework risk factors are identified by their relative proximity to the individual. As such, risk factors are conceptualised as occurring at one of four levels: macrosystem (broad societal/cultural influences); mesosystem (social group influences, such as church/school); microsystem (interpersonal relationship influences), and ontogenetic (individual developmental/internal influences). Table 11.2 provides examples of identified risk factors for IPV and stalking within each level of this model.
Table 11.2 Examples of risk factors associated with male perpetration of intimate partner violence and stalking
Ecological level Intimate partner violence risk factors Stalking risk factors
Macrosystem Patriarchal values/systems
Mesosystem Unemployment
Peer group has pro-violence norms
Unemployment
Microsystem High relationship conflict
Low relationship satisfaction
Controlling behaviours within intimate relationship
Unstable personal relationships
End of relationship
High relationship conflict
Psychological abuse of partner prior to break-up
Controlling behaviours within intimate relationships
Social isolation
Ontogenetic Witnessing IPV as a child
Child abuse
Borderline personality traits
Antisocial personality traits
Drug use/abuse
Alcohol use/abuse
Pro-violence attitudes
Social problem-solving deficits
Negative attitudes towards women
Jealousy
Poor impulse control
Substance abuse
Schizophrenia
Borderline personality traits
Narcissistic personality traits
Attachment
Erotomanic delusions
Low empathy
High trait anger
Jealousy
High education
Poor impulse control
Table 11.2 shows that a very broad range of factors have been implicated in both IPV and stalking, and that these two behaviours appear to be indicated by some common risk factors, which is not surprising given the occurrence of stalking within intimate, and previously violent intimate relationships. Despite the implication of multiple risk factors, the most popular theoretical explanations of IPV have traditionally focused on the role of single factors (see Bowen, 2011aa, for a more in-depth examination of relevant theories of IPV).
Feminist or gendered theories have to date been most influential in accounting for IPV, despite the fact that little empirical support currently exists for this position. This perspective views IPV as predominantly acted out by men toward their female partner, caused by societal rules that support male dominance and female subordination (Dobash & Dobash, 1979; Yllö, 2005). Hence, “patriarchy’” is viewed as a direct cause of men’s violence toward their female partner (Bell & Naugle, 2008). While, it is accepted by the gendered position that some women may be violent to their male partner, it is purported to predominantly occur out of self-defence or retaliation to his aggression (Dobash & Dobash, 2004; Respect, 2008). Consequently, violence towards women is viewed as special, unrelated to other forms of violence or crime (Dixon, Archer, & Graham-Kevan, 2012). Therefore, in the long term feminism seeks to change the root cause of men’s violence to women, by overturning patriarchal social structures, to eradicate violence to women (Dutton, 2006). Despite these admirable aims, there is little empirical support for the expected strong relationship between patriarchy and IPV (e.g. Sugarman & Frankel, 1996; Stith, Smith, Penn, Ward, & Tritt, 2004).
In fact, research that tests the hypothesis that either gender may be perpetrators and/or victims of IPV has found that men and women engage in violent acts at approximately equal rates. This gender symmetry is shown in the results of the National Family Violence Survey (Straus et al., 1980; Straus & Gelles, 1985), described above and in other gold standard pieces of research, such as Archer’s (2000) meta-analysis. Archer examined gender differences in the perpetration of heterosexual IPV in 82 independent studies that examined rates of physical violence by men and women. In total a combined data set of 64,487 people was produced for analysis. Results showed that women were slightly more likely than men to use physical aggression against a partner (Cohen’s d (effect size) = −0.05), yet overall women were slightly more likely to be injured (d = +0.15) and require medical treatment for their injuries than men (d = +0.08). Archer also reported that the sample studied was an important moderator of effect size, with younger and non-clinical samples more likely to be in the female direction. For example, studies using shelter samples produced very high effect sizes in the male direction, community and student samples were slightly more likely to be in the female direction. Such research findings highlight that men and women can be both aggressors and victims of physical violence within their intimate relationships, which undermines the gendered perspective as a complete explanation for IPV.
An equally popular theory that has been applied to understanding IPV is social learning theory (SLT, Bandura, 1977). According to SLT, violent and abusive behaviours and pro-violence beliefs are learned during childhood through either the direct experience or observation of these behaviours and attitudes modelled by others, most typically parents. The likelihood that such behaviours will be exhibited depends on whether they are perceived to be reinforced. Woodin and O’Leary (2009) note that behavioural learning is deemed to occur through processes of both classical and operant conditioning, and also through cognitive med
iational processes (p. 46). At its most basic then, an SLT account of IPV predicts that violence between parents observed by their children, leads their children to use violence in intimate relationships – the so-called intergenerational transmission of violence which is the most widely tested assumption of the SLT account of IPV. However, the resulting empirical evidence suggests that this association is not straightforward.
For example, in a 20-year prospective study of 582 youths and their mothers, Ehrensaft, Cohen, Brown, Smailes, Chen, & Johnson (2003) examined the prospective role of childhood disruptive behaviour disorders, childhood neglect and abuse, parenting practices, and inter-parental violence as risk factors for adult IPV. It was found that a diagnosis of childhood conduct disorder was the single most important risk factor for IPV, increasing the odds of it occurring by seven times. However, exposure to inter-parental violence and childhood abuse both remained significant predictors even when childhood conduct disorder was entered as a predictor in the model, although conduct disorder partially mediated the effect of child abuse. The results of this and other prospective longitudinal studies (e.g. Capaldi & Clark, 1998; Lussièr, Farrington, & Moffitt, 2009; Magdol, Moffitt, Caspie, Fagan & Silva, 1997; White & Widom, 2003) confirm that the relationship between exposure to violent models during childhood and adult IPV is weak, and is influenced by a range of additional factors, most notably childhood conduct disorder and antisocial personality traits. Along with the gendered theory, SLT has been a major influence on the skills-based components of treatment programmes, and many current programmes combine a gendered approach with cognitive-behavioural skills training to a greater or lesser degree (Bowen, 2011a).
Forensic Psychology Page 51
